Encyclopedia of Sustainability Science

and Technology Series

Editor-in-Chief: Robert A. Meyers

Ripudaman Malhotra Editor

Fossil Energy
Second Edition
A Volume in the Encyclopedia of
Sustainability Science and Technology,
Second Edition
Encyclopedia of Sustainability Science
and Technology Series

Editor-in-Chief
Robert A. Meyers
The Encyclopedia of Sustainability Science and Technology series (ESST)
addresses the grand challenge for science and engineering today. It provides
unprecedented, peer-reviewed coverage in more than 600 separate articles
comprising 20 topical volumes, incorporating many updates from the first
edition as well as new articles. ESST establishes a foundation for the many
sustainability and policy evaluations being performed in institutions
worldwide.
An indispensable resource for scientists and engineers in developing new
technologies and for applying existing technologies to sustainability, the
Encyclopedia of Sustainability Science and Technology series is presented at
the university and professional level needed for scientists, engineers, and their
students to support real progress in sustainability science and technology.
Although the emphasis is on science and technology rather than policy, the
Encyclopedia of Sustainability Science and Technology series is also a com-
prehensive and authoritative resource for policy makers who want to under-
stand the scope of research and development and how these bottom-up
innovations map on to the sustainability challenge.

More information about this series at https://link.springer.com/bookseries/

15436
Ripudaman Malhotra
Editor

Fossil Energy
Second Edition

A Volume in the Encyclopedia of Sustainability

Science and Technology, Second Edition

With 173 Figures and 68 Tables

Editor
Ripudaman Malhotra
Malhotra Energy Consultancy
San Carlos, CA, USA

ISBN 978-1-4939-9762-6 ISBN 978-1-4939-9763-3 (eBook)

ISBN 978-1-4939-9764-0 (print and electronic bundle)
https://doi.org/10.1007/978-1-4939-9763-3
1st edition: © Springer Science+Business Media New York 2013
2nd edition: © Springer Science+Business Media, LLC, part of Springer Nature 2020
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software, or
by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with
regard to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Science+Business Media,
LLC, part of Springer Nature.
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Series Preface

Our nearly 1000-member team recognizes that all elements of sustainability

science and technology continue to advance as does our understanding of the
needs for energy, water, clean air, food, mobility, and health, and the relation of
every single aspect of this vast and interconnected body of knowledge to
climate change. Our Encyclopedia content is at a level for university students,
professors, engineers, and other practicing professionals. It is gratifying for our
team to note that our online First Edition has been heavily utilized as evidenced
by over 500,000 downloads which of course is in addition to scientists’
utilization of the Encyclopedia and individual “spin-off” volumes in print.
Now we are pleased to have a Living Reference on-line which assures the
sustainability community that we are providing the latest peer-reviewed con-
tent covering the science and technology of the sustainability of the earth. We
are also publishing the content as a Series of individual topical books for ease
use by those with an interest in particular subjects, and with expert oversight in
each field to ensure that the second edition presents the state-of-the-science
today. Our team covers the physical, chemical and biological processes that
underlie the earth system including pollution and remediation and climate
change, and we comprehensively cover every energy and environment tech-
nology as well as all types of food production, water, transportation and the
sustainable built environment.
Our team of 15 board members includes two Nobel Prize winners (Kroto
and Fischlin), two former Directors of the National Science Foundation (NSF)
(Colwell and Killeen), the former President of the Royal Society (Lord May),
and the Chief Scientist of the Rocky Mountain Institute (Amory Lovins). And
our more than 40 eminent section editors and now book editors, assure quality
of our selected authors and their review presentations.
The extent of our coverage clearly sets our project apart from other publi-
cations which now exist, both in extent and depth. In fact, current compendia
of the science and technology of several of these topics do not presently exist
and yet the content is crucial to any evaluation and planning for the sustain-
ability of the earth. It is important to note that the emphasis of our project is on
science and technology and not on policy and positions. Rather, policy makers
will use our presentations to evaluate sustainability options.
Vital scientific issues include: human and animal ecological support sys-
tems, energy supply and effects, the planet’s climate system, systems of

v
vi Series Preface

agriculture, industry, forestry, and fisheries and the ocean, fresh water and
human communities, waste disposal, transportation and the built environment
in general and the various systems on which they depend, and the balance of all
of these with sustainability. In this context, sustainability is a characteristic of a
process or state that can be maintained at a certain level indefinitely even as
global population increases toward nine billion by 2050. The population
growth, and the hope for increase in wealth, implies something like a 50%
increase in food demand by as early as 2030. At the same time, the proportion
of the population that lives in an urban environment will go up from about 47%
to 60%. Global economic activity is expected to grow 500%, and global
energy and materials use is expected to increase by 300% over this period.
That means there are going to be some real problems for energy, agriculture,
and water, and it is increasingly clear that conflicting demands among biofuels,
food crops, and environmental protection will be difficult to reconcile. The
“green revolution” was heavily dependent on fertilizers which are
manufactured using increasingly expensive and diminishing reserves of fossil
fuels. In addition, about 70% of available freshwater is used for agriculture.
Clearly, many natural resources will either become depleted or scarce relative
to population.

Larkspur, CA, USA Robert A. Meyers, Ph.D.

November 2019 Editor-in-Chief
Volume Preface

To many in the sustainability community, fossil energy is an anathema.

Continued use of fossil resources – oil, coal, natural gas – poses threats to
the environment through the emission of pollutants and greenhouse gases. The
fact that they are a limited or exhaustible resource means that in the future we
could either run out of them or their extraction will get progressively harder to
a point that it takes more energy to extract them than would be derived from
their use. Using fossil energy is clearly not sustainable, and the world has to
look to renewable resources for long-term survival.
This is an encyclopedia about the science and technology of sustainability.
The word, sustainability shares its root with sustenance. Any discussion of
sustainability must therefore include discussions of sustenance, and in the
context of modern society, sustenance stems from the use of energy. We derive
energy from a number of different sources. Annual global consumption of
primary energy in 2010 (the time of publication of the First Edition) was on the
order of 500 exajoules (EJ), about 85% of which comes from fossil resources.
From 500 EJ/year, the global energy demand is expected to rise to somewhere
between 1000 and 1500 EJ/year by the middle of this century. The drivers for
this increased demand are already in place. Large segments of China, India,
and Brazil are poised to increase their standard of living – and the concomitant
energy demand – substantially. If the average energy consumption in the Asia-
Pacific region were to reach the current global average, which is about half of
what is consumed in Europe and Japan, the demand would increase by an
amount equal to the total energy consumption in the North America region.
As we get ready to publish the Second Edition of the Encyclopedia of
Sustainable Science and Technology, I pause to reflect on some of the devel-
opments since the publication of the First Edition. Global energy consumption
has increased 12% to 560 EJ. Contribution of renewable resources like wind
and solar energy has more than doubled, but 85% of global primary energy is
still provided by fossil energy sources of oil, coal, and natural gas. As a result,
34 billion metric tonnes (Gt) of CO2 were emitted from the energy use in 2017.
Production of shale oil by hydraulic fracturing in the USA has had a major
impact on the global supply and price of oil. The relative speed with which this
resource can be produced and shut down has allowed global oil price to
stabilize close to $50/bbl.
The threat to global climate posed by greenhouse gases (GHG) and the role
of humans in it are now clearer than when these concerns first gained

vii
viii Volume Preface

prominence in the 1980s. In its 2013 Summary for Policy Makers the Inter-
governmental Panel on Climate Change (IPCC) spells out the underlying
physical basis of the threat (Climate Change 2013: The Physical Science
Basis, Fifth Assessment Report of the IPCC. www.climatechange2013.org).
Calls for quickly curbing GHG emissions can be heard not just from environ-
mental groups but also from many business communities and national gov-
ernments. Yet, global trends show little progress.
The Papal Encyclical issued in July 2015 also drew attention to the growing
threat of climate change and its disproportionate impact on the impoverished
(Pope Francis, “Encyclical Letter Laudato Si ” July 2015. https://laudatosi.
com/watch). Pope Francis called for “changes of lifestyle, production and
consumption, in order to combat this warming or at least the human causes
which produce or aggravate it.” His statements about combating climate
change received much attention, but there was another deeper message in his
statement, the one about consumerism and social injustice it engenders that
seems to have been largely ignored. The pope recognized the need for vastly
expanding renewable energy sources, but also noted that, “(f)or poor countries,
the priorities must be to eliminate extreme poverty and to promote the social
development of their people.”
Around the same time as the Pope’s Encyclical, the World Bank also issued
Sustainable Development Goals for the world which lists goals and targets in
17 areas: eradicating poverty, providing adequate food and clean water, reduc-
ing gender inequality, taking urgent action to combat climate change, and
ensuring access to affordable reliable energy is listed among the goals.
Achieving most of these goals requires increasing global energy supply.
Speaking about the enormous progress in alleviating poverty around the
world, Dr. Jim Yong Kim, President of World Bank, noted that over a billion
people have been lifted out of poverty in the last 25 years, and he could foresee
lifting another billion in not too distant future.
The progress Dr. Kim noted was made on the backs of coal and oil. Can we
afford to do the same to help the next billion without risking catastrophic
consequences of climate change? The SDG of removing poverty runs up
against the need to curb CO2 emissions. The World Bank and many other
institutions think that widespread deployment of renewables, such and wind
and solar power, can provide the necessary energy. While renewables have a
place in the future mix of global energy sources, we cannot and should not
count on intermittent sources alone to meet the demands of industries and
megacities. Unfortunately, the one CO2-free energy source that is capable of
generating the required scale of power, nuclear, is something that the World
Bank does not support developing.
Amid much fanfare about 200 nations signed the Paris Agreement to curb
global GHG emissions in December 2015. It was an unprecedented achieve-
ment given the previous failed attempts. All nations acknowledged the peril
the world faces from climate change, which is engendered by continued
emission of GHG, principally from the use of fossil fuels. The countries
pledged to cut down their GHG emissions either in absolute numbers or
relative to an expected business-as-usual (BAU) scenario. The individual
countries determine the GHG reductions they pledge to make. However,
Volume Preface ix

there is no mechanism of punitive action to force the countries to stick to the

pledged contributions except a public shame. The intended nationally deter-
mined contributions (INDCs) are reported to the UN, and the emissions of each
country are measured and reported in an agreed-upon standard way, and both
these are available to the public. The lack of enforcement is a recognition of
political realities; any agreement that had forced compliance would not have
had the support of many countries.
The Paris Agreement sets a goal of limiting the rise in global temperature to
2  C above the pre-industrialization level, with a stretch goal of limiting the
rise to 1.5  C. Even achieving the 2  C target is a daunting challenge and
requires limiting total emissions to about 350 Gt CO2. The world currently
emits about 36 Gt CO2 from energy use each year. To meet the stated goal the
world must reduce GHG emissions to 17 Gt-CO2/year by 2035 and achieving a
net zero emissions by 2050. As of this writing, 180 countries have submitted
their INDCs. However, even if all these nations complied with their pledged
INDCs, the global emissions would rise to 55 Gt-CO2/year by 2035! The only
hope for success is if, as envisioned in the Paris Agreement, the signatory
nations review and ramp up their pledges every few years. Will we?
Under BAU the annual energy consumption is expected to rise to more than
double by 2050. Even under an all-renewable, all-electric scenario, which
could conceivably avoid two-thirds of the primary energy, the world would
require generating over 100,000 TWh of electricity relative to 24,000 TWh in
2014; in other words, more than quadrupling the current global electricity
production. There remains a large gap between the targeted reductions in GHG
emissions and those achieved or pledged by nations, chiefly because the
energy needed to alleviate poverty is huge and cannot be met by renewable
power sources like wind and solar.
I find it appalling that the word energy appears only three times in the
31-page Agreement. The word appears twice on page two where the Confer-
ence of Parties “acknowledges the need to promote universal access to sus-
tainable energy in developing countries, in particular in Africa, through the
enhanced deployment of renewable energy.” The third time the word is used is
on page 31 as part of the name of UN’s IAEA: International Atomic Energy
Agency. No wonder then that there is such huge chasm between the target
reductions in CO2 and the pledged INDCs.
Against this rather dreary picture, I struggle to not fall into despair and keep
hope alive. What helps is the recognition that there are thousands of scientists
and engineers, particularly of the younger generation, who are stepping up
their efforts at developing CO2-free sources of energy and technologies that
vastly improve the energy efficiency of a whole host of systems: industrial
processes, appliances, transportation, and communication, among others. Suc-
cessful and widespread deployment of those technologies is the pathway for a
sustainable future.

San Carlos, CA, USA Ripudaman Malhotra

November 2019 Volume Editor
Contents

Fossil Energy: Introduction ............................... 1

Ripudaman Malhotra
Oil and Natural Gas: Global Resources . . . . . . . . . . . . . . . . . . . . . . 5
Peter J. McCabe
Petroleum and Oil Sand Exploration and Production .......... 17
James G. Speight
Oil Shale Processing, Chemistry, and Technology . . . . . . . . . . . . . . 47
Vahur Oja and Eric M. Suuberg
Hydraulic Fracturing .................................... 85
Fred Aminzadeh
Petroleum Refining and Environmental Control and
Environmental Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
James G. Speight
Internal Combustion Engines, Developments in . . . . . . . . . . . . . . . 133
Timothy J. Jacobs
Alaska Gas Hydrate Research and Field Studies . . . . . . . . . . . . . . 185
S. L. Patil, A. Y. Dandekar, and S. Khataniar
Gas to Liquid Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Marianna Asaro, Ronald M. Smith, and Burtron H. Davis
Natural Gas Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Raub W. Smith and S. Can Gülen
Coal and Peat: Global Resources and Future Supply . . . . . . . . . . . 309
Mikael Höök
Coal and Other Mining Operations: Role of Sustainability . . . . . . 333
Sandip Chattopadhyay and Devamita Chattopadhyay
Coal Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Gerald H. Luttrell and Rick Q. Honaker
Coal to Liquids Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Marianna Asaro, Ronald M. Smith, and Burtron H. Davis

xi
xii Contents

CO2 Reduction and Coal-Based Electricity Generation . . . . . . . . . 427

János Beér
Pulverized Coal-Fired Boilers and Pollution Control . . . . . . . . . . . 439
David K. Moyeda
Mitigation of Airborne Pollutants in Coal Combustion:
Use of Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
Bradley R. Adams
CO2 Capture and Sequestration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
Abhoyjit S. Bhown, Grant Bromhal, and Gabriel Barki
Policy Instruments for Mitigating Carbon Dioxide Emissions . . . . 519
Adam Rose and Brandt Stevens
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
About the Editor-in-Chief

Robert A. Meyers
President: RAMTECH Limited
Manager, Chemical Process Technology, TRW Inc.
Post doctoral Fellow: California Institute of Technology
Ph.D. Chemistry, University of California at Los Angeles
B.A., Chemistry, California State University, San Diego

Biography

Dr. Meyers has worked with more than 20 Nobel laureates during his career
and is the originator and serves as Editor in Chief of both the Springer Nature
Encyclopedia of Sustainability Science and Technology and the related and
supportive Springer Nature Encyclopedia of Complexity and Systems Science.

Education

Postdoctoral Fellow: California Institute of Technology

Ph.D. in Organic Chemistry, University of California at Los Angeles
B.A., Chemistry with minor in Mathematics, California State University,
San Diego

Dr. Meyers holds more than 20 patents and is the author or Editor in Chief
of 12 technical books including the Handbook of Chemicals Production
Processes, Handbook of Synfuels Technology, and Handbook of Petroleum
Refining Processes now in 4th Edition, and the Handbook of Petrochemicals

xiii
xiv About the Editor-in-Chief

Production Processes, now in its second edition (McGraw-Hill), and the

Handbook of Energy Technology and Economics, published by John Wiley
& Sons; Coal Structure, published by Academic Press; and Coal Desulfuri-
zation as well as the Coal Handbook published by Marcel Dekker. He served
as chairman of the Advisory Board for A Guide to Nuclear Power Technology,
published by John Wiley & Sons, which won the Association of American
Publishers Award as the best book in technology and engineering.
About the Editor

Ripudaman Malhotra completed his Ph.D. in Chemistry in 1979 at the Uni-

versity of Southern California, Los Angeles, CA, and both M.Sc. and B.Sc. in
Chemistry in 1973 and 1971, respectively, at Delhi University, India. He was a
research chemist at SRI International from 1979 to 2015 (retired), and during his
36-year tenure at SRI, he worked extensively on the chemistry of processing fossil
fuels. His detailed mechanistic studies of these systems have resulted in innovative
processes that achieve desired product selectivity and increased efficiency. As
someone deeply engaged in energy research, he was acutely aware of the looming
energy crisis, which was being exacerbated by the potential of global climate
change. He broadened his research interests into studying alternate resources such
as biomass and application of biotechnology in the areas of energy, chemicals, and
the environment. In 2005, he joined Hew Crane and Ed Kinderman to co-author A
Cubic Mile of Oil: The Looming Energy Crisis and Options for Averting It, which
was published by the Oxford University press in 2010. Among his technical works
are 100 papers in archival literature, co-authorship of a book on nitrations, the
editing of a book on combinatorial materials development, and co-editorship of a
book on advanced materials. He is a section editor of Encyclopedia of Sustainable
Science and Technology and an active member of the Energy and Fuels Division
of the American Chemical Society. In 2005, he was named an SRI Fellow, the
highest award SRI bestows on its employees for excellence in research, and in
2018, he was named a Fellow of the American Chemical Society. He was honored
with Storch Award in Fuel Sciences, Energy and Fuels Division, American
Chemical Society, in 2015; Glenn Award, Division of Fuel Chemistry, American
Chemical Society, in 1991; and Newcomb Cleveland Prize, American Association
for the Advancement of Science, in 1987.

xv
Contributors

Bradley R. Adams Department of Mechanical Engineering, Brigham Young

University, Provo, UT, USA
Fred Aminzadeh University of Southern California, USC Viterbi School of
Engineering, Los Angeles, CA, USA
FACT Inc., Santa Barbara, CA, USA
Marianna Asaro SRI International, Menlo Park, CA, USA
Gabriel Barki NETL’s Mission Execution and Strategic Analysis Site Sup-
port Contract, KeyLogic Systems, Pittsburgh, PA, USA
János Beér MIT, Cambridge, MA, USA
Abhoyjit S. Bhown Electric Power Research Institute, Palo Alto, CA, USA
Grant Bromhal US DOE National Energy Technology Labaratory, Pitts-
burgh, PA, USA
Devamita Chattopadhyay ERG (Eastern Research Group, Inc.), Chantilly,
VA, USA
Sandip Chattopadhyay Tetra Tech Inc., Cincinnati, OH, USA
US Environmental Protection Agency, Office of Pollution Prevention and
Toxics, Washington, DC, USA
A. Y. Dandekar Institute of Northern Engineering, University of Alaska,
Fairbanks, AK, USA
Burtron H. Davis Center for Applied Energy Research, University of Ken-
tucky, Lexington, KY, USA
S. Can Gülen Bechtel Infrastructure & Power Inc., Reston, VA, USA
Mikael Höök Department of Earth Sciences, Uppsala University, Uppsala,
Sweden
Rick Q. Honaker Department of Mining Engineering, Mining and Mineral
Resources Engineering University of Kentucky, Lexington, KY, USA
Timothy J. Jacobs Department of Mechanical Engineering, Texas A&M
University, College Station, TX, USA

xvii
xviii Contributors

S. Khataniar Institute of Northern Engineering, University of Alaska, Fair-

banks, AK, USA
Gerald H. Luttrell Department of Mining and Minerals Engineering,
Holden Hall Virginia Polytechnic Institute and State University, Blacksburg,
VA, USA
Ripudaman Malhotra Malhotra Energy Consultancy, San Carlos, CA, USA
Peter J. McCabe School of Earth, Environmental and Biological Sciences,
Queensland University of Technology, Brisbane, QLD, Australia
David K. Moyeda GE Energy, Irvine, CA, USA
Vahur Oja Department of Chemical Engineering, Tallinn University of
Technology, Tallinn, Estonia
S. L. Patil Institute of Northern Engineering, University of Alaska, Fair-
banks, AK, USA
Adam Rose Price School of Public Policy, and Faculty Fellow,
Schwarzenegger Institute for State and Global Policy, University of Southern
California, Los Angeles, CA, USA
Raub W. Smith GE Power, Schenectady, NY, USA
Ronald M. Smith SRI Consulting, Menlo Park, CA, USA
James G. Speight CD&W Inc., Laramie, WY, USA
Brandt Stevens BK Stevens Consulting, Sacramento, CA, USA
Eric M. Suuberg Division of Engineering, Brown University, Providence,
RI, USA
 40–50 years or more than 60 years even with the
Fossil Energy: Introduction steadily increasing oil consumption. In addition,
there are also unconventional accumulations of
Ripudaman Malhotra these hydrocarbon resources, and extracting
Malhotra Energy Consultancy, San Carlos, them requires development of new technologies.
CA, USA In the case of oil, the unconventional resources are
oil sands, oil shale, and heavy oil. Unconventional
resources of natural gas are coal bed methane,
tight gas, shale gas, and gas hydrates. These
unconventional resources are vast and have the
The challenge of achieving a sustainable future is potential of more than doubling our resource
in being able to balance the energy requirements endowment.
for the billions of people, so they can lead healthy Exploration and production of oil from sedi-
and productive lives against the need to preserve mentary deposits and oil sands is the subject of a
the environment by not running into its limits in chapter by Speight (▶ “Petroleum and Oil Sand
its ability to supply the resources or act as a sink Exploration and Production”). The processes for
for the waste [1]. The collection of chapters in this recovering oil could be a simple matter of drilling
section examines the current status, assesses the into the formation with the oil flowing to the
resource potential of the various fossil fuels, surface under its own pressure, or it may require
examines the technologies for using them, and injection of gases, fluids, and surfactants to coax it
reviews the environmental impact of their use. to flow. In extreme cases, it may even require
Petroleum and natural gas have similar origin underground combustion of a portion of the
and often occur together in geologic formations, resource to release the oil. Speight describes the
and their global distribution and production is chemical and physical factors that govern the flow
discussed together in the chapter by McCabe of oil and the technology options currently avail-
(▶ “Oil and Natural Gas: Global Resources”). able. In a different chapter, Speight (▶ “Petroleum
This chapter makes clear the distinction between Refining and Environmental Control and Envi-
reserves and resources. Reserves represent only ronmental Effects”) provides an account of the
that fraction of the resource base that can be different processes such as distillation, catalytic
economically recovered using current technology. cracking, hydrotreating, reforming, and
These are not fixed quantities as both technology deasphalting used in the refining of crude oil.
and economics change over time. Global annual The chapter also deals with environmental effects
production (and consumption) of oil in 2010 was of the gaseous, liquid, and solid effluents from
31 billion barrels and of natural gas was around these processes. Production of oil from shale is
120 trillion cubic feet. In energy units, they cor- principally achieved by retorting of shale, or other
respond to 180 EJ of oil and 120 EJ of natural gas thermal processes including in situ pyrolysis. Oja
[2]. The current reserves are estimated at 1,236 and Suuberg (▶ “Oil Shale Processing, Chemis-
billion barrels of conventional oil (7,500 EJ) and try, and Technology”) detail the chemistry and
6,545 tcf of natural gas (6,500 EJ). The current technology of these processes in the chapter.
reserves to production ratio (R/P) are about 40 for A new chapter by Aminzadeh (▶ “Hydraulic
oil and about 55 for natural gas. The R/P ratio has Fracturing”) makes a clear distinction between oil
often been mistakenly taken as the time to exhaus- shale and shale oil. The former is the rock-bearing
tion, but new discoveries as well as advances in precursors to oil, kerogen, and requires thermal
technology add to the reserves. In the case of oil, processing to generate oil, while the latter is oil
for example, the R/P ratio has stayed around (or gas) that has already been produced by
© Springer Science+Business Media, LLC, part of Springer Nature 2020 1
R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_920
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2019, https://doi.org/10.1007/978-1-4939-2493-6_920-4
2 Fossil Energy: Introduction

geologic forces but it is trapped in low permeabil- ▶ “Coal to Liquids Technologies” reviews the
ity shale. The chapter describes the recent devel- different approaches like pyrolysis, direct lique-
opments in hydraulic fracturing and its impact on faction, co-processing with petroleum, and indi-
the geopolitics. rect liquefaction that first converts coal into
Gas hydrates represent a particularly notewor- syngas and then uses FT or other conversion pro-
thy resource and are reviewed in the chapter by cess to make liquid fuels.
Patil et al. (▶ “Alaska Gas Hydrate Research and The internal combustion engines that power these
Field Studies”). They are naturally occurring ice- vehicles covert only about 20–30% of the energy in
like substances in which methane or other light the fuel to motive power. Increasing the efficiency of
gases are trapped in the cage structures formed by the engines used in transportation represents a sig-
water molecules. They are stable under certain nificant opportunity to reduce future demand for oil,
conditions of temperature and pressure, and can as well as reduce the carbon footprint of the cars,
be found at many places around the world at trucks, and planes. Jacobs (chapter ▶ “Internal
depths of several hundred meters below the sea- Combustion Engines, Developments in”) pro-
bed and in the permafrost in the polar region. vides a detailed account of the thermodynamics
Global estimates of gas hydrates are on the order of the different kinds of engines and of the various
of 500,000 tcf, about a hundred times the proved technologies under development for improving
reserves of natural gas. Apart from being a poten- the efficiency of the internal combustion engines.
tially very important source for energy in the These include strategies such as engine down-
future, the gas hydrates are also important from sizing, turbocharging, better controls, variable
the perspective of climate change. A general geometry engines, variable valve timing, homo-
warming could release substantial amounts of geneous charge compression engines, and waste
methane, which – being a potent greenhouse heat recovery.
gas – would reinforce any global warming. Höök (chapter ▶ “Coal and Peat: Global
Oil remains the prized fuel. It has a high energy Resources and Future Supply”) reviews the global
density and as a liquid it is well suited for trans- coal and peat resources, which are substantially
portation. The transportation sector relies on oil larger than conventional oil and gas resources.
for over 90% of its energy needs (the remainder Peat and coal are also distributed more evenly
being mostly furnished by coal – via electricity) around the globe. The proved reserves of coal
[3]. Given the importance of liquid fuels, there is are estimated at between 800 billion and a trillion
considerable interest in converting coal and natu- metric tons, representing roughly 20,000 EJ. Most
ral gas, the other hydrocarbon resources, into oil. coal is derived from trees and ferns that grew
In a pair of chapters, Asaro, Smith and Davis some 250–300 million years ago during the car-
(▶ “Gas to Liquid Technologies” and ▶ “Coal to boniferous period. Aerobic bacteria generally
Liquids Technologies”) provide an overview of decompose most plant matter into CO2, but
the various processes for converting natural gas when a tree fell into a swamp and was buried
and coal to liquids. For the second edition, these with limited exposure to oxygen, it could be par-
chapters were revised by Davis, and these reviews tially preserved. That phenomenon occurs even
cover the basic chemistry and catalytic technolo- today and is evidenced in the formation of peat
gies behind the different approaches. The chapter (very young coal) in bogs. Unlike coal, peat has a
on ▶ “Gas to Liquid Technologies” deals with low calorific content and its commercial use as an
steam reforming of methane, autothermal energy resource is limited. However, peat is
reforming, and partial oxidation approaches for important from global carbon cycle. Peat bogs
producing syngas, including strategies for manag- store vast amounts of carbon, but the product of
ing the heat and mass transfer through the use of anaerobic decomposition, methane, contributes to
different reactor technologies. The chapter next greenhouse gases in the atmosphere such that
covers the conversion of syngas into liquids by under steady-state conditions they comprise only
Fischer-Tropsch (FT) synthesis. The chapter on a slight sink of carbon. The situation changes with
Fossil Energy: Introduction 3

fires as peat bogs become major source of carbon solar. In 2010, the global production of energy
in the atmosphere. Human activities, such as was 21,325 trillion kWh, which is equivalent to
drainage of the area for agriculture, exacerbate 77 EJ. However, since 68% of electricity is
the situation, as more peat gets exposed and the derived from burning fossil fuels at an average
swamp is not wet enough to squelch the fire. efficiency of 38%, the total primary energy the
Estimates of the amounts of carbon released world consumed as electricity amounts to 135 EJ,
from peat are on the order of several billion tons, or roughly 27% of the total energy consumed.
the same order of magnitude as from fuels used in Electricity demand in the world is projected to
transportation or power production. rise at a rate greater than for total energy, and
As coal resources are vast, they are likely to meeting that need for additional sources for pro-
continue to contribute substantially to global ducing electricity is critical human welfare.
energy mix. However, their use also presents a Several chapters in this section are devoted to
number of challenges, beginning with mining, production of electrical power. As the fuel with
and through coal preparation and use. Mining the lowest levelized cost of electricity, coal
operations result in excavation of billions of tons remains the dominant fuel for producing electric-
of earth, and depending on the nature of the waste ity, and furnishes over half of the electricity. It
rock and how it is handled, it can lead to contam- results in the emission of about 900 g CO2 per
ination of air, water, and soil systems. The wide- kWh, highest of any fossil resource. The develop-
ranging nature of environmental impact necessi- ment of coal power in future is going to be
tates a commensurately broad portfolio of tech- constrained by the concerns of CO2 emissions,
nologies for the remediation, restoration, and and Beér (chapter ▶ “CO2 Reduction and Coal-
reclamation. Chattopadhyay and Chattopadhyay Based Electricity Generation”) discusses the dif-
(chapter ▶ “Coal and Other Mining Operations: ferent technology options for increasing the effi-
Role of Sustainability”) provide an overview of ciency of coal-fired power plants and reducing the
the environmental impact of coal mining, and CO2 emissions. Ultrasupercritical steam cycles
mining processes in general, as well as review and combined heat and power applications are
the range of the chemical and biochemical strate- relatively straightforward to apply and have the
gies to mitigate the impact. potential to reduce the largest tonnage of
The run-of-mine coal is often laden with non- emissions.
combustible minerals such as shale and clays that Of course, there are other pollutants also that
reduce its heating value. Cleaning and washing of are emitted when coal is used for power produc-
coals removes many of these impurities and tions. Chief among them are oxides of sulfur and
upgrades the coal into a marketable resource, nitrogen and toxic metals such as mercury, sele-
improves the performance of power plants, and nium, and arsenic, as well as carcinogenic PAH-
also reduces potential harmful emissions and dust. containing soot aerosols. Moyeda (chapter
Luttrell and Honaker (chapter ▶ “Coal Prepara- ▶ “Pulverized Coal-Fired Boilers and Pollution
tion”) review these cleaning and preparation oper- Control”) reviews the technologies for scrubbing
ations, and point out the importance of cheap the stack gas emissions. The deployment of scrub-
preparation in extending the resource base. bers for the nitrogen and sulfur oxides, following
Electricity is one of the most useful forms of the Clean Air and Water Act in the 1980s, greatly
energy; its importance to the way we live cannot improved the quality of air and water systems.
be overemphasized. It can be used to perform all Natural gas is the other major fuel for produc-
kinds of useful functions and services that people ing electric power. When used in a gas-fired
desire, and at the point of use, it produces no combined-cycle mode, it produces only 380 g
pollution. Electricity is a secondary source of CO2/kWh. Smith and Gülen (chapter ▶ “Natural
energy, as it must be produced from primary Gas Power”) review the technology for power
sources such as coal, natural gas, oil, nuclear, generation from natural gas. They discuss the
hydro, and, of late, increasingly from wind and different sources of natural gas and provide a
4 Fossil Energy: Introduction

brief history followed by detailed thermodynam- 2. A new chapter by Amizadeh on recovery of oil
ics of gas turbines and modern power plants and gas from tight resources through hydraulic
designs that incorporate combined cycle for max- fracture of shale has been added.
imum overall efficiency. They also review emis- 3. A clear distinction between oil shale and shale
sions from NGCC plants, notably NOx, and oil. The former is the rock-bearing precursors
strategies for minimizing them. The chapter con- to oil, kerogen, and requires thermal pro-
cludes with a discussion of power plant econom- cessing to generate oil, while the latter is oil
ics. In a new chapter, the use of process simulation (or gas) that has already been produced by
in mitigating pollution emissions is reviewed by geologic forces but it is trapped in low perme-
Adams (▶ “Mitigation of Airborne Pollutants in ability shale.
Coal Combustion: Use of Simulation”). 4. Commercial developments in the coal-to-gas
Carbon capture and sequestration (CCS) as a and coal-to-liquid processes have been
strategy for stabilizing atmospheric CO2 levels is updated.
receiving considerable attention. Friedman’s 5. The chapter on mining has been updated to
chapter in the first edition has been revised by describe how to maximize the benefits of min-
Bhown, Bromhal, and Barki (▶ “CO2 Capture ing sector while improving the environmental
and Sequestration”); it provides an overview of and social sustainability.
the state of technology and the different embodi- 6. Technological advances and commercial
ments of CCS. These cover various strategies for developments in carbon capture and sequestra-
first capturing the CO2, which may be performed tion have been added.
pre- or post-combustion or from the atmosphere, 7. A new chapter has been added to describe the
as well as discuss the energy requirements under use of process simulation to mitigating air-
these different scenarios. The chapter also reviews borne pollutants formed during coal
the geochemistry of various options for sequester- combustion.
ing the captured CO2, either in deep saline aqui- 8. A new chapter has been added to describe
fers or in other geological formations. policy options for mitigating CO2 emissions.
In another new chapter, Rose and Stevens
(▶ “Policy Instruments for Mitigating Carbon Diox-
ide Emissions”) present policy options such as car-
Bibliography
bon pricing and cap-and-trade that can be employed
for reducing emissions of carbon dioxide. 1. Meadows D, Randers J, Meadows D (2004) Limits to
The Fossil Energy section of the Second Edi- growth, the 30-year update. Chelsea Green Publishing,
tion of the Encyclopedia of Sustainable Science Hartford
and Technology has been thoroughly revised. 2. BP Magazine (2011) BP statistical review of world
energy, June 2011. www.BP.com
Some salient changes include the following: 3. Energy Information Administration (2010) Annual
energy review. http://www.eia.gov/emeu/aer/pecss_dia
1. Reserves and resources of different fossil fuels gram.html
have been updated to reflect current status.
 Oil sands Sandstones that are naturally impreg-
Oil and Natural Gas: Global nated with bitumen, a highly viscous form of
Resources petroleum. Synonymous with bituminous
sands and tar sands.
Peter J. McCabe Oil shale A rock that contains significant
School of Earth, Environmental and Biological amounts of solid organic chemical compounds
Sciences, Queensland University of Technology, (kerogen) that can generate oil when heated.
Brisbane, QLD, Australia Reserves The discovered, but not yet produced,
amounts of oil or gas that could be extracted
profitably with existing technology under pre-
Article Outline sent economic conditions.
Resources The amounts of oil and gas that have
Glossary been discovered plus the estimated amount that
Definition of the Subject remains to be discovered.
Introduction Shale gas Natural gas that is produced from
Oil and Gas in Conventional Fields shale.
Tight Gas Tight gas Natural gas that is extracted from
Oil Sands rocks with very low porosity and permeability
Coal Bed Methane and which is, therefore, relatively difficult to
Oil Shale and Shale Oil produce.
Shale Gas Unconventional oil and gas Oil and natural gas
Gas Hydrates accumulations that require extraction tech-
Estimated Volumes of Remaining Oil and Gas niques that allow easier flow of oil and gas to
Resources a well (for example, hydraulic fracturing to
Future Directions open pathways or in situ heating to reduce
Bibliography viscosity) or by processing after mining.

Glossary
Definition of the Subject
Coal bed methane (CBM) Natural gas
(methane) that can be extracted from coal Crude oil and natural gas (mostly methane but
beds. Synonymous with coal bed gas and coal including some longer-chain hydrocarbons) have
seam gas (CSG). been used by humans for thousands of years for a
Conventional oil and gas Oil and natural gas variety of purposes including lighting, heating,
that occur in the subsurface and that can be and medicinal uses. However, use was limited
produced using conventional methods of well by access to natural seeps of oil and gas and the
drilling. available technologies to extract and store the
Gas hydrates Accumulations of natural gas that products. The earliest known oil wells were
are trapped in ice-like crystalline solids drilled in China in the fourth century using bam-
consisting of gas molecules surrounded by boo. It was not, however, until the mid-nineteenth
cages of water molecules. Hydrates are stable at century that large-scale production began, when
certain temperatures and pressures within some metal piping allowed deeper drilling into hard
sea-floor sediments and within permafrost in rock. Early commercial production began in
polar regions. Synonymous with gas clathrates. Poland and Romania and this was followed rapidly

© Springer Science+Business Media, LLC 2012 5

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_71
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media LLC, 2012
https://doi.org/10.1007/978-1-4419-0851-3_71
6 Oil and Natural Gas: Global Resources

by drilling successes in the Russian Empire, in share of production, particularly the Middle East
what is now Azerbaijan, and in the United States. and the Soviet Union. The Organization of the
Development of the process of fractional distilla- Petroleum Exporting Countries (OPEC) was
tion at this time fuelled the demand for crude oil, formed in 1960 and by 1972 was responsible for
which could now be economically refined into over 50% of the world’s oil production. The cartel
kerosene for use in oil lamps. The development restricted the supply of oil during the 1970s,
of the internal combustion engine also created a resulting in a fivefold increase in prices. The
demand for oil and this greatly expanded with resulting fall in demand and oversupply led to a
Karl Benz’s invention of the gasoline-powered price collapse in the early 1980s. OPEC has sub-
automobile, patented in 1886. The transport sector sequently never exceeded 42% of world produc-
became the dominant user of oil and currently tion [2] but the rapid economic expansion of many
accounts for 61.4% of oil consumption countries in the first decade of the twenty-first
[1]. Other major uses include lubrication, chemi- century resulted in relatively high prices again as
cal feedstock, and domestic heating. Production demand increased substantially. Today 60% of the
has risen dramatically over the last century world’s oil production is from six countries, of
(Fig. 1) and production of crude oil today is which five belong to OPEC (Table 1).
approximately 73 million barrels per day Although natural gas had long been used in
(a “barrel” is 42 US gallons). limited amounts for illumination, much of the nat-
The United States dominated world oil produc- ural gas that was originally discovered in associa-
tion for the first half of the twentieth century, tion with oil was vented or flared because there was
accounting for over 50% of the world’s annual no way of commercializing the gas. Reliable pipe-
production until 1960. After World War II demand lines to transport gas were not developed until after
for oil increased at a rapid rate and production World War II. Natural gas is now used primarily for
increased fivefold between 1950 and 1972. As the electrical generation in power plants, domestic
world economy rapidly grew, international trade in heating, manufacturing fertilizers, and for other
oil was facilitated by the development of super- industrial purposes. World production of natural
tankers. The dominance of the United States gas grew steadily in the 30 years after World War
decreased as several other regions increased their II, though at a slower rate than oil production.

200

180

160
Oil
140

120
Exajoules

100
Gas
80

60

40

20

0
1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Oil and Natural Gas: Global Resources, Fig. 1 World production of crude oil and natural gas since World War II
expressed in units of energy for comparative purposes. (Based on EIA data [2])
Oil and Natural Gas: Global Resources 7

Oil and Natural Gas: Global Resources, public and private transport systems, are all heavily
Table 1 Major producers of crude oil. (From [1]) dependent on oil and natural gas: they fuel over
Country % of world total 96% of the transport sector. On the negative side
Russia 12.9 they are a major source of CO2 emissions, though
Saudi Arabia 11.8 cleaner than coal per unit of power generated. The
USA 8.3 successful development of cleaner and economi-
Iran 5.4 cally viable alternative fuels is essential in the long
China 5.0 run but the world’s economy today is heavily
Canada 4.0
dependent on the supply of oil and natural gas.
Mexico 3.8
This raises the question as to how much oil and
Venezuela 3.3
natural gas remain in the ground. How long will oil
Kuwait 3.2
and gas resources be available for mankind?
U.A.E. 3.1
Rest of world 39.2

Introduction

Oil and Natural Gas: Global Resources, Table 2 Major Predictions of future shortages of oil began shortly
producers of natural gas. (From [1]) after commercial production started in the late nine-
Country % of world total teenth century. In the first half of the twentieth
USA 19.2 century, there was national concern about imminent
Russia 19.0 and irreversible shortages of oil on at least six occa-
Canada 5.1 sions [3]. In the 1950s M. King Hubbert, a geophys-
Iran 4.6 icist at Shell Development Company in Houston,
Norway 3.4 developed a model of a cycle of production of finite
China 2.9 nonrenewable resources that aimed to predict future
Qatar 2.9 production from analysis of production-to-date and
Algeria 2.6
estimates of the amount of remaining resource in the
Netherlands 2.5
ground [4, 5]. In a series of publications Hubbert
Indonesia 2.5
predicted that US oil production would peak by
Rest of world 35.3
1975 at the latest: actual peak production was in
1970. The peak in US production gave credence to
Production was dominated by the United States, Hubbert’s methodology, which is now commonly
which produced over 50% of the world’s annual referred to as the “Hubbert curve” [6].
production until 1973. Since the early 1970s world Fears arose of global depletion of fossil energy
gas production has risen at a similar rate to oil resources. In 1972 a report for the Club of Rome, a
production (Fig. 1). While the market for natural global think tank, examined known reserves of oil
gas was once limited to the reach of pipelines, and gas and predicted the time of total depletion
natural gas can now be transported globally using under various scenarios [7]. Assuming exponen-
liquefied natural gas (LNG) tankers. This requires tial growth in consumption, the report predicted
removal of impurities and cooling of the gas to depletion of known reserves of oil by 1992 and
approximately – 160  C. LNG now accounts for natural gas by 1994. Even under what was con-
6% of the global market. Today ten countries pro- sidered an optimistic scenario wherein reserves
duce 65% of the world’s natural gas (Table 2). could be increased fivefold by new exploration,
Together oil and natural gas currently comprise oil was predicted to be totally depleted by 2022
54.3% of the world’s total annual consumption of and gas by 2021. These predictions, though influ-
energy and, as such, are a major foundation of the ential at the time, proved unfounded largely
global economy. In particular, the transportation of because consumption did not increase at an
food, raw materials, and goods to market, as well as exponential rate.
8 Oil and Natural Gas: Global Resources

In 1962 Hubbert predicted that world oil pro- decades. Oil and gas resources are often divided
duction would peak in 2000 and, again, in 1969 into conventional or unconventional resources.
predicted the peak would be between 1990 and The definition of what is conventional varies
2000 [5]. As the turning point of the millennium between authors but generally refers to those
approached predictions of “peak oil” increased. resources that can be produced using long-
Most of the predictions suggested a peak in the established methods of well-drilling where the
first years of the twenty-first century. For example, oil and gas can be brought to the surface under
in 1988 Campbell, using the Hubbert curve anal- its own pressure or by pumping without signifi-
ysis, predicted a peak between 2000 and 2005 cant stimulation such as in situ fracturing of the
[8]. Actual production between 2006 and 2010 host rock. Unconventional resources, by contrast,
has been 4.7% higher than in the previous require more complex methods of extraction of
5 years (Fig. 1). However, predictions of immi- the resource from the host rock. They include tight
nent “peak oil” persist and proponents point out gas, oil sands, coal bed methane (CBM), shale
that production has been relatively level for the gas, shale oil, and gas hydrates.
last 6 years (to 2010). Others point to global
political and economic reasons for the recent rel-
atively flat production rate. Oil and Gas in Conventional Fields
The basic problem with the Hubbert curve
analysis is that estimates of remaining resources Oil and gas is generated from organic-rich rocks
change through time. The amount of oil and gas by thermogenic or biogenic processes. When such
on the planet is undoubtedly finite. The amount is “source rocks” are exposed over periods of geo-
large but much of it would be prohibitively expen- logic time to high temperatures the organic mate-
sive to produce using present-day technologies. rial breaks down releasing oil and gas that then
Resource estimates therefore tend to focus on oil migrate toward the surface due to buoyancy. The
and gas accumulations that are anticipated to be temperature required for thermal maturation of a
economically viable for production, either at pre- source rock varies depending on the type of
sent or within the foreseeable future. Advances in organic material but the minimum temperature
technology, however, may make accumulations for oil generation is approximately 50  C and for
technically accessible at lower costs over time, gas generation is 100  C. Most thermogenic oil
thus increasing the size of estimated remaining and gas are generated at depths of 2–6 km. Meth-
resources. In addition, advances in geologic con- ane may also be produced from source rocks by
cepts and subsurface imaging through geophysi- the biogenic breakdown of organic material in
cal techniques have shown oil and gas resources source rocks. Such biogenic gas is produced at
to be substantially more extensive than was at one lower temperatures and at shallower depths.
time thought. As a result, many predictions using Whether of thermogenic or biogenic origin, oil
the Hubbert curve analysis have been far less and gas can be trapped by buoyancy in “reser-
successful than Hubbert’s prediction of US peak voirs” in porous rocks beneath an impermeable
production. Hubbert [5] in 1962, for example, layer of rock typically at depths of over 1 km. The
predicted US natural gas production would peak permeable nature of the host rock, high pressures
in 1975 but production in 2010 was at an all time related to depth of burial, and the concentration of
high. Another example is Campbell’s 1988 pre- oil and gas in discrete reservoirs allow relatively
diction [8] that underestimated US oil production easy extraction of oil and gas by drilling wells.
in 2010 by 70%. Until the 1990s almost all the world’s oil and gas
The following sections describe the various production was produced from such “conven-
types of geologic occurrence of oil and gas and tional” fields. Natural gas in conventional fields
how scientific and technological advances have may occur with (“associated gas”) or without
changed the perception of the size and economic (“nonassociated gas”) oil. In either case the gas
viability of the resources over the last few may contain compounds that can be separated at
Oil and Natural Gas: Global Resources 9

the surface as liquids. Natural gas liquids (NGLs) the compartments. At the same time, technologi-
such as propane, butane, and pentane are some- cal advances allow drilling with much greater
times included in reporting oil production but are precision permitting access to much smaller sub-
not insignificant: they comprise, for example, surface targets over time. While it will never be
3.5% of the total energy production of the United possible to economically drain all the oil and gas
States. out of a field, recovery rates of over 70% may
Estimates of the world’s original endowment eventually be feasible in many areas [12].
of conventional oil and gas resources have tended Much oil and natural gas also remains to be
to increase because of the advances in technology found in undiscovered fields within existing
over time [9]. Advances in geologic concepts, petroleum provinces. The largest fields tend to
drilling technologies, seismic imaging, and com- be discovered early in the exploration history of
puter modeling using large datasets have progres- a basin and progressively smaller ones are found
sively revolutionized the search for oil and gas over time. The infrastructure that is built to
over the last century. Likewise, advances in pro- develop the larger fields allows fields to be devel-
duction technologies, pipeline construction, and oped that would otherwise be too small to justify
tankers have all made resources available that at the construction of pipelines, platforms, ports, and
one time would have been considered economi- processing facilities. By analogy estimates can be
cally unviable. Advances in science and technol- made of the number and size of remaining
ogy allow more oil and gas to be found in old undiscovered fields from the discovery history of
fields, within existing petroleum provinces, and in a petroleum province. This methodology may
new frontier regions. underestimate the total resource if more than one
Existing oil and gas fields that have already type of oil and gas accumulation is present. For
been in production for years may, at first blush, example, the discovery of giant oil and gas fields
seem unusual sites to look for further reserves. in the subsalt regions in offshore regions of the
However, the US Geological Survey (USGS) in Gulf of Mexico added a new “play” in what was
2000 estimated that 48% of the oil and 41% of the thought to be a mature oil province.
natural gas that remains to be added to reserves in There is a strong possibility that further major
the future lie within existing fields [10]. Increases accumulations of conventional oil and gas will
in successive estimates of the estimated recover- also be found in frontier basins where little or no
able oil and gas are known as “reserve growth” On exploration has taken place. Most of the potential
average, only 22% of the oil is currently recovered areas are in regions of deepwater or harsh cli-
from fields worldwide [11]. There are many rea- mates. Over time technological advances have
sons for such a low percentage. Much of the oil made exploration feasible in areas once thought
within a reservoir rock will not easily move inaccessible. This has been particularly true in the
toward wells. Enhanced recovery methods such ability to drill to deeper depths. In 1960 any dril-
as gas reinjection, water-flooding, and flushing ling in water depths over 20 m was regarded as
with polymers and surfactants can free up more “deepwater” but by 1980 oil fields had been found
of the oil. While such secondary and tertiary in over 300 m water depth, by 1990 in over 850 m
recovery methods are commonly used in mature water depth, and by 2010 in over 2400 m water
oil fields in developed countries, they are not yet depth [13]. Total drilling depths also increased
widely used in many other regions. Another rea- dramatically with a record set in 2009 with the
son for reserve growth is that there is rarely a discovery of the giant Tiber oil field in the Gulf of
single reservoir within an oil or gas field as most Mexico which is at 1200 m water depth, with a
are split into numerous compartments, each total drilling depth of 10,680 m. In the last decade
containing amounts of oil/gas in varying amounts. giant oil fields have been found in ultra-deepwater
Advances in seismic imaging over the last areas not only in the Gulf of Mexico but also
20 years have allowed not only good visualization offshore West Africa and Brazil. The giant Brazil-
of these compartments but also the fluids within ian Tupi field lies not only in deepwater
10 Oil and Natural Gas: Global Resources

(>2000 m) but also below a 2000 m thick layer of “sweet spots” is debated as to whether they are
salt. It is only recently that seismic technology has regions with enhanced permeability created by
allowed an understanding of the structures below natural fractures in the rock or are buoyancy
thick salts and drilling technologies have permit- traps similar to conventional gas fields
ted drilling through such strata. [15]. Resource estimates of basin-centered gas
As in the past, technological advances will are typically very large but in some cases may be
permit exploration for oil and gas in frontier reevaluated as geologists develop a better under-
regions currently regarded as inaccessible. Future standing of the nature of sweet spots. Neverthe-
exploration in deepwater will likely lead to more less, tight gas is undoubtedly a major resource
significant discoveries. Outside the Atlantic realm and, historically, it has been the most important
other regions with potential for deepwater explo- component of unconventional gas production in
ration include offshore East Africa, the Great North America but relatively few tight gas fields
Australian Bight, and offshore New Zealand. have yet been developed outside of North Amer-
There is a growing interest in the potential of the ica. Production has risen from 10% of total natural
Arctic region where geologists believe there are gas production in the United States in 1990 to
substantial oil and gas resources but where the 28% in 2009 [2, 16]. Given the magnitude and
severe climate, ice floes, and icebergs all make widespread nature of these accumulations in
exploration and development hazardous and in North America it seems likely that tight gas
many areas either prohibitively expensive or resources occur in many sedimentary basins
impossible using currently available worldwide. Exploration and development of
technologies [14]. such resources must compete with conventional
gas resources that are much cheaper to produce,
and which in many cases may lie at shallower
Tight Gas depths within the same sedimentary basin. How-
ever, it is probably only a matter of time before
In conventional fields gas moves easily through tight gas resources are developed in many regions
the permeable host rock and naturally moves of the world.
toward wells and up to the surface because of
pressure differences. By contrast, sandstones that
have very low permeability require considerably Oil Sands
more effort to produce the gas. These sandstones
are generally thinner than those in conventional Oil sands are rocks that consist predominantly of
fields and have been buried to great depth. The sandstones which contain bitumen within the pore
pores in the sandstone, in which the gas is trapped, spaces that has been produced by the biodegrada-
have been reduced in size by compaction and tion of oils in the subsurface. Most oil reservoirs
cementation during burial. These “tight sands” are sufficiently hot that biogenic activity is
typically occur near the center of sedimentary curtailed or absent. However, oil that migrates
basins and are sometimes referred to as “basin- into shallow reservoir rocks may be altered
centered gas” accumulations. Production of gas completely to bitumen that is very viscous.
from tight sands requires extensive and complex Although found close to the surface, production
drilling. Because each well produces relatively of oil from oil sands is expensive. The bitumen
small volumes of natural gas, many wells must must be heated before it will flow and commercial
be drilled. Hydraulic fracturing of the host rock extraction requires large amounts of energy. Once
can increase the rate of flow of gas to a well. extracted the bitumen must also be upgraded by
Although basin-centered gas accumulations purification and hydrogenation before it can be
can occur over very wide geographic areas, pro- refined like conventional crude oil.
duction is generally from limited regions of a Although oil sands have long been recognized
basin. The geologic nature of these so-called and used in limited ways, it is only within the last
Oil and Natural Gas: Global Resources 11

40 years that commercial production has grown. CBM occurs at much shallower depths than
Canadian oil sands production began in 1967 and most conventional gas fields: generally less than
production has grown steadily. By 2009 produc- 900 m. Extraction of CBM, however, requires
tion from oil sands was equivalent to over 550 mil- many closely spaced wells: over 20,000 wells
lion barrels of oil, accounting for 49% of Canada’s had been drilled in the Powder River Basin [21]
oil production in 2009 [17]. Though production and over 17,000 in Alberta [22]. The gas is
was initially subsidized, production costs have adsorbed onto the pore space in coals and must
fallen with technological advances and is now be released by depressurization, allowing the gas
economically viable at $50/barrel and production to flow toward the well along natural fractures
is forecast to more than triple by 2025 [18]. Vene- known as “cleats” or by pathways produced by
zuela also has substantial oil sand deposits [19] artificial hydraulic fracturing.
and smaller accumulations are known in Russia
and the Middle East.
Oil Shale and Shale Oil

Coal Bed Methane Oil can be produced from some organic-rich

finegrained rocks that are normally referred to as
The presence of methane in coal beds had long “shales” even if the host rock is not strictly a shale
been a hazard for underground coal mining and it by a geologic definition. Production from such
is only recently that this gas has been seen as an sources has a long history. Production of oil
economic resource. Modern mining techniques from shales for illumination preceded the discov-
aimed at mitigating the potential for lethal build- ery of conventional petroleum resources in the
up of methane in mines extracts the gas during or mid-nineteenth century but for most of the twen-
before mining as coal mine methane (CMM), tieth century production of such oil was a very
which in some cases is used for power generation. minor component of global petroleum production.
More significant from an energy resource perspec- However, in the first decade of the twenty-first
tive is the natural gas extracted from coal beds by century there has been a resurgence of interest in
wells in regions where no mining is planned: this these resources because of advances in technol-
is known as coal bed methane (CBM) or coal ogy and rising energy prices.
seam gas (CSG). “Oil shales” are rocks that contain significant
Commercial production of CBM began in the amounts of solid organic chemical compounds
United States in 1989 and annual production had (kerogen) that have not been buried deeply
grown to almost 2 trillion cubic feet (TCF) in enough to allow for oil maturation. Production is
2009, which was 8% of total US gas production generally done by mining the rock and heating it
[2]. Early production was predominantly from in a retort in a processing plant close to the mine
bituminous coals in the San Juan and Raton where the oil and associated gases can be cap-
Basins of New Mexico and Colorado. Production tured. The oil may also be extracted using in situ
from these areas still accounts for almost 50% of methods which require heating the subsurface
US production. In 2000 production began from rock by injection of hot fluids, gases, or steam,
the lower-rank subbituminous and lignite coals of or by the use of heating elements. As the oil is
the Powder River Basin in Wyoming which by expelled from the kerogen it can then be induced
2010 comprised 28% of US CBM production. to flow toward conventional oil wells for
Over the last decade, the technology and extraction.
knowhow for CBM extraction has spread from The leading producer of oil shale in the world
the United States to other major coal-bearing is Estonia, where 90% of the power is generated
regions, particularly Canada, Australia, China, from that source. By far the largest accumulations
and parts of Europe. Australia plans to export of oil shale, however, are in the United States,
CBM gas as LNG [20]. particularly in the Green River Formation of
12 Oil and Natural Gas: Global Resources

Colorado, Utah, and Wyoming, that were depos- technologies. The fractures provide avenues for
ited within ancient saline lake systems some gas migration toward the wells.
40–50 million years ago. There are also major The new technologies for shale gas production
accumulations in Devonian–Mississippian black have revolutionized US gas production. Shale gas
shales in the eastern United States that were production was approximately 1% of total US
deposited in marine environments over 350 mil- production in 2000 but had risen to 14% of total
lion years ago. With the high oil prices of the late production in 2009 and is forecast to be 45% of
1970s a number of pilot projects produced oil production by 2035 [2]. The potential of shale gas
from the Green River Formation in Colorado but has been questioned by some [25]. Like tight gas,
plans for major commercial exploitation were production is from sweet spots whose nature and
abandoned when prices fell in the early 1980s. lateral extent remains to be better defined. The
Interest in the potential for production has been dramatic rise in US production has, however,
rekindled with the high oil prices in recent years. spurred interest in shale gas in many other parts
Other countries with significant oil shale accumu- of the world, especially in Canada, Central
lations include Australia, Brazil, the Democratic Europe, China, India, and Australia.
Republic of the Congo, and Russia [23].
“Shale oil” is oil that is trapped within a
finegrained rock. Extraction of the oil does not Gas Hydrates
require heating but the low permeability of the
rocks requires that the rock be artificially fractured Gas hydrates (also called clathrates) are solid ice-
in situ to allow flow toward a well. Once regarded like substances consisting of rigid cages of water
only of scientific interest, recent advances in molecules that enclose molecules of gas – mainly
hydraulic fracturing have made shale oil econom- methane. Hydrates are stable in a restricted range
ically viable in some areas. The best known shale of temperatures and pressures and occur in two
oil accumulation is the Bakken Formation of the regions: in polar regions, where they are associ-
Williston Basin of Montana and North Dakota and ated with permafrost, and at shallow depths in
adjoining parts of Canada [24]. From a geologic sediment on the outer continental shelves, in
perspective, the Bakken is a petroleum source rock water depths over 300 m [26]. Hydrates can pose
that reached maturation but, unlike most mature an environmental problem because natural break-
source rocks, the oil was never expelled to migrate down of accumulations can result in slumping of
to conventional traps. the seabed and outgassing of methane to the atmo-
sphere. On the other hand, if appropriate technol-
ogies are developed, hydrates may be a future
Shale Gas source of natural gas. Current technologies allow
production only at costs substantially above the
Natural gas is present in some organic-rich market prices that have prevailed over the last
mudrocks. Shale gas has been produced in small decade [27]. Production would require depressur-
quantities for well over a century but recent ization, thermal stimulation, or injection of inhib-
advances in technology allow much higher rates itors to destabilize the hydrate lattices, releasing
of production. Shale beds are relatively thin and the methane to flow toward wells.
vertical wells provide access to a very limited
volume of rock. It is now possible to drill at any
angle and to deviate wells at depth to follow Estimated Volumes of Remaining Oil and
individual shale beds over great distances, allo- Gas Resources
wing access to large volumes of gas-bearing shale.
While production used to be primarily from natu- Proved reserves are the volume of known oil and
rally fractured shales, much recent production gas accumulations that can be produced at a profit
from shales is aided by hydraulic fracturing under existing economic and operating conditions.
Oil and Natural Gas: Global Resources 13

Estimates of world-proved reserves of oil and gas discovering 8174 TCF. The median estimate was
are compiled using best available information by that 4333 of natural gas might be found in the
several organizations including the International studied provinces.
Energy Agency, IHS Energy, and BP. Proved More recent estimates by the USGS that have
reserves of oil and gas are not routinely reported concentrated on specific regions of the world sug-
in many countries. In fact, some of the largest pro- gest that on balance their 2000 estimates are prob-
ducers consider reserves as state secrets. This lack ably reasonable estimates of the global picture
of transparent reporting has led some to question though there is need for revision in some areas.
the size of global reserves giving credence to pre- A 2010 study [30] estimates 17% less
dictions of imminent production declines undiscovered oil and 20% less undiscovered nat-
[28]. However, though differing in detail the ural gas than in the 2000 assessment in eight
reserves estimated by different agencies are similar basins in Southeast Asia; however, there had
in estimated overall size. As of the end of 2009, BP been at least 10 years of discoveries between the
[29] estimated globally proved reserves of 1333 two assessments. Furthermore, the 2010 assess-
billion barrels of oil (BBO), equivalent to over ment examined a number of additional basins in
45 years of production at 2009 rates. With the Southeast Asia resulting in an overall increase in
exception of 237 billion barrels of oil in the Cana- estimated undiscovered oil of 71% and natural gas
dian oil sands, these reserves are entirely of con- by 66%. The 2008 USGS assessment of the Arctic
ventional oil. BP also estimated globally proved revised down the estimate of undiscovered oil by
reserves of natural gas at 6621 trillion cubic feet 48%, largely because of a reappraisal of the West
(TCF), equivalent to 63 years of production. These Siberian Basin, but increased the estimate of nat-
reserves are predominantly conventional gas. ural gas by 21% by including a number of basins
In addition to proved reserves, there are not in the 2000 assessment [14].
undiscovered resources and “contingent The USGS 2000 assessment [10] also esti-
resources” that are currently noncommercial but mated the amount of reserve growth that could
could probably be produced under different eco- be anticipated to 2025 for fields that had been
nomic conditions. The most recent comprehen- discovered prior to 1996. The median estimate
sive global estimate of undiscovered oil and gas for reserve growth for conventional oil was
resources was published by the USGS in 2000 612 BBO and for conventional natural gas 3305
[10]. This was a geologically based study of trillion cubic feet. At the time of writing, the
128 geologic provinces that included the produc- midpoint of the time interval has been reached
ing basins that accounted for more than 95% of over which reserve growth was predicted and a
the world’s known oil and gas outside of the substantial amount of the predicted reserve
United States. The study also examined the dis- growth has already occurred [31]. It seems unrea-
covery history of each province up to 1996. The sonable, however, to assume that reserve growth
report included probabilistic estimates of the vol- will cease in 2025 and interestingly King [32],
umes of conventional oil, gas, and natural gas using a different methodology and separate data
liquids that might be added to proved reserves from that used by the USGS, estimated future
from new field discoveries in the studied prov- reserve growth of oil at between 200 to 1000
inces from the 1996 baseline. The assessment BBO – almost exactly the same range that the
estimated that there was a 95% chance of discov- USGS predicted would occur from 1996.
ering another 334 BBO of conventional oil and a Table 3 provides an estimate of remaining oil
5% chance of discovering 1107 BBO. The median resources at the end of 2007. The proved reserve
estimate, the 50% chance, was that 607 BBO of numbers are from BP [29] and are slightly lower
oil might yet be found in the studied provinces. than those of EIA [2]. The estimated volume of oil
For conventional natural gas, the assessment esti- in undiscovered fields is revised down from the
mated that there was a 95% chance of discovering median estimate in the USGS 2000 assessment
another 2299 TCF and a 5% chance of because 16.5% of that volume had been
14 Oil and Natural Gas: Global Resources

Oil and Natural Gas: Global Resources, development of these resources is largely in its
Table 3 Estimate of remaining oil resources including infancy, especially outside of North America.
NGL (billion barrels). Sources and rationale for estimates
are given in text. For comparison, annual production of oil Recently published estimates of recoverable
was approximately 29.2 billion barrels in 2009 shale gas resources in the United States and
Conventional
Canada vary between 50 and 1000 TCF [34]. Esti-
Proved reserves 1236 mates for the rest of the world would likely have
Reserve growth 600 an even larger range. It is particularly important to
Undiscovered fields 600 discriminate between “in-place resources” and
Total conventional 2436 those that are economically recoverable. The
Unconventional USGS [35] estimated that the United States has
Oil sands proved reserves 237 700 TCF of CBM gas in place but only 100 TCF
Oil sands resources (incl. reserves) 820 that would be economically recoverable. As 20.5
Oil shale resources 2826 TCF was produced over the following 13 years
Total unconventional 3141 under a major drilling effort, 100 TCF may be an
optimistic estimate of ultimate recovery. The
USGS [35] also suggested that global in-place
CBM gas resources may be as high as 7500 TCF
discovered by 2007 [31]. However, it should be but recoverable resources may well be an order of
noted that this may be conservative as the 2000 magnitude lower. Australia, for example, has 9%
assessment did not include the United States and of the world’s coal resources [2] and an estimated
many geologic basins that have relatively small 153 TCF of recoverable CBM gas, of which 90%
volumes of discovered oil or where there has been is currently sub-economic or not yet proven by
little or no exploration. The reserve growth num- drilling [20]. While extraordinarily large esti-
bers are based on King [31] for crude oil, as mates of unconventional gas resources
discussed above, and 50% of the USGS 2000 (particularly for hydrates) should be regarded
estimated median reserve growth for NGL that with some scepticism from an economic perspec-
was predicted to occur between 1996 and 2025. tive, there is ample evidence that unconventional
The oil sand reserve and resource estimates are for gas resources are abundant. In the United States,
Canada [33] and Venezuela [19] only though there which has a wide variety of geologic basins, 50%
are smaller oil sand accumulations in several other of the gas produced in 2010 was unconventional
countries. The shale oil estimate is from Dyni and it is not unreasonable to suppose that globally
[23]. The estimated remaining conventional oil the abundance of unconventional gas resources is
is equivalent to 79 years of annual production at of a similar scale to conventional gas resources.
2009 rates. The estimated unconventional oil adds
up to the equivalent of 125 years of annual
production. Future Directions
Global reserves of natural gas in 2010 were
estimated at a little over 6600 TCF by BP [29] A strong case could be made that the estimates of
and EIA [2]. Taking into account discoveries since remaining oil and gas presented above are on the
1996 [31] and revising down the median estimate low side because, as in the past, unanticipated
of gas in undiscovered fields from the USGS 2000 technological advances may make additional oil
assessment, and also assuming 50% of the esti- and gas accumulations economically viable. Forty
mated reserve growth has taken place, the years ago large-scale economic extraction of oil
remaining conventional natural gas resources are from oil sands, gas from gas shales, gas from coal
estimated at approximately 11,850 TCF, equiva- beds, and gas from tight sandstones all seemed
lent to about 87 years of annual production at improbable, but all are now a significant part of
2009 rates. Estimating global abundance of world production. However, even if the estimates
unconventional gas resources is difficult because are somewhat optimistic, it is clear that there is a
Oil and Natural Gas: Global Resources 15

large volume of remaining recoverable oil and gas Bibliography

resources. How much of that resource will even-
tually be extracted remains to be seen. Primary Literature
There are clearly environmental costs to oil and 1. International Energy Agency (2010) Key world
energy statistics. OECD/IEA, Paris
gas extraction. Pursuing hard-to-get resources
2. Energy Information Agency (2011). http://www.eia.
raises the risk of disasters such as the BP Deep- doe.gov/. Accessed Apr 2011
water Horizon spill in 2010 [13] that can pro- 3. Fanning LM (1950) A case history of oil-shortage
foundly influence public opinion. Ecosystems scares. In: Fanning LM (ed) Our oil resources.
McGraw-Hill, New York, pp 306–406
are impacted by large-scale mining of oil sands
4. Hubbert MK (1956) Nuclear energy and the fossil
and oil shales and some in situ extraction of fuels, Shell Development Company Publication 95.
unconventional oil and gas has the potential to Shell Development Company, Houston, p 40
influence groundwater supplies. Large volumes 5. Hubbert MK (1962) Energy resources: a report to the
committee on natural resources, National Academy of
of water can be produced during CBM develop-
Sciences-National Research Council Publication
ment. The biggest concern, however, is in the 1000-D. National Academy of Sciences-National
consumption of oil and gas with associated rising Research Council, Washington, 141 p
levels of greenhouse gases in the atmosphere and 6. Deffeyes KS (2001) Hubbert’s peak: the impending
world oil shortage. Princeton University Press,
predictions of global climate change. On the other
Princeton, p 208
hand, substitution of natural gas for coal in power 7. Meadows DH, Meadows DL, Randers J, Behrens WW
plants, more efficient vehicles, and carbon capture III (1972) The limits to growth. Universe Books,
and storage could also make substantial reduc- New York, p 205
8. Campbell CJ (1988) The coming oil crisis. Multi-
tions in global greenhouse gas emissions.
Science, Brentwood, p 210
How much of the remaining oil and gas 9. McCabe PJ (1998) Energy resources – cornucopia or
resources will be produced will depend on com- empty barrel? AAPG Bull 82:2110–2134
petition over time from other energy sources, 10. United States Geological Survey (2000) World petro-
leum assessment 2000 – description and results, U.-
especially renewables and nuclear. It is sometimes
S. Geological Survey Digital Data Series DDS-60.
assumed that the depletion of global oil resources U.S. Geological Survey, Denver. http://pubs.usgs.
will inevitably lead to high oil prices that will gov/dds/dds-060/
allow more expensive energy sources to become 11. Sandrea I, Sandrea R (2007) Global oil reserves-1:
recovery factors leave vast target for EOR technolo-
competitive. However, historically substitution of
gies. Oil Gas J 105(5):44–47
one energy source by another has taken place 12. Sandrea I, Sandrea R (2007) Global oil reserves-2:
primarily by reduction in costs of the substituting recovery factors leave EOR plenty of room for growth.
energy source [9]. Rather than a result of long- Oil Gas J 105(12):39–42
13. National Commission on the BP Deepwater Horizon
term depletion of the resource, the high oil prices
Oil Spill and Offshore Drilling (2011) Deep water: the
of the first decade of the twenty-first century argu- Gulf oil disaster and the future of offshore drilling,
ably reflect more restrictions of supply related to p 380. https://s3.amazonaws.com/pdf_final/DEEPWA
conflicts and political problems coinciding with a TER_ReporttothePresident_FINAL.pdf
14. Gautier DL, Bird KJ, Charpentier RC, Grantz A,
rapid rise in demand, especially from developing
Houseknecht DW, Klett TR, Moore TE, Pitman JK,
countries. Oil and natural gas may lose global Schenk CJ, Schuenemeyer JH, Sørensen K, Tennyson
market share in the future either because techno- ME, Valin ZC, Wandrey CJ (2009) Assessment of
logical advances permit cheaper production of undiscovered oil and gas in the Arctic. Science
324:1175–1179
alternate energy sources or because of political
15. Shanley KW, Cluff RM, Robinson JW (2004) Factors
influence on the market, including restrictions on controlling prolific gas production from low-
supplies of oil and natural gas or incentives permeability sandstone reservoirs: implications for
against their consumption, such as carbon taxes. resource assessment, prospect development, and risk
analysis. AAPG Bull 88: 1083–1121
Nevertheless, given the rising global demand for
16. Nehring R (2008) Growing and indispensable: the
energy, it seems likely that oil and gas will be a contribution of production from tight-gas sands to
major part of the energy mix for most, if not all, of U.S. gas production. In: Cumella SP, Shanley KW,
the twenty-first century. Camp WK (eds) Understanding, exploring, and
16 Oil and Natural Gas: Global Resources

developing tight-gas sands – 2005 Vail Hedberg con- 26. Kvenvolden KA (1993) A primer on gas hydrates. In:
ference, AAPG Hedberg Series, No. 3. American Howell DG et al (eds) The future of energy gases, US
Association of Petroleum Geologists, Tulsa, pp 5–12 Geological Survey Professional Paper 1570. U.S. G.P.
17. Canadian Association of Petroleum Producers O, Washington, DC, pp 279–291
(2011) Statistical handbook for Canada’s upstream 27. Walsh MR, Hancock SH, Wilson SJ, Patil SL, Moridis
petroleum industry. Canadian Association of Petro- GJ, Boswell R, Collett TS, Koh CA, Sloan ED
leum Producers, 2011–9999, CAL, p 211 (2009) Preliminary report on the commercial viability
18. The Economist (2011) Muck and brass Economist of gas production from natural gas hydrates. Energy
398(8717):77–80 Econ 31:815–823
19. United States Geological Survey (2009) An estimate 28. Simmons MW (2005) Twilight in the desert. Wiley,
of recoverable heavy oil resources of the Orinoco oil Hoboken, p 422
belt, Venezuela, U.S. Geological Survey Fact Sheet 29. BP (2010) BP Statistical review of world energy, June
2009–3028. U.S. Geological Survey, Reston, p 4 2010. http://bp.com/statisticalreview. Accessed Apr 2011
20. Australian Government (2010) Australian energy 30. United States Geological Survey (2010) Assessment
resource assessment. Department of Resources, of undiscovered oil and gas resources of Southeast
Energy and Tourism, Canberra, p 358 Asia, USGS Fact Sheet 2010–3015. U.S. Geological
21. Swindell GS (2007) Powder River Basin coalbed Survey, Reston
methane wells – reserves and rates Society of Petro- 31. Gautier DL, McCabe PJ, Ogden J, Demayo TN
leum Engineers Report SPE 107308 (2010) Resources, reserves, and consumption of
22. Energy Resources Conservation Board (2010) ST109: energy. In: Graedel TE, van der Voet E (eds) Linkages
Alberta coalbed methane well locations. Energy of sustainability, Strungmann Forum Report. MIT,
Resources Conservation Board, Calgary, p 426 Cambridge, pp 323–340
23. Dyni JR (2006) Geology and resources of some world 32. King KC (2007) Growth; are we underestimating recent
oil-shale deposits, U.S. Geological Survey Scientific discoveries? Abstracts, annual meeting. AAPG 2007:77
Investigations Report 2005–5294. U.S. Geological 33. Government of Canada (2010) Oil sands – a strategic
Survey, Reston, p 42 resource for Canada, North America and the world.
24. U.S. Geological Survey (2008) Assessment of Natural Resources Canada, Ottawa. http://www.nrcan-
undiscovered oil resources in the Devonian- rncan.gc.ca/eneene/pdf/os-sb-eng.pdf
Mississippian Bakken Formation, Williston Basin 34. Mohr SH, Evans GM (2010) Shale gas changes
province, Montana and North Dakota, U.S Geological N. American gas production projections. Oil Gas
Survey Fact Sheet 2008–3021. U.S. Geological Sur- J 108(26):60–64
vey, Reston 35. U.S. Geological Survey (1997) Coalbed methane – an
25. Klump E, Polson J (2009) Shale-gas skeptic's supply untapped energy resource and an environmental con-
doubts draw wrath of Devon (update 2). http://www. cern, USGS Fact Sheet FS-019-97. U.S. Geological
bloomberg.com/ apps/news?pid=newsarchive&sid= Survey, Denver. http://energy.usgs.gov/factsheets/
asEUlpJcuZB4. Accessed Apr 2011 Coalbed/coalmeth.html
 Both oil and gas have a low specific gravity rela-
Petroleum and Oil Sand tive to water and, because of this density differ-
Exploration and Production ence, will pass through the more porous sections
of reservoir rock from their source area to the
James G. Speight surface unless restrained by a trap. A trap is a
CD&W Inc., Laramie, WY, USA reservoir that is overlain and underlain by dense
impermeable cap rock or a zone of very low or no
porosity that restrains migrating hydrocarbon.
Article Outline Reservoirs vary in size up to miles in length and
breadth and can range in thickness from several
Glossary inches to hundreds of feet or more.
Definition of the Subject In general, petroleum is extracted by drilling
Introduction wells from an appropriate surface configuration
Petroleum Exploration and Production into the crude oil-containing reservoir, and the
Oil Sand Exploration and Production wells (production wells) are designed to contain
Future Directions and control all fluid flow throughout the produc-
Bibliography tion operations. The number of wells required for
a reservoir is dependent on a combination of tech-
Glossary nical and economic factors used to determine the
most likely range of recoverable reserves relative
Exploration, recovery, team-based methods, in to a range of potential investment alternatives. In
situ combustion, oil mining, oil sand mining, summary, the drilling program and the crude oil
hot-water process, bitumen, in situ conversion, recovery program are reservoir specific (site
the future. specific).
In terms of site specificity, the crude oil pro-
duction rates from a reservoir depend on several
factors, such as (i) reservoir geometry, which is
Definition of the Subject primarily the thickness of the reservoir formation
and the continuity of the formation, (ii) reservoir
Depletion of the reserves of conventional (easy-
pressure, (iii) reservoir depth, (iv) the rock type
to-refine) crude oil is continuing at a noticeable
and the permeability of the formation, (v) fluid
rate, and other sources of hydrocarbons are
saturation and fluid properties, (vi) the extent of
required – these include crude oil from tight for-
fracturing, (vii) the number of wells and their
mations, heavy oil (a type of petroleum), and
locations in the formation, and (viii) the ratio of
bitumen from oil sand (tar sand) deposits.
the permeability of the formation to the viscosity
of the oil [1, 2]. Furthermore, the geological var-
iability of reservoirs means that production pro-
Introduction files differ from field to field.

Petroleum occurs in the microscopic pores of sed-

imentary rocks that form a reservoir – typically, Petroleum Exploration and Production
reservoir rock consists of sand, sandstone, lime-
stone, or dolomite. However, not all the pores in a Exploration
rock will contain petroleum – some will be filled Exploration for petroleum originated in the latter
with water or brine that is saturated with minerals. part of the nineteenth century when geologists
© Springer Science+Business Media, LLC, part of Springer Nature 2020 17
R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_99
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_99-3
18 Petroleum and Oil Sand Exploration and Production

began to map land features that were favorable for earth’s crust, and the configuration and density
the collection of oil in a reservoir. Of interest to of the underlying rocks, as well as by any abnor-
geologists were outcrops that provided evidence mal mass variation that might be associated with a
of alternating layers of porous and impermeable mineral deposit.
rock. The porous rock (typically a sandstone, In the second group of measurements (those
limestone, or dolomite) provides the reservoir for with depth control), seismic or electric energy is
the petroleum, while the impermeable rock introduced into the ground, and variations in
(typically clay minerals or shale) prevents migra- transmissibility with distance are observed and
tion of the petroleum from the reservoir. By the interpreted in terms of geological quantities.
early part of the twentieth century, most of the Depths to geological horizons having marked dif-
areas where surface structural characteristics ferences in transmissibility can be computed on a
offered the promise of oil had been investigated, quantitative basis and the physical nature of these
and the era of subsurface exploration for oil began horizons deduced. However, geophysical explo-
in the early 1920s. New geological and geophys- ration techniques cannot be applied indiscrimi-
ical techniques were developed for areas where nately. Knowledge of the geological parameters
the strata were not sufficiently exposed to permit likely to be associated with the mineral or subsur-
surface mapping of the subsurface characteristics. face condition being studied is essential both in
In the 1960s, the development of geophysics pro- choosing the method to be applied and in
vided methods for exploring below the surface of interpreting the results obtained. Furthermore,
the earth. not all the techniques described here may be suit-
In petroleum exploration, terms as geophysical able for petroleum exploration.
borehole logging can imply the use of one or more A basic rule of thumb in the upstream
of the geophysical exploration techniques. This (or producing) sector of the oil and gas industry
procedure involves drilling a well and using has been (and maybe still is in some circles of
instruments to log or make measurements at var- exploration technology) that the best place to find
ious levels in the hole by such means as gravity new crude oil or natural gas is near formations
(density), electrical resistivity, or radioactivity. where it has already been found. The financial risk
The principles used are basically magnetism of such a plan is far lower than that associated with
(magnetometer), gravity (gravimeter), and sound drilling a wildcat hole (a drill hole in an unknown
waves (seismograph). These techniques are based area where oil has not been found prior to the
on the physical properties of materials that can be commencement of the new drill hole) in a pro-
utilized for measurements and include those that spective, but previously unproductive, area. On
are responsive to the methods of applied geophys- the other hand, there is a definite trade-off
ics. Furthermore, the methods can be subdivided between rewards for risk. The returns on drilling
into those that focus on gravitational properties, investment become ever leaner as more wells are
magnetic properties, seismic properties, electrical drilled in a particular area because the natural
properties, electromagnetic properties, and radio- distribution of oil and gas field volumes tends to
active properties. These geophysical methods can be approximately log geometric – there are only a
be subdivided into two principal groups: (i) those few large oil fields (giant and supergiant oil
methods without depth control and (ii) those fields), whereas there are a great many small oil
methods having depth control. fields [3].
In the first group of the measurements (those The drilling job is complete when the drill bit
without depth control), the methods incorporate penetrates the reservoir and the reservoir is eval-
effects from both local and distant sources. For uated to see whether the well represents the dis-
example, gravity measurements are affected by covery of a prospect or whether it is a dry hole. If
the variation in the radius of the earth with lati- the hole is dry, it is plugged and abandoned. On
tude. They are also affected by the elevation of the the other hand, if the presence of crude oil has
site relative to sea level, the thickness of the been established, the prospect becomes a live
Petroleum and Oil Sand Exploration and Production 19

prospect, and once the final depth has been the recovery method on the crude oil and on the
reached, the well is completed to allow oil to reservoir must be considered before method appli-
flow into the casing in a controlled manner. First, cation. Furthermore, as might be expected, the
a perforating gun is lowered into the well to the type of exploration technique employed depends
production depth. The gun has explosive charges upon the nature of the site, and as for many envi-
to create holes in the casing through which oil can ronmental operations, the recovery techniques
flow. After the casing has been perforated, a applied to a specific site are dictated by the nature
small-diameter pipe (tubing) is run into the hole of the site and are, in fact, site specific.
as a conduit for oil and gas to flow up the well, and There are three phases for recovering oil from
a packer is run down the outside of the tubing. reservoirs (Fig. 2) that are appropriately named:
When the packer is set at the production level, it is (i) primary recovery, (ii) secondary recovery, and
expanded to form a seal around the outside of the (iii) tertiary recovery. Primary recovery occurs as
tubing. Finally, a multivalve structure – the wells produce because of natural energy from
Christmas tree (Fig. 1) – is installed at the top of expansion of gas and water within the producing
the tubing and cemented to the top of the casing. formation, pushing fluids into the wellbore and
The Christmas tree allows them to control the flow lifting the fluids to the surface. Secondary recov-
of oil from the well. ery requires energy to be applied to lift fluids to
surface – this may be accomplished by injecting
gas down a hole to lift fluids to the surface, instal-
Onshore Production lation of a subsurface pump, or injecting gas or
Recovery, as applied in the petroleum industry, is water into the formation itself. Tertiary recovery
the production of oil from a reservoir. There are occurs when a means is required to increase fluid
several methods by which this can be achieved mobility within the reservoir – this may be accom-
that range from recovery due to reservoir energy plished by introducing additional heat into the
(i.e., the oil flows from the well hole without formation to lower the viscosity (thin the oil)
assistance) to enhanced recovery methods in and improve its ability to flow to the wellbore.
which considerable energy must be added to the Heat may be introduced by either (i) injecting
reservoir to produce the oil. However, the effect of chemicals with water (chemical flood, surfactant

Petroleum and Oil Sand

Exploration and polish rod
Production, Fig. 1 The
Christmas tree
stuffing box

tubing pressure tubing flow valve

gauge

master valve
flow line
casing pressure
gauge
20 Petroleum and Oil Sand Exploration and Production

Petroleum and Oil Sand

Exploration and Primary
Production, Recovery
Fig. 2 Methods for oil
recovery
Natural Flow Artificial Lift
Pump,Gas Lift,Etc.

Secondary
Recovery

Pressure
Waterflood Maintenance
Water,Gas Reinjection

Tertiary
Recovery

Thermal Gas Chemical Microbial

flood), (ii) injecting steam (steam flood), or (iii) saturations, and, last, but certainly not least (viii)
injecting oxygen to enable the ignition and com- fluid properties. Overall, the goal is to increase the
bustion of oil within the reservoir) ( fireflood). mobility of the oil, and once production begins, the
Generally, conventional crude oil reservoirs performance of each well and reservoir is moni-
often exist (i) with an overlying gas cap, (ii) in tored, and a variety of engineering techniques are
communication with aquifers, or (iii) both. The used to progressively refine reserve recovery esti-
crude oil resides together with water and free gas mates over the producing life of the field. The total
in very small holes (pore spaces) and fractures. recoverable reserves are not known with complete
The size, shape, and degree of interconnection of certainty until the field has produced crude oil to
the pores vary considerably from place to place in the point of depletion (where no more crude oil can
an individual reservoir. Below the oil layer, the be produced) or its economic limit and
sandstone is usually saturated with salt water. The abandonment.
oil is released from this formation by drilling a Generally, the first stage in the extraction of
well and puncturing the limestone layer on either crude oil is to drill a well into the underground
side of the limestone dome or fold. If the peak of reservoir. Often, many wells (multilateral wells)
the formation is tapped, only the gas is obtained. If will be drilled into the same reservoir to ensure
the penetration is made too far from the center, that the extraction rate will be economically via-
only salt water is obtained. ble. Also, some wells (secondary wells) may be
Therefore, in designing a recovery project, it is a used to pump water, steam, acids, or various gas
general practice to locate injection and producing mixtures into the reservoir to raise or maintain the
wells in a regular geometric pattern so that a sym- reservoir pressure and so maintain an economic
metrical and interconnected network is formed and extraction rate.
production can be maximized. However, the rela- Directional drilling is also used to reach for-
tive location of injectors and producers depends on mations and targets not directly below the pene-
factors such as (i) reservoir geometry, (ii) reservoir tration point or drilling from shore to locations
lithology, (iii) reservoir depth, (iv) porosity, under water [4]. A controlled deviation may also
(v) permeability, (vi) continuity of reservoir rock be used from a selected depth in an existing hole
properties, (vii) magnitude and distribution of fluid to attain economy in drilling costs. Various types
Petroleum and Oil Sand Exploration and Production 21

of tools are used in directional drilling along with pumps, are used to bring the oil to the surface.
instruments to help orient their position and mea- Other secondary recovery techniques increase the
sure the degree and direction of deviation; two reservoir’s pressure by water injection, natural gas
such tools are the whipstock and the knuckle reinjection and gas lift, which injects air, carbon
joint. The whipstock is a gradually tapered dioxide, or another gas into the reservoir. In addi-
wedge with a chisel-shaped base that prevents tion, reservoir heterogeneity, such as fractures and
rotation after it has been forced into the bottom faults, can cause reservoirs to drain inefficiently
of an open hole. As the bit moves down, it is during the production stage, and any highly
deflected by the taper about 5
from the alignment cemented zones or shale zones (i.e.,
of the existing hole. low-permeability zones) can produce barriers to
If the underground pressure in the oil reservoir the flow of fluids in reservoirs and lead to high
is sufficient, the oil will be forced to the surface residual oil saturation. Reservoirs containing
under this pressure (primary recovery). Natural crude oils with low API gravity often cannot be
gas (associated natural gas) is often present, produced efficiently without application of
which also supplies needed underground pressure enhanced oil recovery (EOR) methods because
(primary recovery). In this situation, it is sufficient of the high viscosity of the crude oil.
to place an arrangement of valves on the Conventional primary and secondary recovery
wellhead – the Christmas tree (Fig. 1) – in order processes are ultimately expected to produce
to connect the well to a pipeline network for about one-third of the original oil-in-place
storage and processing. (OOIP), although recoveries from individual res-
For limestone reservoir rock, acid is pumped ervoirs can range from less than 5% v/v to as high
down the well and out the perforations. The acid as 80% v/v of the original oil-in-place. This broad
dissolves channels in the limestone that lead oil range of recovery efficiency is a result of varia-
into the well. For sandstone reservoir rock, a spe- tions in the properties of the specific rock and
cially blended fluid containing proppants (sand, fluids involved from reservoir to reservoir as
walnut shells, aluminum pellets) is pumped down well as the kind and level of energy that drives
the well and out the perforations. The pressure the oil to producing wells, where it is captured.
from this fluid makes small fractures in the sand- However, there is always the potential for adverse
stone that allow oil to flow into the well, while the effects (Table 1).
proppants hold these fractures open. Once the oil Crude oil is also produced from offshore fields,
is flowing, production equipment is set up to usually from steel drilling platforms set on the
extract the oil from the well. A well is always ocean floor. In shallow, calm waters, these may
carefully controlled at the stage of production be little more than a wellhead and workspace, but
(the flush stage) to prevent the potentially danger- the larger ocean rigs include the well equipment
ous and wasteful gusher. This is a dangerous and processing equipment as well as crew quar-
condition and is (hopefully) prevented by the ters. Such platforms include the floating tension
blowout preventer and the pressure of the drilling leg platform that is secured to the seafloor by giant
mud. In most wells, acidizing or fracturing the cables and drillships. Such platforms can hold a
well starts the oil flow. steady position above a seafloor well using con-
Whatever the nature of the reservoir rock stant, computer-controlled adjustments. In Arctic
(sandstone, SiO2, limestone, or CaCO3), over the areas, islands may be built from dredged gravel
lifetime of the well, the pressure will fall, and at and sand to provide platforms capable of resisting
some point, there will be insufficient underground drifting ice fields.
pressure to force the oil to the surface. Secondary
oil recovery uses various techniques to aid in Primary Recovery Methods
recovering oil from depleted or low-pressure res- Petroleum recovery usually starts with a forma-
ervoirs. Sometimes pumps, such as beam pumps tion pressure high enough to force crude oil into
(horsehead pumps) and electrical submersible the well and sometimes to the surface through the
22 Petroleum and Oil Sand Exploration and Production

Petroleum and Oil Sand Exploration and Production, In dissolved gas drive (solution gas drive) [2,
Table 1 Recovery process parameters and their potential 4], the propulsive force is the gas in solution in the
adverse effects
oil, which tends to come out of solution because
Property Comments of the pressure release at the point of penetration
Carbon dioxide Lowers pH; can change oil of a well. Dissolved gas drive is the least efficient
injection composition leading to phase
separation of sludge or sediment and type of natural drive as it is difficult to control the
blocking of channels gas-oil ratio and the bottom-hole pressure drops
Miscible Hydrocarbon-rich gases lower the rapidly. If gas overlies the oil beneath the top of
flooding solubility parameter and solvent the trap, it is compressed and can be utilized (gas
power of the oil and cause separation cap drive) to drive the oil into wells situated at the
of asphaltene material
bottom of the oil-bearing zone [2, 4]. By produc-
Organic Can lower the solubility parameter
chemicals and solvent power of the oil and ing oil only from below the gas cap, it is possible
cause separation of asphaltene to maintain a high gas-oil ratio in the reservoir
material; blocking of channels until almost the very end of the life of the pool. If,
Acidizing Interaction of crude oil constituents however, the oil deposit is not systematically
upsetting molecular balance and
deposition of sludge or sediment;
developed so that bypassing of the gas occurs,
blocking of channels an undue proportion of oil is left behind.
Pressure Can change composition of oil Usually, the gas in a gas cap (associated natural
decrease medium leading to phase separation gas) contains methane and other hydrocarbons that
of asphaltene material as sludge or may be separated out by compressing the gas.
sediment; blocking of channels
A well-known example is natural gasoline that
Temperature Can change composition of oil
decrease medium leading to phase separation was formerly referred to as casinghead gasoline
of asphaltene material as sludge or or natural gas gasoline. However, at high pres-
sediment; blocking of channels sures, such as those existing in the deeper fields,
the density of the gas increases, and the density of
the oil decreases until they form a single phase in
the reservoir. These are the so-called retrograde
tubing [5]. For a newly opened formation and condensate pools because a decrease (instead of
under ideal conditions, the proportions of gas an increase) in pressure brings about condensation
may be so high that the oil is, in fact, a solution of the liquid hydrocarbons. When this reservoir
of liquid in gas that leaves the reservoir rock so fluid is brought to the surface and the condensate
efficiently that a core sample will not show any is removed, a large volume of residual gas remains.
obvious oil content. A general rough indication of In many cases, this gas is recycled by compression
this situation is a high ratio of gas to oil produced. and injection back into the reservoir, thus
This ratio may be zero for fields in which the rock maintaining adequate pressure within the gas cap,
pressure has been dissipated. The oil must be and condensation in the reservoir is prevented.
pumped out to as much as 50,000 ft3 or more of The most efficient propulsive force in driving
gas per barrel of oil in the so-called condensate oil into a well is natural water drive, in which the
reservoirs, in which a very light crude oil (0.80 pressure of the water forces the lighter recoverable
specific gravity or lighter) exists as vapor at high oil out of the reservoir into the producing wells [2,
pressure and elevated temperature. Once produc- 4]. In anticlinal accumulations, the structurally
tion commenced, the crude oil moves out of the lowest wells around the flanks of the dome are
reservoir into the well by one or more of four the first to come into water. Then, the oil-water
processes. These processes are (i) dissolved gas contact plane moves upward until only the wells
drive, (ii) gas cap drive, (iii) water drive, and at the top of the anticline are still producing oil;
(iv) gravity drive. Early recognition of the type eventually, these also must be abandoned as the
of drive involved is essential to the efficient devel- water displaces the oil. The force behind the water
opment of an oil field. drive may be hydrostatic pressure, the expansion of
Petroleum and Oil Sand Exploration and Production 23

the reservoir water, or a combination of both. Water Secondary Recovery Methods

drive is also used in certain submarine fields. Secondary oil recovery methods use various tech-
Gravity drive is an important factor when oil niques to aid in recovering oil from depleted or
columns of several thousands of feet exist. Fur- low-pressure reservoirs. Surface pumps or sub-
thermore, the last bit of recoverable oil is pro- merged pumps (electrical submersible pumps
duced in many pools by gravity drainage of the (ESPs)) are often used to bring the oil to the
reservoir. Another source of energy during the surface. Other secondary recovery techniques
early stages of withdrawal from a reservoir increase the reservoir pressure by water injection
containing undersaturated oil is the expansion of and/or by gas injection. In fact, the first method
that oil as the pressure reduction brings the oil to recommended for improving the recovery of oil
the bubble point (the pressure and temperature at was a pressure maintenance project which
which the gas starts to come out of solution). involved the reinjection of natural gas, and there
The recovery efficiency for primary production are indications that gas injection was utilized for
is generally low when liquid expansion and solu- this purpose before 1900 [6, 7].
tion gas evolution are the driving mechanisms. The most common follow-up secondary recov-
Much higher recoveries are associated with reser- ery operations usually involve the application of
voirs with water and gas cap drives and with pumping operations or of injection of materials
reservoirs in which gravity effectively promotes into a well to encourage movement and recovery
drainage of the oil from the rock pores. The over- of the remaining petroleum. The pump, generally
all recovery efficiency is related to how the reser- known as the horsehead pump (pump jack, nod-
voir is delineated by production wells. However, ding donkey, or sucker rod pump), provides
for primary recovery operations, no pumping mechanical lift to the fluids in the reservoir. The
equipment is required. If the reservoir energy is up-and-down movement of the sucker rods forces
not sufficient to force the oil to the surface, then the oil up the tubing to the surface. A walking
the well must be pumped. In either case, nothing is beam powered by a nearby engine may supply
added to the reservoir to increase or maintain the this vertical movement, or it may be brought
reservoir energy or to sweep the oil toward the about using a pump jack, which is connected to
well. The rate of production from a flowing well a central power source by means of pull rods.
tends to decline as the natural reservoir energy is Depending on the size of the pump, it generally
expended. When a flowing well is no longer pro- produces up to one-third of a barrel of an oil-water
ducing at an efficient rate, a pump is installed. emulsion at each stroke. The size of the pump is
There are also two processes that can be used to also determined by the depth and weight of the oil
improve formation characteristics, namely, to be removed, with deeper extraction requiring
acidizing and fracturing. Acidizing involves more power to move the heavier lengths of
injecting an acid into a soluble formation, such the rod.
as a carbonate, where it dissolves rock. This pro- Secondary oil recovery operations also involve
cess enlarges the existing voids and increases the injection of water or gas into the reservoir.
permeability. Hydraulic fracturing (fracking) When water is used, the process is called a water-
involves injecting a fluid into the formation flood and with gas, a gas flood. Separate wells are
under significant pressure that makes existing usually used for injection and production. The
small fractures larger and creates new fractures. injected fluids maintain reservoir pressure or
Petroleum production is invariably accompa- repressure the reservoir after primary depletion
nied by a decline in reservoir pressure, and pri- and displace a portion of the remaining crude oil
mary recovery comes to an end as the reservoir to production wells.
energy is reduced. At this stage, secondary recov- In the water-flooding process, water is injected
ery methods are applied to replace produced res- into a reservoir to obtain additional oil recovery
ervoir fluids and maintain (or increase) reservoir through movement of reservoir oil to a producing
pressure. well. Generally, the selection of an appropriate
24 Petroleum and Oil Sand Exploration and Production

flooding pattern for the reservoir depends on the Petroleum and Oil Sand Exploration and Production,
quantity and location of accessible wells. Fre- Table 2 The ratio of injection well production for various
patterns
quently, producing wells can be converted to
injection wells, whereas in other circumstances it Pattern Ratioa
may be necessary or advantageous to drill new Four spot 2
injection wells. Five spot 1
Seven spot 0.5
The mobility of oil is the effective permeability of
Inverted seven spot 2
the rock to the oil divided by the viscosity of the oil:
Nine spot 0.33
Inverted nine spot 3
l ¼ k=m a
Ratio of production wells to injection wells

In this equation, l is the mobility, md/cP; k is

the effective permeability of reservoir rock to a
given fluid, md; and m is the fluid viscosity, the structure of the reservoir can (will) dictate the
cP. Thus, the mobility ratio (M) is the mobility choice of pattern, and mobility ratio has an impor-
of the water divided by the mobility of oil: tant influence on pattern selection. If the ratio is
unfavorable, the injectivity of an injector will
M ¼ Krw mo =Kro mw
exceed the productivity of a producer, and water
injection will be supersede oil production. Hence,
In this equation, Krw is the relative permeability to balance the production with the water injection,
to water, Kro is the relative permeability to oil, mo is more producers than injectors are required. On the
the viscosity of the oil, and mw is the viscosity of other hand, if the mobility ratio is favorable, the
water. The mobility ratio (M) refers that Ko is the injectivity is impaired, and the pattern should have
mobility of oil ahead of the front (measured at Swc), more injectors than producers.
while Kw is the mobility of water at average water In a four-spot pattern, the distance between all
saturation in the water-contacted portion of the like wells is constant. Any three injection wells
reservoir. form an equilateral triangle with a production well
The mobility ratio of a waterflood will remain at the center. The four spots may be used when the
constant before breakthrough, but will increase injectivity is high or the heterogeneity is minimal.
after water breakthrough corresponding to the In a five-spot pattern, the distance between all like
increase in water saturation and relative permeabil- wells is constant. Four injection wells form a
ity to water in the water-contacted portion of the square with a production well at the center. If
reservoir. Furthermore, the mobility ratio at water existing wells were drilled on square patterns,
breakthrough is the term that is of significance in five-spot patterns (as well as nine-spot patterns)
describing relative mobility ratio, i.e., M <1 indi- are most commonly used since they allow easy
cates a favorable displacement as oil moves faster conversion to a five-spot waterflood. In the seven-
than water, and M = 1 indicates a favorable dis- spot pattern, the injection wells are located at the
placement as both oil and water move at equal corner of a hexagon with a production well at its
speed, whereas M >1 indicates an unfavorable center. If the reservoir characteristics yield lower
displacement as water moves faster than oil. than preferred injection rates, either a seven-spot
The well pattern can also contribute to the (or a nine-spot) pattern should be considered
efficiency of crude oil production. Generally, the because there are more injection wells per pattern
choice of pattern (Table 2) for waterflooding must than producing wells.
be consistent with the existing wells. The objec- In the nine-spot pattern, the arrangement is
tive is to select the proper pattern that will provide similar to that of the five spot but with an extra
the injection fluid with the maximum possible injection well drilled at the middle of each side of
contact with the crude oil to minimize bypassing the square. The pattern essentially contains eight
by the water. However, reservoir uniformity (and injectors surrounding one producer. If existing
heterogeneity) must be given consideration such wells were drilled on square patterns, nine-spot
Petroleum and Oil Sand Exploration and Production 25

patterns (as well as five-spot patterns) are most Enhanced oil recovery (tertiary oil recovery) is
commonly used. If the reservoir characteristics the incremental ultimate oil that can be recovered
yield lower injection rates than those desired, from a petroleum reservoir over oil that can be
one should consider using either a nine-spot pat- obtained by primary and secondary recovery
tern or a seven-spot pattern where there are more methods [2, 4, 8, 9]. Enhanced oil recovery
injection wells per pattern than producing wells. methods offer prospects for ultimately producing
In the inverted seven-spot pattern, the arrange- 30–60% v/v, or more, of the original oil-in-place.
ment is similar to the normal seven-spot pattern, Enhanced oil recovery processes use thermal,
except where the position of the producer well chemical, or fluid phase behavior effects to reduce
was in the normal seven-spot pattern, there is or eliminate the capillary forces that trap oil within
now an injector well. Likewise, where the injector pores, to thin the oil or otherwise improve its mobil-
wells were in the normal seven-spot pattern, there ity, or to alter the mobility of the displacing fluids. In
are now producer wells. The inverted seven-spot some cases, the effects of gravity forces, which
pattern may be used when the injectivity is high or ordinarily cause vertical segregation of fluids of
the heterogeneity is minimal. The arrangement of different densities, can be minimized or even used
the wells is similar to the normal nine-spot pattern to advantage. The various processes differ consider-
except the position of the producer well in the ably in complexity, the physical mechanisms
normal nine-spot pattern is occupied by an injec- responsible for oil recovery, and the amount of
tor well. Likewise, where the positions of the experience that has been derived from field applica-
injector wells were in the normal nine spot, there tion. The degree to which the enhanced oil recovery
are now producer wells. If the reservoir is homog- methods are applicable in the future will depend on
enous and the mobility ratio is unfavorable, the development of improved process technology. It
inverted nine-spot pattern may be promising. will also depend on improved understanding of
In the direct line-drive pattern, the lines of fluid chemistry, phase behavior, and physical prop-
injection and production are directly opposite to erties and on the accuracy of geology and reservoir
each other. If the injectivity is low or the heteroge- engineering in characterizing the physical nature of
neity is large, direct line drive is a good option. individual reservoirs [10]. The steam-based
Anisotropic permeability, permeability trends, or methods are the most advanced of all enhanced oil
oriented fracture systems favor line-drive patterns. recovery methods in terms of field experience and
In the staggered line-drive pattern, the wells are in thus have the least uncertainty in estimating perfor-
lines as in the direct line, but the injectors and mance when a good reservoir description is avail-
producers are no longer directly opposed but later- able. Steam processes are most often applied in
ally displaced by a distance specified that is depen- reservoirs containing heavy crude oil, usually in
dent upon the distance between wells of the same place of rather than following secondary or primary
type and the distance between the lines of injector methods. Commercial application of steam pro-
wells and producer wells. The staggered line-drive cesses has been underway since the early 1960s.
pattern is also effective for reservoirs where there is For taxation purposes, the Internal Revenue
anisotropic permeability or permeability trends or Service of the United States has listed the projects
oriented fracture systems. that qualify as enhanced oil recovery projects [11]
and are therefore available for a tax credit, and
these projects are:
Enhanced Oil Recovery Methods 1. Thermal Recovery Methods
Traditional primary and secondary recovery Thermal methods of recovery reduce the vis-
methods typically recover less than half cosity of the crude oil by heat so that it flows more
(sometimes less than one-third) of the oil, only easily into the production well.
one-third of the original oil-in-place. It is at some
point before secondary recovery ceases to remain (a) Steam drive injection – the continuous injec-
feasible that enhanced oil recovery methods must tion of steam into one set of wells (injection
be applied if further oil is to be recovered. wells) or other injection source to effect oil
26 Petroleum and Oil Sand Exploration and Production

displacement toward and production from a reduce the capillary forces that trap the oil in
second set of wells (production wells). the pores of the rock. The surfactant slug
(b) Cyclic steam injection – the alternating injec- displaces most the oil from the reservoir vol-
tion of steam and production of oil with con- ume contacted, forming a flowing oil-water
densed steam from the same well or wells. bank that is propagated ahead of the surfactant
(c) In situ combustion – the combustion of oil or slug. The principal factors that influence the
fuel in the reservoir sustained by injection of surfactant slug design are interfacial proper-
air, oxygen-enriched air, oxygen, or supple- ties, slug mobility in relation to the mobility of
mental fuel supplied from the surface to dis- the oil-water bank, the persistence of accept-
place unburned oil toward producing wells. able slug properties and slug integrity in the
This process may include the concurrent, reservoir, and cost.
alternating, or subsequent injection of water. (b) Microemulsion flooding also known as
surfactant-polymer flooding involves injection
2. Gas Flood Recovery Methods of a surfactant system (e.g., a surfactant, hydro-
carbon, cosurfactant, electrolyte, and water) to
(a) Miscible fluid displacement – the injection of enhance the displacement of oil toward pro-
gas (e.g., natural gas, enriched natural gas, a ducing wells, and [2] caustic flooding – the
liquefied petroleum slug driven by natural gas, injection of water that has been made chemi-
carbon dioxide, nitrogen, or flue gas) or alco- cally basic by the addition of alkali metal
hol into the reservoir at pressure levels such hydroxides, silicates, or other chemicals.
that the gas or alcohol and reservoir oil are (c) Polymer-augmented waterflooding – the
miscible. injection of polymeric additives with water
(b) Carbon dioxide augmented waterflooding – the to improve the areal and vertical sweep effi-
injection of carbonated water, or water and car- ciency of the reservoir by increasing the vis-
bon dioxide, to increase waterflood efficiency. cosity and decreasing the mobility of the
(c) Immiscible carbon dioxide displacement – the water injected; polymer-augmented
injection of carbon dioxide into an oil reser- waterflooding does not include the injection
voir to effect oil displacement under condi- of polymers for modifying the injection pro-
tions in which miscibility with reservoir oil is file of the wellbore or the relative permeability
not obtained; this process may include the of various layers of the reservoir, rather than
concurrent, alternating, or subsequent injec- modifying the water-oil mobility ratio.
tion of water.
(d) Immiscible nonhydrocarbon gas displacement – In keeping with the reservoir characteristics
the injection of nonhydrocarbon gas (such as being a major issue to crude oil production, for
nitrogen) into an oil reservoir, under conditions the enhanced oil recovery processes that involve
in which miscibility with reservoir oil is not the use of chemicals – surfactant/polymer, poly-
obtained, to obtain a chemical or physical reac- mer, and alkaline flooding [13] – it must be recog-
tion (other than pressure) between the oil and the nized that each reservoir has unique fluid and rock
injected gas or between the oil and other reser- properties, and specific chemical systems must be
voir fluids; this process may include the concur- designed for each individual application. The
rent, alternating, or subsequent injection of chemicals used, the concentration of the chemicals,
water. and the number of chemical as well as the amount
of chemicals are also dependent upon the specific
3. Chemical Flood Recovery Methods properties of the fluids and the rocks involved and
upon economic considerations.
(a) Surfactant flooding is a multiple-slug process Certain types of reservoirs, such as those with
involving the addition of surface-active very viscous crude oils and some low-permeability
chemicals to water [12]. These chemicals carbonate (limestone, dolomite, or chert)
Petroleum and Oil Sand Exploration and Production 27

reservoirs, respond poorly to conventional second- injection to increase production from other wells
ary recovery techniques. The viscosity (or the API (indirect steam stimulation). When there is some
gravity) of crude oil is an important factor that must natural reservoir energy, steam stimulation nor-
be considered when heavy oil is recovered from a mally precedes steam drive. In steam stimulation,
reservoir. In these reservoirs, it is desirable to ini- heat is applied to the reservoir by the injection of
tiate enhanced oil recovery operations as early as high-quality steam into the produce well. This
possible, which may mean considerably abbreviat- cyclic process (huff and puff or steam soak) uses
ing conventional primary recovery operations and the same well for both injection and production.
secondary recovery operations to the extent that The period of steam injection is followed by pro-
these operations may be bypassed. duction of reduced viscosity oil and condensed
Thermal enhanced oil recovery processes add steam (water). One mechanism that aids produc-
heat to the reservoir to reduce oil viscosity and/or tion of the oil is the flashing of hot water
to vaporize the oil. In both instances, the oil is (originally condensed from steam injected under
made more mobile so that it can be more effec- high pressure) back to steam as pressure is
tively driven to producing wells. In addition to lowered when a well is put back on production.
adding heat, these processes provide a driving Cyclic steam injection is the alternating injec-
force (pressure) to move oil to producing wells. tion of steam and production of oil with con-
These methods have found most use when the oil densed steam from the same well or wells. Thus,
in the reservoir has a high viscosity. For example, steam generated at surface is injected in a well,
heavy oil is usually highly viscous (hence the use and the same well is subsequently put back on
of the adjective heavy), with a viscosity ranging production. The process includes three stages:
from approximately 100 cP to several million (i) the injection, during which a measured amount
centipoises at the reservoir conditions. In addi- of steam is introduced into the reservoir; (ii) the
tion, oil viscosity is also a function of temperature soak period, which requires that the well be shut in
and API gravity [2, 14]. Thus, for heavy crude oil for a period of time, which is usually measured in
samples with API gravity ranging from 4 to days, to allow uniform heat distribution to reduce
21
API (1.04–0.928 kg/m3): the viscosity of the oil (alternatively, to raise the
reservoir temperature above the pour point of the
loglogðms þ aÞ ¼ A  B logðT þ 460Þ
oil; and (iii) the production stage in which the
now-mobile oil is produced through the injection
In this equation, ms is the oil viscosity in cP, T is
well. The cycle is repeated until the flow of crude
the temperature in
F, A and B are constants, and a
oil diminishes to a point of no returns. Using in
is an empirical factor used to achieve a straight-line
situ combustion to stimulate oil production is
correlation at low viscosity. This equation is usu-
regarded as attractive for deep reservoirs [15]
ally used to correlate kinematic viscosity in
and, in contrast to steam injection, usually
centistokes, in which case an a of 0.6–0.8 is
involves no loss of heat. The duration of the
suggested (dynamic viscosity in cP equals kine-
combustion may be short (<30 days) or more
matic viscosity in cSt times density in g/ml).
prolonged (up to 90 days), depending upon
An alternative equation for correlating viscos-
requirements. In addition, backflow of the oil
ity data (where A and B are constants and T* is the
through the hot zone must be prevented or coking
absolute temperature) is
occurs.
m ¼ aeb=T Both forward and reverse combustion methods
have been used with some degree of success when
Steam drive injection (steam injection) has applied to oil sand deposits. The forward combus-
been commercially applied since the early tion process has been applied to the Orinoco
1960s. The process occurs in two steps: (i) steam deposits [16] and in the Kentucky sands
stimulation of production wells, that is, direct [15]. The reverse combustion process has been
steam stimulation, and (ii) steam drive by steam applied to the Orinoco deposit [17] and the
28 Petroleum and Oil Sand Exploration and Production

Athabasca [2, 4]. In tests such as these, it is essen- formation, which further increases contact of a
tial to control the airflow and to mitigate the well with the reservoir. Directional drilling tech-
potential for spontaneous ignition. A modified nology also enables drilling several wells from a
combustion approach has been applied to the Ath- single well pad, thus cutting costs while reducing
abasca deposit [2, 4]. The technique involved a environmental disturbance. Combining hydraulic
heat-up phase and a production (or blowdown fracturing with directional drilling has improved
phase) followed by a displacement phase using a the production of tight (less permeable) petroleum
fireflood-waterflood (COFCAW) process. and natural gas reservoirs, particularly unconven-
tional gas shales such as the Marcellus Shale
Hydraulic Fracturing formation.
Traditional oil and gas extraction involves drilling In the past three to four decades, hydraulic
through impervious rock that traps concentrated fracturing has been increasingly used in forma-
underground reservoirs of petroleum and natural tions that were known to be rich in natural gas and
gas and may even involve a mining operation for crude oil that were trapped in the impermeable
heavy oil and oil sand bitumen. With conventional reservoir rock that was technologically and eco-
petroleum, extraction occurs simply due to the nomically difficult to produce. The application of
change in pressure caused by the drilling. How- hydraulic fracturing to tight sands revitalized old
ever, not all of the petroleum and natural gas is fields and allowed establishment of new fields.
conveniently located in conventional and accessi- Subsequently, the application of the hydraulic
ble reservoirs. Many oil and gas resources are fracturing process to shale formations has opened
trapped in the pore spaces and cracks within imper- new areas to development, including the Marcel-
meable sedimentary rock formations – shale and lus Shale in the eastern United States, the Barnett
tight sandstone formations are examples of such Shale in Texas, and the Fayetteville Shale in
reservoirs. These reservoirs can vary in Arkansas, among others [18–20].
thickness – the shale formations are relatively thin The process involves the use of a perforating
layers (albeit deep under the ground) but cover gun which is lowered into a newly drilled well and
extensive horizontal areas, and a vertically drilled lined up precisely within the target formation
well will only access a small area of the reservoir using seismic images, well logs, global position-
and, by inference due to the impermeable nature of ing systems, and other indicators to target the
the formation, a minimal part of the resource. spots from which tight gas and oil are most likely
However, when the drilling operation can deviate to occur. When fired, the gun punches small holes
from the conventional vertical plane and move in in the well casing, cement, and rock after which
the horizontal plane, much more of the reservoir fracturing fluid is pushed out the perforations
resource becomes accessible. under high pressure, creating small cracks in the
Thus, conventional petroleum reservoirs formation that allow the gas and oil to flow from
depend on the pressure of their gas cap and the rock. In practice, the well is fractured in stages
oil-dissolved gas to lift the oil to the surface (i.e., and a plug set between each stage. When all of the
gas drive). Water trapping the petroleum from stages have been completed and plugged, the
below also exerts an upward hydraulic pressure plugs are removed (drilled out), which allows
(i.e., water drive). The combined pressure in the gas or oil to flow up through the well to
petroleum reservoirs produced by the natural gas begin production. This fracturing fluid contains
and water drives is known as the conventional proppants such as sand, coarsely ground walnut
drive. As a reservoir’s production declines, lifting shells, and other similarly sized materials to main-
further petroleum to the surface, like the lifting of tain many cracks (fractures) created by the pres-
water, requires pumping or artificial lift. More- sure treatment in the open positon, thus
over, advances in directional drilling technology preventing closure when the pressure treatment
have allowed wells to deviate from nearly vertical is cased. Although the fluid (slickwater) used for
to extend horizontally into the reservoir hydraulic fracturing is predominantly water, it
Petroleum and Oil Sand Exploration and Production 29

does contain chemicals (in addition to the pro- gas. Furthermore, natural fractures, which are
ppant) which can pose an environmental risk. often the primary means for fluid flow in
The goal of hydraulic fracturing is to create a low-permeability reservoirs, can (and often do)
highly conductive fracture system that will allow compromise the ability to predict the geometry
flow of fluids and/or gases through the formation of hydraulic fractures and the effect on production
to the production well. Hydraulic fracturing fluids and drainage. Thus, while it is possible to have
are used to initiate and/or expand fractures, as well knowledge and understanding of existing natural
as to transport proppant into fractures in coalbed fracture systems, geological discontinuities such
formations. Proppants are sand or other granular as fractures and faults will most likely influence
substances injected into the formation to hold or fracture geometry, thereby rendering difficult the
prop open the formation fractures created by prediction of hydraulic fracture behavior.
hydraulic fracturing. The viscosity of fracturing Whether a reservoir is high, moderate, and low,
fluids is considered when they are formulated, to permeability can be defined based on both the
provide for efficient transport and placement of formation permeability and the reservoir fluid vis-
proppant into a fracture. Most of the fracturing cosity (the k/m ratio, k is the formation permeabil-
fluids injected into the formation are pumped back ity in millidarcies, and m is the formation fluid
out of the well along with water, crude oil, and viscosity in centipoises). For a well producing
natural gas. natural gas, the viscosity of the gas is assumed to
Without hydraulic fracturing and horizontal be approximately 0.02 cP. Thus, for a typical gas
drilling, resources such as (i) tight sands well, a low-permeability formation might be k <
containing natural gas and crude oil [11], coalbed 0.1 md, [2] medium-permeability reservoir might
methane, and (iii) shale formation also containing be 10>k>1 md, and [3] a high-permeability res-
natural gas and crude oil would remain ervoir might be 25 md >k. If the formation con-
undeveloped. In fact, many modern oil field pro- tains crude oil with a fluid viscosity of 2 cP, all the
duction operations would not exist without permeability values must be multiplied by a factor
hydraulic fracturing, and as the global balance of of 100 to determine whether the reservoir is a
supply and demand forces the petroleum and nat- low-permeability, moderate-permeability, or
ural gas industry toward more unconventional high-permeability formation. Thus, the definition
resources including shale formations such as the of reservoir permeability depends on the value of
Barnett, Haynesville, Bossier, and Marcellus the viscosity of the reservoir fluid. In heavy-oil
plays, hydraulic fracturing will continue to play reservoirs, in which the viscosity of the fluid is on
a substantive role in unlocking otherwise the order of several thousand centipoises, forma-
unobtainable reserves. tions with a permeability on the order of several
Reservoir characteristics must always be con- Darcys are a low-permeability reservoir. By infer-
sidered when designing hydraulic fracturing. For ence, if the reservoir is a formation containing
example, in moderate- to high-permeability reser- either natural gas or conventional (light) crude
voirs, fractures are designed to improve produc- oil, a permeability on the order of several hundred
tion by bypassing near-wellbore formation millidarcies or more can be considered to be a
damage [21, 22]. In fact, in low-permeability res- high-permeability reservoir.
ervoirs (often referred to as tight shale formations) Furthermore, the main reasons for fracture
and the most common reservoir type to be fracture treating high-permeability formations are (i) to
stimulated, fracture length is the overriding factor improve both the reservoir and wellbore commu-
for increased productivity and recovery. In addi- nication, (ii) to bypass formation damage, (iii) to
tion, and from a reservoir-development stand- reduce the drawdown around the wellbore, (iv) to
point, understanding the fracture geometry and increase the back stress on the formation, (v) to
orientation is crucial for determining well spacing control sand production, (vi) to reduce fines
[23] and field development strategies designed to migration, (vii) to reduce asphaltene deposition,
extract more hydrocarbons petroleum or natural and (viii) to reduce water coning [24]. By creating
30 Petroleum and Oil Sand Exploration and Production

a short, highly conductive fracture connecting the Offshore Production

reservoir to the wellbore, the productivity index is A range of different structures are used offshore
increased; thus, more oil and gas can be produced oil production, depending on size and water
with a lower drawdown. As the drawdown is depth. In the several decades, there has been an
reduced, the tendency of a poorly consolidated evolution of the offshore installations, including
reservoir to produce sand is also reduced. In cer- those that rest on the sea bottom. These include
tain formations, the reduction in drawdown also installations with multiphase piping to the shore
helps to deter (i) fines migration, (ii) deposition of with no offshore topside structure. In addition,
asphaltene deposition, and (iii) water coning – the deviation drilling (directional drilling) is used to
phenomenon in which bottom water gradually and reach different parts of the reservoir from a few
frequently suddenly displaces a part or all of the wellhead cluster locations which have replaced
oil production when a certain rather critical pro- the outlying wellhead towers.
duction rate from the well is exceeded. A shallow water complex (typically found in
Finally, hydraulic fracturing is utilized off- water depths up to 350 ft) is characterized by
shore primarily during the well completion phase several independent platforms with different
of developing a well for production to enhance parts of the process and utilities linked with gang-
safety and security of the well while optimizing way bridges. Individual platforms include the
production. This constitutes most hydraulic frac- wellhead platform, the riser platform, the pro-
turing activities that are conducted offshore. cessing platform, the accommodations platform,
Hydraulic fracturing can also be used to prepare and the power generation platform.
a well for enhanced oil recovery or to work over A gravity base complex consists of enormous
the well to increase production when the well has concrete fixed structures placed on the bottom,
been under production for some time. typically with oil storage cells in the skirt that
Coupled with hydraulic fracturing, the crude oil rest on the sea bottom. The large deck receives
and natural gas resources in shale formation and in all parts of the process and utilities in large mod-
tight formations (often referred to as tight oil and ules. Large fields at 350–1600 ft water depth are
tight gas, respectively) become accessible and typical of such complexes. The concrete was
recoverable. Furthermore, horizontal drilling poured at an onshore location, with enough air in
makes it possible for a well to be drilled vertically the storage cells to keep the structure floating until
several thousand feet or meters and then curved to it is towed out to sea and lowered on to the seabed.
extend at an angle parallel to the earth’s surface, A compliant tower platform is similar to a fixed
threading the well through the horizontal gas for- platform and consists of a narrow tower, attached
mation to capture more pockets of gas. On the other to a foundation on the seafloor and extending up
hand, in some geological settings, it is more appro- to the platform. This tower is flexible, as opposed
priate to directionally drill s-shaped wells from a to the relatively rigid legs of a fixed platform. This
single pad to minimize surface disturbance. These flexibility allows them to operate in much deeper
types of wells are drilled vertically several thou- water, as they can “absorb” much of the pressure
sand feet and then extend in arc shapes beneath the exerted by the wind and sea. Compliant towers are
surface of the earth. Whatever the type of well, used between 1600 and 3300 ft water depth.
multiple wells can be drilled from a central location A floating production platform has all topside
to proceed in different directions within the reser- systems located on a floating structure with dry or
voir. During drilling, mobile drilling units are subsea wells. The floating production, storage,
moved between wells on a single drilling location and offloading platform has the advantage of
(pad), which avoids dismantling and reassembling being a stand-alone structure that does not need
drilling equipment for each well, making the pro- external infrastructure such as pipelines or stor-
cess shorter. This procedure limits the required age. Crude oil is offloaded to a shuttle tanker at
number of drilling pads and, therefore, leaves a regular intervals, from days to weeks, depending
smaller environmental footprint on the surface. on production and storage capacity. The platform
Petroleum and Oil Sand Exploration and Production 31

Petroleum and Oil Sand Exploration and Production, Table 3 Types of ocean tankers used for crude oil
transportation
Type/cargo Length (m) Beam (m) Draft (m) dwt* Speed (knot)
1. VLCC/crude 332 58 22 298,000 15.5
2. Multiproduct/crude-heated tanks 244 42 14.6 105,000 14
3. Clean/dirty product 183 32 12.2 47,000 15.6
4. LPG carrier 209 31.4 12.5 47,000 –
5. LNG spherical tanks 272 47.2 11.4 67,000 18.5
6. LNG “oblong” tanks 287 41.8 11.3 71,470 19.2
dwt* (deadweight tons) is the cargo capacity of the vessel

is typically a tanker-type hull or barge, often SPAR is not an acronym, but refers to its likeness
converted from an existing crude oil tanker to a ship’s spar. SPARs can support dry comple-
(VLCC or ULCC) (Table 3). tion wells, but are more often used with subsea
The wellheads or subsea risers from the sea wells.
bottom are located on a central or bow-mounted Subsea production systems are wells located on
turret so that the ship can rotate freely to point into the seafloor, as opposed to the surface. As in a
wind, waves, or current. The turret has wire rope floating production system, the petroleum is extra-
and chain connections to several anchors (position cted at the seabed and can then be “tied-back” to
mooring), or it can be dynamically positioned an already existing production platform or even an
using thrusters (dynamic positioning). Most onshore facility, limited by horizontal distance or
installations use subsea wells. The main process “offset.” The well is drilled by a movable rig, and
is placed on the deck, while the hull is used for the extracted oil and natural gas are transported by
storage and offloading to a shuttle tanker. undersea pipeline and riser to a processing facility.
A tension leg platform consists of a structure This allows one strategically placed production
held in place by vertical tendons connected to the platform to service many wells over a reasonably
seafloor by pile-secured templates. The structure large area. Subsea systems are typically in use at
is held in a fixed position by tensioned tendons, depths of 500 m or more and do not have the
which provide for use of the platform in a broad ability to drill, only to extract and transport. Dril-
water depth range up to about 6500 ft. The ten- ling and completion are performed from a surface
dons are constructed as hollow high-tensile rig. Horizontal offsets of up to 250 km/150 miles
strength steel pipes that carry the spare buoyancy are currently possible.
of the structure and ensure limited vertical motion.
A semisubmersible platform has a similar design
but without taut mooring. This arrangement per- Oil Sand Exploration and Production
mits more lateral and vertical motion and is gen-
erally used with flexible risers and subsea wells. Heavy oil and bitumen (the component of interest
A SPAR platform consists of a single tall float- in an oil sand deposit) are often defined (loosely
ing cylindrical hull, supporting a fixed deck. The and incorrectly) in terms of API gravity. A more
cylinder does not however extend all the way to appropriate definition of bitumen, which sets it
the seabed, but is tethered to the bottom by a series aside from heavy oil and conventional petroleum,
of cables and lines. The large cylinder serves to is based on the recovery process as presented in
stabilize the platform in the water and allows for the definition of oil sand (tar sand) offered by the
movement to absorb the force of potential hurri- US government as the extremely viscous hydro-
canes. SPARs can be quite large and are used for carbon which is not recoverable in its natural
water depths from 350 ft and up to 16,000 ft. state by conventional oil well production methods
32 Petroleum and Oil Sand Exploration and Production

including currently used enhanced recovery tech- acid number, can give meaningful results to use in
niques [2, 4]. By inference, conventional petro- predictions of problems. An alternative way of
leum and heavy oil (recoverable by conventional defining the relative organic acid fraction of
oil well production methods including currently crude oils is therefore a real need in the oil indus-
used enhanced recovery techniques) are different try, both upstream and downstream.
to oil sand bitumen. Be that as it may, some stage Opportunity crude oils are either new crude
of production, conventional petroleum (in the oils with unknown or poorly understood proper-
later stages of recovery) and heavy oil (in the ties relating to processing issues or are existing
earlier stages of recovery) may require the appli- crude oils with well-known properties and pro-
cation of enhanced oil recovery methods for cessing concerns [27]. Opportunity crude oils are
recovery. Both heavy oil and oil sand bitumen often, but not always, heavy crude oils but in
are difficult to refine, and, in this context of this either case are more difficult to process due to
section, there are three other types of feedstocks high levels of solids (and other contaminants)
that are worthy of mention: extra heavy oil, high- produced with the oil, high levels of acidity, and
acid crude oil, and opportunity crude oil (Table 4). high viscosity (Table 5). These crude oils may
Extra heavy oil is a nondescript term (related to also be incompatible with other oils in the refinery
viscosity) of little scientific meaning which is feedstock blend and cause excessive equipment
usually applied to oil sand bitumen, which is fouling when processed either in a blend or
generally incapable of free flow under reservoir separately.
conditions. The general difference is that extra There is also the need for a recovery operations
heavy oil, which may have properties similar to to be on guard to accommodate opportunity crude
oil sand bitumen in the laboratory, unlike oil sand oils and/or high-acid crude oils which, for many
bitumen in the deposit, has some degree of mobil- general refining purposes, are often included in
ity in the reservoir or deposit [25]. Extra heavy the term heavy feedstocks.
oils can flow at reservoir temperature and can be
produced economically, without additional Oil Mining
viscosity-reduction techniques, through variants Oil mining is the term applied to the surface or
of conventional processes such as long horizontal subsurface excavation of petroleum-bearing for-
wells or multilaterals. This is the case, for mations for subsequent removal of the heavy oil
instance, in the Orinoco basin (Venezuela) or in or bitumen by washing, flotation, or retorting
offshore reservoirs of the coast of Brazil, but once treatments. Oil mining includes recovery of oil
outside of the influence of the high reservoir tem- and/or heavy oil by drainage from reservoir beds
perature, these oils are too viscous at surface to be to mine shafts or other openings driven into the
transported through conventional pipelines and rock or by drainage from the reservoir rock into
require heated pipelines for transportation. Alter- mine openings driven outside the reservoir but
natively, the oil must be partially upgraded or fully connected with it by boreholes or mine wells.
upgraded or diluted with a light hydrocarbon Oil mining methods should be applied in reser-
(such as aromatic naphtha) to create a mix that is voirs that have significant residual oil saturation
suitable for transportation. and have reservoir or fluid properties that make
High-acid crude oil is crude oil that contain production by conventional methods inefficient or
considerable proportions of naphthenic acids impossible.
which, as commonly used in the petroleum indus- The high well density in improved oil mining
try, refers collectively to all of the organic acids usually compensates for the inefficient production
present in the crude oil [26]. In many instances, caused by reservoir heterogeneity. However, close
the high-acid crude oils are the heavier crude oils. spacing of the production wells can also magnify
The total acid matrix is therefore complex, and it the deleterious effects of reservoir heterogeneity.
is unlikely that a simple titration, such as the If a high-permeability streak exists with a lateral
traditional methods for measurement of the total extent that is less than the inter-well spacing of
Petroleum and Oil Sand Exploration and Production 33

Petroleum and Oil Sand Exploration and Production, Table 4 Simplified differences between the various types of
crude oil and tar sand bitumena
Conventional crude oil
Mobile in the reservoir; API gravity: >25

High-permeability reservoir
Primary recovery and secondary recovery
Tight oil
Similar properties to the properties of conventional crude oil; API gravity: >25

Immobile in the reservoir
Low-permeability reservoir (tight sands, tight shale formation)
Horizontal drilling into reservoir and fracturing (typically multi-fracturing) to release fluids/gases
Medium crude oil
Similar properties to the properties of conventional crude oil; API gravity: 20–25

High-permeability reservoir
Primary recovery and secondary recovery
Opportunity crude oil
May have similar properties to the properties of conventional crude oil
API gravity: 20–25

High-permeability reservoir
Primary recovery and secondary recovery
Typically needs cleaning (processing) before refining because of contaminants
Heavy crude oil
More viscous than conventional crude oil; API gravity: 10–20

Mobile in the reservoir
Recovery method: typically, enhanced oil recovery such as a steam-based process
High-acid crude oil
Can be equivalent to medium crude oil or heavy crude oil and recovered accordingly
API gravity: (typically) 15–25

Total acid number (TAN) >0.5 mg KOH/ml of crude
Extra heavy oil
Similar properties to the properties of tar sand bitumen; API gravity: <10

Mobile in the reservoir/deposit
Reservoir temperature is higher than the melting (liquefaction) point of the material
Recovery method: typically, enhanced oil recovery such as a steam-based process
Tar sand bitumen
Immobile in the deposit; API gravity: <10

Mining (often preceded by explosive fracturing) followed by bitumen separation from the sand
Non-mining methods: steam-assisted gravity draining (SAGD) and solvent methods (VAPEX)
Extreme heating methods and innovative methodsb
a
API gravity data are general estimates only – the recovery method is more indicative of the type of feedstock that is
produced from a reservoir or deposit
b
Innovative methods exclude tertiary recovery methods and methods such as SAGD, VAPEX but does include variants or
hybrids thereof

conventional wells but is comparable to that of heterogeneous reservoirs. Other reservoirs that
improved oil mining, the channeling is more unfa- are good candidates for oil mining are those that
vorable for the improved oil mining method. are shallow, have high oil saturation, have a
Under favorable circumstances, higher produc- nearby formation that is competent enough to
tion rates may be achieved in fractured reservoirs support the mine, and cannot be efficiently pro-
by improving mining methods in less duced by conventional methods.
34 Petroleum and Oil Sand Exploration and Production

Petroleum and Oil Sand Exploration and Production, Table 5 Feedstock contaminants and their effects
Contaminants Effect on Catalyst Mitigation
Sulfur Catalyst fouling Hydrodesulfurization
Deactivation of active sites
Nitrogen Adsorption of basic nitrogen Hydrodemetallization
Destruction of active sites
Metals Fouling of active sites Demetallization
Fouling of pores
Particulate matter Deactivation of active sites Filter/pretreatment
Pore plugging
Coke precursors Formation of coke Remove resin and asphaltene constituents
Catalyst fouling
Deactivation of active sites
Pore plugging

Oil Sand Mining (i) Primary properties, which are those proper-
The immobile bitumen occurring in oil sand ties that have an influence on the fluid flow
deposits poses a major recovery problem. The and fluid storage properties and include rock
material is notoriously immobile at formation and fluid properties, such as porosity, perme-
temperatures and must therefore require some ability, wettability, crude oil viscosity, and
stimulation (usually by thermal means) to ensure pour point
recovery. Alternatively, proposals have been (ii) Secondary properties, which are those prop-
noted which advocate bitumen recovery by sol- erties that significantly influence the primary
vent flooding or by the use of emulsifiers. There is properties, including pore size distribution,
no doubt that with time one or more of these clay mineral type, as well as content
functions may come to commercial fruition, but (iii) Tertiary properties, which are those proper-
for the present the two major commercial opera- ties that influence oil production operation
tions rely on the mining technique. and include fracture breakdown pressure,
The alternative to in situ processing is to mine hardness, and thermal properties) and the
oil sand, transport the mined material to a pro- mining operations (e.g., temperature, subsi-
cessing plant, extract the bitumen, and dispose of dence potential, and fault distribution)
the waste sand. Such a procedure is often referred
to as oil mining. This is the term applied to the Also, there are also important rock mechanical
surface or subsurface excavation of petroleum- parameters of the formation in which a tunnel is to
bearing formations for subsequent removal of be mined and from where all oil mining operations
the heavy oil or bitumen by washing, flotation, will be conducted. These properties are mostly
or retorting treatments. related to the mining aspects of the operations,
Engineering a successful oil mining project and not all are of equal importance in their influ-
must address several items because there must ence on the mining technology. Their relative
be sufficient recoverable resources, the project importance also depends on the individual reser-
must be conducted safely, and the project should voir. In addition, there is also hydrotransport
be engineered to maximize recovery within eco- which is a combination of ore transport and pre-
nomic limits. For example, the reservoir proper- liminary extraction. Underground mining is also
ties, which can affect the efficiency of heavy oil or an option, but extra caution is needed because of
bitumen production by mining technology, can be the fear of collapse of the unconsolidated sand-
grouped into three classes: bitumen formation.
Petroleum and Oil Sand Exploration and Production 35

After the bituminous sand has been recovered, the surface and is removed as froth. A middling
it is mixed with water and caustic soda to form a stream (mostly water with suspended fines and
slurry and is pumped along a pipeline to the some bitumen) is withdrawn from approximately
extraction plant. The extraction process thus midway up to the side of the cell wall. Part of the
begins with the mixing of the water and agitation middlings is recycled to dilute the conditioning
needed to initiate bitumen separation from the drum effluent for pumping. Clay minerals do not
sand and clay minerals. settle readily and generally accumulate in the mid-
dlings layer. A high concentration of clay minerals
Hot-Water Process increases the viscosity and can prevent normal
The hot-water process is a successful commercial operation in the separation cell.
process that has been applied to bitumen recovery The separation cell acts like two settlers – one
from mined oil sand deposits in North America on top of the other – and in the lower settler, the
[28–32]. The process utilizes: sand settles down, whereas in the upper settler, the
bitumen floats. The bulk of the sand in the feed is
(i) The film of water that coats most of the removed from the bottom of the separation cell as
mineral matter, which permits extraction by tailings. A large portion of the feed bitumen floats
the hot-water process to the surface of the separation cell and is removed
(ii) The linear and the nonlinear variation of as bituminous froth. A middling stream consists
bitumen density and water density, respec- mostly of water with some suspended fine min-
tively, with temperature so that the bitumen erals and bitumen particles, and a portion of the
that is heavier than water at room tempera- middlings may be returned for mixing with the
ture becomes lighter than water at 80
C conditioning drum effluent to dilute the
(180
F) separation-cell feed for pumping. The remainder
(iii) Natural surface-active materials (surfactants) of the middlings is withdrawn from the separation
in the bitumen which also contribute to free- cell to be rejected after processing in the scaven-
ing the bitumen from the sand ger cells.
The combined froth from the separation cell
In the process, the oil sand is introduced into a and scavenging operation contains an average of
conditioning drum where the sand is heated and about 10% w/w mineral material and up to 40%
mixed with water to encourage agglomeration of w/w water. The dewatering and demineralizing
the oil particles. Conditioning is carried out in a are accomplished in two stages of centrifuging;
slowly rotating drum that contains a steam- in the first stage, the coarser mineral material is
sparging system for temperature control as well removed but much of the water remains. The feed
as mixing devices to assist in lump size reduction then passes through a filter to remove any addi-
and a size ejector at the outlet end. The oil sand tional large-size mineral matter that would plug up
lumps are reduced in size by ablation and mixing the nozzles of the second-stage centrifuges. In the
action. The conditioned pulp has the following scavenging cell, froth flotation with air is usually
characteristics: (i) solids 60–85% w/w and employed to recover more bitumen. The scaven-
(ii) 7.5–8.5 pH. ger froth is combined with the separation-cell
Lumps of as-mined oil sand are reduced in size froth to be further treated and upgraded to syn-
by ablation, and the conditioned pulp is screened thetic crude oil. Tailings from the scavenger cell
through a double-layer vibrating screen. Water is join the separation-cell tailing stream and go to
then added to the screened material (to achieve waste.
more beneficial pumping conditions), and the In summary, the hot-water separation process
pulp enters the separation cell through a central involves extremely complicated surface chemis-
feed well and distributor. The bulk of the sand try with interfaces among various combinations of
settles in the cell and is removed from the bottom solids (including both silica sand and aluminosil-
as tailing, but the majority of the bitumen floats to icate minerals), water, bitumen, and air. The
36 Petroleum and Oil Sand Exploration and Production

control of pH is critical with the preferred range the oil sand to provide a feed suitable for a fluid
being 8.0–8.5, which is achievable by the use of coking process; ca 80% of sand is removed. Nom-
any of the monovalent bases. Polyvalent cations inal composition of the oil product is 58% by
must be excluded because they tend to flocculate weight (bitumen, 27% w/w mineral matter and
the clay minerals and thus raise the viscosity of the 15% w/w water.
middlings in the separation cell. The spherical agglomeration process resem-
bles the sand-reduction process. Water is added to
Other Processes oil sands and the mixture is ball milled. The bitu-
The issues arising from bitumen mining and bitu- men forms dense agglomerates of 75–87% w/w
men recovery may be alleviated somewhat by the bitumen, 12–25% w/w sand, and 1–5% w/w
development of process options that require con- water.
siderably less water in the sand/bitumen separa- The oleophilic sieve process [35, 36] offers the
tion step. Such an option would allow a more potential for reducing tailing pond size because of
gradual removal of the tailing ponds. a reduction in the water requirements. The process
A cold-water process for bitumen separation is based on the concept that when a mixture of an
from mined oil sand has also been recommended oil phase and an aqueous phase is passed through
[33, 34]. The process uses a combination of cold a sieve made from oleophilic materials, the aque-
water and solvent, and the first step usually ous phase and any hydrophilic solids pass through
involves disintegration of the oil sand charge the sieve, but the oil adheres to the sieve surface
that is mixed with water, diluent, and reagents. on contact. The sieve is in the form of a moving
The diluent may be a petroleum distillate fraction conveyor, the oil is captured in a recovery zone,
such as aromatic naphtha or kerosene and is added and the recovery efficiency is high.
in approximately a 1:1 weight ratio to the bitumen An anhydrous solvent extraction process for
in the feed. The pH is maintained at 9–9.5 by the bitumen recovery has been attempted and usually
addition of wetting agents and approximately involves the use of a low-boiling hydrocarbon.
0.77 kg of soda ash per ton of oil sand. The The process generally involves up to four steps.
effluent is mixed with more water, and in a raked In the mixer step, fresh oil sand is mixed with
classifier, the sand is settled from the bulk of the recycle solvent that contains some bitumen and
remaining mixture. The water and oil overflow the small amounts of water and mineral and the
classifier and are passed to thickeners where the solvent-to-bitumen weight ratio is adjusted to
oil is concentrated. Clay minerals in the oil sand approximately 0.5. The drain step consists of a
feed have a distinct effect on the process; it forms three-stage countercurrent wash. Settling and
emulsions that are hard to break and are wasted draining time is approximately 30 min for each
with the underflow from the thickeners. stage. After each extraction step, a bed of sand is
The sand-reduction process is a cold-water formed, and the extract is drained through the bed
process without solvent. In the first step, the oil until the interstitial pore volume of the bed is
sand feedstock is mixed with water at approxi- emptied. The last two steps of the process are
mately 20
C (68
F) in a screw conveyor in a ratio devoted to solvent recovery from the bitumen
of 0.75–3 ton per ton of oil sand (the lower range and from the solids, which holds the key to the
is preferred). The mixed pulp from the screw economic success of the process.
conveyor is discharged into a rotary drum screen, Another aboveground method of separating
which is submerged in a water-filled settling ves- bitumen from mined oil sand involves direct
sel. The bitumen forms agglomerates that are heating of the oil sand without previous separa-
retained by an 840-m (20-mesh) screen. These tion of the bitumen [37]. Thus, the bitumen is not
agglomerates settle and are withdrawn as oil prod- recovered as such but is an upgraded overhead
uct. The sand readily passes through the 840 m product. In the process, the sand is crushed and
(20-mesh) screen and is withdrawn as waste introduced into a vessel, where it is contacted with
stream. The process is called sand reduction either hot (spent) sand or with hot product gases
because its objective is the removal of sand from that furnish part of the heat required for cracking
Petroleum and Oil Sand Exploration and Production 37

and volatilization. The volatile products are 50,000–1,000,000 cP or more at formation temper-
passed out of the vessel and are separated into atures (approximately 0–10
C, 32–50
F depending
gases and (condensed) liquids. The coke that is upon the season). This offers a formidable (but not
formed as a result of the thermal decomposition of insurmountable) obstacle to bitumen recovery.
the bitumen remains on the sand, which is then In principle, the non-mining recovery of bitu-
transferred to a vessel for coke removal by burn- men from oil sand deposits is an enhanced recov-
ing in air. The hot flue gases can be used either to ery technique and requires the injection of a fluid
heat incoming oil sand or as refinery fuel. As into the formation through an injection wall. This
expected, processes of this type yield an upgraded leads to the in situ displacement of the bitumen
product but require various arrangements of pneu- from the sand followed by bitumen production at
matic and mechanical equipment for solid move- the surface through an egress well (production
ment around the refinery. well). Thus, in oil sand deposits, it is often desir-
In improved mining, directional (horizontal or able to initiate enhanced oil recovery (EOR) oper-
slant) wells are drilled into the reservoir from a ations as early as possible, which means
mine in an underlying formation to drain oil by considerably abbreviating conventional second-
pressured depletion and gravity drainage. In the ary recovery operations or bypassing them alto-
process of gravity drainage extraction of liquid gether. Thermal floods using steam and controlled
crude oil, the wells are completed so that only in situ combustion methods are also used. Ther-
the forces acting within the reservoir are used. mal methods of recovery reduce the viscosity of
A large number of closely spaced wells can be the crude oil by heat so that it flows more easily
drilled into a reservoir from an underlying tunnel into the production well.
more economically than the same number of wells The technologies applied to oil recovery
from the surface. In addition, only one pumping involve different concepts, some of which can
system is required in underground drainage, cause changes to the oil during production. Tech-
whereas at the surface each well must have a nologies such as alkaline flooding, microemulsion
pumping system. The objective of using a large (micellar/emulsion) flooding, polymer-
number of wells is to produce each well slowly so augmented waterflooding, and carbon dioxide
that the gas/oil and water/oil interfaces move miscible/immiscible flooding do not require or
toward each other efficiently. By maintaining the cause any change to the oil. The steaming tech-
reservoir pressures because of forces acting on the nologies may cause some steam distillation that
reservoir, it is then assured that the oil production can augment the process when the steam-distilled
is provided by the internal forces due to gravity material moves with the steam front and acts as a
(the buoyancy effect) and capillary effects. solvent for oil ahead of the steam front. Again,
Large vertical shafts sunk from the surface are there is no change to the oil although there may be
generally the means through which underground favorable compositional changes to the oil insofar
openings can be excavated. These shafts are one as lighter fractions are recovered and heavier
means of access to offer an outlet for removal of materials remain in the reservoir [38–41].
excavated rock, provide sufficient opening for Thermal recovery methods (Fig. 2) have found
equipment, provide ventilation, and allow the most use when heavy oil or bitumen has an
removal of oil and gas products during later pro- extremely high viscosity under reservoir condi-
duction. These requirements plus geological con- tions [2, 4]. For example, bitumen is highly vis-
ditions and oil reservoir dimensions determine the cous, with a viscosity ranging up to million
shaft size. It is expected that an access shaft will centipoises or more at the reservoir conditions.
range from 8 to 20 ft in diameter. Thus, the thermal recovery processes (such as
cyclic steam injection, steam flooding, and in
Non-mining Methods situ combustion) add heat to the reservoir to
Whereas conventional crude oils may have a viscos- reduce oil viscosity and/or to vaporize the oil. In
ity of several poise (at 40
C, 105
F), the oil sand both instances, the oil is made more mobile so that
bitumen has a viscosity on the order of it can be more effectively driven to producing
38 Petroleum and Oil Sand Exploration and Production

wells. In addition to adding heat, these processes Steam-Based Processes

provide a driving force (pressure) to move oil to Steam-based processes are the most advanced of
producing wells. all enhanced oil recovery methods in terms of field
In the modified in situ extraction processes, com- experience and thus have the least uncertainty in
binations of in situ and mining techniques are used estimating performance, provided that a good res-
to access the reservoir. A portion of the reservoir ervoir description is available. Steam processes
rock must be removed to enable application of the in are most often applied in reservoirs containing
situ extraction technology. The most common heavy oil, which is mobile at reservoir tempera-
method is to enter the reservoir through a large- ture. Commercial application of steam processes
diameter vertical shaft, excavate horizontal drifts has been underway since the early 1960s.
from the bottom of the shaft, and drill injection Steam drive injection (steam injection) has
and production wells horizontally from the drifts. been commercially applied since the early
Thermal extraction processes are then applied 1960s. The process occurs in two steps: (i) steam
through the wells. When the horizontal wells are stimulation of production wells, that is, direct
drilled at or near the base of the oil sand reservoir, steam stimulation, and (ii) steam drive by steam
the injected heat rises from the injection wells injection to increase production from other wells
through the reservoir, and drainage of produced such as by indirect steam stimulation. Steam drive
fluids to the production wells is assisted by gravity. requires sufficient effective permeability (with the
There are, however, several serious constraints immobile bitumen in place) to allow injection of
that are particularly important and relate to bulk the steam at rates sufficient to raise the reservoir
properties of the oil sand and the bitumen. While temperature to mobilize the bitumen and drive it
the Canadian deposits are unconsolidated sands to the production well.
with a porosity ranging up to about 45% v/v, Cyclic steam injection (also called huff and
other deposits may range from predominantly puff or steam soak) is the alternating injection of
low-porosity, low-permeability consolidated sand steam and production of oil with condensed steam
to, in a few instances, unconsolidated sands. In from the same well or wells. This process is pre-
addition, the bitumen properties are not conducive dominantly a vertical well process, with each well
to fluid flow under deposit conditions. Neverthe- alternately injecting steam and producing heavy
less, where the general nature of the deposits pro- oil and steam condensate. In practice steam is
hibits the application of a mining technique, a injected into the formation at greater than fractur-
non-mining method may be the only feasible bitu- ing pressure followed by a soak period after which
men recovery option. production is commenced. The heat injected
Another general constraint to bitumen recov- warms the heavy oil and lowers its viscosity.
ery by non-mining methods is the relatively low A heated zone is created through which the
injectivity of oil sand formations. Thus, it is usu- warmed heavy oil can flow back into the well.
ally necessary to inject displacement or recovery This is a well-developed process; the major limi-
finds at a pressure such that fracturing (parting) is tation is that less than 30% (usually less than 20%
achieved. Such a technique therefore changes the v/v) of the initial oil-in-place can be recovered.
reservoir profile and introduces a series of chan- In the fracture-assisted steam technology
nels through which fluids can flow from the injec- (FAST) process, steam is injected rapidly into an
tion well to the production well. On the other induced horizontal fracture near the bottom of the
hand, the technique may be disadvantageous inso- reservoir to preheat the reservoir.
far as the fracture occurs along the path of least
resistance, giving undesirable (i.e., inefficient) Combustion Processes
flow characteristics within the reservoir between In situ combustion (fireflood) is normally applied
the injection and production wells, leaving a large to reservoirs containing low-gravity oil but has
part of the reservoir relatively untouched by the been tested over perhaps the widest spectrum of
displacement or recovery fluids. conditions of any enhanced oil recovery process.
Petroleum and Oil Sand Exploration and Production 39

In the process, heat is generated within the reser- reservoir to be plugged by the mobilized fluids
voir by injecting air and burning part of the crude ahead of a forward combustion front. In this pro-
oil. This reduces the oil viscosity and partially cess, the vaporized and mobilized fluids move
vaporizes the oil in place, and the oil is driven through the heated portion of the reservoir behind
out of the reservoir by a combination of steam, hot the combustion front. The reverse combustion
water, and gas drive. Forward combustion partially cracks the bitumen, consumes a portion
involves movement of the hot front in the same of the bitumen as fuel, and deposits residual coke
direction as the injected air; reverse combustion on the sand grains. In the process, part of the
involves movement of the hot front opposite to the bitumen will be consumed as fuel and part will
direction of the injected air. be deposited on the sand grains as coke leaving
During the process, energy is generated in the 40–60% w/w recoverable. This coke deposition
formation by igniting bitumen in the formation and serves as a cementing material, reducing move-
sustaining it in a state of combustion or partial ment and production of sand.
combustion. The high temperatures generated The addition of water or steam to an in situ
decrease the viscosity of the oil and make it more combustion process can result in a significant
mobile. Some cracking of the bitumen also occurs, increase in the overall efficiency of that process.
and an upgraded product rather than bitumen itself Two major benefits may be derived. Heat transfer
is the fluid recovered from the production wells. in the reservoir is improved because the steam and
The relatively small portion of the oil that condensate have greater heat-carrying capacity
remains after the displacement mechanisms have than combustion gases and gaseous hydrocar-
acted becomes the fuel for the in situ combustion bons. Sweep efficiency may also be improved
process. Production is obtained from wells offset- because of the more favorable mobility ratio of
ting the injection locations. In some applications, steam-bitumen compared with gas-bitumen.
the efficiency of the total in situ combustion oper- The THAI (toe-to-heel air injection) is based
ation can be improved by alternating water and air on the geometry of horizontal wells that may solve
injection. The injected water tends to improve the the problems that have plagued conventional in
utilization of heat by transferring heat from the situ combustion. The well geometry enforces a
rock behind the combustion zone to the rock short flow path so that any instability issues asso-
immediately ahead of the combustion zone. ciated with conventional combustion are reduced
The forward combustion process involves or even eliminated [2, 4, 42–50].
movement of the hot front in the same direction In situ conversion, or underground refining, is
as the injected air, while reverse combustion a promising new technology to tap the extensive
involves movement of the hot front opposite to reservoirs of heavy oil and deposits of bitumen.
the direction of the injected air. In the process, the The new technology [44, 51] features the injection
hydrocarbon products released from the zone of of high-temperature, high-quality steam and hot
combustion move into a relatively cold portion of hydrogen into a formation containing heavy
the formation. Thus, there is a definite upper limit hydrocarbons to initiate conversion of the heavy
of the viscosity of the liquids that can be recovered hydrocarbons into lighter hydrocarbons. In effect,
by a forward combustion process. On the other the heavy hydrocarbons undergo partial under-
hand, since the air passes through the hot forma- ground refining that converts them into a synthetic
tion before reaching the combustion zone, burn- crude oil (or syncrude) [52, 53]. The higher boil-
ing is complete; the formation is left completely ing fraction of the synthetic crude oil is treated to
cleaned of hydrocarbons. provide the fuel and hydrogen required by the
On the other hand, the reverse combustion process, and the lighter portion is marketed as a
process is particularly applicable to reservoirs conventional crude oil.
with lower effective permeability (in contrast Thus, below ground, superheated steam and
with forward combustion). It is more effective hot hydrogen are injected into a heavy oil or
because the lower permeability would cause the bitumen formation, which simultaneously
40 Petroleum and Oil Sand Exploration and Production

produces the heavy oil or bitumen and converts it Long horizontal wells with several multilateral
in situ (i.e., within the formation) into syncrude. branches have been used widely in the develop-
Above ground, the heavier fraction of the syn- ment the heavy oils of Venezuela using aggregate
crude is separated and treated on-site to produce horizontal lengths as large as 10,000 m in oil of
the fuel and hydrogen required by the process, 1200–5000 cP viscosity.
while the lighter fraction is sent to a conventional Inert gas injection (IGI) is a technology for
refinery to be made into petroleum products [44, conventional oils in reservoirs where good verti-
51]. Furthermore, the potential advantages of an cal permeability exists or where it can be created
in situ process for bitumen and heavy oil include through propped hydraulic fracturing. It is gener-
(i) leaving the carbon-forming precursors in the ally viewed as a top-down process with nitrogen
ground, (ii) leaving the heavy metals in the or methane injection through vertical wells at the
ground, (iii) reducing sand handling, and top of the reservoirs, creating a gas-oil interface
(iv) bringing a partially upgraded product to the that is slowly displaced toward long horizontal
surface. The extent of the upgrading can, hope- production wells [2, 4]. As with all gravity drain-
fully, be managed by adjusting the exposure of the age processes, it is essential to balance the injec-
bitumen of heavy oil to the underground thermal tion and production volumes precisely so that the
effects [54–57]. system does not become pressure driven, but
In situ recovery processes (although less effi- remains in the gravity-dominated flow regime.
cient in terms of bitumen recovery relative to SAGD (steam-assisted gravity drainage), using
mining operations) may have the added benefit paired horizontal wells (Fig. 3), was developed
of leaving some of the more obnoxious constitu- first in Canada for deposits containing the immo-
ents (from the processing objective) in the bile bitumen [58]. In the process, low-pressure
ground. steam continuously injected through the upper
well creates a steam chamber along the walls of
Other Processes which the heated bitumen flows and is produced
Many innovative concepts in heavy oil production in the lower well. The heat reduces the oil viscos-
have been developed in the last 10 years [2, ity to values as low as 1–10 cP (depending on
4]. However, there are varying degrees of success temperature and initial conditions) and develops a
and all are dependent on the properties of the steam chamber that grows vertically and laterally.
deposit. Due to the variation in bitumen properties The steam and gases rise because of their low
and the geology of the different deposits, there is density, and the oil and condensed water are
no panacea for bitumen recovery that can be removed through the lower well. The gases pro-
applied on a worldwide basis. The geological duced during application of the SAGD process
properties and the properties of the bitumen must tend to be methane with some carbon dioxide
be treated on a site-specific basis. and traces of hydrogen sulfide.

Petroleum and Oil Sand

Exploration and
Production, Fig. 3 The
SAGD process

Steam Injection Well

Production Well

Reservoir Boundary
Petroleum and Oil Sand Exploration and Production 41

The SAGD process, as for all gravity-driven dilation and contraction so that oil and water flow
processes, is extremely stable because the process into and out of pores, leading to periodic fluid
zone grows only by gravity segregation, and there accelerations in the pore throats. As the porosity
are no pressure-driven instabilities such as channel- dilation wave moves through the porous medium
ing, coning, and fracturing. SAGD seems to be at a velocity of about 50–100 ft/s (40–80 m/s), the
relatively insensitive to shale streaks and similar small expansion and contraction of the pores with
horizontal barriers, even up to several meters thick the passage of each packet of wave energy help
(3–6 ft), that otherwise would restrict vertical flow unblock pore throats, increase the velocity of liquid
rates. The combined processes of gravity segrega- flow, overcome part of the effects of capillary block-
tion and shale thermal fracturing make SAGD so age, and reduce some of the negative effects of
efficient that recovery ratios of 60–70% are instability due to viscous fingering, coning, and
claimed. Nevertheless, the process is not univer- permeability streak channeling.
sally applicable to all reservoirs and deposits. Vapor-assisted petroleum extraction (VAPEX)
Cold heavy-oil production with sand (CHOPS) is a new process in which the physics of the
is also used as a production approach in unconsol- process are essentially the same as for SAGD
idated sandstones. The process results in the devel- and the configuration of wells is generally similar.
opment of high-permeability channels The process involves the injection of vaporized
(wormholes) in the adjacent low cohesive strength solvents such as ethane or propane to create a
sands, facilitating the flow of oil foam that is vapor chamber through which the oil flows due
caused by solution gas drive. Instead of blocking to gravity drainage [59–62]. The process can be
sand ingress by screens or gravel packs, sand is applied in paired horizontal wells, single horizon-
encouraged to enter the wellbore by aggressive tal wells, or a combination of vertical and hori-
perforation and swabbing strategies. Vertical or zontal wells. The key benefits are significantly
slightly inclined wells (vertical to 45
) are operated lower energy costs, potential for in situ upgrading
with rotary progressive cavity pumps (rather than and application to thin reservoirs, with bottom
reciprocating pumps), and old fields are converting water or reactive mineralogy.
to higher-capacity progressive cavity pumps, giv- Because of the slow diffusion of gases and
ing production boosts to old wells. Because mas- liquids into viscous oils, this approach, used
sive sand production creates a large disturbed zone, alone, perhaps will be suited only for less viscous
the reservoir may be positively affected for later oils although preliminary tests indicate that there
implementation of thermal processes. are micro-mechanisms that act so that the VAPEX
Typically, a well placed on CHOPS production dilution process is not diffusion rate limited, and
will initially produce a high percentage of sand, the process may be suitable for the highly viscous
greater than 20% by volume of liquids. However, oil sand bitumen [63, 64].
this generally drops after some weeks or months. Nevertheless, VAPEX can undoubtedly be
The huge volumes of sand are disposed of by used in conjunction with SAGD methods. As
slurry fracture injection or salt cavern placement with SAGD and IGI, a key factor is the generation
or by sand placement in a landfill in an environ- of a three-phase system with a continuous gas
mentally acceptable manner. Obviously, the pro- phase so that as much of the oil as possible can
duction of excessive amounts of sand is a cause be contacted by the gaseous phases, generating
for mechanical and environmental concern. the thin oil film drainage mechanism. As with IGI,
Pressure-pulsing technologies (PPT) involve a vertical permeability barriers are a problem and
radically new aspect of porous media mechanics must be overcome through hydraulic fracturing to
that was discovered and developed into a production create vertical permeable channels, or undercut by
enhancement method in the period 1997–2003. The the lateral growth of the chamber beyond the
mechanism by which PPT works is to generate a lateral extent of the limited barrier, or “baffle.”
porosity dilation wave (a fluid displacement wave As with any solvent process, the loss of solvents
similar to a tsunami); this generates pore-scale in geological formations (such as by adsorption
42 Petroleum and Oil Sand Exploration and Production

onto clay minerals and other minerals) drastically Future Directions

affects process economics and raises many serious
environmental issues. With the current energy problems, the motivation
Hybrid approaches that involve the simulta- for recovering as much as possible of the in-place
neous use of several technologies are evolving reserves is greater than ever. There is a potentially
and will see greater applications in the future. In uncomfortable and politically disastrous situation
addition to hybrid approaches, the new production where the gap between energy requirements and
technologies, along with older, pressure-driven available energy supplies is widening quickly.
technologies, will be used in successive phases Consequently, the search for new domestic
to extract more oil from reservoirs, even from supplies has shifted in large measure to increas-
reservoirs that have been abandoned after primary ingly hostile environments such as the Alaskan
exploitation. These hybrid approaches hold North Slope and the offshore waters along the
(on paper at least) the promise of significantly outer continental shelves of the United States
increasing recoverable reserves worldwide, not and to the tight formation that contain natural
just in heavy oil cases. gas and crude oil. These changes in production
Microbial enhanced oil recovery (MEOR) pro- operations have resulted in both significantly
cesses involve the use of reservoir microorgan- higher costs (not to mention the associated envi-
isms or specially selected natural bacterial to ronmental disasters that often accompany such
produce specific metabolic events that lead to venture) and fewer new commercial discoveries.
enhanced oil recovery. In terms of petroleum recovery, steam-based
In microbial enhanced oil recovery processes, processes will remain the processes of choice for
microbial technology is exploited in oil reservoirs the recovery of much of the oil in the ground over
to improve recovery [65–67]. From a microbiolo- the next two to three decades. In some instances, in
gist’s perspective, microbial enhanced oil recovery situ combustion (fireflooding) will be rejuvenated
processes are somewhat akin to in situ bioremedi- as the need to recover residual oil from depleted
ation processes. Injected nutrients, together with reservoirs, heavy oil, extra heavy oil, and oil sand
indigenous or added microbes, promote in situ bitumen becomes more important. With the pre-
microbial growth and/or generation of products ponderance of heavy oil and oil sand bitumen, it is
which mobilize additional oil and move it to pro- likely that efforts will be made to emphasize partial
ducing wells through reservoir repressurization, (or full) upgrading as an integral part of the recov-
interfacial tension/oil viscosity reduction, and ery process. Any type of upgrading during recov-
selective plugging of the most permeable zones ery will enhance the quality of the recovered oil
[68, 69]. This technology requires consideration leaving some of the undesirable constituents of the
of the physicochemical properties of the reservoir heavy oil or bitumen in the ground as thermal
in terms of salinity, pH, temperature, pressure, and products that are incompatible with the heavy oil
nutrient availability [70, 71]. or bitumen. Enhancement of the quality of the
Finally, recent developments in upgrading of recovered oil by partial upgrading during the
heavy oil and bitumen [2, 4] indicate that the near recovery operation will facilitate upgrading of the
future could see a reduction of the differential cost recovered product in the refinery.
of upgrading heavy oil. These processes are based Biotechnology will play a more significant role
on a better understanding of the issues of in enhancing crude oil recovery and refining.
asphaltene solubility effects at high temperatures Enhanced oil recovery processes (microbial
as well as the adverse effects on feedstock char- enhanced oil recovery (MEOR)) involve stimulat-
acter, which is especially true for the heavy oil and ing indigenous reservoir microbes or injecting
bitumen feedstocks (Table 5), incorporation of a specially selected consortia of natural bacteria
catalyst that is chemically precipitated internally into the reservoir to produce specific metabolic
during the upgrading, and improving hydrogen events that lead to improved oil recovery and
addition or carbon rejection. refining operations such as desulfurization
Petroleum and Oil Sand Exploration and Production 43

[14]. However, like all recovery and refining pro- 11. CFR 1.43-2 (2004) Internal Revenue Service, Depart-
cesses, a complete evaluation and assessment of ment of the Treasury, Government of the United
States, Washington, DC
microbial from a scientific and engineering stand- 12. Reed RL, Healy RN (1977) In: Shah DO, Schechter
point must be performed and must be based on RS (eds) Improved oil recovery by surfactant and
economics, applicability, and the performance polymer flooding. Academic, New York
standards required to further improve the process 13. OTA (1978) Enhanced oil recovery potential in the
United States. Office of Technology Assessment,
efficiency. Washington, DC. January. NTIS order #PB-276594
Moreover, with respect to refining operations, 14. Speight JG (2000) Desulfurization of heavy oils and
new processes for the conversion of residua and residua, 2nd edn. Marcel Dekker, New York
oil sand bitumen will probably be used perhaps in 15. Terwilliger PL (1975) Paper 5568. In: Proceedings of
50th annual fall meeting. Society of Petroleum Engi-
conjunction with (and not in place of) visbreaking neers, American Institute of Mechanical Engineers
and coking options with some degree of 16. Terwilliger PL, Clay RR, Wilson LA, Gonzalez-Gerth
hydroprocessing or visbreaking as a primary con- E (1975) J Pet Technol 27:9
version step [72–74]. In addition, other processes 17. Burger J (1978) In: Chilingarian GV, Yen TF (eds)
Developments in petroleum science, No. 7, bitumens,
may replace or, more likely, augment the asphalts and tar sands. Elsevier, New York, p 191
deasphalting units in many refineries. An excep- 18. Fisher K (2012) Trends take fracturing back to the
tion, which may become the rule, is the upgrading future: American oil and gas. Reporter 55(8):86–97
of bitumen from oil sands. 19. Scanlon BR, Reedy RC, Nicot JP (2014) Comparison
of water use for hydraulic fracturing for oil and gas
Above all, and with special relevance to this versus conventional oil. Environ Sci Technol
section, there is the necessity of managing recov- 48:12386–12393
ery operations in such a manner that environmen- 20. US EIA (2014) United States Energy Information
tal issues are given serious consideration and Administration, May 7, 2014, Annual Energy Outlook
2014. Energy Information Administration, US Depart-
protection of the environment is assured. ment of Energy, Washington, DC. http://www.eia.gov/
forecasts/aeo/index.cfm. Accessed 21 Mar 2015
21. Veatch RW Jr (1983) Overview of current hydraulic
fracturing design and treatment technology – part 1. J
Bibliography Pet Technol 35(4):677–687. Paper SPE-10039-PA,
Society of Petroleum Engineers, Richardson
1. Taber JJ, Martin FD (1983) Technical screening 22. Reinicke A, Rybacki E, Stanchits S, Huenges E, Dresen
guides for the enhanced recovery of oil. SPE 12069. G (2010) Hydraulic fracturing stimulation techniques
In: Proceedings of 58th SPE annual technical confer- and formation damage mechanisms: implications from
ence and exhibition, San Francisco, October 5–8 laboratory testing of tight sandstone–proppant systems.
2. Speight JG (2016) Introduction to enhanced recovery Chem Erde 70(S3):107–117
methods for heavy oil and tar sands, 2nd edn. Gulf 23. Holditch SA, Jennings JW, Neuse SH (1978) The opti-
Professional Publishing, Elsevier, Oxford, UK mization of well spacing and fracture length in low
3. Drew LJ (1997) Undiscovered mineral and petroleum permeability gas reservoirs. Paper No. SPE-7496. In:
deposits: assessment & controversy. Plenum Publish- Proceedings of SPE annual fall technical conference
ing Corporation, New York. Chapter 3 and exhibition, Houston, October 1–3. Society of
4. Speight JG (2014) The chemistry and technology of Petroleum Engineers, Richardson
petroleum, 5th edn. CRC Press, Boca Raton 24. Valko PP, Oligney RE, Economides MJ (1998) High
5. Lake LW, Walsh MP (2004) Primary hydrocarbon permeability fracturing of gas wells. Pet Eng Int
recovery. Elsevier, Amsterdam 71(1):75–88
6. Craft BC, Hawkins MF (1959) Applied petroleum res- 25. Delbianco A, Montanari R (2009) Encyclopedia of
ervoir engineering. Prentice-Hall, Englewood Cliffs hydrocarbons, volume III/new developments: energy,
7. Frick TC (1962) Petroleum production handbook, transport, sustainability. Eni S.p.A, Rome
vol II. McGraw-Hill, New York 26. Shalaby HM (2005) Refining of Kuwait’s heavy crude
8. Lake LW (1989) Enhanced oil recovery. Prentice-Hall, oil: materials challenges. In: Proceedings of workshop
Englewood Cliffs on corrosion and protection of metals. Arab School for
9. Arnarnath A (1999) Enhanced oil recovery scoping Science and Technology. December 3–7, Kuwait
study. Report No. TR-113836. Electric Power 27. Ohmes R (2014) Characterizing and tracking contam-
Research Institute, Palo Alto inants in opportunity crudes. Digital refining. http://
10. Borchardt JK, Yen TF (1989) Oil field chemistry, www.digitalrefining.com/article/1000893,Charaterisi
Symposium series no. 396. American Chemical Soci- ng_and_tracking_contaminants_in_opportunity_cru
ety, Washington, DC des_.html#.VJhFjV4AA. Accessed 1 Nov 2014
44 Petroleum and Oil Sand Exploration and Production

28. Clark KA (1944) Trans Can Inst Min Met 47:257 48. Jiang S, Liu X, Liu Y, Zhong L (2005) In situ
29. Carrigy MA (1963a) Bulletin No. 14. Alberta upgrading heavy oil by aquathermolytic treatment
Research Council, Edmonton under steam injection conditions. Paper No. SPE
30. Carrigy MA (1963b) The oil sands of Alberta. Infor- 91973. In: Proceedings of international symposium
mation Series No. 45. Alberta Research Council, on oilfield chemistry, Woodlands, February 204. Soci-
Edmonton ety of Petroleum Engineers, Richardson
31. Fear JVD, Innes ED (1967) In: Proceedings of 7th 49. Xia TX, Greaves M (2006) In situ upgrading of Ath-
World Petroleum Congress, vol 3, p 549 abasca tar sand bitumen using THAI. Chem Eng Res
32. Speight JG, Moschopedis SE (1978) Fuel Process Des 84(9):856–864
Technol 1:261 50. Greaves M, Xia TX (2008) Stability of THAI
33. Miller JC, Misra M (1982) Fuel Process Technol 6:27 process – theoretical and experimental observations.
34. Misra M, Aguilar R, Miller JD (1981) Sep Sci Technol J Can Pet Technol 47:65
16(10):1523 51. Gregoli AA, Rimmer DP (2000) Production of syn-
35. Kruyer J (1982) In: Proceedings of second interna- thetic crude oil from heavy hydrocarbons recovered by
tional conference on heavy crude and tar sands, Cara- in situ hydrovisbreaking. US Patent 6,016,868,
cas, February 7–17 January 25
36. Kruyer J (1983) Preprint No. 3d. Summer National 52. Motaghi M, Saxena P, Ravi R (2010) Partial upgrading
Meeting, American Institute of Chemical Engineers, of heavy oil reserves. Pet Technol Q Q4:55–64
Denver, August 28–31 53. Shah A, Fishwick R, Wood J, Leeke G, Rigby S,
37. Gishler PE (1949) Can J Res 27:104 Greaves M (2010) A review of novel techniques for
38. Richardson WC, Fontaine MF, Haynes S (1992) Com- heavy oil and bitumen extraction and upgrading.
positional changes in distilled, steam-distilled, and Energy Environ Sci 3:700–714
steam-flooded crude oils. Paper No. SPE 24033. West- 54. Reichert C, Fuhr B, Sawatzky H, Lafleur R,
ern Regional Meeting, Bakersfield, March 30–April 1 Verkoczy B, Soveran D, Jha K (1989) Fire flood
39. Sharpe HN, Richardson WC, Lolley CS (1995) Repre- recovery process effects upon heavy oil properties.
sentation of steam distillation and in situ upgrading Pet Sci Technol 7:851–878
processes in heavy oil simulation. Paper No. SPE 55. Graue DJ (2001) Upgrading and recovering heavy
30301. In: Proceedings of international heavy oil sym- crude oils and natural bitumens by in situ hydrovis-
posium, Calgary, June 19–21 breaking. US Patent 6,328,104, December 11
40. Weissman JG, Kessler RV (1996) Downhole heavy 56. Fan H, Liu Y (2002) Downhole catalysts upgrades
crude oil hydroprocessing. Appl Catal A Gen heavy oil. Oil Gas J 100(11):60–62
140(1):16 57. Greaves M, Xia TX (2004) Downhole catalytic pro-
41. Weissman JG (1997) Review of processes for down- cess for upgrading heavy oil: produced oil properties
hole catalytic upgrading of heavy crude oil. Fuel Pro- and composition. J Can Pet Technol 43(9):25–30
cess Technol 50(2–3):199 58. Dusseault MB, Geilikman MB, Spanos TJT
42. Zou J, Gray MR, Thiel J (1989) Structural changes in (1998) J Pet Technol 50(9):92–94
cold lake bitumen due to in situ combustion. AOSTRA 59. Butler RM, Mokrys IJ (1991) J Can Pet Technol
J Res 5:75 30(1):97–106
43. Mokrys IJ, Butler RM (1993) In situ upgrading of 60. Butler RM, Mokrys IJ (1995) Process and apparatus
heavy oils and bitumen by propane deasphalting: the for the recovery of hydrocarbons from a hydrocarbon
Vapex process. Paper No. SPE 25452. In: Proceedings deposit. US Patent 5,407,009, April 18
of production operations symposium, Oklahoma City, 61. Butler RM, Mokrys IJ (1995) Process and apparatus
March 21–13. Society of Petroleum Engineers, for the recovery of hydrocarbons from a hydrocarbon
Richardson deposit. US Patent 5,607,016, March 4
44. Gregoli AA, Rimmer DP, Graue DJ (2000) Upgrading 62. Butler RM, Jiang Q (2000) J Can Pet Technol
and recovery of heavy crude oils and natural bitumen 39:48–56
by in situ hydrovisbreaking. US Patent 6,016,867, 63. Yang C, Gu Y (2005a) A novel experimental tech-
January 25 nique for studying solvent mass transfer and oil swell-
45. Kuhlman M (2000) The benefits of in situ upgrading ing effect in a Vapor Extraction (VAPEX) process.
reactions to the integrated operations of the Orinoco Paper No. 2005–099. In: Proceedings of 56th annual
Heavy-Oil Fields and Downstream Facilities. Paper technical meeting. The Canadian International Petro-
No. SPE 62560. In: Proceedings of SPE/AAPG west- leum Conference, Calgary, June 7–9
ern regional meeting, Long Beach, June 19–22. Soci- 64. Yang C, Gu Y (2005b). Effects of solvent-heavy oil
ety of Petroleum Engineers, Richardson interfacial tension on gravity drainage in the VAPEX
46. Xia TX, Greaves M, Werfilli WS, Rathbone RR process. Paper No. SPE 97906. In: Society of petro-
(2002) Downhole conversion of lloydminster heavy leum engineers international thermal operations and
oil using THAI-CAPRI processes. SPE 78998 heavy oil symposium, Calgary, November 1–3
47. Greaves M, Xia TX, Ayasse C (2005) Underground 65. Banat IM (1995) Biosurfactant production and possi-
upgrading of heavy oil using THAI – toe-to-heel air ble uses in microbial enhanced oil recovery and oil
injection, SPE 97728 pollution remediation. Bioresour Technol 51:1–12
Petroleum and Oil Sand Exploration and Production 45

66. Clark JB, Munnecke DM, Jenneman GE (1981) In-situ 79. Devold H (2013) Oil and gas production handbook: an
microbial enhancement of oil production. Dev Ind introduction to oil and gas production, transport, refin-
Microbiol 15:695–701 ing and petrochemical industry, 3rd edn. ABB Oil Gas,
67. Stosur GJ (1991) Unconventional EOR concepts. Crit Oslo
Rep Appl Chem 33:341–373 80. El-Gendy NS, Speight JG (2015) Handbook of refin-
68. Bryant RS, Lindsey RP (1996) World-wide applica- ery desulfurization. CRC Press, Boca Raton
tions of microbial technology for improving oil recov- 81. Gudmestad OT, Zolotukhin AB, Jarlsby ET
ery, pp 27–134. In: Proceedings of the SPE (2010) Petroleum resources, with emphasis on off-
symposium on improved oil recovery. Society of shore fields. WIT Press, Billerica
Petroleum Engineers, Richardson 82. Holloway MD, Rudd O (2013) Fracking: the opera-
69. Bryant RS, Donaldson EC, Yen TF, Chilingarian GV tions and environmental consequences of hydraulic
(1989) Microbial enhanced oil recovery. In: fracturing. Scrivener Publishing, Beverly
Donaldson EC, Chilingarian GV, Yen TF (eds) 83. Nersesian RL (2010) Energy for the 21st century: a
Enhanced oil recovery II: processes and operations. comprehensive guide to conventional and alternative
Elsevier, Amsterdam, pp 423–450 sources, 2nd edn. M.E. Sharpe, Armonk
70. Khire JM, Khan MI (1994a) Microbially enhanced oil 84. Speight JG (2008) Synthetic fuels handbook: proper-
recovery (MEOR). Part 1. Importance and mecha- ties, processes, and performance. McGraw-Hill,
nisms of microbial enhanced oil recovery. Enzym New York
Microb Technol 16:170–172 85. Speight JG (2011) The refinery of the future. Gulf
71. Khire JM, Khan MI (1994b) Microbially enhanced oil Professional Publishing, Elsevier, Oxford, UK
recovery (MEOR). Part 2. Microbes and the subsur- 86. Speight JG, Arjoon KK (2012) Bioremediation of
face environment for microbial enhanced oil recovery. petroleum and petroleum products. Scrivener Publish-
Enzym Microb Technol 16:258–259 ing, Beverly
72. Dickenson RL, Biasca FE, Schulman BL, Johnson HE 87. Speight JG (2013) Oil sand production processes. Gulf
(1997) Refiner options for converting and utilizing Professional Publishing, Elsevier, Oxford, UK
heavy fuel oil. Hydrocarb Process 76(2):57–62 88. Speight JG (2013) Heavy oil production processes.
73. Radovanović L, Speight JG (2011) Visbreaking: a Gulf Professional Publishing, Elsevier, Oxford, UK
technology of the future. In: Proceedings of first inter- 89. Speight JG (2013) Shale gas production processes.
national conference – process technology and environ- Gulf Professional Publishing, Elsevier, Oxford, UK
mental protection (PTEP 2011). University of Novi 90. Speight JG (2014) The chemistry and technology of
Sad, Technical Faculty “Mihajlo Pupin,” Zrenjanin, petroleum, 5th edn. CRC Press, Boca Raton
Republic of Serbia, December 7, 2011. pp 335–338 91. Speight JG (2015) Handbook of offshore oil and gas
74. Speight JG (2012) Visbreaking: a technology of the operations. Gulf Professional Publishing, Elsevier,
past and the future. Sci Iran 19(3):569–573 Oxford, UK
92. Speight JG (2016) Handbook of hydraulic fracturing.
Wiley, Hoboken
Other Information Sources
93. Speight JG, Islam MR (2016) Peak energy – myth or
75. Ancheyta J, Speight JG (2007) Hydroprocessing of
reality. Scrivener Publishing, Beverly
heavy oils and residua. CRC-Taylor & Francis
94. Speight JG (2017) Deep shale oil and gas. Gulf Pro-
Group, Boca Raton
fessional Publishing, Elsevier, Oxford, UK
76. Bower T (2009) Oil: money, politics, and power in the
95. Wihbet PM (2009) The rise of the new oil order.
21st century. Grand Central Publishing, Hachette
Academy & Finance, Geneva
Book Group, New York
96. Yergin D (2011) The quest: energy security and the
77. Chaudhuri UR (2010) Fundamentals of petroleum &
remaking of the modern world. Penguin Press,
petrochemical engineering. CRC Press, Boca Raton
New York
78. Crane HD, Kinderman EM, Malhotra R (2010)
97. Wang X, Economides M (2009) Advanced natural gas
A cubic mile of oil: realities and options or averting
engineering. Gulf Publishing Company, Houston
the looming global energy Crisic. Oxford University
Press, Oxford, UK
 Kerogen Macromolecular and organic solvent-
Oil Shale Processing, insoluble organic matter of oil shale, but more
Chemistry, and Technology broadly applied to these sorts of organic com-
ponents of sedimentary rocks
Vahur Oja1 and Eric M. Suuberg2 Oil shale Oil shale has a wide range of defini-
1
Department of Chemical Engineering, Tallinn tions. A general definition that encompasses
University of Technology, Tallinn, Estonia most is that it consists of a sedimentary rock
2
Division of Engineering, Brown University, containing various amounts of solid organic
Providence, RI, USA material bound dispersedly in a mineral matrix
and yielding principally a shale oil upon
pyrolysis
Article Outline Pyrolysis A thermochemical decomposition pro-
cess of organic matter, affected by heating the
Glossary material in the absence of oxygen
Definition of the Subject Retort gas Noncondensable combustible gas, a
Introduction mixture of gases from an oil shale retort that
General Characteristics of Oil Shale remains after particulate matter and condensed
Oil Shale Conversion Processes: An Overview liquids have been removed
Shale Oil Production Retort water Water collected during oil shale
Technological Aspects of Shale Oil Recovery retorting from off-gas stream together with oil
Final Considerations and Future Directions Retorting residue or spent shale Solid waste
Bibliography material remaining after retorting. This mate-
rial contains most of the mineral matter of the
Glossary original shale (possibly partially decomposed)
and a carbon-rich (char) residue from the
Bitumen Organic solvent-soluble portion of oil organic fraction (which may sometimes be
shale organic matter referred to as a semicoke)
Direct retorting Retorting in which the process Retorting In shale oil recovery technologies,
heat requirement is accomplished by a com- retorting is a pyrolytic process designed to
bustion of a portion of the oil shale within the produce oil. Generally, this process is
retort performed by placing oil shale in a sealed
Ex situ retorting Oil shale retorting performed vessel heated in order to promote pyrolysis;
aboveground (requiring mining of the oil this reactor is called a retort
shale) Shale oil Room temperature liquid phase organic
In situ retorting Heating of oil shale under- material, oil product, recoverable from the
ground in order to release shale oil (also called thermal decomposition of kerogen or separated
underground retorting). Confinement of the from the off-gas stream of an oil shale retort.
shale is provided by the surrounding earth, in Shale oil is not the same as tight oil, which is
this case conventional oil contained in a shale or sand-
Indirect retorting Retorting in which the pro- stone formation of low or very low permeabil-
cess heat requirement is provided by heated ity. This tight oil is also sometimes called shale
gaseous or solid heat carriers produced outside oil, and this has led to the confusion
the retort and then brought into contact with the Thermobitumen A bitumen not contained in the
retort contents original oil shale but produced by the action of

© Springer Science+Business Media, LLC, part of Springer Nature 2020 47

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_102
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_102-3
48 Oil Shale Processing, Chemistry, and Technology

heat on the shale, a thermal degradation good scientific sense, but in other cases, it is
(pyrolysis) product of kerogen important to note that there might be inadvertent
Tight oil Petroleum that is trapped in petroleum- changes induced in the material, and in its poten-
bearing rocks of lower permeability than in tial conversion chemistry, by the separation. This
traditional petroleum reservoirs. To recover summary will only provide a brief overview of
this oil, advanced drilling technologies some of the aspects of research performed on the
(hydraulic fracturing) must be used different phases present in the oil shale and offer
only this general warning to be cautious regard-
ing the possibility of artifacts induced by separa-
Definition of the Subject tions of the raw material.
The main focus of this entry is on the chemistry
Oil shale is a complex mixture of organic and of oil shale conversion. Virtually all oil shale
inorganic phases, in which the organic phase is conversion processes begin with a thermal degra-
normally that of the greater economic and prac- dation of the organic macromolecular structure.
tical value. In nature there exists a wide range of The organic fraction of oil shale is largely
materials that include these same components, unrecoverable in any useful form without such a
and not all of these materials are candidates for thermal degradation, since the extractable bitu-
actual use; in some instances, the organic content men content in unheated oil shale is normally
is simply too low to make processing of the oil but a few percent at most. The thermal degrada-
shale of any economic interest. Thus, one of the tion is critical whether the oil shale is processed to
first things that needs to be answered in asking recover oil or burned directly to generate heat and,
what oil shales it makes sense to study is the from that, electricity. In the case of combustion,
question of what sorts of oil shale should be the thermal degradation is what releases the small
considered as the subject of practical use. This organic molecular fragments that can burn in the
summary begins with a consideration of this fun- vapor phase. The less that remains in a solid
damental question and is guided by what has semicoke phase after the initial heating in a boiler,
historically been accepted as definitions of oil the more quickly and smoothly the combustion of
shales of practical interest. It is recognized that the shale can take place and the more efficiently
these definitions come from different locations used will be the organic fraction for its heating
and time periods, with differing demands for the value content. Most shale oil combustion pro-
energy or raw materials that the oil shale could cesses do operate at very high degrees of burnout
provide. This, in turn, means that these defini- of the organic matter.
tions can change with time and place, but here In the case of oil recovery from the oil shale,
they offer a framework for the discussion that there is a much-studied set of conditions used to
follows. maximize the recovery of oil. There is a balance
Because in oil shale the organic and inorganic between the need to achieve sufficient break-
phases form such an intimate mixture, it is criti- down of the macromolecular structure of the
cal to consider how this might influence the prac- organic phase, which will permit fragments to
tical conversion processes to which the oil shale escape as oil, and the further pyrolysis of the
is ultimately subjected. At the same time, it is organic phase to release light gases and leave
essential to recognize how the experimental behind a hydrogen-depleted semicoke that has
research used to develop these practical pro- much lower value than oil (and that is embedded
cesses can be influenced by decisions on how to in the remaining mineral matrix, making it diffi-
prepare the oil shale for testing. The isolation of cult to utilize). This entry summarizes several
the kerogen material from the oil shale offers aspects of what is known regarding the chemistry
advantages in terms of simplifying or enabling of thermal conversion and its kinetics and offers
certain types of organic fraction analysis, or the an overview of some practical technologies in
evaluation of proposed processing steps in the which the thermal conversion of the material
laboratory. In many instances, this might make takes place.
Oil Shale Processing, Chemistry, and Technology 49

Introduction material dispersedly bound in a mineral matrix.

Oil shales from different deposits vary in their
This entry focuses on the chemical nature of oil mineral contents and types, chemical composition
shale, and some key aspects of its conversion to oil. of organic matter, geological period of deposition,
One of the first difficulties encountered in this field is depositional condition, etc. [1, 2]. Therefore, there
deciding what starting materials qualify as “oil are a number of definitions of oil shale based upon
shale.” A brief review of various practical definitions application, operational, or scientific point of
is first offered. Different practical definitions arise view [3]. Similarly there are various oil shale
from the different applications of the material. Next, classification schemes [3–6]. Moreover, the term
the chemical structure of the shale is considered, “oil shale” is itself somewhat misleading as the
including both the inorganic and organic phases organic matter is not oil-like, and the mineral
contained therein. In this section, there is also a matter is not always classifiable as shale [7].
consideration of how the organic phase can be ana- The following are some selected definitions of
lyzed, including how it may be separated from the oil shale, proposed by several authors in the con-
whole shale. nection with its energy production or other tech-
In section “Definition of the Subject,” the chem- nological utilization:
ical processes that result in the practical conversion
processes of oil shale are introduced in general terms. 1. Gavin [8] defined oil shale as a “compact lam-
In this section it is noted that not all oil shale is inated rock of sedimentary origin, yielding
necessarily used to produce an oil product. In some over 33% of ash and containing organic matter
instances, the oil shale may be directly burned in that yields oil when distilled, but not apprecia-
devices similar to those used for combustion of coal. bly when extracted with ordinary solvents for
In section “Shale Oil Production,” attention is turned petroleum.”
toward retorting to produce oil products. Except for 2. Ozerov and Polozov have described oil shale
recent decades in Estonia, where direct firing of oil as “a hard foliated combustible rock formed by
shale for producing electrical power and heat has joint accumulation of pelagic plants and ani-
been practiced, retorting for oil has been the main mals and mineral mass which had been trans-
driver for worldwide interest in oil shale. This section formed by the action of geographic conditions
reviews the parameters that influence the results of and chemical, biochemical and hydro-
retorting, as well as providing an overview of some chemical processes” [9].
chemical aspects of the process. This review closes 3. Schlatter has characterized oil shale as “a het-
in section “Technological Aspects of Shale Oil erogeneous mixture of organic and mineral
Recovery” with a brief consideration of different matter. It is a fine-grained, tight rock with
classes of technology developed for carrying out essentially no permeability or porosity” [10].
the retorting process.
It is important to recognize that there is a much It is important to note that in defining oil shale
more voluminous literature on oil shale chemistry for energy technological purposes, the following
and conversion than has been discussed here. The oil shale characteristics have been emphasized:
authors maintain that the present selection does,
however, provide a reasonably comprehensive • A high mineral matter content, which may be
overview of the topic, while emphasizing some about two to five times higher than the organic
of the newer contributions to the literature. matter content (generally significantly higher
in comparison to coals, which contain mineral
matter, by definition, of less than 40%).
General Characteristics of Oil Shale • The major portion of oil shale organic matter is
insoluble in organic solvents (this in contrast to
Oil Shale Classification tar sands).
Oil shale can be defined as a sedimentary rock • The organic matter has hydrogen/carbon
containing various amounts of solid organic atomic ratios typically in the range 1.2–1.6
50 Oil Shale Processing, Chemistry, and Technology

(which is significantly enriched in hydrogen generation. The heating value of typical oil shale
compared with coals, but hydrogen deficient kerogen is about 40 MJ/kg. For example, the
relative to crude oil). heating value of an Estonian kukersite oil shale
• The capacity of oil shales to ignite and burn kerogen has been reported to be 37.3 MJ/kg
without separation of the organic matter from [11], while for US Green River Formation
the mineral matter. kerogen, the value is around 41.1 MJ/kg and
• They undergo thermal decomposition that for eastern US Devonian kerogen around
results in production of a significant amount 37.5 MJ/kg [14]. This is, of course, not the
of liquid organic product or shale oil (higher- same as the heating value of the oil shale as
grade oil shales can yield 100 L or more per ton would be determined for the whole rock, using
of dry oil shale). a standard calorimeter (in which the dilution
• They have a moisture content of less than effect of the zero, or negative, heating value
10–13%. However, depending on the oil mineral matter is important). A combustion
shale, the moisture content can be up to 40%. grade oil shale is defined as that having a
Oil shale with a high moisture content must be minimum upper caloric value (for the whole
dried before processing, and this can require a oil shale rock) of 3.1 MJ/kg (dry basis) [12],
large amount of energy relative to the energy though a heating value limit of not less than
content of the oil shale. about 4.2–6.3 MJ/kg has been suggested for
• They exhibit a low permeability of the rock to practical purposes [7, 15].
gases, vapors, and oils. 2. Average oil yield, or oil production potential, is
determined using a “modified” Fischer assay
The organic matter (often termed kerogen, see method (the most widely used standardized
below) content of oil shale varies widely, ranging technique, ASTM D-3904-80) or an equivalent
typically from 10% to 40%. As the organic matter analytical technique in which the shale is sub-
of oil shale is the source of energy and/or synthetic jected to heating in order to liberate oil from the
fuels value, there have been several attempts to kerogen. The oil yield depends mostly on kero-
classify oil shale by minimum and maximum gen per mass of oil shale but may also depend
organic matter content. The upper limit of organic upon the fraction of kerogen convertible to oil
matter content has been proposed to be about 50% (there are differences between different kero-
[6, 11], or even higher [7]. The lower organic gens). On the basis of this assay, an “oil shale”
matter limit is usually defined in relation to the is a shale that yields at least 42 L/t of dry shale
minimum energy requirement of oil shale as an (or 10 US gal/t, the basis used by the US Geo-
industrial raw material. For example, a lower limit logical Survey) [14], though a limit as low as
of 5% of organic matter has been proposed to 25 L/t has also been suggested. Similarly, a
define “commercial deposits.” This is because crude dividing line between lower- and higher-
the organic matter content of an oil shale should grade shale has been defined at about 100 L of
be at least roughly about 2.5 wt% for the latter to oil/t of shale (with those yielding below 90 L/t
provide the calorific requirements necessary to considered “low grade,” 90–150 L/t “moderate
heat the rock to 500
C [12, 13]. grade,” and those above 150 L/t “high grade”
Oil shale grade or energy potential is currently [11, 12]). Commercial grade oil shales range
classified or characterized for commercial pur- from about 100 to 200 L/t of shale. However,
poses on either a heating value or an average oil the Fischer assay does not actually indicate the
yield basis. maximum oil yield that can be actually pro-
duced by a given oil shale. Yields greater than
1. The heating value is used to evaluate the qual- Fischer assay have been obtained by
ity of an oil shale in the context of direct firing hydropyrolysis, donor solvents, and rapid
of oil shale in a power plant to produce heat, heating, especially for lower oil yield, non-
commonly used to raise steam for electricity softening oil shales [16, 17].
Oil Shale Processing, Chemistry, and Technology 51

More than 600 oil shale deposits have been dolomite), silicates, sulfides, and others. The inor-
discovered around the world [11]. Oil shales are ganic composition of oil shales varies widely. The
found in many countries around the world, with oil shale mineral composition can change from
nearly 100 major deposits in 27 countries [18]. mostly calcareous to mostly aluminosilicate. For
There exist many surveys/reviews on history of oil example, a typical mineral composition (in weight
shale development throughout the world [18–20]. percent of the mineral matter alone) of carbonate-
Oil shale resources are typically expressed rich kukersite oil shale from Estonia [22] is dolo-
in either of two different ways: (1) as tons of mite, 3%; calcite, 65%; quartz, 8%; illite, 15%;
oil shale and (2) as crude shale oil equivalents orthoclase, 5%; chlorite, 1%; and pyrite, 3%. For
contained in the oil shale. Global deposits have the carbonate-rich Mahogany-zone US Green
been estimated to be about 411 metric gigatons or to River Formation oil shale [14], the composition
range from 2.8 to 3.3 trillion barrels (4.5  1011 to is somewhat different: dolomite, 32%; calcite,
5.2  1011 m3) of recoverable oil equivalent [21]. 16%; quartz, 15%; illite, 19%; albite, 10%; micro-
When talking about known oil shale deposits, cline, 6%; pyrite, 1%; and analcite, 1%. In the case
there is a need to distinguish between two terms – of siliceous eastern US Devonian oil shale [14], a
“resources” and “reserves.” Resources typically typical mineral composition is quartz, 28%; illite,
refer to estimates of all deposits of oil shales, 40%; feldspar, 12%; pyrite, 14%; and other min-
while reserves refer only to those from which oil erals, 6%. Quite often, mineral matter composi-
extraction can be economically profitable with the tion is represented by the chemical composition of
use of existing technologies. It should be noted oil shale ash, in which the mineral components
that these technologies are constantly developing, have been converted to their oxides during com-
and because of that, estimates of reserves will not bustion. This is illustrated in Table 1. It is, how-
necessarily remain constant over time, irrespect- ever, very important to keep in mind that the
ive of use or discovery of new resources. minerals are fundamentally altered by the com-
bustion process, and the ash composition is only
Oil Shale Composition an approximate guide to what minerals existed in
Oil shale is composed of mineral and organic the raw oil shale.
matter whose proportions vary with the grade of In addition to the main minerals, oil shales
oil shale. contain a large number of trace elements (also
referred to as minor elements). The content of the
Mineral Matter trace elements is typically in the parts per million
Oil shale mineral matter contains several classes (ppm) range. Trace element data for several
of minerals such as carbonates (calcite and deposits are available in the literature [7, 24, 25].

Oil Shale Processing, Chemistry, and Technology, Table 1 Mineral matter composition in some oil shales presented
as oxides in ash, wt%
Green river [USA] Kukersite [Estonia] El-lajjun [Jordan] Dictyonema [Estonia]
Compound [14] [23] [23] [23]
SiO2 40–60 28.25 28.86 63.72
Al2O3 10–15 9.84 7.43 16.43
Fe2O3 5–10 5.44 3.70 8.80
CaO 10–25 41.02 52.79 0.85
MgO 5–10 5.05 1.30 0.50
K2O 3.60 8.72
Na2O 0.3
SO3 6.07 1.30 0.70
P2O5 5.9
52 Oil Shale Processing, Chemistry, and Technology

Organic Matter somewhat higher than the calculated “conditional

Oil shale organic matter is composed of solvent- organic matter,” the following equation has been
soluble and insoluble organic materials. The recommended: Orgd = 100  Ad + 0.125FeO
organic matter in oil shales is mostly in the form  (H2O)M  0.625Spyr  (CO2)Md, where FeO
of solvent-insoluble cross-linked macromolecular is iron oxide content in ash, (H2O)M is the content
structure or kerogen. However, the term “kero- of mineral water of hydration, and Spyr is the
gen” is often used for whole organic matter of pyritic sulfur content [27]. The difference between
oil shale, including a small solvent-extractable true organic matter and “conditional organic mat-
fraction of the oil shale (termed bitumen). ter” is generally not greater than 1–3%.
Because organic matter is not easily extract- It is clear that for most technological uses of oil
able from the whole oil shale, there is an issue of shale, there will be a need to go beyond simple
how to quantify it. One simple way is to assume estimates of organic weight fraction in character-
that it is the fraction that can be burned in air. Of izing the material.
course any loss of mineral components during
burning (as by dehydration or decarbonylation
reactions) can lead to an overestimate of the Kerogen Kerogen is the major fraction of oil
organic matter. shale organic matter, and it is insoluble in organic
The organic matter content is thus often solvents. It has a solid cross-linked macromolec-
expressed on a dry shale basis as “conditional ular structure and in this respect is somewhat
organic matter,” Orgcondd = 100  Ad  (CO2)Md, similar to coal. Accordingly, kerogen is swellable,
in which Ad is ash yield upon combustion, but not soluble, in organic solvents. The term
(CO2)Md is carbonate-derived CO2 released by “kerogen” was introduced by Professor Crum
the mineral matter at high temperatures, and a Brown to describe the insoluble organic matter
dry basis is indicated by the superscript d [7]. in oil shales that gives shale oil on heating and
Basically, this formula embodies the concept decomposition [28]. Nowadays the term kerogen
that what is not left after combustion as ash, or is used to characterize solid organic matter in all
released by carbonate decomposition, must have sedimentary rocks [29].
been organic matter. Both the ash and decompos- There are several excellent publications on kero-
able carbonate are available from separate ana- gens available in the literature [4, 30–32]. Oil shale
lyses. In reality, the actual organic matter content kerogens possess relatively high atomic hydrogen to
can be somewhat different than that obtainable carbon atomic ratios (typically 1.2–1.6), at least in
from the formula above since, for example, the comparison to coals, for which this ratio is typically
formation of SO42 ion from combustible sulfur less than unity. It is the high hydrogen to carbon
can contribute to the mass of ash, even if the sulfur ratio (combined with low oxygen to carbon ratio)
originated in the organic matter. Also, any other that makes oil shale kerogens of interest as a poten-
volatilization of mineral components, beyond the tial source of liquid fuels. Heating values of kero-
decomposition of the carbonates, would also lead gens have been shown to be in the range of
to errors in estimates of organic matter. 2440 MJ/kg, though as noted above the typical
Therefore for specific oil shales, more values are around 35–40 MJ/kg [9, 14].
detailed equations have been developed. For The macromolecular structure of oil shale ker-
example, for kukersite oil shale, which has a ogens is complex – kerogen is not a “homopoly-
true organic matter content somewhat lower mer.” Instead of being built of repeating regular
than that typically calculated from the “condi- structural units as are synthetic polymers or natu-
tional organic matter,” the following formula has ral materials such as cellulose, it is composed of
been proposed: Orgd = 100  1.028Ad  0.964 highly variable structural units, both in terms of
(CO2)Md + 0.637(SO42  0.07), where SO42 size and chemical character. The exact chemical
is the sulfate content [26]. For dictyonema structure of an oil shale kerogen is difficult to
shale, with a true organic matter content describe; therefore studies of “kerogen structure”
Oil Shale Processing, Chemistry, and Technology 53

can provide data only on average structural Kerogens can be classified in different ways on
parameters. This is the same situation as the basis of elemental composition. One of the
exists for such natural materials as coals or most common is based on the van Krevelen dia-
lignins. gram originally developed for classifying coals
Due to kerogen’s insolubility in organic sol- (Fig. 1). Kerogens fall into certain groups on the
vents, it was historically somewhat difficult to basis of the atomic ratios of hydrogen to carbon
characterize with respect to its chemical structure. and oxygen to carbon. Kerogens can be grouped
Attempts to liberate it from the oil shale rock as Type I, II, III, or IV with further subgrouping
would necessarily involve the risk of chemical according to sulfur content. Oil shale kerogens are
degradation. Even if the approach involves predominantly Type I (Green River oil shale from
dissolving away the mineral matter (see below), Uinta Basin is a typical kerogen of this group) and
there are concerns regarding whether the kerogen Type II (Toarcian oil shale from the Paris Basin is
has been altered, and of course, there remains the an example from this group). The less common
problem of reliable characterization of the chem- coaly shales belong to Type III (Upper Cretaceous
ical structure of a solid organic macromolecular sedimentary rock from the Doula Basin, as one
substance once it is isolated. This latter point is example) [32, 34]. The following numerical H/C
similar to the situation in the field of coal chemis- and O/C ratio ranges have been suggested: for
try, where decades of effort have been expended in Type I H/C > 1.25 and O/C < 0.15, for Type II
establishing structures on which there is now rea- H/C < 1.25 and 0.03 < O/C < 0.18, and for Type
sonably general, though not yet total, agreement. III H/C < 1.0 and 0.03 < O/C < 0.3 [32]. Some-
Aspects that kerogen structure have in recent what different H/C and O/C, and also O/S, ratios
decades also yielded to more sophisticated char- are found for kerogens at the end of diagenesis,
acterization methods, though, again, one can do but the above are the values for the materials of
little more than characterize average structures or practical importance.
structural elements.
As a result of the intractability of the raw Kerogen Isolation An important tool in the
kerogen structure, much has been done using its field of kerogen study is that of kerogen isola-
elemental composition as the simplest indicator of tion (or concentration) from the oil shale matrix.
properties or predictor of behavior. The composi- This involves the removal of solvent
tion of kerogens is typically presented based on solubles (bitumen) and mineral matter, leaving
carbon (C), hydrogen (H), sulfur (S), nitrogen (N), behind the kerogen. Several reviews are
and oxygen (O) content, since these are by far and available on isolation or concentration procedures
away the major elements of interest. For more (for information readers are referred to [18, 32,
precise compositional characterization of kukersite 35–37]).
oil shale kerogen, chlorine (Cl) is also included Solvent solubles are removable by ordinary sol-
[33]. Table 2 presents some sample kerogen vent extraction. Since all unextracted organics report
(or organic matter) elemental compositions of oil as kerogen, the efficacy of extraction is important in
shales from major deposits. establishing actual kerogen content. The amount of

Oil Shale Processing, Chemistry, and Technology, Table 2 Examples of kerogen [7] (or organic matter,* Ref. [14])
elemental compositions
Oil shale C H S N O H/C O/C
Kukersite (Estonia) 77.3 9.8 1.7 0.4 10.8 1.52 0.105
Green River (USA) 78.3 9.9 1.6 2.1 8.1 1.52 0.078
Dictyonema (Estonia) 70.5 7.4 4.2 2.5 15.4 1.26 0.164
Kasphir (Russia) 67.1 8.0 10.2 1.2 13.5 1.43 0.151
Timahdit (Morocco) 68.9 8.5 7.0 3.0 12.6 1.48 0.137
Eastern US Devonian* 82.0 7.4 2.3 2.0 6.3 1.08 0.058
54 Oil Shale Processing, Chemistry, and Technology

Oil Shale Processing, 2.00

Atomic H/C
Chemistry, and
Technology, Fig. 1 The
van Krevelen diagram,
showing the approximate
locations of different I
kerogen types, based
upon atomic ratios (From
Ref. [32], with permission) 1.50
II
en esis
D iag

Immature
zone
Principal

Catagenesis
1.00 zone of
oil formation
III

0.50
t

Zone
We
.
tag

of gas
Dry
Me

formation

0 010 020 030

Atomic O/C

Oil Shale Processing, Chemistry, and Technology, Table 3 Yield of bitumen per conditional organic matter. Atomic
H/C ratios of some oil shale kerogens and bitumens are shown for comparison [7]
Yields per conditional organic matter (%) (solvents H/C H/C
Oil shale used) bitumen kerogen
Kukersite (Estonia) 0.71 (benzene:methanol (1:1)) 1.75 1.52
Kultak-Zevardy 12.01 (benzene:methanol (3:1)) 1.42 1.2
(Uzbekistan)
Uktha (Russia, Komi) 21.42 (benzene:methanol (3:1)) 1.58 1.21
Green River (USA) 5.4 (chloroform)
Dictyonema (Estonia) 3.36 (benzene:methanol (3:1))

extractable obtained, and its properties, may vary A number of physical separation methods used
considerably depending upon the solvent used for to separate kerogen from the mineral part of oil
extraction. The amount of the extracted matter also shale are described in the literature [39–41]. These
increases with an increase in extraction process methods involve primarily such physicochemical
temperature and with the polarity and chemical processes as flotation and centrifugation. Combi-
activity of the extractant [38]. Typical extraction nation techniques, involving such processes as
yields are low, representing only a few percent of ultrasonication or electrostatic and magnetic sep-
organic matter, but varying significantly from arations are also physical methods used in kero-
shale to shale (see Table 3, for examples). gen enrichment [38]. It is worthwhile to note that
Oil Shale Processing, Chemistry, and Technology 55

for certain oil shales, some physical methods are provided some insights into the size of the systems
of value, since particle size reduction is accompa- being examined. There was a recent study of
nied by an increase in particle organic matter kukersite (Type I) and dictyonema (Type II) oil
concentration. shale samples involving field ionization mass
The most widely used laboratory techniques spectrometry (FIMS) pyrolysis under high vac-
for kerogen isolation from inorganic matter are uum. This investigation indicated number-average
based on chemical methods. In this regard, oil molecular weight values of 408 and 383 Da with
shale minerals can be grouped as carbonates standard deviation values of 141 and 128 for
(which are soluble in dilute acids), silicates kukersite and dictyonema oil shale pyrolysis prod-
(soluble in HF), and sulfides (which can be ucts, respectively (estimated using a three-parameter
removed by oxidizing and reducing agents). Gaussian distribution function to describe the
Mineral composition determines the reagents results) [46].
(and the efficiency of reagents) to be used for Earlier studies have, however, pointed to the
demineralization [37]. For oil shale deminerali- possibility of higher molecular weights for the
zation with chemical methods, readers are decomposition products. Suuberg and Sherman
referred to [37]. [47] studied pyrolysis of Green River oil shale
under vacuum conditions and obtained a weight-
Kerogen Degradative Analysis Although ele- average molecular weight of around 800 Da,
mental analysis offers valuable information for using a gel permeation liquid chromatographic
classification, other destructive and nondestruc- technique. This compared favorably with a weight-
tive techniques have been used to obtain infor- average molecular weight of around 850 Da
mation on functional groups and on structural obtained for the same Green River shale, using
functionalities of different oil shales [30, 42]. the FIMS pyrolysis method [48]. The latter studies
Earlier quantitative characterizations of kerogen showed the importance of pressure on molecular
functionalities were based on destructive weight distributions. Oils obtained using a Fischer
techniques [43]. The destructive techniques assay method showed a much narrower molecular
include thermal degradation (pyrolysis, weight distribution, shifted to much lower molec-
hydrogenolysis, supercritical extraction) and ular weights. Hence, it is dangerous to compare
chemical degradation with different chemical the molecular weight distributions from different
reagents based on the specificity of the reagents studies unless the mass transport conditions are
(for more information concerning this topic, see known to be identical. In any event, all of the
[18, 43]). Destructive analysis techniques are above studies point to the typical characteristics
based on obtaining smaller identifiable com- of a macromolecular system being thermally
pounds that have a structural relationship to the degraded to provide fragments exhibiting a distri-
kerogen [43]. bution of molecular weights (which are all, of course,
To describe the pyrolytic behavior of kerogen, lower than the original matrix molecular weight).
detailed knowledge of kerogen structure is critical
since the structural properties and functionalities
of kerogen strongly influence its pyrolytic behav- Nondestructive Analysis of Kerogen
ior. It has been shown in coal pyrolysis studies that Structure During the last three decades, there
devolatilization rates and product compositions have been a number of qualitative and quantitative
can be correlated with coal organic matter chem- studies aimed at characterizing the structure of oil
ical structure [44, 45]. So while certain types of shale kerogens using nondestructive techniques
functional groups may be directly related to their such as solid-state nuclear magnetic resonance
pyrolytic products, the use of this fairly non- (NMR) techniques (cross polarization, magic-
specific method is of limited value. angle spinning), electron spin resonance (ESR),
The analysis of kerogen thermal degradation Fourier transform infrared spectroscopy, and
product molecular weight distributions has X-ray photoelectron spectroscopy [13, 49–51].
56 Oil Shale Processing, Chemistry, and Technology

One recent comprehensive study on the struc- emphasized that what is shown are “averaged” struc-
ture of kerogens was carried out using both X-ray tures that accurately depict elemental ratios and to
photoelectron spectroscopy and solid-state 13C some extent functional groups, but which cannot be
nuclear magnetic resonance techniques [49]. held out as necessarily representing the actual link-
This study determined the average chemical com- ages between different groups and which certainly
positions of 18 oil shale kerogens (hydrogen- do not represent the different molecular weight
carbon skeleton, oxygen functionalities, nitrogen distributions present in the kerogens. It should
functionalities, sulfur functionalities). The data of also be emphasized that kukersite oil shale kero-
this study revealed that the average carbon- gen is a somewhat unique Type I kerogen due to its
hydrogen skeletal units of immature Type I and high hydroxyl (phenolic) group content [43]. This
Type II oil shale kerogens (according to H/C and is apparent in comparing Fig. 2a and b. Hence,
O/C ratios) are generally composed of two- to care should be exercised in using any published
three-ring hydroaromatic/aromatic clusters with model of kerogen structure in trying to
a significant number of short attachments and understand the chemical behavior of any other
long, roughly linear, alkyl chains. These data type of kerogen. Due to its high oxygen content,
also indicated that the number-average molecular kukersite kerogen is also sometimes categorized
weight of the “average repeating structural unit” as straddling Type I and Type II, since the lines
of oil shale kerogens might be around 400 amu. between the different kerogen types are not nec-
These values compare favorably with the molec- essarily as hard and fast as might be inferred from
ular weight distributions from the actual degrada- Fig. 1.
tion studies cited above.
Several other studies have been concerned
with the characterization of kerogen macromo- Bitumen or Bitumoid Bitumen (bitumoid) is a
lecular structures using solvent swelling in dif- minor fraction of oil shale organic matter that is
ferent solvents or solvent mixtures [46, soluble in common organic solvents (representing
52–54]. These studies were performed to evalu- up to 20% of organic material [38]) and is there-
ate at least semiquantitatively the number- fore recoverable by solvent extraction. Some
average molecular weight between cross-links authors define the bitumen as a benzene-soluble
(or cross-link density), a fundamental property matter, while others define this material as soluble
of a cross-linked macromolecular network. in petroleum-based solvents, yet others define it
These studies indicate that oil shale kerogen more broadly as material soluble in any organic
solvent swelling ratios are lower (therefore solvents, and finally, some define it as material
cross-link density higher and molecular weight entirely soluble in CS2. According to [31], there
between cross-links smaller) than those seen in are no standard protocols for extraction proce-
bituminous coals [6]. dures, either in terms of the solvent polarity, sam-
Extensive structural data are available for a few ple grinding, temperature, duration of extraction,
oil shale kerogens, especially for oil shales of major or stirring efficiency. Perhaps the main reason
deposits. For several kerogens the structural infor- why there has not been greater emphasis on this
mation has been the basis for detailed molecular point is that from a practical point of view, the
models of a portion of the structures (e.g., for Rundle amounts involved typically represent only a few
Ramsay Crossing [51], Green River [51], and percent by mass of the organic matter.
Kukersite [33], see also [18]). Reviews of the history These extractables, bitumen, contain bio-
of the development of hypothetical structural markers or organic compounds whose structures
models of the Green River Formation kerogen can are related to the chemical structures of the
be found in the literature [18, 28]. For illustrative biological material from which the organic
purposes, a recent two-dimensional model of matter was derived [55, 56]. The atomic H/C
kukersite oil shale kerogen chemical structure is ratio of bitumen has been shown to be somewhat
shown in Fig. 2a [33] and another model of Green higher than of corresponding kerogen
River shale in Fig. 2b [51]. Again, it needs to be (see Table 3).
Oil Shale Processing, Chemistry, and Technology 57

Oil Shale Processing, Chemistry, and Technology, formula C421H638O44S4NCl (From Ref. [33], with
Fig. 2 Two-dimensional models of kerogen chemi- permission). (b) Model of Colorado Green River
cal structure. (a) Model of Estonian kukersite kerogen (From Ref. [51], with permission)
kerogen with molecular mass 6581 Da and empirical
58 Oil Shale Processing, Chemistry, and Technology

Oil Shale Conversion Processes: An Integrated retorting techniques (both ex situ

Overview and in situ) have been under consideration to
recover shale oil together with nahcolite, alumina,
Oil shale resources are abundant and widespread and soda ash from some oil shale deposits [60].
throughout the world, and oil shale has been used However, the major emphasis in oil shale upgrading
for energy technological purposes for hundreds of has remained on kerogen thermochemical conver-
years (e.g., there is a shale oil extraction patent sion methods (processes using heat to liberate oil).
dating from 1694, granted by the British Crown, There have also been a few studies (laboratory
[57]). Despite this, currently only few countries research) on the use microbial activities to break
(Estonia, China, Brazil, Germany, and Israel) up kerogen in its mineral matrix [38, 61].
exploit oil shale on a commercial scale. The main In principle the general thermochemical
reason for this is that oil shale is a low-grade solid upgrading processes that have been developed
fuel. It cannot compete with coal as a solid fuel, and utilized for coals (combustion, gasification,
since the latter is available in richer (lower ash liquefaction, and pyrolysis) can also be used for
content) easily accessible deposits. Oil shale is a oil shale upgrading. The goal is, broadly speaking,
potential source of liquid fuels but suffers there to transform the chemical energy locked within oil
from the fact that the world oil market situation shale into more concentrated and conveniently
has historically been economically unfavorable in usable forms. However, due to oil shale’s thermal
terms of fostering the development of upgrading behavior and character, some types of processes
technologies. Generally, there has been need of are more favorable than others. It is important
government subsidies in order to consider the to note that the hydrogen/carbon ratio of oil
development of an industry based upon this shale organic matter (typically 1.2–1.6) is more
resource, and the energy market itself (influenced favorable to the production of liquids than that
mainly by crude oil price range and electricity cost) found in coal (typically 0.7–1), and this reduces
has made oil shale upgrading truly economical only the pressure to increase H/C ratio, which has
in the few countries mentioned. Moreover, the eco- economically been one of the major stumbling
nomic feasibility of oil shale development is certain blocks of coal-to-liquid conversion processes.
to be strongly affected by further strengthening On the other hand, with mineral content of
environmental regulations (air, water, and land between 70% and 90%, considerably higher than
use), socioeconomic issues, and market and invest- that of commercial coal deposits, oil shale pre-
ment risks. Taken purely on the basis of greenhouse sents the problem of processing a solid with a high
gas emission per unit of energy, oil shale is unat- mineral rock content, whose handling and heating
tractive in much the same way as is any other solid in any thermal process is economically quite
fossil fuel, such as coal. Therefore oil shale devel- unattractive.
opment in recent decades has been quite chaotic and
episodic, and there have been many processes/tech- Direct Combustion
nologies proposed and many have been abandoned Oil shale can be utilized as a fuel for thermal
due to economic reasons. power plants, where it is burned directly as is
Just as is coal, oil shale is also a multiuse coal. Currently there are several countries that
organic-mineral raw material whose economic utilize oil shale as fuel for power plant fuel: Esto-
value has sometimes extended beyond its fuel nia, Israel, Germany, and China [62, 63]. Some
value. From some oil shales, uranium and other countries have closed their oil shale-fueled power
minerals and rare elements can be extracted. As plants (Romania) or turned to other fuels (Russia).
an example, in 1946–1952, a marine type of Other countries are planning to build such plants
dictyonema shale [58] was used for uranium pro- (Jordan and Egypt) or burn oil shale together with
duction in Sillamäe, Estonia, and in 1950–1989, coal (Canada and Turkey) [62, 64].
alum shale was used in Sweden for the same There are three commercial technologies in
purpose [59]. use for oil shale combustion [65]: (1) pulverized
Oil Shale Processing, Chemistry, and Technology 59

combustion or pulverized firing (PF), as is used in Low-Temperature Pyrolysis

the older units in Estonia; (2) fluidized bed com- Low-temperature pyrolysis up to at about 500
C
bustion (FBC), used by Rohrbach Zement in (also referred as retorting or semicoking or low-
Dotternhausen, Germany; and (3) circulating flu- temperature carbonization) has been historically
idized bed combustion (CFBC), used in two new the most favored thermochemical conversion pro-
units at the Narva power plant in Estonia, Huadian cess for high-grade oil shales. In this pyrolysis or
power plant in China, and PAMA power plant retorting process, the organic matter is converted
at Mishor Rotem in Israel. Generally speaking, to oil (termed retort oil, shale oil, shale crude oil),
PF technology is no longer in favor, because it gas (typically called retort gas), and solid residue
operates at such a high temperature that there (typically termed semicoke in the oil shale indus-
can be no sulfur capture by the carbonates present try). The processes can be carried out either by
in the mineral matter, and therefore it is environ- mining the oil shale and then producing oil
mentally much more problematic. The thermal by heating it in aboveground retorting facilities
efficiency of the new Estonian CFBC units has (termed ex situ processing) or by heating the oil
been reported to be in the neighborhood of shale underground without mining it and bring-
about 36%, which is typical of this sort of fluid- ing it to the surface (termed in situ processing).
ized bed combustion system. The older PF units Limiting the processing temperature to below
operated at only about 30% thermal efficiency to 500
C is desirable, in terms of maximizing oil yield.
electricity [66]. Oil shales are commonly retorted in commer-
An exceptionally extensive study of the corre- cial ex situ retorts without beneficiation (organic
lation between the heating value and various com- matter is not concentrated). Commercial ex situ
positional parameters of Estonian kukersite has shale oil production facilities are in operation in
been presented by [67]. While the quantitative Estonia, Brazil, and China. Commercially pro-
results of that study are applicable only to this duced shale oils are presently used as heating
particular oil shale, the methodology may well be fuels or as feedstocks for producing chemicals
generalizable. That study suggested that a good [69]. Due to their broad boiling point distribu-
working formula for the estimation of the higher tions, high pour points, high viscosities, and
heating value of the organic part of oil shale is high heteroatom contents (nitrogen, sulfur, and
oxygen), raw shale oils need to be upgraded to
QH ¼ 0:3566C þ 1:0623H  0:1339O be considered for higher value uses. Retorts could
þ0:0658S þ 0:0106N  0:0280Cl be included within a petroleum refinery. There
have been a number of investigations on the use
in which the heating value is in MJ/kg, and the of the shale oils as refinery feedstocks for manu-
element symbols represent the mass percentages facture of motor fuels [70].
of the respective elements in the organic fraction
of the oil shale. As part of the same original study High-Temperature Pyrolysis and Gasification
by Saar, the lower heating value of the typical High-temperature pyrolysis (or gasification)
dried whole Estonian shale is (see [68]). under self-generated atmospheres was historically
used to produce town gas (e.g., commercial scale
QL ¼ 34:45  0:302 A  0:4369ðCO2 Þc chamber ovens used in Estonia, [71]). Just as in
the case of town gas produced from coal feed-
in which A is the dry oil shale ash content, and stocks, gas production from oil shale was aban-
(CO2)c represents the content of carbonate min- doned due to economic reasons, largely having to
eral CO2 in the dry oil shale. do with the low cost of production of natural gas.
Since direct combustion of oil shale is not the Modern gasification technologies, applied to
main focus of this review, the reader is referred to conversion of coals or biomass, use partial oxida-
the comprehensive volume recently published by tion (steam, oxygen) or hydrogen environments.
Ots on this topic [68]. In these gasification processes, the solid feed
60 Oil Shale Processing, Chemistry, and Technology

material is converted to synthetic gases (composed The thermal dissolution processes generally
mainly of hydrogen and carbon monoxide) that involve heating the oil shale to temperatures
can be further used as combustion fuel in, for below those characterized by active devolatilizat-
example, integrated gasification combined cycle ion, as there is a desire to maximize oil yields and
(IGCC) plants designed to produce electrical minimize gas production. The temperature range
power, or further catalytically converted (via of active devolatilization depends somewhat upon
Fischer-Tropsch reactions) to produce chemicals, gas pressure applied to the system.
including liquid fuels. However, existing com- The thermal dissolution of oil shales has been
mercial gasification technologies that would lend mainly of academic or research interest as the H/C
themselves to integration with upgrading plants ratio of oil shale organic matter is already much
are not economically/technically suitable for oil more favorable than of coal. For more informa-
shale feed materials, unless the mineral matter tion, on oil shale thermal dissolution, there are
is reduced before gasification. There have been recent brief reviews covering the topic [79, 80]
no demonstrations of large-scale oil shale benefi- and, separately, supercritical extraction [81, 82].
ciation technologies for concentrating oil shale One step toward commercial development is
organic matter (to reduce mineral matter) to levels the proposed 1 ton/h so-called Rendall process
suitable for proven gasification technologies. demonstration plant under construction in Australia
Although in principle, oil shale gasification by Blue Ensign Technologies Limited. This pro-
could be more attractive than shale oil extraction cess is designed to use oil shale from the Julia
via retorting, especially for low oil yield oil Creek deposit in Queensland and to hydrotreat it
shales, this thermochemical conversion process in a two-stage process using a petroleum-based
has not been a subject of commercial interest. solvent. Again, this proposal is only at a labora-
Over the course of time, a number of laboratory tory stage.
scale to pilot scale studies have been carried out A considerable amount of work (up to process
on this topic [72–76]. development unit scale) has been published on
hydropyrolysis or pyrolysis under high hydro-
gen pressures, including catalytic hydropyrolysis
Other Thermochemical-Based Conversion and hydrogasification [72, 83]. Indeed, this tech-
Technologies nology was already explored in Estonia in the
Thermal dissolution or heating in the presence of 1930s [84].
solvent has been considered as an alternative to
“dry” thermochemical conversion such as classi-
cal retorting. These attempts have included those Shale Oil Production
in subcritical and supercritical fluids, reactive and
nonreactive fluids, and of special interest have General Considerations Regarding Process
been hydrogen-donor solvents that can provide Temperatures
hydrogen in the conversion process. In the latter Technologies to recover oil from oil shale are
case, gaseous molecular H2 can also be intro- based on thermal breakup of kerogen, the cross-
duced. Again, there is strong analogy to methods linked macromolecular organic matter of oil shale.
that have been suggested for direct liquefaction or All present technological approaches require heat
solvent refining of coals. Examples include pro- to take kerogen to temperatures at which thermo-
cesses such as solvent-refined coal (SRC-I and the chemical conversion processes are sufficiently
improved version, SRC-II) processes, the Exxon fast to be of interest from a practical point of
Donor Solvent (EDS) process, and the H-Coal view. As noted above, depending on process con-
process [14, 72, 77]. A brief review on recent ditions, the thermochemical conversion results in
advances in direct coal liquefaction and a list of three general classes of products with different
major direct coal liquefaction processes is given quantities and varying qualities. These products
by Shui et al. [78]. are the oil (more generally non-polymeric organic
Oil Shale Processing, Chemistry, and Technology 61

compounds with a wide molecular weight distri- moving bed coal gasifier-type retorting equip-
bution), a carbon-rich solid residue, and gas ment, which utilizes particles with size in the
(including water vapor). order of 100 mm (e.g., 10–125 mm for the Kiviter
Kerogen decomposition rates increase with process, [86]). The latter high heating rate condi-
temperature, subject to the Arrhenius rate law. tions are encountered in fluidized bed reactors,
The rates will be quantitatively explored in a which use small particles with sizes below a few
later section. Kerogen decomposition is slow at millimeters. In all these retorting processes (which
low temperatures and occurs on a timescale of take place in an inert gas environment or under
days to years below 350
C, which is why practi- self-generated gas environment with very low
cal processes are generally carried out at higher oxygen content, or in a reactive, non-oxygen envi-
temperatures. The decomposition becomes fast ronment), the oil shale is heated up to the active
enough for practical purposes at temperatures of pyrolytic reaction region from 450
C to 550
C,
400
C and above. Cummins and coworkers [85] where oils and gases evaporate and hydrogen-
studied Green River oil shale low-temperature deficient organic matter or semicoke is left in the
isothermal pyrolysis from 150
C to 350
C. solid phase.
They showed that 2% of kerogen was converted In situ processes generally involve heating
to benzene solubles in 90 days at 150
C, and in rates below 1
C/min, because of the impractical-
360 days, 7.7% of kerogen was converted. At ity of rapid heating in an in situ environment. The
200
C in 90 days 3% was converted and in relatively low-temperature range of 340–370
C,
360 days 8.9%. When the temperature was raised with timescales of years, is utilized by Shell in its
to 300
C in 0.5 days, 3.3% was converted, and in in situ (underground) conversion process (Shell
4 days 12.8% yield was obtained. Finally, at Oil Shale Test project) [60]. The above-cited time-
350
C in 0.5 days, 27.6% was converted, and in scales for kerogen conversion show that it is
4 days, 65.1% of kerogen was converted into not the kinetics of conversion alone that determine
benzene solubles. these long times; rather, they are also influenced
The above demonstrates that there is actually by the time needed to heat the large volumes
no lower temperature limit for kerogen decompo- of rock and to allow the released oil to flow out
sition, consistent with what would be concluded of the shale.
from Arrhenius kinetics; rather, the question lies
in establishing the timescale needed at any partic- Other Processes Taking Place during
ular temperature. It is the fact that practical ex situ Retorting
(i.e., in retort) conversion processes must be car- The destruction of kerogen by heating is accom-
ried out on the timescale of minutes or hours, at panied by processes in other oil shale constituents,
most, that sets the commonly cited “lower tem- that is, in bitumen and mineral matter. These pro-
perature limit” for decomposition. These limits cesses include:
may be extended for in situ techniques, where
confinement volume in an aboveground retort is • Solvent-extractable organic matter or bitumen
no longer a consideration, but there is still a lower (non-covalently linked non-polymeric com-
limit at which the process would simply take too pounds) can vaporize or decompose, depend-
long to be practical. ing on their thermal stability. Depending on the
Real processes can be run under conditions oil shale, the bitumen content can constitute up
varying from slow heating rate and long residence to 20% of organic matter [38], though as noted
time to fast heating rate and short residence time. above it is generally much less than this. If able
For example, in aboveground (ex situ) retorting to quickly escape the hot oil shale and be
equipment, heating rates and residence times can cooled, these compounds should be captured
vary from the order of 1
C/min and hours to the in a relatively minimally altered form.
order of 1000
C/min and seconds, respectively. • Mineral matter present in oil shale can have
The former is encountered, for example, in catalytic effect on thermochemical conversion
62 Oil Shale Processing, Chemistry, and Technology

processes of kerogen [87, 88]. Also, the min- heat requirement for heating oil shale to 500
C,
erals themselves can decompose at tempera- including allowing for vaporizing volatiles, is
tures of thermochemical conversion of kerogen. observed to be about 650–750 kJ/kg of oil shale
Some minerals can partly or fully decompose [14]. The heat of the pyrolysis reactions is by itself
even under typical retorting conditions (e.g., relatively small [14, 72]. A review of some recent
nahcolite, dawsonite, analcite, smectite, kao- applications of differential scanning calorimetry
linite, siderite, and pyrite) [6, 24, 68]. These (DSC), mainly aimed at providing kinetic infor-
decomposition reactions are generally endo- mation, also provides reference to some recent
thermic. At temperatures around 450
C, mag- work on thermal effects of pyrolysis [90].
nesium carbonate starts to decompose [6]. The
interaction of a mineral with organic matter or Assay Methods: Determining Potential Oil
other minerals present in the oil shale can cause Yield and Its Relationship to Other Properties
altered thermal behavior, relative to the pure Procedures for determining oil production poten-
mineral [57]. Rajeshwar and coworkers have tials are known as assay methods. Assay methods
summarized thermal characteristics/behavior generally involve laboratory scale slow heating
of minerals (17 minerals) commonly found in rate (order of 10
C/min) batch pyrolysis systems
oil shale deposits [57]. Thermochemical prop- or so-called carbonization tests. The procedures
erties and mineral reactions of oil shale min- were developed to provide a simple and reproduc-
erals have been reviewed by many authors ible basis for fossil fuel (coal, lignite, oil shale)
(e.g., [6, 23, 24, 57, 60]). characterization and comparison, as opposed to
• Trace elements of oil shale can distribute providing scientific insights into the material.
between retorting gas, oil, solid residue, and The modified Fischer assay method has
retort water [89]. For example, there may be been widely used (since 1946) to evaluate the
the presence of trace elements of environmen- approximate oil potential of oil shale deposits
tal concern (arsenic, selenium, mercury) in and compare different oil shales. The term “mod-
retort oil, gas, or water [24, 72], or trace ele- ified” denotes that the original Franz Fischer
ments in oil (vanadium, iron, nickel, arsenic) method [91], used in Germany to characterize
can poison refinery catalysts [24, 89]. Actual coals, was taken over to characterize oil shales
concentration values depend on the retorting and has gone through improvement in this pro-
process and, of course, the trace element com- cess. The technique has been standardized as the
position of specific oil shale. Most of the trace American Society for Testing and Materials
elements are retained in the solid spent shale method (ASTM method) D-3904-80, D3904-85,
itself [72]. and D3904-90 or as the International Organiza-
tion of Standardization (ISO) standard ISO
Thermal Effects during Retorting 647-74 or as the GOST 3168-93 (the GOST is
Large quantities of heat are required to manufac- the former Soviet Union Standard).
ture a usable fuel (liquid or gaseous) from oil The standardized modified Fischer assay
shale. Heat requirements include (1) heating the method (ISO 647-74; GOST 3168-93) consists
mineral and organic matter to reaction tempera- of heating a 50-g dried and crushed oil shale
ture, (2) providing the heat of reaction related to (90% particles passing through a 1-mm mesh,
organic matter decomposition and heat effects with not more than 50% particles passing through
covering chemical changes in mineral matter, a 0.2-mm mesh) in a specially designed cast alu-
and (3) providing the heat of vaporization of vol- minum retort (made of 99.5% aluminum) to
atiles (oil and water). The thermal management of 520
C in 70 min with decreasing heating rate,
processes must also include provision for cooling followed by holding the sample at that tempera-
the products of the process back to ambient tem- ture for 10 min. The process is performed in a self-
peratures. The major amount of process heat is generated atmosphere. The distilled vapors of oil,
consumed with heating mineral matter. The total gas, and water are passed through a condenser
Oil Shale Processing, Chemistry, and Technology 63

cooled with ice water and collected into a gradu- detector, for example. The basic method was
ated centrifuge tube. The oil and water are then designed to measure the amount of free hydrocar-
separated by centrifuging. The quantities reported bons in the sample (S1), the evolution peak tem-
are the weight percentages of shale oil (and its perature (Tmax) and the amount of hydrocarbons
specific gravity), water, shale residue, and “gas generated through thermal cracking of nonvolatile
plus loss” by difference. organic matter (S2), and the amount of CO2 pro-
The above procedures were developed to pro- duced during pyrolysis of kerogen (S3). Based on
vide a reproducible basis for oil shale characteri- these, the hydrogen index (HI) as mg S2 per gram
zation/comparison. Estimated errors are not of total organic carbon, oxygen index (OI) as mg
expected to be greater than 5% (relative) under S3 per gram of total organic carbon, and produc-
the standardized conditions used [3]. However, tivity index (PI) as S1/(S1 + S2) can be calculated.
the Fischer assay is a static assay under self- Rock-Eval pyrolysis temperature Tmax and HI are
generated “inert” atmosphere and factors, such used to evaluate maturity of oil shales. HI and OI
as cracking of volatiles due to long residence have been used to construct a modified van
time in a hot zone, and char forming reactions Krevelen diagram.
[12] can limit the oil yield. Therefore, for exam- Thus the approximate grade of oil shale
ple, the use of an inert gas sweep [12] provides (average oil yield) has been determined histori-
greater oil yields than from Fischer assay under a cally in laboratory batch retorts such as the “mod-
self-generated atmosphere. It is not uncommon to ified” Fischer assay method or the Rock-Eval
encounter laboratory retorting results in which oil technique. There are high and low oil yield kero-
yields exceed Fischer assay results. gens based on modified Fischer assay results.
Other assays, developed for carbonization tests Tables 4, 5, and 6 are compiled to show how the
of fossil fuels, have also been applied to character- composition of kerogen affects the product yield
izing oil shales and included the Heinze retort [92], and product composition. Table 4 also confirms a
the Gray and King method (developed in England) general observation that the amount volatiles gen-
[91, 93], the original Franz Fischer method [91], erated during thermochemical conversion under
the Bureau of Mines oil shale method [91], the an inert environment up to about 500
C decreases
Tosco material balance assay (with collection of from kerogen Type I to Type II, and gas to oil ratio
gaseous products) [72], and the “material balance” and semicoke to oil ratio increase from Type I to
Fischer assay method (modification for collection Type II [18]. It should be noted that a van
of gaseous products) [94]. Also, results based upon Krevelen diagram-based classification of oil
various thermogravimetric analyses (TGA) have shale behavior (based upon H/C and O/C ratios)
been used by many investigators and are not nec- provides only a rough prediction of behavior, and
essarily directly comparable to the results from better correlations between pyrolysis yields and
Fischer assay methods. organic matter types might need additional indi-
One widely applied oil shale rapid characteri- cators or different classification approaches [98].
zation and screening method is Rock-Eval analy- As the modified Fischer assay has been a
sis [95, 96]. Rock-Eval pyrolysis has been used to standardized method for evaluating oil yields of
evaluate the maturity of the rock, as well as its different oil shales, the chemical and physical prop-
richness (organic carbon content) and quality (oil erties of different oil shale deposits have been cor-
and gas potential) [97]. The technique is based on related with Fischer assay results. Properties that are
temperature-programmed heating of a small sam- successfully correlated include, for example, den-
ple (under 100 mg) in an inert atmosphere (helium sity, thermal conductivity, organic carbon, total
or nitrogen) up to 600
C (to 800
C using version hydrogen, aliphatic/aromatic carbon, and unpaired
Rock-Eval 6). The evolution of hydrocarbon vol- electron concentration [13]. In general, each spe-
atiles is monitored by a flame ionization detector, cific deposit exhibits its own correlations, and these
and carbon dioxide (CO2) by other means, such correlations need to be derived for the specific
as an infrared analyzer or a thermal conductivity deposit, as the relationships are mostly empirical.
64 Oil Shale Processing, Chemistry, and Technology

Oil Shale Processing, Chemistry, and Technology, Table 4 Fischer assay yields, dry organic matter basis, of some
selected Type I and II oil shales. Parent organic matter H/C and O/C ratios are also shown
Green River (Utah, El-lajjun Kukersite Eastern USA Devonian Dictyonema
USA) [7] [23] [7] (Antrim) [7] [7]
H/C ratio of organic 1.52 1.42 1.51 1.14
matter
O/C ratio of organic 0.08 0.05 0.11 0.23
matter
Kerogen type I I I/II II II
Oil yield, wt% 59.3 61.9 60.0 22.2 19.8
Gas yield, wt% 22.7 20.9 21.5 25.1 16.5
Semicoke yield, wt 11.8 10.8 9.1 42.5 45.4
%
Retorting water 6.2 6.4 9.4 10.2 18.3

Oil Shale Processing, Chemistry, and Technology, Table 5 Fischer assay gas compositions of the selected Type
I and Type II oil shales shown in Table 4
Gas composition, Green River (Utah, El-lajjun Kukersite Eastern US Devonian Dictyonema
vol% USA) [7] [23] [7] (Antrim) [7] [7]
CO2 30.3 14.5 25.0 9.7 13.6
H2S 3.3 25.2 8.8 29.1 20.3
CmHn 2.0 9.3 16.1 2.6 7.5
CO 16.8 1.0 7.9 14.3 5.7
H2 24.2 21.6 8.2 22.4 27.1
C n H2n+2 21.2 28.4 34.0 19.2 25.8

Oil Shale Processing, Chemistry, and Technology, Table 6 Fischer assay oil elemental compositions of the some
selected Type I and Type II oil shales shown in Table 4 [7]
Green River (Utah, USA) Kukersite Eastern US Devonian (Antrim) Dictyonema
C 84.7 83.0 83.8 83.3
H 12.0 9.9 10.6 9.3
S 0.6 1.1 1.8 2.6
N 2.1 0.1 0.7 1.0
O (by difference) 0.6 5.9 3.1 3.8
H/C 1.7 1.43 1.52 1.34
O/C 0.005 0.05 0.03 0.03

Thermochemical conversion of oil shale kero- observation that shale oil aromaticity can be
gen to oil produces changes in oil chemical com- greater or lower than that of the parent
position relative to the original oil shale organic kerogen [13].
matter. The analysis of data on hundreds of oil There are a number of papers [7, 9] concerned
shales (compiled by Urov and Sumberg [7]) indi- with the characterization of Fischer assay oils,
cates that the Fischer assay oil H/C ratio can be examining, for example, elemental composition,
either lower or higher than that of the parent molecular weight (varying from 190 to 310 Da),
organic matter. This is consistent with an density (or specific gravity varying from 0.8 to
Oil Shale Processing, Chemistry, and Technology 65

1.04, with only a few exceptions <1), heating Pressure Effects on Retorting Yields
value (39–44 MJ/kg), viscosity, boiling range The mass transfer of oil out of pyrolyzing particles
(30–70% boiling over up to 350
C), and chemical is affected by the pressure difference between the
groups (nonaromatic hydrocarbons, aromatic particle interior, where the products are produced,
hydrocarbons, heteroatomic compounds, etc.). and the surrounding atmosphere. Delayed escape
It was already noted that the Fischer assay is a of the volatiles can result in secondary cracking
standard method, and it does not provide reactions of higher molecular weight products,
either the maximum or minimum oil yield that resulting in lower oil yields and increase gas and
can be produced from a given oil shale. Changes char yields.
in parameters such as reactor configuration, par- With an increase in pressure under an inert
ticle size, heating rate, pressure, or pyrolysis environment (or self-generated environment), the
environment also can affect product yields and oil yield decreases, the gas yield increases, the
compositions [16, 72, 99]. In real processes char yield increases, and oil specific gravity and
these factors are not always independent, as, for viscosity decrease. Also, the amount of lighter
example, high heating rates require fine particles distillate oil fraction increases [72]. All of these
or hydrogen environment is used with high phenomena are consistent with a process of crack-
pressure. ing of higher molecular weight oils to lighter oils
(with the deposition of coke). Different results can
be seen when reactive pyrolysis atmospheres
Heating Rate Effects on Retorting Oil Yields (such as hydrogen or steam) are used. On the
Fluidized bed retorting, where heating rates reach other hand, these conclusions are not necessarily
1000
C/min and particle sizes below a few mil- general, as in high heating rate (1000 K/s), small
limeters are used, produces higher oil yields than particle pyrolysis, oil yields were observed to be
obtainable from Fischer assay. For example, yields quite similar, decreasing only a small amount with
about 110% relative to Fischer assay have been increasing pressure above a temperature range of
reported in the case of Green River oil shale and 800–900 K [47]. The oil loss at higher pressures
about 150% of Fischer assay for eastern US Devo- was associated with the occurrence of cracking
nian shale [16]. Fluidized bed pyrolysis shows, reactions. The molecular weight distributions of
in addition to increased oil yield and decreased the shale oil were relatively insensitive to pres-
gas yield, different oil properties such as higher sure, though at higher temperatures, higher pres-
aromaticity, molecular weight, density, and vis- sures tended to slightly favor lower molecular
cosity [16], as compared to Fischer Assay oils. It weights. The results of this work also showed
should also be noted that laboratory scale fluid- clearly that there was an evaporative equilibrium
ized bed retorting experiments usually resulted in responsible for determining the molecular weight
higher oil yield than commercial scale design distribution in the oil, as the average molecular
retorts [16]. weights in the oil were significantly below those
In large-scale retorting systems, high heating of the extractable thermobitumen left in the shale.
rates require high heat transfer rates and, thus, a Hence the effects of pressure are associated with
finely ground feed. Therefore differences in prod- evaporative processes, though these effects are
ucts from low and high heating rate thermochem- not as significant as they are in determining the
ical conversion processes are influenced also by yields of tars from coal pyrolysis [45, 100].
the effect of particle size. The larger the particle,
the slower the mass transfer rate of volatiles to the Effect of Gas Atmosphere on Retorting
outer surface of particle and, thus, the more oppor- An inert gas sweep of a retort provides greater oil
tunity for the products to undergo secondary reac- yields than obtained from the Fischer assay proce-
tions that would alter their nature. Hence, it is dure under self-generated atmospheres [12]. Pyrol-
difficult in most cases to separate the effects of ysis in a steam atmosphere has been shown to have
heating rate and particle size. favorable effect on products by improving yield
66 Oil Shale Processing, Chemistry, and Technology

and quality of shale oil [12] relative to Fischer communities was to improve the fundamental
assay. In fluidized bed experiments, steam has understanding of pyrolysis chemistry with an ulti-
been studied as one possible fluidization medium mate objective of developing realistic simulation
[12, 16]. However, the beneficial effect of steam models. The examination of coal pyrolysis tended
relative to nitrogen is the subject of dispute among to place great emphasis on the importance of the
different researchers [12]. interplay between chemical and physical pro-
Pyrolysis under high hydrogen pressure cesses encountered in pyrolysis, something that
(hydroretorting, hydropyrolysis) is shown to be a received less attention in the shale pyrolysis work.
potential way to upgrade low-grade oil shales to Coal pyrolysis is an important phenomenon
oil. Hydroretorting results in increased oil yield inasmuch as it is an initial step in all major coal
and more aromatic oil with lower heteroatom con- thermochemical conversion processes such as
tent and lower molecular weight distribution combustion, gasification, carbonization, and liq-
[17]. Hydrogen pressures of ca 5–10 MPa were uefaction. The past four decades resulted remark-
sufficient to have an effect [17]. In some studies, able advances in the modeling of coal pyrolysis.
higher pressures, 15 MPa, have been used [92]. A There was a shift away from simple, empirical
hydroretorting assay unit has been developed kinetic models to use of more complex “network”
along with a standardized procedure for its use devolatilization models. Those models provide
[17]. In this unit (538
C and hydrogen pressure detailed predictions of pyrolytic behavior for a
6.9 MPa) for low oil yield oil shales (Type II wide variety of coals under variable temperature,
kerogens), oil yields can be increased several pressure, and heating rate conditions. There are
times relative to Fischer assay, even 400% over many extensive reviews and books published on
Fischer assay yield [17]. coal pyrolysis [6, 45], covering the chemistry of
pyrolysis, mass transfer effects, comparison of
Chemical Aspects of Oil Recovery by Retorting experimental data and models, etc.
In some ways oil shale is similar to coal as the
Shale Pyrolysis in Comparison to Coal Pyrolysis organic matter is mostly a macromolecular three-
The process of recovering shale oil is concerned dimensional structure, however, with different
with the decomposition of kerogen, the major organic compositions and hydrogen-carbon skel-
portion of oil shale organic matter. Kerogen is a etons. Therefore, oil shale pyrolysis is character-
cross-linked macromolecular structure whose ized by different product characteristics and by
thermal behavior depends upon the severity of different rates and extents of pyrolytic process
heating, yielding either mostly hydrogen-rich events (cross-linking, swelling, softening, resolid-
solvent-soluble material or light volatiles (oil, ification, and devolatilization). The pyrolytic
gas), and a carbon-rich solid residue. The overall devolatilization processes of oil shales occur at
pyrolysis process is extremely complex with a somewhat lower temperatures than do the same
variety of physical-chemical mechanisms. The sorts of processes in coals. The active devolat-
composition and yields of pyrolysis products ilization starts at about 350–400
C with a maxi-
depend on the type and characteristics of oil mum in the range of 400–500
C and completes at
shale (primarily of kerogen) and, as noted above, about 500–550
C. The oily pyrolysis products
the particle size, final temperature, heating rate, are more volatile (with higher hydrogen/carbon
residence time of vaporizing products in hot zone, ratio and lower molecular weight distribution)
pressure, and pyrolysis atmosphere. The possibly than the coal tarry products. Oja has demonstrated
important role of secondary reactions has been that the kukersite oil shale primary tar (oil) pro-
noted in several places above. duced during moderate rate pyrolysis (5
C/s)
In the period from 1970 to 1990, there were under a nitrogen sweep in a tubular furnace is
extensive theoretical and experimental studies on easily volatile below the temperatures of active
coal pyrolysis, and relatively fewer such studies devolatilization [101]. This is not generally the
on oil shale. The aim of both research case for coal liquid (tar) pyrolysis products.
Oil Shale Processing, Chemistry, and Technology 67

It can be assumed that the general principles importance of such processes in oil shale pyroly-
used to model the mechanisms of coal pyrolysis sis will be considered below.
could also be used to describe oil shale kerogen
pyrolysis. For coal, the following pyrolysis event General Observations of Oil Shale Pyrolysis
sequence was proposed [45]: Phenomena
A large number of studies (including reviews) on
• Disruption of hydrogen bonds pyrolysis of oil shales have been conducted and
• Vaporization and transport of non-covalently reported over the course of history. The pyrolytic
bonded “guest” molecules behavior of oil shales have been investigated gen-
• Low-temperature cross-linking (in low-rank erally based on individual oil shales or in some
coals) that is accompanied with CO2 and H2O cases generalized to oil shale kerogen types
evolution (Type I, II, or III in accordance of van Krevelen
• Bridge breaking and fragmentation of the mac- diagram) [42, 102, 103]. Although kerogen types
romolecular network (depolymerization) (similar to coal ranks) have been shown to be
• Hydrogen utilization to stabilize free radicals useful for generalizing oil shale kerogen (organic
• Vaporization and gas phase transport of light matter) behavior, it should be kept in mind that
fragments differences between kerogens make a large differ-
• Moderate temperature cross-linking reactions ence in behavior, and the trends with kerogen type
to resolidify the macromolecular network have not been firmly established.
• Decomposition of functional groups to pro- There have been a number of studies of the
duce light species or gases changes in kerogen structure (in the solid organic
• High-temperature condensation of the macro- residue) during pyrolysis/heating. These changes
molecular network involving hydrogen involve elemental composition, functional group
elimination losses, and other structural modifications, as well
as changes in the number of free radicals. Tech-
In most respects, the same processes must take niques employed for tracking the changes include,
place in oil shale pyrolysis. for example, elemental analysis, Fourier trans-
The importance of the extent and timing of form infrared spectroscopy, solid-state nuclear
cross-linking in the pyrolytic processes of coals magnetic resonance (NMR) spectroscopy, elec-
has been emphasized in coal pyrolysis models. tron microscope, or electron spin resonance
A number of studies have been performed on [104, 105]. There are of course also a great num-
swelling of thermally pretreated coals indicating ber of studies on volatile product evolution during
loosening and/or tightening of coal structures pyrolysis. This includes the evolution of volatile
in the temperature region prior to tar evolution products correlated with kerogen functional group
or active pyrolysis. This phenomenon was found decomposition. Techniques used have included
to depend on coal rank: changes in lower rank Fourier transform infrared spectroscopy, mass
coals indicate low-temperature cross-linking, spectrometry, or gas chromatography [106].
whereas trends in higher rank coals vary from For properties of crude shale oils, readers are
no observable changes to structural relaxation. referred to [14, 69, 72, 107].
It was observed that the extent of low-temperature
cross-linking was associated with decomposi- Observations Regarding Changes in the
tion of carboxyl groups and possibly hydroxyl Macromolecular Structure of Oil Shale Kerogens
groups as formation of new cross-links before during Pyrolysis
devolatilization happened together with CO2 and In relation to the different oil yields observed in
water release. It has been concluded that low- the case of pyrolysis of Type I and Type II
temperature cross-linking results in low tar yield kerogens (see Tables 4, 5, and 6), solvent swell-
and can be a reason for the non-softening pyro- ing experiments on thermally pretreated oil
lytic behavior seen in low-rank coals. The shales have indicated that high oil yield
68 Oil Shale Processing, Chemistry, and Technology

kukersite kerogen (Type I), which exhibits soft- variation in reported kinetic parameters
ening pyrolysis behavior during pyrolysis, (activation energies and frequency factors). Con-
shows a tendency toward structural relaxation in sequently, the results are often of limited general-
the low-temperature pre-pyrolysis region ity, applicable to the particular experimental
[46]. This observation is qualitatively comparable conditions studied, often potentially influenced
to that seen in high tar yield softening coals. On by heat and mass transfer limitations on observed
the other hand, low oil yield dictyonema kerogen kinetics.
(Type II), which exhibits non-softening pyrolysis A sampling of kinetic results is shown in
behavior, shows a tendency for pre-pyrolysis Fig. 3. These results have been taken from studies
cross-linking or structural tightening [108], simi- in which both activation energies and pre-
lar to what happens in low-rank non-softening exponential factors have been reported. Except
coals [109]. as noted, they are based upon a model of pyrolysis
that is first order in kerogen, and normally the rate
Mechanisms and Kinetics of Kerogen data are those either for total mass loss or the
Decomposition production of oil. Hence, the “model” can be
Mechanisms of kerogen decomposition have been simply represented as:
the subject of numerous studies, including many
different experimental techniques and have dX=dt ¼ AexpðE=RTÞX
involved both isothermal and non-isothermal pro-
cedures. Techniques used have included thermo- in which X is a conversion parameter, expressed
gravimetry, differential scanning calorimetry, in terms of remaining unconverted fraction. In
differential thermal analysis, Rock-Eval, and var- many of these studies, a variety of different fitting
ious other types of reactors. These studies tracked techniques were employed (e.g., Friedman,
changes in at least one pyrolyzing kerogen-related Coates-Redfern, and simple Arrhenius). The
component (parent material or product) as a func- point here is not to evaluate the different methods
tion of time and temperature. In many such stud- or studies one against the other but to provide a
ies, complex materials, such as pyrolysis oil, are rough sense of how much agreement there is in the
considered as single pseudo-components. There literature on pyrolysis kinetics.
are many papers that consider the mechanisms It is immediately apparent that despite the clus-
and kinetics of kerogen decomposition, and only tering of results on the graphs, the actual rates in
a sampling of these is offered here [6, 57, 80, 103, the temperature regions of most interest can
110–130]. Despite the existence of a relatively diverge from one another by significant amounts.
extensive literature on oil shale pyrolytic behav- In the case of the kukersite samples of Fig. 3a, the
ior, very few analyses and reviews are avail- difference near 400
C (1000/T[K] = 1.49) is
able that offer comparison between the available about two orders of magnitude. The spread in
information or comparison to coal and biomass reported values for Green River oil shale is not
research results. as great near 400
C (see Fig. 3b, which shows an
Different mechanisms, with various degrees order of magnitude spread), but there are still
of complexity, have been proposed to describe oil significant differences in reported activation ener-
shale pyrolytic behavior. Generally, these are gies that lead to differences of even greater mag-
based on tracking intermediate, primary, and sec- nitude near temperatures of interest. Likewise,
ondary pyrolysis products, often without a clear there is a significant disparity in the results
identification of all of the products in any of the reported for Rundle Australian oil shale kerogen
classes being tracked. Kinetic parameters have in Fig. 3c. Also, these results point out that there
been calculated by fitting experimental data to an can be significant differences between the kero-
assumed semiempirical model of the pyrolysis gens from different samples. The kerogen from
process. There is, thus, not surprisingly a large kukersite shows typically higher reactivity than is
Oil Shale Processing, Chemistry, and Technology 69

Oil Shale Processing, Chemistry, and Technology, [138], heavy dashed-dotted line [139], thin line with small
Fig. 3 Comparisons of kinetics of kerogen pyrolysis. black points [140], thin dotted line with crosses [134], thin
(a) A comparison of rates on Estonian kukersite kerogens: dashed-dotted line with circles [141], and thin dashed line
thick lines [131–133], thin lines from Oja (unpublished [142]. (c) A comparison of rates obtained on kerogens
results), lines with crosses [134], and line with small circles from Rundle Australian oil shale: heavy line with crosses
[135]. (b) A comparison of rates obtained on Colorado [143], thin line [134], heavy dashed line [142], and dotted
Green River oil shale: dark line with crosses [136], thin line [144]
solid line [137], heavy dashed line [85], heavy dotted line

shown by the Green River or Rundle samples. The Nature and Role of Thermobitumen
Hence, there is not surprisingly a dependence of At many pyrolysis conditions, thermobitumen
kinetics on the origin of the shale. (also referred as pyrolytic bitumen or “metaplast”
This comparison of kinetics is not offered to in coal chemistry) is an important intermediate of
evaluate the relative merits of any one study over oil shale pyrolysis. The formation of the thermo-
another. It is to point out that there remain signif- bitumen has historically caused technological
icant unresolved differences in reported kinetics problems in retort design and development due
and that this is an issue that will seemingly require to its sticky and plugging character. The thermo-
further work to resolve. bitumen can be defined as an intermediate organic
70 Oil Shale Processing, Chemistry, and Technology

substance, which is organic solvent soluble but There are few studies on thermobitumen char-
nonvolatile at its formation temperatures. Sev- acterization, but generally its chemical, physical,
eral researchers define the bitumen more specif- and thermodynamic properties have been little
ically as a benzene-soluble pyrolysis intermediate studied. The properties and composition (elemental
[145]. The nonvolatile thermobitumen is used in composition, molecular weight distribution) of
models as a single pseudo-component that pre- thermobitumen is temperature and time depen-
cedes oil generation. In contrast to the thermo- dent. Figure 6 shows the average molecular weights
bitumen, the oil is defined as a volatile organic of oil and bitumen as a function of time and
product of the pyrolysis process. The kinetics of temperature for western US Colorado (Exxon-
thermobitumen formation have been described in Colony) and eastern US Kentucky (New Albany)
several papers, for Green River oil shale in [113, oil shales, and Table 7 presents thermobitumen
145] and for Kukersite oil shale in [80, 119, 133]. characteristics of Estonian kukersite oil shale.
The observation of thermobitumen formation The results of Table 7 show the expected decrease
has been incorporated into a number of mecha- in bitumen molecular weight as conversion pro-
nisms [6, 72, 146]. Generally pseudo-first-order gresses to greater extents at higher temperatures.
formation reactions are assumed. For example, a These and other available studies [80, 150] indi-
qualitative thermal degradation mechanism pro- cate that the thermobitumen relative to oil produced
posed by [147], and now fairly widely accepted in at same temperature has lower H/C atomic ratio (and
general form by many other researchers, is shown higher aromaticity) and higher number-average
in Fig. 4. It embodies the well-established feature molecular weight (depending on oil shale and time-
that there is a breakdown of kerogen to extract- temperature history, the thermobitumen can exceed a
able bitumen that precedes a significant amount value of 1500 Da). It was noted by Miknis and
of oil release. The thermobitumen formation, its Turner [145] that atomic H/C and number-average
composition, and yield are related to the time- molecular weights of the thermobitumen produced
temperature history of the process and to the oil were strongly time and temperature dependent,
shale kerogen structure. It is now more generally while oil showed remarkably constant values at
believed that it is the second stage of volatiles all times and temperatures. These results on bitu-
formation (from the thermobitumen) that gives men are in general agreement with the results of
rise to most of the oil product. Light gases may [47] on Colorado oil shale, which showed that this
be released in both volatiles’ release periods. fraction had a very high molecular weight
Figure 5 shows the time and temperature (a significant fraction >1000 Da). In this latter
dependence of thermobitumen yields for western study, the oil itself had a higher number-average
US high oil yield Colorado oil shales (Type I molecular weight than observed by Miknis and
kerogens) and eastern US low oil yield Kentucky Turner [145], presumably due to the difference in
New Albany oil shale (Type II kerogens). Figure 5 pyrolysis conditions, which favored the evapora-
indicates that low oil yield non-softening oil tion of higher molecular weight fractions.
shales cannot form considerable amounts of inter- In addition to direct pyrolysis (retorting), other
mediate pyrolytic bitumen. This has been explained technological solutions, such as supercritical gas
by cross-linking behavior of these oil shale kero- extraction, hydropyrolysis, or thermal dissolu-
gens and direct kerogen conversion to residue but tion with solvents, take advantage of the chemical
also by autocatalytic reactions between bitumen processes that occur at thermobitumen forma-
and kerogen [145]. tion stage.

Kerogen Pyrolytic bitumen Solid residue

Volatiles Volatiles

Oil Shale Processing, Chemistry, and Technology, Fig. 4 A generally accepted mechanistic scheme for pyrolysis of
kerogen
Oil Shale Processing, Chemistry, and Technology 71

Oil Shale Processing, Chemistry, and Technology, Fig. 5 Time and temperature dependence of pyrolytic bitumen
for Colorado and Kentucky oil shales (Ref. [145], with permission)

Oil Shale Processing,

Chemistry, and
Technology,
Fig. 6 Molecular weights
of oil and bitumen versus
time and temperature for
(a) Colorado (Exxon-
Colony) and (b) Kentucky
(New Albany) oil shale
(Ref. [145], with
permission)
72 Oil Shale Processing, Chemistry, and Technology

Oil Shale Processing, Chemistry, and Technology, 82.1%; H, 10.63%; S, 0.75%; O + N, 6.5%; and atomic
Table 7 Thermobitumen characteristics of kukersite oil H/C, 1.55) from thermochemical conversion in a Fischer
shale (elemental composition of organic matter C, retort [148, 149]

Process conditions Yield, % thermobitumen per Elemental composition, % Mol.

(temperature, time) organic matter C H S O+N H/C weight
275
C, 456 h 17.9 83.0 10.1 0.50 6.7 1.46 710
300
C, 387 h 69.8 85.1 8.77 0.32 5.8 1.24 1300
340
C, 12 h 59.8 83.8 9.78 0.52 5.9 1.40 1240
360
C, 4 h 63.8 83.6 9.56 0.50 6.3 1.37 790
380
C, 4 h 35.6 85.3 8.85 0.53 5.3 1.25 651

Competitive Reaction Processes and Hydrogen Technological Aspects of Shale Oil

An important concept in coal thermochemical Recovery
decomposition is the formation of volatile and
nonvolatile reactive radical components as a The fundamental chemical and heat and mass
result of bond breaking in the macromolecular transfer processes in oil shale retorting have
structure. Reactions between radical components been described to a limited degree in comparison
can lead to char formation, if the radical species to similar coal and biomass processes. By com-
involved are located on large fragments of the parison, there is a rich literature on practical oil
network structure. The stabilization of radicals shale retorting technologies (including flow
by hydrogen can result in volatiles (or potential sheets, process descriptions, and product charac-
volatiles), if the hydrogen capping of the struc- terizations) [14, 71, 72]. There are also reviews/
tures prevents their incorporation into a new net- reports covering processes from laboratory scale
work structure. through process development unit scale to com-
On the other hand, the formed volatiles can mercial scale: [69, 86, 107, 151–155]. The above
undergo secondary reactions prior to escaping books and reviews also cover a number of other
the reacting particles. This may happen in con- important topics associated with oil shale pro-
tact with solid material or in the vapor phase. cessing, such as availability of oil shale reserves
Such secondary reactions are promoted by long [60, 69, 107], oil shale mining [69, 72], research
residence times in high-temperature zones and development needs [60, 69], socioeconomical
within a particle or a reactor. Thermochemical problems [69], environmental impact [60, 69,
conversion in the presence of hydrogen can 107], oil shale upgrading economics and shale
result in an increase in non-polymeric volatile oil cost components [12], shale oil refining [20,
hydrocarbons, if a radical secondary reaction 69] and future perspectives [20, 107].
pathway can be intercepted by hydrogen cap- Over hundreds of years of oil shale develop-
ping of a radical. In addition to helping reduce ment, hundreds of retorts (retorting processes)
the macromolecular structure to smaller species have been invented [69]. Through the long history
by the above routes, hydrogen is also consumed of oil shale utilization, ex situ (above the ground
in reducing sulfur, nitrogen, and oxygen that or surface) retorting technologies have been prac-
exist in the kerogen structure [14]. Hydrogen tically the only commercial technology for oil
can be introduced into these roles either directly production from oil shale. In ex situ processing,
from the gas phase or from a hydrogen-donor the oil shale must be mined (using surface mining
solvent [14]. There are effects of temperature, or underground mining), crushed, and transported
pressure, and catalysts on these hydrogen reac- to the processing plant for retorting. By compari-
tion pathways [14, 40]. son, in situ processes cause less land surface
Oil Shale Processing, Chemistry, and Technology 73

disturbance associated with mining. This may 1. By conduction through a retort wall. To this
involve limited excavation to access oil shale group belonged ex situ technologies such as
sources or to create permeability by fracturing or the vertical Pumpherston retort (Bryson retort)
drilling heater holes or production wells. Because [156] or Davidson horizontal rotary retort
the oil shale is retorted in the ground, in situ [157]. In these types of ex situ technologies,
processing requires that the produced oil be extra- heat is transferred through the retort wall. Due
cted through an oil well in the same way as is to the slow conduction that characterizes such
conventional crude oil [152]. Therefore, com- retorts, many such processes are mainly of
monly cited advantages of in situ process are historical interest as their throughputs per unit
limited mining and solids’ (oil shale and spent of vessel volume are low. Development phase
shale) handling, and no expense associated with processes of this genre are the Oil-Tech elec-
aboveground retorting equipment. The disadvan- trically heated vertical retort process (Ambre
tage of in situ processing involves the complexity Energy Limited) [155] and the A.F.S.K. Hom
in controlling process parameters. There are also Tov co-pyrolysis process (Mishor Rotem).
concerns raised regarding the ability to capture all The principle of wall conduction can also be
of the oil produced underground. applied to in situ technologies. In that case, it
In addition to the issues of technical viability involves placing heating elements or heating
(including process controllability) and capital cost, pipes into the oil shale deposit. Representative
the choice to use ex situ as opposed to in situ of this approach is the Shell in situ conversion
processes depends on factors such as richness of process (ICP) technology (Shell Mahogany
the oil shale, the depth or thickness of the shale Project) [155], which uses electric-resistance
bed, the accessibility to deposits, and environmen- heaters combined with a surrounding “freeze
tal considerations (e.g., associated with escape of wall” to confine the retorting products. Another
oil into groundwater or extraction of materials example is the EGL oil shale process [155] in
from spent shale). Again, ex situ processes are which heat is transferred by superheated steam
often also regarded as environmentally problem- (in a closed system) through a series of pipes
atic beyond their mining impacts, due to problems serving as heat exchangers below the oil shale
of solid residue disposal and potential for ground- bed. The Geothermic Fuel Cell™ (GFL™)
water pollution from that source. approach developed by Independent Energy
Besides the simple ex situ and in situ classifi- Partners, Inc. (IEP) uses a high-temperature
cations, there are many other retorting classifica- fuel cell stack [155].
tion schemes used in the literature as technologies 2. By gases from shale combustion in the retort
can differ in various aspects such as process (referred to as the directly heated retorts
principle. These include process modes such as approach). These technologies utilize heat
batch, semi-batch, or continuous, as well as pro- that originates from burning materials (char
cess flows including cocurrent, countercurrent, or and oil shale gas) inside the retort and is there-
crosscurrent. An important classification is based fore also called internal combustion retorting.
upon heat transfer mechanism to the shale (direct The heat is transferred through the retort by
or indirect, and what the carrier is in direct) as well gases generated in the combustion process. In
as heat generation method. In addition, retort combustion retorts the retorting gas is diluted
style, complexity of technology, or by-product and has low calorific value (for syngas classi-
processing can be bases of classification. fication depending on the heating values, see
A vitally important design consideration is [60]). A currently used commercial ex situ
always thermal management, both in terms of retorting technology of this type is the Fushun
heat generation and heat transfer to the shale to process [86] in China. Other examples of thor-
be retorted. Heat can be transferred to oil shale by oughly investigated ex situ technologies based
different means: upon this approach are the Kiviter process
74 Oil Shale Processing, Chemistry, and Technology

(early versions), Nevada-Texas-Utah (NTU), accomplished by mixing with spent shale/

Paraho Direct, Union “A,” and Superior multi- sand) [60], the Shell Spher process (Shell Pel-
mineral processes [60, 72]. In situ retorting let Heat Exchange Retorting) [159], and the
also utilized this principle in technologies that Lawrence Livermore National Laboratory Hot
burned oil shale underground. An example of Recycled Solid (LLNR HRS) process [152].
this was the Occidental Petroleum Corporation 5. By other means such as radio frequency [151],
Oxy process [69] (modified in situ combustion microwave [160], solar [161] or nuclear explo-
retorting process). sion (e.g., Project Bronco) [72], etc. A number
3. By gaseous heat carrier heated outside retort of recent shale oil in situ recovery schemes
(referred as indirectly heated retorts). In this (e.g., Raytheon/Schlumberger; PyroPhase)
case, heat is transferred by hot gases heated in have been proposed based on volumetric
an outside furnace and then injected into the heating by means such as radio frequency or
retort (or retorting bed). Again, in these sys- microwave heating [155]. These have thus far
tems, the retorting gas is diluted and has low been mostly of interest on the research and
calorific value. Currently used commercial ex development level [151].
situ technologies of this group are the current 6. By combinations of the above. For example, a
Kiviter process in Estonia and Petrosix process combination of solid and gaseous heat carriers
in Brazil [72]. Other main ex situ technology is utilized in hot recycle solids fluidized bed
examples are Union “B,” Paraho Indirect, retorting systems (e.g., KENTORT II, [16,
Superior Indirect [72], and fluidized bed- 162]). Since gas density is about 1000 times
based processes. Similarly, in situ technologies lower than that of oil shale, assuming that
can utilize gases that are heated aboveground roughly same mass heat carrier must be in
(such as by combustion), which are then contact with the shale, then a large volume of
injected into the oil shale formation as heat gas is needed to heat the oil shale [113]. The
carriers. Examples of such hot gas injecting KENTORT II [16, 162] uses recirculating hot
methods are the Chevron CRUSH process spent shale from a gasification section.
and the Petro Probe process and the in situ
gas extraction technology (IGE) of Mountain The abovementioned Shell Spher process and
West Energy [155]. LLNR HRS process could arguably also be
4. By solid heat carrier heated outside the retort included in this class of processes.
(referred indirectly as heated retort). In this An example of an in situ combined heat trans-
instance, heat is transferred by mixing hot fer mode process could be Electrofrac™ (Exxon
solid particles (heated outside the retort) with Mobil Corporation) [155], which combines wall
the oil shale. Reactive heat carriers, such as oil conduction and volumetric heating through elec-
shale ash, or nonreactive, such as sand or trically conductive material (used to fill fractures).
ceramic balls, have been used. In these pro- There are a large number of different retorting
cesses retort gases with high calorific value are solutions. Generally, the performance of retorts
produced. These are ex situ processes that typ- (in situ or ex situ reactors) are typically evaluated
ically take place in rotating kiln-type retorts. or compared based on the percentage of Fischer
Currently used commercial ex situ technology assay yield achieved and an overall thermal effi-
of this type includes the Galoter process [158] ciency (as energy efficiency). Typical commercial
in Estonia, which utilizes hot shale ash as the scale retorts, with the exception of fluidized bed
heat-carrying medium. Other main technolo- retorts, have lower oil yields than obtainable from
gies of this type include the Alberta Taciuk Fischer assay. For example, for modified in situ
Process (ATP), TOSCO II (developed by the internal combustion retorting, it is around 60%
Oil Shale Corporation, utilizing externally [14] of Fischer assay. For a Fushun generator
heated ceramic balls 15 mm in diameter as type of retort, this value is 65% [86] and 70%
heat carrier), Lurgi-Ruhrgas (retorting is for the Kiviter type of retort [86]. Higher values
Oil Shale Processing, Chemistry, and Technology 75

are obtained with Petrosix retorting, 85–90% [86]; Oil Shale Processing, Chemistry, and Technology,
Galoter retorting, 85–90% [86]; and Taciuk Table 8 Some selected ex situ retorting technologies
retorting, 82% [86]. Process/retort References
The thermal efficiency is defined as heat of Fushun retort [86]
desired product divided by the total heat of the Brazilian Petrosix [60, 69]
process [14]. For example, some comparisons The Nevada-Texas-Utah retort or NTU [69, 72]
(also known as the Dundas-Howes
may involve using the equation
retort)
The Paraho direct/indirect process [14, 69]
Thermal efficiency
Gross Galorific value of ðoil þ gasÞ The Union Oil retort “A” (Union Oil [6, 38, 60,
¼ Company) 72]
Gross Galorific value of shale þ heat of retorting
The Union Oil retort “B” indirectly [6, 38, 60,
heated retort (Union Oil Company) 69, 72]
which can be found in Probstein and Hicks [14]. The Union Oil Steam Gas Recirculation [72]
More detailed evaluation of processes comes (RSG) retorting process (Union Oil
down to desired product yields (oil and gas) and Company)
energy consumption per unit of desired product The Superior Oil process [69, 72]
produced, including additional energy require- TOSCO II (Tosco corporation) [14, 60, 69]
ment such as for gas purification, spent shale The Lurgi-Ruhrgas process [6, 60, 69,
72, 107]
disposal, or water treatment.
Kiviter process [38, 72, 86]
References offering insights regarding the con-
Galoter process [38, 72, 86,
struction details and operational principles of the 158]
retorting equipment referred to the above, includ- AOSTRA Taciuk Process (ATP) [72, 86, 107]
ing product characteristics, are found in Tables 8 Chamber oven (similar to coke [38]
and 9. As there are hundreds of retorting options batteries)
that have been considered, this is far from being HYTORT (Institute of Gas Technology, [12, 17]
a complete list of possibilities. Nonetheless, it IGT)
offers some insight into some significant processes. Chattanooga process (pressurized fluid [155]
bed reactor)
The next sections offer a few more details on
Gas combustion retort, US Bureau of [6, 38, 60,
selected ex situ and in situ processes. Mines 72, 107]
Oil-Tech vertical surface retort [155]
technology (Millennium Synfuels, Llc)
Ex Situ Technologies: An Example and some
KENTOR II [162]
Further Considerations Shell Spher processes (Shell Pellet Heat [159]
There are many ex situ retorting technologies that Exchange Retorting)
differ by reactor types (fluidized bed, moving
packed bed, solid mixers), by operating condi-
tions (temperature, residence time), and by tech- horizontal rotary kiln-type retort, with contact
nical details, such as drying of feed, feed of hot oil shale ash obtained from solid retorting
distributor, discharge of spent shale, oil and gas residue combustion outside of the retort. The
separation (oil and gas separation from off-gas), process operates in continuous mode with oil
heat recovery, or by-product utilization. Therefore shale and ash in cocurrent flows. The properties
in addition to the pyrolysis reactor, the heart of of the shale oil produced (specific gravity, H/C
retorting process, a particular retorting technology ratio, viscosity, boiling range, and hydrocarbon
includes various other components (unit opera- types) depend upon retorting temperature and
tions) in the plant. residence time, of course, but can also depend
Figure 7 shows a principal schematic diagram on other factors such as condenser construction
of the Galoter process [158]. The principle of and temperature of recovery and separation of
the Galoter process is heating fine-grained oil shale oil and the reactive properties of the hot
shale (particles with size below 25 mm) in a shale residue heat carrier.
76 Oil Shale Processing, Chemistry, and Technology

Oil Shale Processing, Chemistry, and Technology, It needs to be emphasized again that an impor-
Table 9 Some selected in situ retorting processes and tant parameter in ex situ retorting design is the
technologies
heat up time of the retorted shale. This determines
Process/retort References overall process time and therefore reactor volume
Occidental Petroleum “Oxy” [60, 69, and thus retorting process cost [113] but also
72]
impacts retort oil quality. The time to transfer
Sinclair oil and gas company process [6, 60]
heat to shale depends on the feed oil shale particle
Equity Oil company process [6, 60]
Shell Oil’s thermally conductive in situ [60, 107,
size. Often, operational details require specific
conversion process 155] particle size ranges [72]. It has been indicated
Osborne’s in situ process [60] that [113] the heat up time of the oil shale particles
Dow Chemical Co.’s process [60, 72] larger than 1 cm scales roughly as diameter
Equity Oil company process [72] squared and thus a heat up time of 2.5 min can
Chevron’s Technology for the Recovery [155] be estimated for 10-mm particles and 4 h for
and Upgrading of Oil from Shale 100-mm particles. Accordingly, the ex situ tech-
(CRUSH) process
nologies have been divided in three groups, based
EGL oil shale process [155]
upon particle size:
Electrically enhanced oil production [155]
(EEOP) (Electro-Petroleum, Inc.)
PetroProbe process (Petro Probe, Inc.) [155] 1. Processes with large feed particle sizes
Electrofrac™ (Exxon Mobil Corporation) [155] (so-called lump oil shale). These processes
Geothermic Fuel Cell™ technology [155] take place over the longest timescale; the
(GFL™) shale residence time in retorts is measured in
In situ gas extraction technology (IGE), [155] hours. For example, the industrial Kiviter pro-
Mountain West Energy
cess [72] uses a feed size of 10–125 mm and
Radio frequency/critical fluid oil [155]
extraction technology (Raytheon and CF
the Fushun process of 10–75 mm [11].
technology) 2. Processes that use coarsely ground particles.
EchoShale in-capsule process (Red Leaf [155] The shale residence timescale is of minutes.
Resources, Inc.) For example, the industrial Galoter process
PyroPhase [155] [72] uses a feed size of 0–25 mm [11].
American Shale Oil Company (AMSO) [155] 3. Processes fed with pulverized oil shale with
process: Conduction, Convection, Reflux
size below 5 mm. The residence time is shortest,
(CCR) process (former EGL Resources
process) with the timescale of seconds. This is used in
fluidized bed retorting units [164]. The fluid-
ized bed retorting systems are less developed
A heat transfer medium (the heat carrier) can than the two former types, and no commercial
generally be reactive or nonreactive. Examples of units are yet in use.
reactive heat carriers include the gaseous heat
carriers such as hydrogen or solid heat carriers In Situ Technologies: Further Considerations
such as oil shale ash. Oil shale ash from carbonate The common principle of in situ processes is
oil shales has chemisorptive properties, due to heating the oil shale deposit underground. Differ-
its alkaline nature, and can reduce the concentra- ent versions use different methods of heat transfer,
tion of H2S in the product gas as well as phenolic as already described. With in situ technologies,
compounds in the oil [23]. Also the iron oxide there is an opportunity to utilize poorer and deeper
in ash shows potential to reduce H2S in the gas oil shale deposits than can be used for conven-
but has little effect on phenolic compounds tional aboveground processing [165]. Although
[23]. It should, however, be emphasized that heat a number of field-scale experiments have been
carrier ash used in solid heat carrier retorts does conducted over the last decades in several coun-
not affect significantly the sulfur concentration tries [60, 69, 151, 153], no in situ technology-
in oils (i.e., organic sulfur) [23, 163]. based commercial plant has been placed in
Oil Shale Processing, Chemistry, and Technology 77

Ash separator Shale oil

Dry oil shale
separator vapors from oil
9
shale pyrolysis
Waste heat
By-pass
Boilers
boiler 3
of power
Electric station
2 AFC
filter
Heat carrier
separation Dust removal 15
5 8
chamber
14
6 Gas and oil vapors
separation

7 Dry oil shale

endlessscrew
AFD
Mixer
DP-1
13 Reactor
4
SP
Semicoke

Pulp tank
1
DP
P-1
G-1 12
10
11

Oil Shale Processing, Chemistry, and Technology, for oil shale pyrolysis process; 4, raw oil shale feed; 5, ash
Fig. 7 Galoter retorting process (Ref. [158], with permis- separated by first-, second-, and third-stage ash cyclones;
sion). The key features of this process are the “Reactor,” 6, dried oil shale mixture with stack gas after AFD; 7, dried
the drum reactor in which oil shale pyrolysis takes place, oil shale separated by first-, second-, and third-stage
after mixing of raw and hot spent shale; “AFC,” cyclones prior to entering the system of dry oil shale
aerofountain combustor for spent shale (semicoke) burn- screw conveyers; 8, retort gas for firing in power station
out; “AFD,” aerofountain dryer for raw shale; P-1, centrif- boilers; 9, stack gas to ESP for final purification; 10, ash
ugal air blower; DP, dredger pump; DP-1, small dredger slurry for storage/disposal; 11, shale oils to storage; 12,
pump of electric filters; SP, pump for settled recycle water; light oil or gasoline fraction (separated from water) to
G-1, centrifugal gas blower; 1, compressed air given to the storage; 13, recycle water to ash hydro removal system;
retort; 2, solid heat carrier mixture with stack gas after 14, phenolic water to incineration in power station boilers;
combustion of semicoke in the AFC; 3, mixture of ash 15, oil vapors from retort, after cyclone separators, to
and stack gas after separation of solid heat carrier required condensation system

operation to date. However, several of the latest situ processes uses natural rock permeability or
innovative technological approaches involve in permeability created by artificial fracturing
situ processes [155]. (by explosives or hydrostatic pressure) in place.
There are different possible ways of classify- The true in situ processes follow a general
ing in situ technologies (see [60]). The most sequence involving dewatering and establishing
common is classification into true in situ pro- measures to minimize contact between groundwa-
cesses and modified in situ processes. Some ter and the reaction zone; fracturing or rubblizing
define the true in situ processes as involving to create permeability (if natural permeability is
minimal disturbance of the oil shale ore bed and not sufficient); heating the oil shale by ignition,
the modified in situ processes as any that use hot fluids, or other means; and recovering oil and
rubblizing either through blasting or partial min- gas through wells [69].
ing. Therefore the distinction between true and The modified in situ processes involve partial
modified in situ processes is a matter of degree of mining under the oil shale deposit to create void
the bed disturbance. space followed by rubblizing the rest into the
True in situ processes thus do not generally created void. To create voids about 10–25% of
require significant mining. This category of in overall space is mined out [60].
78 Oil Shale Processing, Chemistry, and Technology

Final Considerations and Future have been studied, but the general impression is
Directions that more work is needed to better establish gen-
eral trends. Likewise, there have been an enor-
This review was intended to capture many aspects mous number of studies performed on elucidating
of the current state of knowledge and basic the kinetics of pyrolysis, but rarely has there been
research directions in the field of oil shale a critical examination of results obtained on the
research. It did so in the context provided by same material examined by different means in
currently and historically explored technologies different laboratories. As a result, there is a
for oil shale conversion. Generally speaking, crude understanding of rate processes, but one
there is more new development interest in oil that for predictive purposes cannot be claimed to
production processes, but the direct combustion be better than an order of magnitude. Despite large
processes will remain in use in particular instances numbers of carefully performed studies, there has
(such as in Estonia). While the pyrolytic phenom- not yet been closure on the overall nature and rates
ena that shale undergoes are common to both oil of the many processes that determine behavior in
production and combustion processes, there has practical processes.
not been as much emphasis on fundamental
exploration of these phenomena from the combus-
tion side, and in that case, there are other techno-
Bibliography
logically important issues that have captured more
basic research interest (e.g., the behavior of ash in
Primary Literature
the boilers, the high-temperature corrosion of 1. Cane RF (1976) The origin and formation of oil
boilers, the problems of ash handling and dis- shale. In: Yen TF, Chilingarian GV (eds) Oil shale.
posal, the problems of sulfur capture). For the Elsevier, Amsterdam
combustion community, the behavior of the 2. Duncan DC (1976) Geologic settings of oil shale
deposits and world prospects. In: Yen TF,
organic portion of the ash has been somewhat Chilingarian GV (eds) Oil shale. Elsevier,
secondary, and even the final phases of organic Amsterdam
materials’ burnout, which has captured the atten- 3. Hutton AC (1995) Organic petrography of oil shales.
tion of many researchers in the coal combustion In: Snape C (ed) Geochemistry and conversion of oil
shales. NATO ASI series. Kluwer, The Netherland
community, has been of less interest, because of 4. Durand B (1980) Kerogen. Insoluble organic matter
the inherently higher reactivity of oil shale chars from sedimentary rocks. Technip, Paris
as compared to coal chars. 5. Hutton AC (1987) Petrographic classification of oil
From the perspective of those who seek to shales. Int J Coal Geol 8:203–231
6. Lee S (1991) Oil shale technology. CRC, Boca Raton
produce oil, the major problems have been those 7. Urov K, Sumberg A (1999) Characteristics of oil
of thermal management and the handling of large shales and shale like rocks of known deposits and
amounts of solids, as well as dealing with a range outcrops. Oil Shale 16:1–64
of environmental problems associated with the 8. Gavin JM (1924) Oil shale. U.S. Government Print-
ing Office, Washington, DC
processing (whether aboveground or in situ). 9. Ozerov IM, Polozov VI (1968) Principles of oil shale
What can be economically done with oil shale in commercial classification. In: United Nations sym-
light of the above has dictated the nature of the posium on the development and utilization of oil
processes more than the consideration of the shale resources, USSR All-Union Research Institute
of Petroleum Chemistry, Tallinn
chemical aspects. Nonetheless, it has been observed 10. Schlatter LE (1968) Definition, formation and classi-
that there can be factors whose importance may fication of oil shales. In: United Nations symposium
be sometimes overlooked. Issues of mass trans- on the development and utilization of oil shale
port can materially impact both yields and quality resources, Tallinn
11. EASAC (European Academies Science Advisory
of the product oils and may be sensitive to factors Council) (2007) A study on the EU oil shale
such as pressure, particle size, heating rate, tem- industry – viewed in the light of the Estonian experi-
perature, and residence time in a hot environment ence. European Academies Science Advisory Coun-
and gaseous environment. All of these factors cil, Halle (Saale), Germany
Oil Shale Processing, Chemistry, and Technology 79

12. Ekinci E, Yürüm Y (1995) Steam and coprocessing 30. Rullkötter J, Michaelis W (1990) The structure of
of oil shales. In: Snape C (ed) Composition, geo- kerogen and related materials. A review of recent
chemistry and conversion of oil shales. NATO ASI progress and future trends. Org Geochem
series. Kluwer, The Netherland 16:829–852
13. Miknis FP (1995) Solid state 13C NMR in oil shale 31. Vandenbroucke M (2003) Kerogen: from types to
research: an introduction with selected applications. models of chemical structure. Oil Gas Sci Technol
In: Snape C (ed) Composition, geochemistry and 58:243–269
conversion of oil shales. NATO ASI series. Kluwer, 32. Vandenbroucke M, Largeau C (2007) Kerogen ori-
The Netherland gin, evolution and structure. Org Geochem
14. Probstein RF, Hicks RE (2006) Synthetic fuels. 38:719–833
Dover, New York 33. Lille Ü, Heinmaa I, Pehk T (2003) Molecular model
15. Hamarneh YM (1984) Oil shale deposits in Jordan. of Estonian kukersite kerogen evaluated by 13C MAS
Symposium on characterization and chemistry of oil NMR spectra. Fuel 82:799–804
shales. Am Chem Soc 29(3):41–50 34. Durand B, Monin JC (1980) Elemental analysis of
16. Carter SD, Graham UM, Rubel AM, Robl TL (1995) kerogens (C, H, O, N, S, Fe). In: Durand B (ed)
Fluidized bed retorting of oil shale. In: Snape C (ed) Kerogen. Insoluble organic matter from sedimentary
Geochemistry and conversion of oil shales. NATO rocks. Technip, Paris
ASI series. Kluwer, London 35. Arbiter N (1983) Concentration of oil shale. Miner
17. Roberts MJ, Snape CE, Mitchell SC (1995) Process Technol Rev 1:207–248
Hydropyrolysis: fundamentals, two-stage processing 36. Durand B, Nicaise C (1980) Procedures of kerogen
and PDU operations. In: Snape C (ed) Geochemistry isolation. In: Durand B (ed) Kerogen. Insoluble
and conversion of oil shales. NATO ASI series. organic matter from sedimentary rocks. Technip,
Kluwer, The Netherland Paris
18. Speight JG (2007) The chemistry and technology of 37. Robl TL, Taulbee DN (1995) Demineralization and
petroleum, 4th edn. CRC, Boca Raton kerogen macerals separation and chemistry. In:
19. Prien CH (1976) Survey of oil shale research in the Snape C (ed) Geochemistry and conversion of oil
last three decades. In: Yen TF, Chilingarian GV (eds) shales. NATO ASI series. Kluwer, London
Oil shale. Elsevier, Amsterdam 38. Yen TF, Chilingarian GV (1976) Introduction to oil
20. Speight JG (2008) Synthetic fuels handbook. Prop- shales. In: Yen TF, Chilingarian GV (eds) Oil shale.
erties, processes and performance. McGraw-Hill, Elsevier, Amsterdam
New York 39. Forsman JP (1963) Geochemistry of kerogen. In:
21. World Energy Council (2007) Survey of energy Breger IA (ed) Organic geochemistry. Pergamon,
resources 2007. World Energy Council 2007, Lon- Oxford, pp 148–182
don, UK 40. Robinson WE (1969) Isolation procedures for kero-
22. Puura E (1999) Technogenic minerals in the waste gens and associated soluble organic materials. In:
rock heaps of Estonian oil shale mines and their use Eglinton G, Murphy MTJ (eds) Organic geochemis-
to predict the environmental impact of the waste. Oil try. Springer, Berlin, pp 181–195
Shale 16:99–107 41. Thomas RD, Lorentz PB (1970) Use of centrifugal
23. Oja V, Elenurm A, Rohtla I, Tali E, Tearo E, separation to investigate how kerogen is bound to the
Yanchilin A (2007) Comparison of oil shales from minerals in oil shale. U. S. Bur Mines Rep Invest
different deposits: oil shale pyrolysis and copyrolysis 7378
with ash. Oil Shale 24:101–108 42. Gonzalez-Vila FJ, Ambles A, del Rio JC, Grasset L
24. Patterson HJ (1994) A review of the effects of (2001) Characterization and differentiation of kero-
minerals in processing of Australian oil shales. gens by pyrolytic and chemical degradation tech-
Fuel 73:321–327 niques. J Anal Appl Pyrolysis 58-59:315–328
25. Shpirt MY, Punanova SA, Strizhakova YA (2007) 43. Vitorovic D (1980) Structure elucidation of kerogen
Trace elements in black and oil shales. Solid Fuel by chemical methods. In: Durand B (ed) Kerogen.
Chem 41:119–127 Insoluble organic matter from sedimentary rocks.
26. Raudsepp H (1953) About the method for determin- Technip, Paris
ing of organic mass in Baltic oil shales. Proc Tallinn 44. Smith LK, Smoot LD, Fletcher TH, Pugmire RJ
Polytech Inst A 46:3–22. (in Russian) (1994) The structure and reaction processes of coal.
27. Krym YS (1932) About laboratory methods for deter- The plenum chemical engineering series. Plenum,
mination the tendency of selfignition for coals. Khim New York
Tverd Topl 2-3:7–22. (in Russian) 45. Solomon PR, Serio MA, Suuberg EM (1992) Coal
28. Yen TF (1976) Structural aspects of organic compo- pyrolysis: experiments, kinetic rates and mechanism.
nents in oil shales. In: Yen TF, Chilingarian GV (eds) Prog Energy Combust Sci 18:133–220
Oil shale. Elsevier, Amsterdam 46. Savest N, Hruljova J, Oja V (2009) Characterization
29. Hutton A, Bharati S, Robl T (1994) Chemical and of thermally pretreated kukersite oil shale using
petrographic classification of kerogen/macerals. the solvent-swelling technique. Energy Fuel
Energy Fuel 8:1478–1488 23:5972–5977
80 Oil Shale Processing, Chemistry, and Technology

47. Suuberg EM, Sherman J, Lilly WD (1987) Product 63. Qian J, Wang J, Li S (2003) Oil shale development in
evolution during rapid pyrolysis of Green River for- China. Oil Shale 20:356–359
mation oil shale. Fuel 66:1176–1184 64. Hamarneh Y (2006) Oil shale resources development
48. Solomon PR, Carangelo RM, Horn E (1986) The in Jordan. Natural Resources Authority of Jordan,
effects of pyrolysis conditions on Israeli oil shale Amman. 1998. Revised and Updated by Dr. Jamal
properties. Fuel 65:650–662 Alali, Amman
49. Kelemen SR, Afeworki M, Gorbaty ML, Sansone M, 65. Alali J, Abdelfattah AS, Suha MY, Al-Omari W
Kwiatek PJ, Walters CC, Freund H, Siskin M, (2006) Oil shale. In: Sahawneh J, Madanat M (eds).
Bence AE, Curry DJ, Solum M, Pugmire RJ, Natural Resources Authority of Jordan
Vandenbroucke M, Leblond M, Behar F (2007) 66. Hotta A, Parkkonen R, Hiltunen M, Arro H,
Direct characterization of kerogen by x-ray and Loosaar J, Parve T, Pihu T, Prikk A, Tiikma
solid-state 13C nuclear magnetic resonance methods. T (2005) Experience of Estonian oil shale combus-
Energy Fuel 21:1548–1561 tion based on CFB technology at Narva power plants.
50. Schmitt KD, Sheppard EW (1984) Determination of Oil Shale 22:381–397
carbon center types in solid fuel by CP/MAS NMR. 67. Arro H, Prikk A, Pihu T (2003) Calculation of
Fuel 63:1241–1244 qualitative and quantitative composition of Estonian
51. Siskin M, Scouten CG, Rose KD, Aczel T, oil shale and its combustion products. Fuel
Colgrove SG, Pabst RE Jr (1995) Detailed structural 82(2179–2195):2197–2204
characterization of the organic material in rundle 68. Ots A (2006) Oil shale fuel combustion. Tallinn
Ramsay crossing and Green River oil shales. In: Book Printers, Tallinn
Snape C (ed) Composition, geochemistry and con- 69. Office of Technology Assessment (1980) An
version of oil shales. NATO ASI series. Kluwer, The assessment of oil shale technology. Congress of the
Netherland United States. Office of Technology Assessment,
52. Larsen JW, Li S (1994) Solvent swelling studies of Washington, DC
Green River Kerogen. Energy Fuel 8:932–936 70. Oja V (2006) A brief overview of motor fuels from
53. Larsen WJ, Parikh H, Michels R (2002) Changes in shale oil of kukersite. Oil Shale 23:160–163
the cross-linking density of Paris Basin Toarcian 71. Dinneen GU (1976) Retorting technologies of oil
kerogen during maturation. Org Geochem shales. In: Yen TF, Chilingarian GV (eds) Oil shale.
33:1143–1152 Elsevier, Amsterdam
54. Savest N, Oja V, Kaevand T, Lille Ü (2007) Interac- 72. Baughman GL (1978) Synthetic fuels data hand-
tion of Estonian kukersite with organic solvents: a book, 2nd edn. Cameron Engineering, Denver
volumetric swelling and molecular simulation study. 73. Bjerle I, Ecklund H, Svensson O (1980) Gasification
Fuel 86:17–21 of Swedish black shale in the fluidized bed. Reactiv-
55. Gonzalez-Vila FJ (1995) Alkane biomarkers. Geo- ity in steam and carbon dioxide atmosphere. Ind Eng
chemical significance and application in oil shale Chem Process Des Dev 19:345–351
geochemistry. In: Snape C (ed) Geochemistry and 74. Ecklund H, Svensson O (1983) Reactor model for the
conversion of oil shales. NATO ASI series. Kluwer, gasification of black shale in the fluidized bed. Com-
London parison with the pilot plant data. Ind Eng Chem
56. Vandenbroucke M (1980) Structure of kerogens Process Des Dev 22:396–401
as seen by investigations on soluble extracts. In: 75. Jaber JO (2000) Gasification potential of Ellajjun oil
Durand B (ed) Kerogen. Insoluble organic matter shale. Energy Convers Manag 41:1615–1624
from sedimentary rocks. Technip, Paris 76. Schachter Y (1979) Gasification of oil shale. Isr
57. Rajeshwar K, Nottenburg R, Dubow J (1979) J Technol 17:51–57
Thermophysical properties of oil shales. J Mater Sci 77. Perry RH, Green DW (1997) Perry’s chemical engi-
14:2025–2052. (Review) neering handbook, 7th edn. McGraw-Hill, New York
58. Maalmann I (2003) Historical survey of nuclear non- 78. Shui H, Cai Z, Xu C (2010) Recent advances in direct
proliferation in Estonia 1946–1995. Estonian Radia- coal liquefaction. Energies 3:155–170
tion Protection Centre, Tallinn, Estonia 79. Gorlov EG (2007) Thermal dissolution of solid fossil
59. Dyni JR (2006) Geology and resources of some fuels. Solid Fuel Chem 41:290–298
world oil shale deposits. Scientific investigations 80. Tiikma L, Johannes I, Luik H, Zaidentsal A, Vink N
report 2005–5294. United States Department of the (2009) Thermal dissolution of Estonian oil shale.
Interior, United States Geological Survey, Reston, J Anal Appl Pyrolysis 85:502–507
Virginia, United States 81. Koel M, Ljovin S, Hollis K, Rubin J (2001) Using
60. Lee S, Speight JG, Loyalka SK (2007) Handbook of neoteric solvents in oil shale studies. Pure Appl
alternative fuel technology. CRC, Boca Raton Chem 73:153–159
61. Bryant RS, Bailey SA, Stepp AK, Evans DB, Parli 82. Tucker JD, Masri B, Lee S (2000) A comparison of
JA, Kolhatkar AR (1998) Biotechnology for heavy retorting and supercritical extraction techniques on
oil recovery. No 1998.110 el-Lajjun oil shale. Energy Sources 22:453–463
62. Brendow K (2003) Global oil shale issues and per- 83. Rocha JD, Brown SD, Love GD, Snape CE (1997)
spectives. Oil Shale 20:81–92 Hydropyrolysis: a versatile technique for solid fuel
Oil Shale Processing, Chemistry, and Technology 81

liquefaction, sulphur specification and biomarker 101. Oja V (2005) Characterization of tars from Estonian
release. J Anal Appl Pyrolysis 40-41:91–103 Kukersite oil shale based on their volatility. J Anal
84. Kogerman PN, Kopwillem JJ (1932) Hydrogenation Appl Pyrolysis 74:55–60
of Estonian oil shale and shale oil. J Inst Pet Technol 102. Durand-Souron C (1980) Thermogravimetric analy-
18:833–845 sis and associated techniques applied to kerogens. In:
85. Cummins JJ, Robinson WE (1972) Thermal degra- Durand B (ed) Kerogen. Insoluble organic matter
dation of Green River Kerogen at 150
C to from sedimentary rocks. Technip, Paris
350
C. U.S. Bur Mines Rep Invest 7620 103. Skala D, Korica S, Vitorovic D, Neumann H-J (1997)
86. Qian J, Wang J (2006) World oil shale retorting Determination of kerogen type by using DSC and TG
technologies. In: International conference on oil analysis. J Therm Anal 49:745–753
shale: recent trends in oil shale, Jordan, 7–9 Nov 104. Derenne S, Largeau C, Casadevall E, Damste JSS,
2006 Tegelaar EW, Leeuw JW (1989) Characterization of
87. Espitalie J, Madec M, Tissot B (1980) Role of min- Estonian Kukersite by spectroscopy and pyrolysis:
eral matrix in kerogen pyrolysis: influence on petro- evidence for abundant alkyl phenolic moieties in an
leum generation and migration. AAPG Bull Ordovician, marine, type II/I kerogen. Org Chem
64:59–66 16:873–888
88. Sert M, Ballice L, Yüksel M, Saglam M (2009) Effect 105. Silbernagel BG, Gebhard LA, Siskin M, Brons G
of mineral matter on product yield and composition (1987) ESR study of kerogen conversion in shale
at isothermal pyrolysis of Turkish oil shales. Oil pyrolysis. Energy Fuel 1:501–506
Shale 26:463–474 106. Reynolds JG, Crawford RW, Burnham AK (1991)
89. Patterson JH, Dale LS, Chapman JF (1988) Analysis of oil shale and petroleum source rock
Partitioning of trace elements during the retorting of pyrolysis by triple quadrupole mass spectrometry:
Australian oil shales. Fuel 67:1353–1356 comparisons of gas evolution at the heating rate of
90. Kök MV (2008) Recent developments in the appli- 10
C/min. Energy Fuel 5:507–523
cation of thermal analysis techniques in fossil fuels. 107. US DOE (2004) Strategic significance of America’s
J Therm Anal Calorim 91:763–773 oil shale resources, vol II: oil shale resources, tech-
91. Davis JD, Galloway AE (1928) Low-temperature nology, and economics. U.S. Department of Energy,
carbonization of lignites and sub-bituminous coals. Washington, DC
Ind Eng Chem 20:612–617 108. Kilk K, Savest N, Yanchilin A, Kellogg DS, Oja V
92. Citiroglu M, Türkay S, Cepni ZI, Snape CE (2010) Solvent swelling of Dictyonema oil shale: low
(1996) Fixed bed pyrolysis and hydropyrolysis of temperature heat-treatment caused changes in swell-
an immature type I Turkish oil shale. Turk J Chem ing extent. J Anal Appl Pyrolysis 89:261–264
20:175–185 109. Solomon PR, Serio MA, Despande GV, Kroo E
93. Watson NC (1984) A modified gray-king assay method (1990) Cross-linking reactions during coal conver-
for small oil shale samples. Fuel 63:1455–1458 sion. Energy Fuel 4:42–54
94. Goodfellow L, Haberman CE, Atwood MT (1968) 110. Al-Ayed OS, Matouq M, Anbar Z, Khaleel AM,
Modified Fischer assay equipment, procedures and Abu-Nameh E (2010) Shale pyrolysis kinetics and
product balance determinations. In: Joint symposium variable activation energy principle. Appl Energy
on oil shale, tar sand, and related materials. American 87:1269–1272
Chemical Society. San Francisco National Meeting, 111. Allred VD (1966) Kinetics of oil shale pyrolysis.
2–5 April 1968 Chem Eng Prog 62:55–60
95. Behar F, Beaumont V, Penteado HL (2001) Rock- 112. Braun RL, Rothman AJ (1975) Oil shale pyrolysis:
Eval 6 technology: performances and developments. kinetics and mechanism of oil production. Fuel
Oil Gas Sci Technol 56:111–134 54:129–131
96. Lafargue E, Marquis F, Pillot D (1998) Rock-Eval 113. Burnham AK (1995) Chemical kinetics and oil shale
6 applications in hydrocarbon exploration, produc- processing design. In: Snape C (ed) Composition,
tion and soils contamination studies. Oil Gas Sci geochemistry and conversion of oil shales. NATO
Technol 53:421–437 ASI series. Kluwer, The Netherland
97. Peters KE (1986) Guidelines for evaluating petro- 114. Degirmenci L, Durusoy T (2002) Effect of heating
leum source rock using programmed pyrolysis. rate on pyrolysis kinetics of Gönyük oil shale. Energy
AAPG Bull 70:318–329 Sources 34:931–936
98. Rahman M, Kinghorn RF, Gibson PJ (1994) The 115. Degirmenci L, Durusoy T (2005) Effect of heating
organic matter in oil shales from the lowmead rate and particle size on the pyrolysis of Gönyük oil
basin, Qeensland, Australia. J Pet Geol 17:317–326 shale. Energy Sources 27:787–795
99. Yang HS, Shon HY (1985) Effect of reduced pressure 116. Dieckmann V, Schenk HJ, Horsfield B, Welte DH
on oil shale retorting. Ind Eng Chem Process Des (1998) Kinetics of petroleum generation and cracking
Dev 24:271–273 by programmed-temperature closed-system pyrolysis
100. Suuberg EM, Unger PE, Lilly WD (1985) Experi- of Toarcian Shales. Fuel 77:23–31
mental study on mass transfer from pyrolyzing coal 117. Ekstrom A, Callaghan G (1987) The pyrolysis kinet-
particles. Fuel 64:956–962 ics of some Australian oil shales. Fuel 66:331–337
82 Oil Shale Processing, Chemistry, and Technology

118. Huss EB, Burnham AK (1982) Gas evolution during 136. Braun RL, Burnham AK (1986) Kinetics of Colorado
pyrolysis of various Colorado oil shales. Fuel oil shale pyrolysis in a fluidized bed reactor. Fuel
66:1188–1196 65:218–225
119. Johannes I, Zaidentsal A (2008) Kinetics of low- 137. Li S, Yue C (2003) Study of pyrolysis kinetics of oil
temperature retorting of kukersite oil shale. Oil shale. Fuel 82:337–342
Shale 25:412–425 138. Campbell JH, Koskinas GH, Stout ND (1978) Kinet-
120. Johannes I, Tiikma L (2004) Kinetics of oil shale ics of oil generation from Colorado oil shale. Fuel
pyrolysis in an autoclave under non-linear increase 57:372–376
of temperature. Oil Shale 21:273–288 139. Wang CC, Noble RD (1983) Composition and kinet-
121. Johannes I, Tiikma L, Zaidenstal A, Luik L (2009) ics of oil generation from non-isothermal oil shale
Kinetics of kukersite low-temperature pyrolysis in retorting. Fuel 62:529–533
autoclaves. J Anal Appl Pyrolysis 85:508–513 140. Sohn HY, Yang HS (1985) Effect of reduced pressure
122. Kök MV (2002) Oil shale: pyrolysis, combustion, on oil shale retorting. 1. Kinetics of oil generation.
and environment: a review. Energy Sources Ind Eng Chem Process Des Dev 24:265–270
24:135–143 141. Burnham AK, Braun RL, Coburn TT, Sandvik EI,
123. Strizhakova YA, Usova TV (2008) Current trends in Curry DJ, Schmidt BJ, Noble RA (1996) An appro-
the pyrolysis of oil shale: a review. Solid Fuel Chem priate kinetic model for well-preserved algal kero-
24:197–201 gens. Energy Fuel 10:49–59
124. Thankur DS, Nutall HE (1987) Kinetics of pyrolysis 142. Nutall HE, Guo TM, Schrader S, Thakur DS (1983)
of Moroccan oil shale by thermogravimetry. Ind Eng Pyrolysis kinetics of several key world oil shales. In:
Chem Res 26:1351–1356 Miknis FP, McKay JF (eds) Geochemistry and chem-
125. Torrente MC, Galán MA (2001) Kinetics of the ther- istry of oil shales. ACS symposium series, vol 230.
mal decomposition of oil shale from Puertollano pp 269–300
(Spain). Fuel 80:327–334 143. Wall GC, Smith SJC (1987) Kinetics of production of
126. Wang Q, Liu H, Sun B, Li S (2009) Study on pyrol- individual products from the isothermal pyrolysis of
ysis characteristics of Huadian oil shale with iso- seven Australian oil shales. Fuel 66:345–350
conversional method. Oil Shale 26:148–162 144. Parks TJ, Lynch LJ, Webster DS (1987) Pyrolysis
127. Wang Q, Sun B, Hu A, Bai J, Li S (2007) Pyrolysis model of rundle oil shale from in-situ 1H NMR data.
characteristics of Huadian oil shales. Oil Shale Fuel 66:338–344
24:147–157 145. Miknis FP, Turner TF (1995) The bitumen interme-
128. Williams PT, Ahmad N (2000) Investigation of oil diate in isothermal and nonisothermal decomposition
shale pyrolysis processing conditions using thermo- of oil shales. In: Snape C (ed) Composition, geo-
gravimetric analysis. Appl Energy 66:113–133 chemistry and conversion of oil shales. NATO ASI
129. Wiser WH, Anderson LL (1975) Transformation of series4. Kluwer, The Netherland
solids to liquid fuels. Annu Rev Phys Chem 146. Miknis FP, Turner TF, Berdan GL, Conn PJ (1987)
26:339–357 Formation of soluble products from thermal decom-
130. Xue HQ, Li SY, Wang HY, Zheng DW, Fang CH position of Colorado and Kentucky oil shales.
(2010) Pyrolysis kinetics of oil shale from northern Energy Fuel 1:477–483
Songliao Basin in China. Oil Shale 27:5–16 147. Aarna AY (1954) Isothermal destruction of Baltic oil
131. Johannes I, Kruusment K, Veski R (2007) Evaluation shale. Trans Tallinn Polytech Inst A 57:32–34.
of oil potential and pyrolysis kinetics of renewable (in Russian)
fuel and shale samples by rock-Eval analyzer. J Anal 148. Kask KA (1955) About bituminization of kerogen of
Appl Pyrolysis 79:183–190 oil shale kukersite. Report I. Trans Tallinn Polytech
132. Johannes I, Kruusment K, Veski R, Bojesen-Koefoed JA Inst A 63:51–64. (in Russian)
(2006) Characterisation of pyrolysis kinetics by rock- 149. Kask KA (1956) About bituminizing of kerogen of
Eval basic data. Oil Shale 23:249–257 oil shale-kukersite. Report II. Trans Tallinn Polythec
133. Johannes I, Tiikma L, Zaidenstal A (2010) Compar- Inst A 73:23–40. (in Russian)
ison of the thermobituminization kinetics of Baltic 150. Tiikma L, Zaidentsal A, Tensorer M (2007) Forma-
oil shales in open retorts and autoclaves. Oil tion of thermobitumen from oil shale by low-
Shale 27:17–25 temperature pyrolysis in an autoclave. Oil Shale
134. Petersen HI, Bojesen-Koefoed JA, Mathiesen A 24:535–546
(2010) Variations in composition, petroleum 151. Bridges JE (2007) Wind power energy storage for in
potential and kinetics of Ordovician-Miocene type situ shale oil recovery with minimal CO2 emission.
I and type I-II source rocks (oil shales): implications IEEE Trans Energy Convers 22:103–109
for hydrocarbon generation characteristics. J Pet 152. Burnham AK, McConaghy JR (2006) Comparison of
Geol 33:19–42 the acceptability of various oil shale processes. In:
135. Burnham AK, Braun RL (1999) Global kinetic anal- Twenty-sixth oil shale symposium, Golden, 16–18
ysis of complex materials. Energy Fuel 13:1–22 Oct 2006. UCRL-CONF-226717
Oil Shale Processing, Chemistry, and Technology 83

153. Prien CH (1951) Pyrolysis of coal and oil shale. Ind 161. Burnham AK (1989) On solar thermal processing
Eng Chem 43:2006–2015 and retorting of oil shale. Energy 14:667–674
154. Smith MW, Shadle LJ, Hill DL (2007) Oil shale 162. Taulbee DN, Carter SD (1992) Investigation of prod-
development from the perspective of NETL’s uncon- uct coking by hot recycle solids in the KENTORT II
ventional oil resource repository. DOE/NETL-IR- fluidized bed retort. American Chemical Society, San
2007-022, Pittsburgh Francisco, pp 800–809
155. U.S. Department of Energy (2007) Secure fuels from 163. Elenurm A, Oja V, Elenurm A, Tali E, Tearo E,
domestic resources. The continuing evolution of Yanchilin A (2008) Thermal processing of
American’s oil shale and tar sands industries. U.S. dictyonema argillite and kukersite oil shale: transfor-
Department of Energy, Maryland, United States mation and distribution of sulfur compounds in pilot-
156. Louw SJ, Addison J (1985) Studies of the Scottish oil scale Galoter process. Oil Shale 25:328–334
shale industry, vol 1: history of the industry, working 164. Wang WD, Zhou CY (2009) Retorting of pulverized
conditions, and mineralogy of Scottish and Green oil shale in fluidized-bed pilot plant. Oil Shale
River formation shales. Final Report on US Depart- 26:108–113
ment of Energy. Institute of Occupational Medicine, 165. Kök MV, Guner G, Bağci AS (2008) Application of
Edinburgh EOR techniques for oil shale fields (in-situ combus-
157. Luts K (1944) Estonian oil shale kukersite, its chem- tion approach). Oil Shale 25:217–225
istry, technology and analysis. Revalen Buchverlag,
Reval. (in German)
158. Golubev N (2003) Solid oil shale heat carrier tech- Books and Reviews
nology for oil shale retorting. Oil Shale 20:324–332 Ogunsola OI, Hartstein AM, Ogunsola O (2010) Oil shale:
159. Gwyn JE, Roberts SC, Hardesty DE, Johnson GL, a solution to the liquid fuel dilemma. Oxford University
Hinds GP (1980) Shell pellet heat exchange retorting: Press, New York
the SPHER energy-efficient process for retorting oil Qian J (2010) Oil shale – petroleum alternative. China
shale. American Chemical Society, San Francisco, Petrochemical Press, Beijing
pp 59–69 Yen TF, Chilingarian GV (1976) Oil shale. Elsevier Scien-
160. Bradhurst DH, Worner HK (1996) Evaluation of tific, Amsterdam
oil production from the microwave retorting of
Australian shales. Fuel 75:285–288
 such as their length, whether they are open or
Hydraulic Fracturing closed, among other things.
Geomechanics The science of understanding the
Fred Aminzadeh mechanical impact of hydraulic fracturing and
University of Southern California, USC Viterbi nature of stress, rock failure/damage and reser-
School of Engineering, Los Angeles, CA, USA voir pressure created as a result of it.
FACT Inc., Santa Barbara, CA, USA Hydraulic fracturing A process to recover oil
and gas from tight formations by injecting large
volumes of water, sand, and small volumes of
Article Outline chemical additives (proppant) to increase oil or
natural gas flow.
Glossary Induced seismicity Seismic events (mostly
Definition of the Subject microseismic) caused by humans such as
Summary hydraulic fracturing, fluid injection to or pro-
What Is Hydraulic Fracturing? duction from an oil and gas field, CO2 seques-
Why Hydraulic Fracturing Is Important tration, or geothermal operation.
Fracture Characterization Micro-seismic event Small seismic (micro
Geomechanics of Hydraulic Fracturing earthquake) activities (usually under 2 Richter
Environmental Aspects of Hydraulic Fracturing scale, and mostly between 10 and 100,000
Induced Seismicity times smaller) sometimes caused by hydraulic
Assessment of Global Oil and Gas Resources fracturing. The “felt Microseismic event” is
Amenable for Extraction Via Hydraulic referred to as an event that does not does not
Fracturing cause any damage, though the shaking of the
Economics of Hydraulic Fracturing earth is detectable.
Future Work Natural versus stimulated fractures Natural
Bibliography fractures are small cracks in the rock that are
created over a long period of time, versus stim-
Glossary ulated fractures that are caused by SFIP of
hydraulic fracturing.
Environmental impact of hydraulic SFIP Any Subsurface Fluid (or gas) Injection of
fracturing Any unintended consequences of Production operation.
hydraulic fracturing operation on the environ- Shale resources Shale oil and gas resources are
ment such as sizable induced seismicity, leak- unconventional oil and gas produced from
age of fracturing fluid to the water table or the shale source rock fragments by pyrolysis.
surface, emission of methane gas, excessive use
of water, and the footprint of the operation such
as the damage caused to the road to bring sup- Definition of the Subject
plies (sand/water, propane) to the hydraulic
fracturing site or disposal of the fracturing fluid. Hydraulic fracturing is an oil and gas operation
Fracture characterization Use of different used to recover hydrocarbon resources trapped in
techniques to create a three dimensional distri- low-permeability shale. Water, along with sand, is
bution of natural or stimulated fractures. In injected into the shale to fracture it; the sand acts
some cases, fracture characterization also as a proppant, keeps the pores from collapsing
involves determining the nature of the fractures back, and allows the hydrocarbons to flow.

© Springer Science+Business Media, LLC, part of Springer Nature 2020 85

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_1052
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2019, https://doi.org/10.1007/978-1-4939-2493-6_1052-1
86 Hydraulic Fracturing

Coupled with horizontal drilling, the hydraulic Section “Environmental Aspects of Hydraulic
fracturing has made it possible to recover substan- Fracturing” addresses different environmental
tial quantities of hydrocarbons. Because the shale concerns about hydraulic fracturing. This includes
has low permeability, the production of oil the potential ground water contamination, amount
declines rapidly, and the shale may have to be of water used for hydraulic fracturing the water
fractured multiple times, and often multiple hori- disposal process, and methane emission concerns.
zontal wells may be drilled in conjunction with a Another environmental concern is the risk of
single well head. The commonly used term, induced seismicity or man-made earthquakes.
fracking, refers to the combined practice of Given the much publicized controversy on
hydraulic fracturing and horizontal drilling. whether and to what extent hydraulic furcating
creates induced seismicity section “Induced Seis-
micity” is dedicated to deal with this issue.
Summary

Here, we provide a brief introduction to hydraulic What Is Hydraulic Fracturing?

fracturing (HF) and various controversies that
have been associated with it in recent years. Glob- Hydraulic fracturing (HF) is an oil and gas oper-
ally in general, and in North America and Europe ation used to recover hydrocarbon resources that
in particular, much attention has been given to the are trapped in low-permeability shale and other
negative aspects of HF by environmentalists. This lithologies. Over geologic time periods, these
has been countered by the energy industry, resources were formed by the maturation of kero-
pointing to its economic benefits and the role it gen, the organic precursor of petroleum. How-
has played in securing energy independence. ever, unlike conventional oil and gas where
Many, in particular those in the fossil energy hydrocarbons migrate into the reservoir from a
industry, consider shale oil/gas, with the associ- separate source, the hydrocarbon source and res-
ated horizontal drilling and hydraulic fracturing, ervoir rock share low permeability, forming an
as one of the major developments in the oil and unconventional system. While there are signifi-
gas industry of the past two decades. Others, cant volumes of hydrocarbon trapped in uncon-
especially many environmentalists, consider ventional reservoirs, the extremely low natural
fracking proponents as public enemy number one. permeability of the formations impedes commer-
Different sections of this entry attempt to high- cial production using conventional techniques.
light different scientific facts about hydraulic frac- HF is a process that involves injection of large
turing, the common-sense environmental volumes of water (several million gallons), sand,
concerns, and the respective economic ramifica- and small volumes of chemical additives to
tions. After a brief overview of the principles of increase oil or natural gas flow from low perme-
hydraulic fracturing in section “What Is Hydraulic ability formations. The large pressure associated
Fracturing?”, we discuss the importance of with injection of “fracturing fluid” creates new
hydraulic fracturing in section “Why Hydraulic fractures and extends existing fractures that
Fracturing Is Important.” This is followed by frac- enhance hydrocarbon flow, while sand mixed
ture characterization (section “Fracture Character- with injected fluid holds the new and existing
ization”) and geomechanics (section fractures open. Some of the injected fluid flows
“Geomechanics of Hydraulic Fracturing”). They back to the wellbore and is pumped to the surface,
examine the natural fractures in the subsurface or is injected back to the reservoir. Figure 1 is an
and how one can characterize them, how hydrau- example of a typical HF configuration.
lic fracturing helps to expand the natural fractures HF is usually performed on horizontal or direc-
and/or create new (stimulated) fractures as well as tional wells, oftentimes with the well track being
the underlining rock mechanics properties and the perpendicular to the direction of maximum hori-
related stress regime. zontal in situ stress. Figure 1 shows a schematic
Hydraulic Fracturing 87

Fracturing
Equipment
&
Treatment
Shale Fractures
Domestic Wellhead
Well

stage 1

stage 2

stage 3
2,000 – 10,000 ft

Shale
stage 10 stage 9 stage 8 stage 7 stage 6 stage 5 stage 4 stage 3 stage 2 stage 1 Reservoir

4,000 – 10,000 ft

Hydraulic Fracturing, Fig. 1 An example of a typical hydraulic fracturing configuration (National Academy of
Sciences 2012) [1]

display of the horizontal well, and the enlarged in the US. According to the US Energy Informa-
shale fractures are shown in the top right with the tion Administration [5], 95% of new US wells
multistage (3 stages) fracturing. drilled in 2016 were hydraulically fractured.
Although hydraulic fracturing has been used This accounts for two-thirds of total US natural
since the 1950s, over the last decade it has been gas production and about half of US crude oil
the subject of intense public debate. Some of the production. While shale oil and gas has contrib-
concern has been over its potential impacts on uted substantially to US reserves and greatly
drinking water, the potential for emission of gas, impacted US oil production (increasing by about
and the associated induced seismicity or man- 3.2 million barrels/day), it is sobering to note that
made earthquakes. For more on various aspects its contribution to global reserves is rather modest.
of hydraulic fracturing debate and its benefits and US reserves of shale oil amount to 25 billion
drawbacks, see Hydraulic Fracturing 101 [2], barrels, an amount comparable to the annual
California Council on Science and Technology global consumption of oil [6]. Thus, the availabil-
(CCST) report on well stimulation [3], and ity of this resource does end the dominance of
hydraulic fracturing in Colorado [4]. conventional oil as the primary source of global
energy, and it limits the political and economic
leverage major oil producers have enjoyed in
Why Hydraulic Fracturing Is Important the past.
It should be noted that since 2015, due to lower
Hydraulic fracturing and horizontal drilling have oil and gas prices, we have seen a noticeable
played a key role in making shale a significant part decline in shale related activities, from explora-
of the fossil energy resources globally, especially tion to drilling and production. According to the
88 Hydraulic Fracturing

Baker Hughes rig count data, at peak in March of the price sensitivities in different plays. Since
2015 there were 1900 active oil rigs operating in 2017, production increased in all plays.
the various US basins [7]. Following the price Modern HF combined with horizontal drilling
collapse, the rig count dropped to about 400 in allows multiple wells to be drilled from one sur-
June 2016 and then began climbing again in face location, reducing the size of the drilling area
response to higher oil prices. As of early 2018 above ground by as much as 90%. Fracking is the
the number of oil rigs stood at about 1000. The key to unlocking vast US shale resources, freeing
relative speed with which these operations can be up oil and natural gas that previously was inac-
turned on and off has a profound influence on the cessible while protecting groundwater supplies
ability of oil producers to raise oil prices by lim- and the environment. America’s shale energy rev-
iting their own production. olution began with extensive US government
The increased production of shale oil in the agencies and contractors (e.g., Department of
USA has made it easier for the world oil market Energy, DOE; Gas Technology Institute, GTI).
to withstand several incidents when production This was further accelerated by privately financed
and export from certain countries was halted efforts (e.g., oil and gas operators and service
from war or other political upheavals [8]. companies). In both cases, the primary driver
Producing shale oil can cost upwards of $60/ was new technology.
barrel. Prior to 2006 oil pricing hovered around The notion of fracturing rocks to improve per-
$40/barrel, making hydraulic fracturing cost pro- meability is an old one. Based on an idea proposed
hibitive. Subsequently, largely in response to by Floyd Farris of Stanolind Oil and Gas Corpo-
increasing global demand, the oil price rose to ration (later, Amoco and then BP), attempts at
above $100/barrel (with some spikes as high as producing oil in tight formations were made by
$140/barrel). At these higher prices, shale oil pro- using mixtures of naphthenic acid and palm oil
duction by hydraulic fracture was profitable and it (napalm). The process, “Hydrafrac,” was patented
spurred drilling activities in many US basins. The in 1949 and licensed to Haliburton.
profitability can vary significantly from one play The Devonian shale basins of the Western
to the next. Figure 2 shows an earlier peak for Appalachian, Michigan, and Illinois Basins were
shale production for different US plays that known to contain vast quantities of gas, but the
arrived in 2015, after which it dipped through low permeability of the rock limited production
2017, with the rate of decline being an indication from this resource. The declining US reserves of

Hydraulic Fracturing,
Fig. 2 The explosion of
shale production since
2010 [5]
Hydraulic Fracturing 89

natural gas prompted the DOE and the GTI to well as stimulating economic activity and positive
sponsor research and development efforts for impact on the labor market for a variety of busi-
technologies for assessing and producing gas nesses like service and supply companies, HF has
from this resource. These efforts led to techniques supported growth across an economy that other-
such as the slick-water fracture, horizontal dril- wise has struggled in recent years. According to
ling, and microseismic technology for character- an IMF Report, the shale gas revolution has had a
izing and exploiting these resources. This macroeconomic impact between 0.3% and 1% of
included several DOE supported demonstration US GDP [10]. Many performance factors are used
projects in Ohio, Texas, and New Mexico to val- to assess the effectiveness and safety of
idate these technologies (API 2018) [9]. HF. Among them are better understanding of frac-
The development of commercial fracking is ture azimuth/geometry/length/distribution, Stim-
largely due to the efforts of Mitchell Energy, who ulated Reservoir Volume, and potential hazards
drew on the government-funded work, and despite (induced seismicity and hydraulic fracturing
numerous setbacks persisted for decades in fluid leakage). We will revisit the environmental
improving the technologies and driving down implications of hydraulic fracturing in a different
their costs. Of course, the rise in the cost of crude section.
oil in 2008 helped in the process. Once hydraulic To summarize, one of the key factors
fracking became economically feasible, it was impacting shale production is the economics and
widely deployed by many companies, and experi- oil and gas prices, as discussed earlier. However,
ential learning drove down the costs. Unleashing there are other factors that have impacted the
natural gas from shale resulted in a sharp decline of economics of shale resources, including the evo-
its cost from about $15/Mbtu in 2008 to around $3/ lution of the technologies, including subsurface
Mbtu in 2014. About one-third of the natural gas is characterization and advances in hydraulic frac-
used for producing hydrogen, which in turn is used turing, ability to monitor creation of new frac-
industrially to refine fuels or produce fertilizers. tures, and associated microseismic events. The
The large availability of natural gas at relatively next two sections provide some insight on the
low cost has allowed US refineries to increase their science behind hydraulic fracturing and how the
output and increase exports of finished products. advances in the related science and technology
While those against hydraulic fracturing may can further improve the economics of production
argue otherwise, the industry proponents maintain from shale and other tight formations.
shale gas has also had a positive environmental
impact. Since natural gas is not easily exported, its
increased availability in the US has resulted in a
decline in its price and made it the fuel of choice Fracture Characterization
for electric power production. Many US power
producers switched from burning coal to burning The basic idea behind new fracture generation is
natural gas. In 2005, electricity generation from illustrated by Fig. 3. Induced seismicity and stim-
coal and gas in the US amounted to 2,000 TWh ulated fractures are caused in most cases by
and 760 TWh, respectively; by 2017 the two changes in pore fluid pressure and/or change in
sources were each generating about 1,200 TWh, stress in the subsurface, especially in the presence
while the total electricity generation has held of (a) faults with specific properties and orienta-
steady at 4,000 TWh. Thus, over 700 TWh of tions and (b) a critical state of stress in the crust.
coal generation has switched to gas with the atten- The factor that appears to have the most direct
dant environmental benefit of reduced carbon correlation in regard to stimulated fracture is the
dioxide (CO2) emissions. net fluid balance or the total balance of fluid
The positive economic impact of shale-related introduced into or removed from the subsurface.
natural gas and oil resources is undeniable. By Additional factors may also influence the way
helping to lower power and materials costs, as fluids affect the subsurface.
90 Hydraulic Fracturing

In certain cases, the observed microseismicity is

found to correlate to shear deformation, although
HR is generally considered a tensile fracturing
process. The corresponding timeframe is also tied
to the relative high frequency content of the seis-
mic recording equipment and much more rapid
than the relative slow and hence aseismic
tensile fracture dilation occurring throughout the
injection period. Fracture movements can also
potentially occur at some distance away from the
HF associated with pressure and/or pressure
Hydraulic Fracturing, Fig. 3 Generation of fractures changes resulting from the fracturing. Although
due to application of horizontal and vertical stress and
fluid injection during hydraulic fracturing
the fracture geometry can often be inferred directly
from the microseismicity, attempts to understand
the fracture deformation can be a challenge. Maity
Microseismic mapping associated with and Aminzadeh [15] report on the use of Micro-
hydraulic fracture stimulations is used to extract seismic, 3D seismic, and well log data combination
the fracture geometry information, such as heights to characterize subsurface fractures by developing
and length. Since a large amount of shear stress a neural network-based “Fracture Zone Identifier”
near the fracture tip is created, it allows one to (FZI) attributes. Equation 1 shows the relationship
accurately determine the length and the height of between different FZI attributes and the porosity,
the fracture [11]. The assessment of fracture com- different types of velocities and density. Such an
plexity using microseismicity has yielded good implicit relationship is established using a specially
results since it reflects the general area of stimu- designed neural network.
lation treatment. It is almost impossible to deter-
mine the interaction between created fractures FZI n ¼ F ffn , Z n , V Pn , V Sn , rn , V En g (1)
with different levels of complexity. Furthermore,
the technology to differentiate between stimu- Figure 4 shows a three-dimensional display
lated, unchanged, and damaged fracture systems of subsurface characterization of the natural
is still evolving. Microseismic imaging is an fracture zones. Obviously, after HF, these frac-
important technology for tracking fracture crea- ture zones would be altered and the newly cre-
tion and reactivation of preexisting fractures ated or expanded fractures can be characterized
occurring during various industrial operations, using similar techniques.
particularly hydraulic fracture treatments of With the aid of repeated microseismic
unconventional reservoirs. Often, microseismic measurement data and possibly 4D seismic, the
locations are used to interpret the fracture geom- FZI attribute can also be used to quantify changes
etry, although additional insights into the fracture in the size and nature of the fractures and their
deformations can also be extracted through source growth. Other techniques such as the use of statis-
characterization. The resulting source deforma- tical distributions in seismicity can also shed valu-
tion investigations provide additional information able insights into the fracturing process and its
about the fracturing and are sometimes interpreted relation to the structural features within the reser-
as representing the entire spectra of deformations. voir. Such methods examine seismicity as a func-
Furthermore, new, unconventional, passive seis- tion of the in situ stress state within the reservoir.
mic imaging techniques such as coherency-based
stacking methods and emission tomography [12,
13] provide the potential for identification of low Geomechanics of Hydraulic Fracturing
signal-to-noise ratio (SNR) emissions such as
smaller microseismic events or low frequency Geomechanical evaluation and modeling in the oil
creep failure [14]. and gas industry is used to predict reservoir
Hydraulic Fracturing 91

Hydraulic Fracturing,
Fig. 4 Fracture mapping
using MEQ, Seismic &
Petrophysical Data using
artificial neural network
ANN-based hybrid FZI
attribute mapping, from [15]

parameters, such as bottom hole pressure, in situ model for such significant fracture propagation
rock stresses, modulus of elasticity, leak-off coef- complexities.
ficient, formation porosity, and permeability. It is Geomechanical models are usually very sensi-
no coincidence that, with increased demand for tive to input parameters, including stress anisot-
hydraulic fracturing, geomechanical simulations ropy, completion effectiveness to fracture
have gained popularity as well. For example, such initiation, and the DFN characteristics. Correlat-
evaluations allow the determination of reservoir ing the microseismic data with the corresponding
response to fracturing fluid injections. Starting geomechanical deformation would increase the
with a discrete fracture network (DFN), one can confidence in the geomechanical simulation and
model both the aseismic and microseismic modes provides an important framework to interpret the
of deformation. This helps interpret the resulting microseismicity as the geomechanical response of
injection pressures, rates, volumes in conjunction the reservoir. A linear elastic fracture mechanics
with seismicity data, improving the interpretation (LEFM) model can describe stress distribution
of stimulated fracture networks. around a single hydraulic fracture. The full stress
Preexisting fractures and the in situ stress state field solution is valid everywhere in the domain
have a significant bearing on hydraulic fracturing and is a function of stress boundary conditions
operations. In particular, hydraulic fracture prop- and fracture geometry.
agation and hydraulic vs. natural fracture interac- Predicting and controlling the initiation and
tions can be correlated with the observed cessation of a hydraulic fracture remains a chal-
microseismicity distribution. Recent studies have lenge due to compositional and poromechanical
shown strong correlation between microseismic- heterogeneity, which causes stress concentration,
ity distribution and in situ reservoir properties and due to inelasticity at the grain scale. The
such as faults and rock brittleness [16, 17]. Recent presence of organic matter also affects the rock-
studies involving ground truth measurements mechanical properties and directionality in frac-
from the Hydraulic Fracturing Test Site in the ture initiation and propagation processes. Overall,
Permian Basin [18] have shown that hydraulic understanding the fracturing behavior of a rock at
fracture growth and complexity can be significant the microscale plays a critical role in predicting
and the propped rock is relatively small in com- the performance of HF. Fracture initiation and
parison with the overall stimulated rock [19]. New propagation behaviors and estimation of fracture
mathematical techniques are being developed to toughness and directionality could be
92 Hydraulic Fracturing

accomplished through the use of microscale energy, and magnitude which are used to charac-
geomechanical scratch tests [20]. terize the size of microseismic events, incremental
We discussed microseismic monitoring in the fracture surface areas, and lengths. This will lead
“Fracture Characterization” section, providing the to the construction of the effective stimulated
data on the effectiveness of stimulation volume (ESV) providing the information on the
(HF) operation. During fracture stimulation oper- complexity of the hydraulic fracture network.
ations, a change of the in situ stresses and rock Consequently, seismic moment or cumulative
volume generates microseismicity. Mapping the moment can provide insights into fracture behav-
spatial and temporal pattern of these microseismic ior during stimulations. Energy-based approach is
events is used for tracking the progress of HF used to calibrate fracture model with microseis-
propagation. Microseismic monitoring can also mic data. The approach presents an energy-based
serve as a complementary indicator of changes numerical algorithm which stimulates a vertical
in fluid pathways and mechanical changes in the hydraulic fracture by parameter selection of elas-
treatment well. The deformation of the fracture tic, stress, and material properties defined in a
wall opening relates to the fluid pathway through layered 3D geologic medium [22].
P-waves. In addition, at the fracture tips, the shear To provide the information about the mecha-
stress regime may create S-waves which represent nism of failure at fracture sites, the moment tensor
the extent of the fracture wall. inversion (MTI) concept can be used. MTI
The primary data used for evaluation of micro- describes the orientation, magnitude, and slip of
seismic events are waveform measurements individual microseismic events. The decomposi-
obtained from the surface, downhole receivers, tion of each moment tensor into different compo-
or both. A “hybrid” microseismic survey design, nents provides us with the estimation of the
comprised of both surface and borehole geo- different failure mode: shear slip, tensile opening,
phones, allows for optimum coverage for effec- expansion, and their corresponding ratios. It also
tive characterization of the subsurface [21]. These gives other processes the orientation of local frac-
data provide information related to strain and ture planes and the direction of shear slip. MTI
stress changes in the reservoir as well as the nature concept allows constructing DFNs, representing
of the physical processes such as induced fractur- the distribution, orientation shape, connectedness,
ing of the rock or slippage on preexisting frac- and fluid properties of hydraulically induced frac-
tures. Waveform measurements can also indicate tures [23]. In addition, they can also detect com-
undesired fracture growth or fault activation. plex fracturing behavior such as fault activation or
The processed microseismic waveform record- fracture propagation alongside bedding
ing describes the magnitude and direction of the plans [24].
rock movements. Vector and spectral fidelity Given an estimate of the fracture network in
define the accuracy of the measuring content, the subsurface, our models will ascertain whether
leading to the construction and calibration of a a particular Subsurface Fluid Injection and
model of seismic P-wave and S-wave. The accu- Production (SFIP) strategy will cause failure at a
racy of microseismic interpretation depends on particular site. A more accurate statement would
estimated event hypocenters, the accuracy of the refer to the probability of failure occurring at a
velocity model, and the monitoring geometry. given site, reflecting on the observation that this
Thus, low signal-to-noise ratio is one of the inference is predicated on a particular (simplified)
greatest challenges for the processing of micro- model and a finite set of data, introducing uncer-
seismic data. Indirectly, field and surface deploy- tainty in the calibration and ensuing prediction.
ment challenges can be significant and generally Hazard maps are themselves a function of the
act as the primary constraint in survey geometries selected model and data, and their sensitivity
and resulting detectability. with respect to additional data and/or model
One of the most important concepts in analyz- refinement can be used to design data acquisition
ing microseismic data is the seismic moment, efforts and model refinement through multiscale
Hydraulic Fracturing 93

modeling. Accordingly, a hierarchical probabilis- instances of potable water contamination

tic model should be considered, where its param- from subsurface releases in California Well
eters reflect subscale effects, and whose sampling stimulation technologies, as currently prac-
distributions can be accurately estimated from ticed in California, do not result in a signifi-
data. Sensitivity of the hazard maps with respect cant increase in seismic hazard.” Overall, in
to these parameters will then reflect the value of California, for industry practice of today, the
additional measurements and additional model direct environmental impacts of well stimula-
complexity. tion practice appear to be relatively limited. It
should be emphasized that these conclusions
may not be easy to generalize elsewhere for
Environmental Aspects of Hydraulic various reasons, including: Difference in geol-
Fracturing ogy and reservoir parameters, difference in the
volume of hydraulic fracturing jobs in Cali-
HF has been the subject of intense debate for its fornia compared to many other States with
potential environmental risk factors. This has pro- more extensive shale production, and stricter
mpted much scrutiny of HF by regulators and the California State regulations than those of Fed-
general public. Many in the environmental com- eral and many other States.
munity believe biodiversity may suffer as a result Some of these studies have been used by
of land development and water supply damage. the American Petroleum Institute and other oil
Methane leakage may be much higher than previ- and gas industry advocates to point out no
ously reported by the EPA [25] and this has severe additional regulations on hydraulic fracturing
implications for the climate. Fracking could exac- are required and many of the existing ones
erbate the issue, which outweighs the potential should be scaled back. For example, API
CO2 pollution reduction by using clean gas in argues its ongoing workshop series “Commit-
place of oil/coal. Also, magnitude 2+ earthquakes ment to Excellence in Hydraulic Fracturing”
have increased in frequency in Texas and Okla- is one of the tools that the oil and natural gas
homa as well as the more common induced seis- industry uses to reinforce with regulators,
micity below magnitude 2, although much of the remind lawmakers, and educate the public on
seismicity is attributed to the waste water injection industry’s commitment to and leadership on
not to HF directly. While it is difficult to predict, safety, health, and environmental protection.
the ill effects of HF could be mitigated by research Its website continuously revises the standards
and policy in changes. related to HF. They can be seen on the
The four main areas of concern are considered Hydraulic Fracturing section of API’s website
to be (a) ground water contamination, (b) too (http://www.api.org/oil-and-natural-gas/wells-
much water used in the process and its disposal to-consumer/exploration-and-production/hydr
process, (c) methane emission, and (d) induced aulic-fracturing).
seismicity. The first three issues will be addressed Others may argue that many compounds
in this section, while the induced seismicity used for HF may be harmful, especially if
related concerns will be the subject of section enough care is not given to avoid entry of
“Induced Seismicity.” the fracking fluid to the water column.
Among compounds used in HF operation
(a) Ground-Water Contamination. The CCST are: acids, sodium chloride, polyacrylamide,
conducted an in-depth study of the issue and ethylene glycol, borate salts, sodium/potas-
released three-volume CCST Independent sium carbonate, glutaraldehyde, guar gum,
Report in 2016 titled: Advanced Well Stimu- citric acid, and isopropanol. Each of these
lation Technologies in California. While there compounds is used for specific purposes
are several key findings, the general conclu- such as helping to dissolve minerals and initi-
sions are: “There are no publicly reported ating fissures in rock before fracturing,
94 Hydraulic Fracturing

delaying breakdown of the gel polymer 2 re-fractured wells, there was no migration
chains, minimizing the friction between fluid of hydrocarbon or fracturing fluid into the
and pipe, preventing scale deposits in the pipe, nearby Edward-Trinity Plateau aquifer.
maintaining fluid viscosity with temperature (b) Amount of Water Used. A recent report
increase, ensuring effectiveness of other com- by Ceres highlighted the concern that water
ponents, eliminating bacteria in the water, used for fracking may be depleting water
thickening the water for sand suspension, pre- resources, particularly in arid regions, and
venting precipitation of metal oxides, and thus exacerbating the water shortage
increasing the viscosity of the fracture fluid. [29]. Fracking of wells typically requires the
After the wells on a pad are drilled, cased, use of several million gallons of water. How-
and cemented, the horizontal part of the pro- ever, the amount of water per se is not very
duction pipe is perforated with small holes in large when compared with other uses of water.
the casing, to expose the wellbore to the shale. In the 10 years from 2005 and 2014, fracking
Then, fracking fluid is pumped into the well operations in Texas – the state with most frack
under high pressure to create microfractures in wells – used 250 billion gallons of water, or
the shale and free the natural gas or oil. The on average about 25 billion gallons a year
sand in fracking fluid keeps the fractures open [30]. However, overall annual water usage in
after the pressure is released, and the Texas is 22 million acre-feet, or roughly 7,200
chemicals are chief agents to reduce friction billion gallons of water. In other words, the
and prevent corrosion. water used for fracking in Texas represents
Federal and State regulatory agencies nor- less than 0.5% of water used each year.
mally provide guidelines for energy develop- Viewed from this perspective, water use for
ment in the United States, including HF. fracking does not appear to cause shortages.
Among such regulations related to shale Often though, the challenge with water use
resources are: Clean Water Act, Clean Air arises from the fact that it must be sourced
Act, Safe Drinking Water Act, National Envi- locally – trucking in millions of gallons of
ronmental Policy Act, Resource Conservation water is expensive. Thus, if local groundwater
and Recovery Act, Emergency Planning and supplies are limited, fracking operations
Community Right to Know Act. Some of could aggravate the water shortage. Other
these regulations have recently been under efforts to minimize the negative impact of
re-assessment and possibly revisions since HF due to the excessive amount of water
mid-2017. usage, especially in the areas with severe
Debate continues on whether, and to what water shortage problems, include recycling
extent, HF could be found responsible for the waste water and HF fluid [31].
ground water contamination. It is understood (c) Methane Emissions. Similar concerns have
that a healthy debate should be based on sci- been raised on the potential emissions of CO2
ence and best practices in engineering as well from power generation and methane emission
extensive modeling and simulation, combined from HF. Others argue such emissions are at
with field monitoring. Some examples of such their lowest point in nearly 30 years, in large
research and development (R&D) efforts are part due to the widespread use of natural gas.
in [26] and [27]. Figure 5 shows an example of Furthermore, it is argued that methane emis-
such work. sions from the oil and natural gas industry
Recent studies at the Hydraulic Fracturing make up just 4% of total US greenhouse gas
Test Site in Midland Basin highlight the min- emissions. Nevertheless, the debate can be
imal environmental impact of large pad scale settled by installing monitoring stations in
fracturing operations on ground water and air major HF operations to provide an accurate
quality [28]. During this test, which involved assessment of the potential risk and develop
fracturing of 11 new horizontal wells and operational procedures to minimize and
 Hydraulic Fracturing

Hydraulic Fracturing, Fig. 5 Numerical modeling to quantify risk factors for water contamination associated with hydraulic fracturing [26]
95
96 Hydraulic Fracturing

mitigate such risk. For additional details on There are however positive aspects of induced
the Methane emission and other air pollution seismicity. Creation of induced seismicity, as a
impacts of hydraulic fracturing, see the work result of HF, is directly related to the opening up
by different US national laboratories and of the cores which helps with the stimulation of
others [32]. shale reservoirs, increase in permeability, and pro-
duction increase. This can be tested in a laboratory
environment to mimic a field setting. Figure 6
shows seismicity distribution in a field in Califor-
Induced Seismicity
nia with the main faults and subsidiary faults
(left). The laboratory CT-scan of a rock sample
The potential for earthquake-like activity induced
under stress, simulating a HF experiment, with
due to production of oil and natural gas has
slice of the sample in the center, with the associ-
become a topic of recent interest, owing to envi-
ated induced seismicity is shown on the right [34].
ronmental concerns. The National seismic hazard
We can also use the information on the fre-
maps in the United States are given by Petersen
quency of generated microseismicity (referred to
et al. [33]. It has been suggested that subsurface
as b-value) to distinguish between the induced
injection and production activities might factor
(from fluid injection or HF) and the tectonic
into the risk of induced seismicity. The extensive
related seismicity. This has been demonstrated
use of HF to stimulate production from shale plays
by many including Goebel et al. [35] as illustrated
has led many to speculate that the rise in hydraulic
in Fig. 7. Figure 7 (left) shows spatial variations in
stimulation operations may have sparked a con-
b-value in the study region in San Joaquin Valley.
current increase in seismicity. As the majority of
The faulting (tectonics)-related b-values generally
mapped faults in California are considered by the
seem to be lower (normally below 1, highlighted
US Seismic Hazard Map to be of higher risk,
in blue colors). This can be compared against the
understanding the potential seismic sensitivity of
higher b-values (in red colors) mostly in the areas
the region to hydraulic stimulation takes on
with a number of active oil and gas fields. Figure 7
greater importance.

36.5°N
1 cm 1 cm

36.3°N

36.1°N a
c
1 cm
35.9°N

35.7°N

b d e
35.5°N
120.8°W 120.6°W 120.4°W 120.2°W 120°W

Hydraulic Fracturing, Fig. 6 A comparison of induced seismicity in a field and laboratory setting with the associated
faulting
Hydraulic Fracturing 97

36.5°N

36.0°N

36.0°N

35.5°N 35.5°N

Magnitude
M=3–4 35.0°N
M=4–5 Oil Fields (4)
M>5
Oil Fields (5)
35.0°N
km
0.7 0.9 1.1 1.3 km
0 25 50 b-value 0 5 10 15
0 25 50 No Frac. Wells
34.5°N
121.0°W 120.5°W 120.0°W 119.5°W 119.0°W 121.5°W 121.0°W 120.5°W 120.0°W 119.5°W 119.0°W 118.5°W

Hydraulic Fracturing, Fig. 7 Spatial variations in b-value in the study region (Left), color coded number of hydraulic
fracturing in a few sizable oil and gas fields in San Joaquin Valley (Right)

(right) shows a few large oil fields with the color wet shale gas and tight oil in 46 countries is
coded number of hydraulic fracturing jobs. 7,577 tcf and 419 billion barrels (Gbbl), respec-
The probability of activating faults due to tively. These values are larger than the proved
anthropogenic influences is likely highest in global reserves of 6,832 tcf of natural gas and
regions where active faults, fluid injection wells, 240 Gbbl of oil [37]. Shale gas resources in the
and low b-values have been encountered. This is USA, Canada, and Mexico are 623, 573, and
the case, e.g., within the southern part of the study 545 tcf, respectively. Outside North America, sig-
region. In the proximity of active faults, both nificant resources are present in China (1100 tcf),
different types of seismicity may occur and more Argentina (800 tcf), Algeria (707 tcf), and
detailed studies are required to differentiate Australia (430 tcf). Technically recoverable
induced from natural (tectonic) seismicity. Areas global tight oil resources are estimated at
of relatively high b-values are likely connected to 419 bbl. Substantial tight oil resources are found
smaller ambient stresses and seismic events are in the USA (78 Gbbl), Russia (75 Gbbl), China
less likely to grow to large sizes. (32 Gbbl), Libya (26 Gbbl), and the UAE
(23 Gbbl).
Whereas production of shale gas and oil in the
Assessment of Global Oil and Gas US has been commercially deployed for over a
Resources Amenable for Extraction Via decade now, it has not been widely deployed
Hydraulic Fracturing elsewhere. Factors limiting the development
include public resistance to the technology, avail-
Shale resources as well as tight oil and gas are ability of adequate water, and lack of suitable
known to exist in parts of Europe, Australia, geologic conditions. If the experience in the US
Russia, China, and elsewhere, but this resource is replicated in other countries, tight gas and oil
has not been fully characterized by test wells. resources could play a very large role in the global
National Energy Technology Laboratory (NETL) energy scene.
highlighted the technology developments that
have led to the emergence of shale gas develop-
ment in a large scale [36]. The US Energy Infor- Economics of Hydraulic Fracturing
mation Agency has published assessments of
shale gas and oil resources in various countries The total investment for a hydraulic fracturing
and basins; according to the compilation well must cover not only the cost of drilling and
published in 2015, the technically recoverable completing the well, but also of obtaining land
98 Hydraulic Fracturing

rights and permits. These costs vary with location, See additional references on: What can micro-
and there is a broad range of costs. According to seismic tell us about HF [38], Geomechanical
the Energy Information Agency, the costs can approach for microseismic fracture mapping [39],
range between three and ten million dollars a Finite element method based modeling of HF [40],
well. Depending on the geology, the wells pro- Flowback of Fracturing Fluids with upgraded visu-
duce different amounts of oil and gas, with oil alization of HF and its Implication of on Overall
being the more valuable product, and the one Well Performance [41], Simulation of Hydraulic
that determines the economics. There are wells Fracturing-Induced Permeability Stimulation
that only produce gas, but, because there is using Coupled Flow and Continuum Damage
already an oversupply of gas such that no new Mechanics [42] and Oil, the Next Revolution [43].
dry wells are being developed, only those with
takeoff contracts are in production. It is a shame
that the associated gas from oil-producing wells in
Bakken and other basins are being flared for lack Future Work
of pipeline infrastructure.
For new HF wells, the average breakeven cost In spite of many advances made in different
of crude oil from different basins currently ranges aspects of HF, many new technologies need to
between $30 and $55 per bbl. The total cost be developed and implemented in order to make
depends on the geographic location, depth of the the operation more efficient, reduce cost, and
target, horizontal length of the well (offset), num- minimize its environmental footprint. A few
ber of HF stages, and many other factors. For areas requiring further attention are highlighted
example, in the Wolfcamp Delaware basin, the below:
average breakeven cost is $42/bbl, while in the
Bakken basin it is $52. These costs are consider- (a) To improve the operational aspects of HF, a
ably lower than they were in 2008 (~$70/bbl), better understanding of the reservoir parame-
when fracking first became commercially viable, ters and geology is important. More effective
thanks to the spike in crude oil prices. Innovations use of machine learning techniques for opti-
such as methods for better locating resources, mum well placement and determining the best
more efficient drilling, and multipad drilling parameters for multistage fracking is crucial.
have markedly reduced costs. In face of losing One such idea involves DNA fingerprinting
market share, in 2012, the Organization of the where different types of data are fused to
Petroleum Exporting Countries (OPEC) and Rus- improve the HF operation [44].
sia decided to increase their production and drive (b) Given the short lived impact of HF, “re-
down the price of oil from $100/bbl to $50/bbl. At fracking” is often required after a period of
$50/bbl, producing conventional oil is still profit- time (1–3 years from the original operation).
able. The expectation was that the competition This presents another opportunity for future
would be driven out of business at this reduced research to make optimum re-fracking
price. However, because of the improved technol- decisions.
ogies, producing oil from many of tight forma- (c) “Shock-Waves” or super waves [45] and Pulsed
tions was still profitable. Operations at those that Power Plasma Stimulation [46] are being con-
were not profitable were simply turned off tempo- sidered as an alternative approach to create sim-
rarily and could be brought on-line quickly when ilar or better results as the conventional HF.
conditions became favorable. The rapidity with These approaches aim at different stimulation
which HF wells can be turned off or brought techniques that do not require the injection of
back on stream is a game changer in the geopol- water or the use of chemical additives. They
itics of oil. Furthermore, the new focus on “re- promise to reduce cost with more favorable
fracking” to extend the life of the shale oil and gas operational conditions and less negative envi-
fields may become a new normal. ronmental impacts.
Hydraulic Fracturing 99

(d) Aside from the waterless/chemical-less possi- 5. U.S. Energy Information Administration (2018) Petro-
bilities for HF, many artificial neural network leum and other liquids data. https://www.eia.gov/petro
leum/data.php#crude
and AI based approaches are being considered 6. BP Statistical Review of Global Energy (2017) https://
for an optimum fresh water use management www.bp.com/en/global/corporate/energy-economics/
and recycling the frack-fluid. Best management statistical-review-of-world-energy.html
practices must be developed that will reduce 7. Baker Hughes (2017) Baker Hughes rig count. http://
phx.corporate-ir.net/phoenix.zhtml?c=79687&p=irol-
fresh water requirements, disposal challenges, rigCountOverview
and environmental impacts including truck traf- 8. FT (2018) https://www.ft.com/content/a8314d06-
fic, site congestion, and exposure to harmful 0741-11e8-9650-9c0ad2d7c5b5
chemicals by promoting recovery and re-use 9. DOE’s Unconventional Gas Research Programs
1976–1995. https://geographic.org/unconventional_
of produced water for well drilling and comple- gas_research/eastern_gas.html
tion applications using best practices. This is of 10. Minh-Thong Le (2018) An assessment of the potential
special importance where water supplies are for the development of the shale gas industry in coun-
limited and environmental regulations are tries outside of North America. Heliyon 4:e00516.
https://doi.org/10.1016/j.heliyon.2018.e00516
limited. 11. Warpinski NR, Mayerhofer MJ, Vincent MC, Cipolla
(e) In many situations, a good portion of the CL, Lolon EP (2009) Stimulating unconventional
injected HF liquid (FP) returns to the surface reservoirs: maximizing network growth while
with the oil and natural gas being produced. optimizing fracture conductivity. J Can Pet Technol
48(10):39–51. SPE-114173-PA. https://doi.org/
Research should continue to further character- 10.2118/114173-PA
ize these returned liquids and establish safe 12. Lacazette A, Vermilye J, Fereja S, Sizking C (2013)
methods for treating/disposing of them. In the Ambient fracture imaging: a new passive seismic
meantime, less hazardous HF fluids should method. In: Unconventional resources technology
conference (URTeC), Denver, 12–14 Aug 2013
continue to be explored iteratively, with their 13. Maity D, Salehi I, Ciezobka J (2014) Semblance
use potential characterized within a lifecycle weighted emission mapping for improved hydraulic
framework. Energy intensive aspects of treat- fracture treatment characterization. In: SEG annual
ment of FP to safely return water to the local meeting, Denver, 26–31 Oct 2014
14. Das I, Zoback MD (2013) Long-period long-duration
environment should also be addressed. seismic events during hydraulic stimulation of shale
and tight-gas reservoirs – Part 2: location and mecha-
nisms. Geophysics 78(6):KS97–KS105
15. Maity D, Aminzadeh F (2015) Novel fracture
Bibliography zone identifier attribute using geophysical and well
log data for unconventional reservoirs. Interpretation
1. National Academy of Sciences (2012) Induced Seis- 3(3):T155–T167. https://pdfs.semanticscholar.org/
micity Potential in Energy Technologies. https://www. 0ccb/4a21a8dd7d4a4b1bb2be4f8ce0192811aa9e.pdf
nap.edu/resource/13355/Induced-Seismicity-Report- 16. Wessels S, Kratz M, Pena ADL (2011) Identifying fault
Brief-Final.pdf activation during hydraulic stimulation in the Barnett
2. King GE (2012) Hydraulic fracturing 101: what every shale: source mechanisms, b values, and energy release
representative, environmentalist, regulator, reporter, analysis of microseismicity. In: SEG annual meeting,
investor, university researcher, neighbor and engineer San Antonio, 18–23 Sept
should know about estimating frac risk and improving 17. Maity D, Ciezobka J (2018) Correlating microseismic-
frac performance in unconventional gas and oil. In: ity with relevant geophysical and petrophysical data to
SPE 152596, SPE hydraulic fracturing technology understand fracturing process during hydraulic stimu-
conference lation: a case study from the Permian Basin. In: SEG
3. California Council on Science and Technology annual meeting, Anaheim, 15–19 Oct
(2016) Advanced well stimulation technologies in 18. Ciezobka J, Courtier J, Wicker J (2018) Hydraulic
California: an independent review of scientific and fracturing test site (HFTS) – project overview and
technical information, July 2016. https://ccst.us/wp- summary of results. In: Unconventional resources
content/uploads/160708-blm-report.pdf technology conference (URTeC), Houston, 23–25 July
4. Colorado Oil & Gas Conservation Commission 19. Maity D, Ciezobka J, Eisenlord S (2018) Assessment
(2011) Frequently asked questions about hydraulic of in-situ proppant placement in SRV using through-
fracturing. https://cogcc.state.co.us/Announcements/ fracture core sampling at HFTS. In: Unconventional
Hot_Topics/Hydraulic_Fracturing/Frequent_Questions_ resources technology conference (URTeC), Houston,
about_Hydraulic%20Fracturing.pdf 23–25 July
100 Hydraulic Fracturing

20. Ante MA, Manjunath GL, Aminzadeh F, Jha B (2018) United States national seismic hazard maps.
Microscale laboratory studies for determining fracture U.S. Geological Survey open-file report 2014–1091,
directionality in tight sandstone and shale during 243p. https://doi.org/10.3133/ofr20141091. URTeC:
hydraulic fracturing. In: Unconventional resources 2903021
technology conference (URTeC), Houston, 34. Aminzadeh F, Goebel THW (2013) Identifying
23–25 July 2018. https://doi.org/10.15530/urtec- induced seismicity in active tectonic regions: a case
2018-2903021 study of the San Joaquin Basin, California. AGU fall
21. Maity D, Aminzadeh F (2013) A new approach meeting, Abstract: S31F-06, San Francisco
towards optimized passive seismic survey design 35. Goebel THW, Hosseini SM, Cappa F, Hauksson E,
with simultaneous borehole and surface measure- Ampuero JP, Aminzadeh F, Saleeby JB (2016) Waste-
ments. In: Joint PSAAPG/SPE/PSSEPM/PCS-SEG water disposal and earthquake swarm activity at the
conference, Monterey, 19–25 Apr 2013 southern end of the Central Valley, California.
22. Boroumand N, Eaton D (2016) Interpreting microseis- Geophys Res Lett 43. https://doi.org/10.1002/
mic data from hydraulic fracturing: understanding 2015GL066948
subsurface deformation through numerical modeling, 36. NETL (2011) https://www.netl.doe.gov/file%20library/
recorder. Can Soc Explor Geophys 41(09) research/oil-gas/Shale_Gas_March_2011.pdfNetl
23. Eyre TS, Van Der Baan M (2015) Introduction to 37. https://www.eia.gov/analysis/studies/worldshalegas/,
moment tensor inversion of microseismic events. In: BP 2018
GeoConvention, Calgary, 4–8 May 2015 38. Maxwell S (2014) What can microseismic tell us about
24. Tan Y, Engelder T (2016) Further testing of the hydraulic fracturing? In: AAPG/SEG/SPWLA
bedding-plane-slip model for hydraulic-fracture open- Hedberg conference, Austin, 7–11 Dec 2014
ing using moment-tensor inversions. Geophysics 81(5). 39. Hosseini M, Aminzadeh F (2014) A geomechanical
https://library.seg.org/doi/10.1190/geo2015-0370.1 approach for microseismic fracture mapping. In:
25. EPA (2018) http://science.sciencemag.org/content/ AAPG/SEG/SPWLA Hedberg conference, Austin,
early/2018/06/20/science.aar7204 7–11 Dec 2014
26. Jabbari N, Aminzadeh F, de Barros F (2016) Hydraulic 40. Khodabakhshnejad A, Aminzadeh F (2014) An
fracturing and the environment, risk assessment for extended finite element method based modeling of
groundwater contamination. Stoch Environ Res Risk hydraulic fracturing. In: AAPG/SEG/SPWLA
Assess 30:10. https://doi.org/10.1007/s00477-016- Hedberg conference, Austin, 7–11 Dec 2014
1280-0 41. Desi K, Aminzadeh F (2016) Flowback of fracturing
27. Jabbari N, Aminzadeh F, de Barros F (2015) Assessing fluids with upgraded visualization of hydraulic frac-
the groundwater contamination potential from a well turing and its implication of on overall well perfor-
in a hydraulic fracturing operation. J Sustain Energy mance. J Sustain Energy Eng 4(3–4):250–261
Eng 3(1):66–79 42. Samnejad M, Aminzadeh F, Jha B (2017) Simulation
28. Eisenlord S, Hayes T, Perry K (2018) Environmental of hydraulic fracturing-induced permeability stimula-
impact analysis on the hydraulic fracture test site tion using coupled flow and continuum damage
(HFTS). In: Unconventional resources technology mechanics, SPE-187250-MS. In: ATCE conference
conference (URTeC), Houston, 23–25 July in San Antonio, 9–11 Oct 2017
29. Ceres (2016) Hydraulic fracturing & water stress: 43. Maugeri L (2012) Oil, the next revolution. Discussion
water demand by the numbers. https://www. paper 2012-10, Belfer Center for Science and Interna-
ceres.org/resources/reports/hydraulic-fracturing-water- tional Affairs, Harvard Kennedy School, June 2012
stress-water-demand-numbers 44. Aminzadeh F, Ranjith R, Lingareddy M, Suhag A,
30. Fracking in Texas, Ballotpedia. https://ballotpedia. Jha B, Alimi H (2017) DNA fingerprinting to optimize
org/Fracking_in_Texas reservoir characterization and hydraulic fracturing,
31. Lebas R, Lord P, Luna D et al (2013) Development and using machine learning based integration of geophys-
use of high-TDS recycled produced water for ics, geochemistry and geomechanics with machine
crosslinked-gel-based hydraulic fracturing. Presented learning. Provisional patent, D2018-0109, University
at the SPE hydraulic fracturing technology conference, of Southern California
Woodlands, 4–6 Feb. SPE-163824-MS. https://doi. 45. Biennier L, Jayaram V, Suas-David N, Georges R,
org/10.2118/163824-MS Kiran Singh M, Arunan E, Kassi S, Dartois E, Reddy
32. Hydralicfracturing.com (2018) http://www.hydraulic KPJ (2017) Shock-wave processing of C60 in hydro-
fracturing.com/#/?section=air-emissions gen. Astron Astrophys. https://doi.org/10.1051/0004-
33. Petersen MD, Moschetti MP, Powers PM, Mueller CS, 6361/201629067
Haller KM, Frankel AD, Zeng Y, Rezaeian S, 46. Xiao Y, House W, Unal E, Soliman M (2017) Pulsed
Harmsen SC, Boyd OS, Field N, Chen R, Rukstales power plasma stimulation technique – experimental
KS, Luco N, Wheeler RL, Williams RA, Olsen AH study on single pulse test for fracture initiation. ISH
(2014) Documentation for the 2014 update of the J Hydraul Eng 4(3):67–74
 Recovery Recovery of petroleum at the surface
Petroleum Refining and using primary, secondary, and tertiary recovery
Environmental Control and methods.
Environmental Effects Tar sand mining Recovery of tar sand by min-
ing (digging) tar sand from the formations at or
James G. Speight close to the surface.
CD&W Inc., Laramie, WY, USA

Definition of the Subject

Article Outline
The work summarizes the various processes and
Glossary the process emissions that occur during crude oil
Definition of the Subject refining. There are also general descriptions of the
Introduction various pollution, health, and environmental
Refinery Processes problems especially specific to the crude oil
Environmental Regulations and Pollutants industry and places in perspective government
Refinery Emissions laws and regulations as well as industry efforts
Entry into the Environment to control these problems.
Toxicity
Managing Wastes and Future Directions
Bibliography Introduction

Glossary Despite the numerous safety protocols that are in

place and the care taken to avoid incidents that
Bitumen A semi-solid to solid hydrocar- affect the environment, virtually every industry
bonaceous material found filling pores and suffers accidents that lead to environmental prob-
crevices of sandstone, limestone, or argilla- lems, to complexities, and to chemical contamina-
ceous sediments such as tar sand. tion. And this does not include environmental
Exploration The search for petroleum using contamination from domestic source! Furthermore,
a variety of physical and spectrographic it must be recognized that the capacity of the envi-
methods. ronment to absorb effluents from crude oil pro-
Hot water process The recovery of bitumen cessing and the use of the products are not
from tar sand by use of hot water whereby the unlimited. In fact, the environment is an extremely
bitumen floats and the sand sinks. limited resource, and discharge of chemical emis-
In situ conversion Partial or complete conver- sions into it should be subject to severe constraints.
sion of heavy oil or tar sand bitumen in the Thus, it is necessary to understand the nature and
reservoir or deposit as part of the recovery magnitude of the problems involved [1]. Hence, in
process. the current context, there is an essential need for
Oil mining The recovery of petroleum using a control over the amounts and types of emissions
mining method whereby an underground from the use of crude oil and its products.
chamber is produced by mining and the oil is Crude oil (petroleum) is a major energy source
allowed or encouraged to drain into the in many parts of the world and will continue to be
chamber. a primary source of energy for the next several

© Springer Science+Business Media, LLC, part of Springer Nature 2020 101

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_70
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_70-3
102 Petroleum Refining and Environmental Control and Environmental Effects

C1 To C4

Crude Unit Naphtha Hydro– Reformate

Reforming
Treating

Alkylate
Alkylation

Hydro– Diesel and Jet Fuel

Atmospheric
Distillation

Treating

Gasoline
Hvy Atm Gas Oil FCC Feed Fluidized
Hydro– Catalytic
Refining Cracking Fuel Oil

Lt Vac Gas Oil

Distillation
Vacuum

Hvy VGO Hydrocracking

Resid Fuel Gas and Coker Gasoline

Thermal
Processing Coke

Hydrogen
Sulfur Sulfur
Sulfide–containing
Complex
Gas

Petroleum Refining and Environmental Control and Environmental Effects, Fig. 1 Schematic overview of a
refinery

decades. However, considering the composition Refinery Processes

of crude oil and the products arising therefrom, it
is not surprising that crude oil itself and crude Crude oil refining is a complex sequence of chemi-
oil-derived products can have adverse effects on cal and physical changes that result in the production
the environment. Crude oil, in the unrefined or of a variety of products (Table 1) (Fig. 1). In fact,
crude form, like many industrial feedstocks, has crude oil refining might be considered as a collection
little or no direct use, and its value as an industrial of individual, yet related, processes that are each
commodity is only realized after the production of capable of producing effluent streams. Moreover,
salable products by a series of refining steps refining crude oil, as it is currently known, will
(Fig. 1). Each refining step is, in fact, a separate continue at least for the next several decades.
process and thus a refinery is a series of integrated Crude oil refining consists of, initially unless
process units that generate the desired products to properties dictate otherwise, dividing the petro-
meet the demands of the market. Therefore, the leum into fractions of different boiling ranges by
value of petroleum is directly related to the yield distillation. Other forms of treatment are utilized
of products and is subject to the call of the market. during the refining process to remove undesirable
In general, crude oil, once refined by a variety of components of the crude oil. The fractions them-
process units, yields salable products. selves are often further distilled to produce the
The purpose of this article is to summarize the desired commercial product. A variety of addi-
typical refinery processes and to present the vari- tives may be incorporated into some of the refined
ous pollution, health, and environmental prob- products to adjust the octane ratings or improve
lems that are especially specific to the crude oil engine performance characteristics.
refining industry and to briefly present the gov- In terms of individual processes, the potential for
ernment laws and regulations (of the United waste generation and, hence, leakage of emissions
States) as well as industry efforts to control these rests with all refining units, and special care must be
problems [2–5]. taken to ensure the security of the environment.
Petroleum Refining and Environmental Control and Environmental Effects 103

Petroleum Refining and Environmental Control and Environmental Effects, Table 1 Summary of refinery
processes
Process name Action Method Purpose Feedstock(s) Product(s)
(i) Separation processes
Desalting Separation Nonthermal Remove brine Crude oil Clean crude oil
Atmospheric Separation Thermal Separate fractions Desalted crude Gas, gas oil, distillate,
distillation without cracking oil residuum
Vacuum Separation Thermal Separate fraction Atmospheric Gas oil, lube stock,
distillation without cracking residuum residuum
(ii) Conversion processes
Catalytic Alteration Catalytic Upgrade gasoline Gas oil, Naphtha, petrochemical
cracking residuum feedstocks
Coking Polymerize Thermal Convert residuum Residuum Naphtha, petrochemical
feedstocks
Hydrocracking Hydrogenation Catalytic Produce lower- Gas oil, Lower-boiling, high-
boiling products cracked oil, quality products
residual
Visbreaking Decomposition Thermal Reduce viscosity Atmospheric Distillate, cracked
tower resid residuum
(iii) Skeletal alteration processes
Alkylation Combining Catalytic Combine olefin Iso-butane/ Iso-octane (alkylate)
and iso-paraffin olefin
Polymerization Polymerize Catalytic Combine two or Cracked High-octane naphtha,
more olefins olefins petrochemical
feedstocks

Desalting 350
F) to reduce viscosity and surface tension

Crude oil often contains water, inorganic salts, for easier mixing and separation of the water, but
suspended solids, and water-soluble trace metals. the temperature is limited by the vapor pressure of
As a first step in the refining process, to reduce the crude oil constituents. In both methods,
corrosion, plugging, and fouling of equipment and other chemicals may be added – ammonia is
to prevent poisoning the catalysts in processing often used to reduce corrosion, and caustic
units, these contaminants must be removed by or acid may be added to adjust the pH of the
desalting (dehydration). The two most typical water wash. Wastewater and contaminants are
methods of crude oil desalting are (i) chemical discharged from the bottom of the settling tank
separation and (ii) electrostatic separation. In to the wastewater treatment facility, while the
chemical desalting, water and chemical surfactant desalted crude is continuously drawn from the
(demulsifiers) are added to the crude oil, heated so top of the settling tanks and sent to the crude
that salts and other impurities dissolve into the distillation tower.
water or attach to the water, and then held in a The process creates an oily sludge that may,
tank where they settle out. Electrical desalting is depending upon the constituents, be a hazardous
the application of high-voltage electrostatic waste and a high-temperature salt wastewater
charges to concentrate suspended water globules stream (i.e., sent for treatment with other refinery
in the bottom of the settling tank. Surfactants are wastewaters). The primary polluting constituents in
added only when the crude oil contains suspended desalter wastewater include hydrogen sulfide,
solids. A third and less common process involves ammonia, phenol, high levels of suspended solids,
filtering heated crude oil using diatomaceous earth. and dissolved solids, with a high biochemical oxy-
In the desalting process, the feedstock crude oil gen demand (BOD). In some cases, it is possible to
is heated to between 65 and 177
C (150
and recycle the desalter effluent water back into the
104 Petroleum Refining and Environmental Control and Environmental Effects

Petroleum Refining and Environmental Control and Petroleum Refining and Environmental Control and
Environmental Effects, Table 2 Different generic boil- Environmental Effects, Table 3 Properties of the vari-
ing fractions separated from crude oil by distillation ous fuel oils
Boiling rangea Fuel oil Properties



Fraction C F No. 1 fuel Similar to kerosene or range oil (fuel used
Light naphtha 1 to 150 30–300 oil in stoves for cooking)
Medium naphtha 50–180 120–355 Defined as a distillate intended for
Heavy naphthab 150–205 300–400 vaporizing in pot-type burners and other
burners where a clean flame is required
Kerosenec 205–260 400–500
No. 2 fuel Often called domestic heating oil
Light gas oil 260–315 400–600
oil Properties are similar to diesel and higher-
Heavy gas oil 315–425 600–800
boiling jet fuels
Lubricating oil >400 >750 Defined as a distillate for general purpose
Vacuum gas oil 425–600 800–1100 heating in which the burners do not require
Residuum >510 >950 the fuel to be completely vaporized before
a
For convenience, boiling ranges are converted to the burning
nearest 5
No. 4 fuel A light industrial heating oil that is intended
b
Sales gasoline is a product produced by blending several oil where preheating is not required for
streams, including naphtha handling or burning
c
Kerosene destined for conversion to diesel fuel; crude Two grades that differ primarily in safety
kerosene is often quoted as having a wider boiling range (flash) and flow (viscosity) properties
(150–350
C, 302–662
F) before separation into various No. 5 fuel A heavy industrial oil that often requires
fuel and/or solvent products oil preheating for burning and, in cold
climates, for handling
No. 6 fuel A heavy residuum oil
desalting process, depending upon the type of oil Commonly referred to as Bunker C oil
crude being processed. when it is used to fuel ocean-going vessels
Preheating is required for both handling
Regulations are in place that set the maximum
and burning this oil
contamination concentration levels that are designed
to ensure environmental safety. The inorganic
chemicals that occur in desalter effluents include
salts of calcium, magnesium, sodium, and potas- concentration process on nonvolatile constituents
sium with lesser amounts of copper, iron, manga- and contain significantly less hydrocarbon con-
nese, and zinc that include bicarbonates, carbonates, stituents than the original crude oil. The constitu-
chlorides, and sulfates. Other factors include the pH ents of residua may, depending on the crude oil, be
(alkalinity/acidity of the stream), hardness, electrical molecular entities of which the majority contains
conductivity, total dissolved solids, and surfactants. at least one heteroatom.
A high level of any of these three chemicals in the Air emissions from a crude oil distillation unit
soil or in the water is an indication that one or more include emissions from the combustion of fuel
specific processes (identified from the chemicals oils (Table 3) in process heaters and boilers, fugi-
that have been released) or pollution prevention tive emissions of volatile constituents in the crude
processes are not performing as required by opera- oil and fractions, and emissions from process
tional specifications. vents. The primary source of emissions is com-
bustion of fuels in the crude preheat furnace and in
Distillation boilers that produce steam for process heat and
Atmospheric and vacuum distillation units are stripping. When operating in an optimum condi-
closed processes, and exposure to hazardous tion and burning cleaner fuels (e.g., natural gas,
materials is expected to be minimal. The output refinery gas), these heating units create relatively
of both units are product streams of various boil- low emissions of sulfur oxides (SOx), nitrogen
ing ranges (Table 2). The residua that are pro- oxides (NOx), carbon monoxide (CO), hydrogen
duced by distillation are the results of a sulfide (H2S), particulate matter, and volatile
Petroleum Refining and Environmental Control and Environmental Effects 105

hydrocarbons. If fired with lower-grade fuels (condensed steam containing hydrogen sulfide
(e.g., refinery fuel pitch, coke) or operated ineffi- and ammonia) from side stripping fractionators
ciently (incomplete combustion), heaters can be a and reflux drums. Many refineries now use vac-
significant source of emissions. uum pumps and surface condensers in place of
Fugitive emissions of volatile hydrocarbons barometric condensers to eliminate the generation
arise from leaks in valves, pumps, flanges, and of the wastewater stream and reduce energy con-
other similar sources where crude and its fractions sumption. Reboiler side-stripping towers rather
flow through the system. While individual leaks than open steam stripping can also be utilized on
may be minor, the combination of fugitive emis- the atmospheric tower to reduce the quantity of
sions from various sources can be substantial. sour water condensate.
Those potentially released during crude distilla-
tion include ammonia, benzene, toluene, and Visbreaking and Coking
xylenes, among others. These emissions are con- Visbreaking (viscosity reduction, viscosity break-
trolled primarily through leak detection and repair ing), a mild form of thermal cracking, is used to
programs and occasionally using special leak produce more desirable and valuable products.
resistant equipment. The process is a relatively mild, liquid-phase ther-
When sour (high-sulfur) crude oil is processed, mal cracking process used to convert heavy, high-
there is an increased potential for exposure to viscosity feedstocks to lower viscosity fractions
hydrogen sulfide in the preheat exchanger and suitable for use in heavy fuel oil. A secondary
furnace, tower flash zone and overhead system, benefit from the visbreaking operation is the pro-
vacuum furnace and tower, and bottoms duction of naphtha and gas oil streams that usually
exchanger. Hydrogen chloride (arising from the have higher product values than the visbreaker
brine in the original crude oil) may be present in charge. Visbreaking also produces a small quan-
the preheat exchanger, tower top zones, and over- tity of hydrocarbon gases and a larger amount of
heads. Wastewater may contain water-soluble sul- gasoline and remains a process of promise for
fides in high concentrations and other water- heavy feedstocks [6, 7].
soluble compounds such as ammonia, chlorides, The process can also be used as the preliminary
phenol, mercaptans, etc., depending upon the (pretreatment) step for upgrading heavy feed-
crude feedstock and the treatment chemicals. stocks [8]. In this type of process, the heavy feed-
Safe work practices and/or the use of appropriate stock is first thermally cracked using visbreaking
personal protective equipment may be needed for or hydrovisbreaking technology to produce a
exposures to chemicals and other hazards, such as product that is lower in molecular weight and
heat and noise, and during sampling, inspection, boiling point than the feed. The product is then
maintenance, and turnaround activities. deasphalted using an alkane solvent at a solvent to
The primary source of emissions is combustion feed ratio of less than two wherein separation of
of fuels in the crude preheat furnace and in boilers solvent and deasphalted oil from the asphaltenes
that produce steam for process heat and stripping. is achieved using a two-stage membrane separa-
In addition to the expected emissions from the tion system in which the second stage is a centrif-
crude oil, the distillation units generate consider- ugal membrane.
able wastewater. The process water used in distil- Like many thermal cracking processes,
lation often comes in direct contact with oil and can visbreaking tends to produce a relatively small
be highly contaminated. The typical constituents of amount of fugitive emissions and sour wastewater
sour wastewater streams from crude distillation [9, 10]. Usually some wastewater is produced
include hydrogen sulfide, ammonia, suspended from steam strippers and the fractionator. Waste-
solids, chlorides, mercaptans, and phenol, charac- water is also generated during unit cleanup and
terized by a high pH (i.e., high alkalinity). cooling operations and from the steam injection
Both atmospheric distillation and vacuum dis- process to remove organic deposits from the
tillation produce an oily sour wastewater soaker or from the coil. Combined wastewater
106 Petroleum Refining and Environmental Control and Environmental Effects

flows from thermal cracking, and coking pro- steam injection process to cut coke from the coke
cesses are approximately three gallons per barrel drums. Combined wastewater flows from thermal
of process feed. Propylene (CH3CH=CH2; cracking, and coking processes are approximately
b.p. –48
C, 54
F), another source of toxic three gallons of water per barrel of feedstock.
releases from refineries, is produced during crack- Like most separation processes in the refinery,
ing and coking processes. In addition to being a the process water used in coker fractionators (as is
gas, propylene has a measurable solubility in also the case in other product fractionators) often
water which increases the potential for release to comes in direct contact with oil and can have a
both air and water during processing. high oil content (much of that oil can be recovered
Coking is a thermal process for the continuous through wastewater oil recovery processes). Thus,
conversion of residua into lower-boiling products. the main constituents of sour water from catalytic
The feedstock can be atmospheric residuum, vac- cracking include high levels of oil, suspended
uum residuum, or cracked residuum, and the prod- solids, phenols, cyanides, hydrogen sulfate, and
ucts are gases, naphtha, fuel oil, gas oil, and coke. ammonia. Typical wastewater flow from catalytic
Coking processes generally utilize longer reaction cracking is about 15.0 gallons per barrel of feed
times than thermal cracking processes, and to processed (more than one-third of a gallon of
accomplish this, drums or chambers (reaction ves- wastewater for every gallon of feed processed)
sels) are employed, but it is necessary to use two and represents the second largest source of waste-
or more such vessels so that coke removal can be water in the refinery.
accomplished in those vessels not on-stream with- Particulate emissions from the decoking oper-
out interrupting the semicontinuous nature of the ations can be considerable. Coke-laden water
process. Gas oil may be the major product of a from decoking operations in delayed cokers
coking operation and serves primarily as a feed- (hydrogen sulfide, ammonia, suspended solids)
stock for catalytic cracking units. The coke can be and coke dust (carbon particles and hydrocarbons)
used as fuel, but processing for specialty uses, occurs.
such as electrode manufacture, production of
chemicals, and metallurgical coke, is also possi- Fluid Catalytic Cracking
ble. For these latter uses, the coke may require Catalytic cracking is widely used to convert heavy
treatment to remove sulfur and metal oils into more valuable naphtha (a blend stock for
impurities – calcined petroleum coke can be gasoline manufacture) and other low-boiling
used for making anodes for aluminum manufac- products. As the demand for gasoline increased,
ture and a variety of carbon or graphite products catalytic cracking replaced thermal cracking with
such as brushes for electrical equipment. the evolution of catalytic cracking. Fluid catalytic
Delayed coking is the oldest, most widely used cracking (FCC) refers to the behavior of the cata-
process and has changed very little in the five or lyst during this process insofar as the fine, pow-
more decades in which it has been on stream in dered catalyst (typically zeolites, which have a
refineries. Fluid coking is a continuous fluidized particle size on the order of 70 m) takes on the
solids process that cracks feed thermally over properties of a fluid when it is mixed with the
heated coke particles in a reactor vessel to gas, vaporized feed. Fluidized catalyst particles circu-
liquid products, and coke. Heat for the process is late continuously between the reaction zone and
supplied by partial combustion of the coke, with the regeneration zone.
the remaining coke being drawn as product. The In terms of process parameters, catalytic cracking
new coke is deposited in a thin fresh layer on the is typically performed at temperatures ranging from
outside surface of the circulating coke particle. 485 to 540
C (900–1,000
F) and pressures up to
Coking processes produce a relatively small 100 psi. Feedstocks for the process have typically
amount of sour wastewater from steam strippers been gas oil fractions, but the focus is shifting to gas
and fractionators. Wastewater is generated during oil-residua blends, gas oil-heavy oil blends, and gas
coke removal and cooling operations and from the oil-bitumen [11]. In some cases, heavy oils have
Petroleum Refining and Environmental Control and Environmental Effects 107

been blended with the minimum amount of gas oil wastewater from the fractionator/gas concentration
(added as a flux) as the feedstock to catalytic crack- units and steam strippers contain oil, suspended
ing units. In the process, the feedstock enters the unit solids, phenols, cyanides, hydrogen sulfide,
at temperatures on the order of 485–540
C ammonia, spent catalysts (metals) and hydrocarbons.
(900–1,000
F), and the circulating catalyst pro- Catalytic cracking generates significant quanti-
vides heat from the regeneration zone to the oil ties of spent process catalysts (containing metals
feed. Carbon (coke) is burned off the catalyst in from crude oils and hydrocarbons) that are often
the regenerator, raising the catalyst temperature sent off-site for disposal or recovery or recycling.
from 620 to 735
C (1,150–1,350
F), before the Management options can include land filling, treat-
catalyst returns to the reactor. ment or separation and recovery of the metals.
Fluid catalytic cracking units are major sources Metals deposited on catalysts are often recovered
of air emissions in refineries [5]. Air emissions are by third-party recovery facilities. Spent catalyst
released in process heater flue gas, as fugitive fines (containing aluminum silicate and metals)
emissions from leaking valves and pipes, and from electrostatic precipitators are also sent
during regeneration of the cracking catalyst. If off-site for disposal and/or recovery options.
not controlled, catalytic cracking is one of the Catalytic crackers also produce a significant
most substantial sources of carbon monoxide amount of fine catalyst dust that results from the
and particulate emissions in the refinery. In constant movement of catalyst grains against each
non-attainment areas where carbon monoxide other. This dust contains primarily alumina (Al2O3)
and particulates are above acceptable levels, car- and small amounts of nickel (Ni) and vanadium
bon monoxide waste heat boilers and particulate (V) and is generally carried along with the carbon
controls are employed. Carbon monoxide pro- monoxide stream to the carbon monoxide waste
duced during regeneration of the catalyst is heat boiler. The dust is separated from the carbon
converted to carbon dioxide either in the regener- dioxide stream exiting the boiler using cyclones,
ator or further downstream. Catalytic crackers are flue gas scrubbing, or electrostatic precipitators.
also significant sources of sulfur oxides and nitro-
gen oxides. The nitrogen oxides produced by cat- Hydrocracking and Hydrotreating
alytic crackers are expected to be a major target of Hydrocracking is a refining technology in which
emission reduction in the future. the outcome is the conversion of a variety of
Catalytic cracking units, like coking units, usu- feedstocks to a range of products, and units to
ally include some form of fractionation or steam accomplish this goal can be found at various
stripping as part of the process configuration. points in a refinery (Fig. 1) [12, 13, 14]. The
These units all produce sour waters and sour concept of hydrocracking allows the refiner to
gases containing some hydrogen sulfide and produce products having a lower molecular
ammonia. Like crude oil distillation, some of the weight with a higher hydrogen content and a
toxic releases reported by the refining industry are lower yield of coke. In summary, hydrocracking
generated through sour water and gases, notably facilities add flexibility to refinery processing and
ammonia. Gaseous ammonia often leaves frac- to the product slate. Hydrocracking is more severe
tionating and treating processes in the sour gas than hydrotreating there being the intent, in
along with hydrogen sulfide and fuel gases [5]. hydrocracking processes, to convert the feedstock
Catalytic cracking (primarily fluid catalytic crack- to lower-boiling products rather than to treat the
ing) generates considerable sour wastewater from feedstock for heteroatom and metal removal only.
fractionators used for product separation, from Hydrocracking generates air emissions
steam strippers used to strip oil from catalysts, and through process heater flue gas, vents, and fugi-
in some cases from scrubber water. The steam strip- tive emissions [5]. Unlike fluid catalytic cracking
ping process used to purge and regenerate the cata- catalysts, hydrocracking catalysts are usually
lysts can contain metal impurities from the feed in regenerated off-site after months or years of oper-
addition to oil and other contaminants. Sour ations, and little or no emissions or dust are
108 Petroleum Refining and Environmental Control and Environmental Effects

generated. However, the use of heavy oil as feed- hydrodesulfurization. The sulfides of hydrogenat-
stock to the unit can change this balance. ing metals are particularly used for hydrodesul-
Hydrocracking produces less sour wastewater furization, and catalysts containing cobalt,
than catalytic cracking. Hydrocracking, like cata- molybdenum, nickel, and tungsten are widely
lytic cracking, produces sour wastewater at the used on a commercial basis. Hydrotreating cata-
fractionator. These processes include processing lysts are usually cobalt-molybdenum catalysts,
in a separator (API separator, corrugated plate inter- and under the conditions whereby nitrogen
ceptor) that creates sludge [5]. Physical or chemical removal is accomplished, desulfurization usually
methods are then used to separate the remaining occurs as well as oxygen removal. Indeed, it is
emulsified oils from the wastewater. Treated waste- generally recognized that fullest activity of the
water may be discharged to public wastewater hydrotreating catalyst is not reached until some
treatment, to a refinery secondary treatment plant interaction with the sulfur (from the feedstock)
for ultimate discharge to public wastewater treat- has occurred, with part of the catalyst metals
ment, or may be recycled and used as process converted to the sulfides. Too much interaction
water. The separation process permits recovery of may of course lead to catalyst deactivation.
usable oil and creates a sludge that may be recycled The hydrotreating process generates air emis-
or treated as a hazardous waste. sions through process heater flue gas, vents, and
In addition, oily sludge from the wastewater fugitive emissions [5]. Unlike fluid catalytic crack-
treatment facility that result from treating sour ing catalysts, hydrotreating catalysts are usually
wastewaters may be hazardous wastes (unless regenerated off-site after months or years of oper-
they are recycled in the refining process). These ations, and little or no emissions or dust are gener-
include API separator sludge, primary treatment ated from the catalyst regeneration process at the
sludge, sludge from various gravitational separa- refinery. The off-gas stream from hydrotreating is
tion units, and float from dissolved air flotation usually very rich in hydrogen sulfide and light fuel
units. Like catalytic cracking, hydrocracking pro- gas. This gas is usually sent to a sour gas treatment
cesses generate toxic metal compounds, many of and sulfur recovery unit along with other refinery
which are present in spent catalyst sludge and sour gases. Fugitive air emissions of volatile com-
catalyst fines generated from catalytic cracking ponents released during hydrotreating may also be
and hydrocracking. These include metals such as toxic components. These include benzene, toluene,
nickel (Ni), cobalt (Co), and molybdenum (Mo). and the xylene isomers (BTX), and other volatiles
Hydrotreating is the less severe removal of that are considered to be toxic chemical releases
heteroatomic species by treatment (milder than under the EPA Toxics Release Inventory [15, 16].
hydrocracking) of a feedstock or product in the Hydrotreating also produces some residual
presence of hydrogen [5]. The character of the materials in the form of spent catalyst fines, usu-
hydrotreating processes is chemically very simple ally consisting of aluminum silicate and some
since they essentially involve removal of sulfur metals (e.g., cobalt, molybdenum, nickel, tung-
and nitrogen as hydrogen sulfide and ammonia, sten). Spent hydrotreating catalyst is now listed
respectively: as a hazardous waste (except for most support
material). Hazardous constituents of this waste
Sfeedstock þ H2 ! H2 S include benzene and arsenia (arsenic oxide,
As2O3). The support material for these catalysts
2Nfeedstock þ 3H2 ! 2NH3 is usually an inert ceramic (e.g., alumina, Al2O3).
Hydrotreating generates sour wastewater from
However, nitrogen is the most difficult contam- fractionators used for product separation. Like
inant to remove from feedstocks, and processing most separation processes in the refinery, the process
conditions are usually dictated by the require- water used in fractionators often comes in direct
ments for nitrogen removal. contact with oil and can be highly contaminated. It
In general, any catalyst capable of participating also contains hydrogen sulfide and ammonia and
in hydrogenation reactions may be used for must be treated along with other refinery sour
Petroleum Refining and Environmental Control and Environmental Effects 109

waters. Oily sludge from the wastewater treatment temperatures, typically operates at reactor temper-
facility that result from treating oily and/or sour atures of 21–38
C (70–100
F). Iso-butane con-
wastewaters from hydrotreating and other refinery centrations are maintained very high (i.e., at ratios
processes may be hazardous wastes, depending on of 4:1 or more above the reaction requirements) in
how they are managed. These include API separator the reactor vessels to ensure that the olefins are
sludge, primary treatment sludge, sludge from var- reacted. The reactor effluent is distilled to separate
ious gravitational separation units, and float from the propane, iso-butane, and alkylate boiling frac-
dissolved air flotation units. tions. The propane is routed to propane product
treating, the isobutane is recycled back to the
alkylation reactors, and the alkylate is routed to
Alkylation and Polymerization
gasoline blending, or in some cases to additional
In the alkylation process, a low-molecular-weight
solvents refinery processing.
olefin with isobutane in the presence of a catalyst,
The polymerization process (more correctly,
either sulfuric acid or hydrofluoric acid. The
the oligomerization process), as practiced in the
chemistry of the combination of olefins with par-
petroleum industry, is a process by which olefin
affins to form higher iso-paraffins is simple:
gases are converted to liquid condensation prod-
ucts that may be suitable for gasoline (hence poly-
ðCH3 Þ3 CH þ CH2 ¼ CH2 mer gasoline, polymerate) or other liquid fuels.
! ðCH3 Þ3 CHCH2 CH3 The feedstock for the process usually consists of
propylene (propene, CH3.CH=CH2) and butyl-
The high-octane value of the product (alkylate) enes (butenes, various isomers of C4H8) from
is an excellent blending stock for premium grades cracking processes or might even be selective
of gasoline. Furthermore, since alkylate contains olefins for dimer, trimer, or tetramer production:
no olefins, no aromatics, and no sulfur, it is also an
excellent blending stock for use in reformulated nCH2 ¼ CH2 ! HðCH2 CH2 Þn H
gasoline. The alkylation reaction is catalyzed by
the presence of very strong acid, either sulfuric In this process, n is usually 2 (dimer), 3 (tri-
acid or hydrofluoric acid. mer), or 4 (tetramer); the molecular size of the
In the sulfuric-acid-based alkylation process, product is limited to give products boiling in the
the acid is continually cycled through the process; gasoline range constituents. This contrasts with
but as it cycles, it becomes diluted and contami- polymerization that is carried out in the polymer
nated from impurities in the hydrocarbon feeds. industry where n may be on the order of several
The alkylation reactors typically operate at temper- hundred. The four-carbon to 12-carbon com-
atures 2–21
C (35–70
F, maximum) to minimize pounds that are required as the constituents of
polymerization of the olefins to form undesirable liquid fuels are the prime products. However, in
hydrocarbons) for the sulfuric acid process. The the petrochemical section of a refinery (Tables 4
concentration of the sulfuric acid catalyst is impor- and 5), polymerization, which results in the pro-
tant to the efficiency of the alkylation reaction; duction of, for example, polyethylene, is allowed
when the concentration of the acid decreases to to proceed until the products having the required
approximately 88% v/v, a portion of the contami- high molecular weight have been produced.
nated acid is withdrawn and replaced with fresh The process may be accomplished thermally or
acid. The contaminated, dilute sulfuric acid is then in the presence of a catalyst at lower temperatures.
regenerated to its original purity and concentration. Thermal polymerization is regarded not as effec-
Hydrofluoric acid exists in a vapor state at tive as catalytic polymerization but has the advan-
ambient conditions, and this dictates that extreme tage that it can be used to polymerize saturated
precaution is necessary to ensure that this toxic materials that cannot be induced to react by cata-
substance is contained inside the process equip- lysts. The process consists essentially of vapor-
ment. The hydrofluoric acid process, which is less phase cracking of, say, propane and butane
sensitive to polymerization at warmer followed by prolonged periods at high
110 Petroleum Refining and Environmental Control and Environmental Effects

Petroleum Refining and Carbon number Hydrocarbon type

Environmental Control
Saturated Unsaturated Aromatic
and Environmental
Effects, 1 Methane
Table 4 Examples of 2 Ethane Ethylene
hydrocarbon intermediates Acetylene
used in the petrochemical 3 Propane Propylene
industry
4 Butanes n-Butenes
Isobutene
Butadiene
5 Pentanes Isopentenes
(Isoamylenes)
Isoprene
6 Hexanes Methylpentenes Benzene
Cyclohexane
7 Mixed heptenes Toluene
8 di-Isobutylene Xylenes
Ethylbenzene
Styrene
9 Cumene
12 Propylene tetramer
tri-Isobutylene
18 Dodecylbenzene
6–18 n-Olefins
11–18 n-Paraffins

Petroleum Refining and Environmental Control and temperature (510–590
C, 950–1,100
F) for the
Environmental Effects, Table 5 Sources of petrochem- reactions to proceed to near completion.
ical intermediates
Emissions from alkylation processes and poly-
Hydrocarbon Source merization processes include fugitive emissions
Methane Natural gas of volatile constituents in the feed and emissions
Ethane Natural gas that arise from process vents during processing.
Ethylene Cracking processes
These can take the form of acidic hydrocarbon
Propane Natural gas, catalytic reforming,
cracking processes
gases, non-acidic hydrocarbon gases, and fumes
Propylene Cracking processes that may have a strong odor (from sulfonated
Butane Natural gas, reforming and cracking organic compounds and organic acids, even at
processes low concentrations). To prevent releases of
Butene(s) Cracking processes hydrofluoric acid, refineries install a variety of
Cyclohexane Distillation mitigation and control technologies (e.g., acid
Benzene Catalytic reforming inventory reduction, hydrogen fluoride detection
Toluene Catalytic reforming systems, isolation valves, rapid acid transfer sys-
Xylene(s) Catalytic reforming tems, and water spray systems).
Ethylbenzene Catalytic reforming In hydrofluoric acid alkylation processes,
Alkylbenzenes Alkylation acidic hydrocarbon gases can originate anywhere
>C9 Polymerization
hydrogen fluoride is present (e.g., during a unit
Petroleum Refining and Environmental Control and Environmental Effects 111

upset, unit shutdown, or maintenance) [5]. fluoride can also be routed back to a hydrofluoric
Hydrofluoric acid alkylation units are designed acid manufacturer.
to pipe these gases from acid vents and valves to A basic step in hydrofluoric acid manufacture
a separate closed-relief system where the acid is is the reaction of sulfuric acid with fluorspar
neutralized. (calcium fluoride) to produce hydrogen fluoride
Another source of emissions is combustion and calcium sulfate. Spent alumina is also gener-
of fuels in process boilers to produce steam for ated by the defluorination of some hydrofluoric
strippers. As with all process heaters in the refinery, acid alkylation products over alumina. It is dis-
these boilers produce significant emissions of sul- posed of or sent to the alumina supplier for recov-
fur oxides, nitrogen oxides, carbon monoxide, par- ery. Other solid residuals from hydrofluoric acid
ticulate matter, and volatile hydrocarbons. alkylation include any porous materials that may
Alkylation generates relatively low volumes of have contacted the hydrofluoric acid.
wastewater, primarily from water washing of the
liquid reactor products. Wastewater is also gener-
ated from steam strippers, depropanizers and Catalytic Reforming
debutanizers, and can be contaminated with oil Catalytic reforming consists of two types of chem-
and other impurities. Liquid process waters ical reactions that are catalyzed by two different
(hydrocarbons and acid) originate from minor types of catalysts: (i) isomerization of straight-
undesirable side reactions and from feed contam- chain paraffins and isomerization (simultaneously
inants and usually exit as a bottoms stream from with hydrogenation) of olefins to produce
the acid regeneration column. The bottom layer is branched-chain paraffins and (ii) dehydrogenation-
an acid-water mixture that is sent to the neutraliz- hydrogenation of paraffins to produce aromatics and
ing drum. The acid in this liquid eventually ends olefins to produce paraffins. The process is used to
up as insoluble calcium fluoride. convert alkanes (paraffins) to cycloalkanes
Sulfuric acid alkylation generates considerable (cycloparaffins) and to aromatics, and emissions
quantities of spent acid that must be removed and from catalytic reforming include fugitive emissions
regenerated. Nearly all the spent acid generated at of volatile constituents in the feed and emissions
refineries is regenerated and recycled and, although from process heaters and boilers [5].
technology for on-site regeneration of spent sulfu- As with all process heaters in the refinery, com-
ric acid is available, the supplier of the acid may bustion of fossil fuels produces emissions of sulfur
perform this task off-site. If sulfuric acid produc- oxides (SOx), nitrogen oxides (NOx), carbon mon-
tion capacity is limited, acid regeneration is often oxide (CO), particulate matter, and volatile hydro-
done on-site. The development of internal acid carbons (volatile organic chemicals, VOCs).
regeneration for hydrofluoric acid units has virtu- Furthermore, benzene, toluene, and the xylene iso-
ally eliminated the need for external regeneration, mers are toxic aromatic chemicals that are pro-
although most operations retain one for start-ups or duced during the catalytic reforming process and
during periods of high feed contamination. used as feedstocks in chemical manufacturing. Due
Both sulfuric acid and hydrofluoric acid alkyl- to their highly volatile nature, fugitive emissions of
ation units generate neutralization sludge from these chemicals are a source of their release to the
treatment of acid-laden streams with caustic solu- environment during the reforming process. Point
tions in neutralization or wash systems. Sludge air sources may also arise during the process of
from hydrofluoric acid alkylation neutralization separating these chemicals.
systems consists largely of calcium fluoride and In a continuous reformer, some particulate and
unreacted lime and is usually disposed of in a dust matter can be generated as the catalyst moves
landfill. It can also be directed to steel manufactur- from reactor to reactor and is subject to attrition.
ing facilities, where the calcium fluoride can be However, due to catalyst design, little attrition
used as a neutral flux to lower the slag-melting occurs, and the only outlet to the atmosphere is
temperature and improve slag fluidity. Calcium the regeneration vent, which is most often
112 Petroleum Refining and Environmental Control and Environmental Effects

scrubbed with a caustic to prevent emission of Petroleum Refining and Environmental Control and
hydrochloric acid (this also removes particulate Environmental Effects, Table 6 Component streams
for gasoline
matter). Emissions of carbon monoxide and
hydrogen sulfide may occur during regeneration Producing Boiling range



of catalyst. Stream process C F
(i) Paraffins
Butane Distillation 0 32
Isomerization
Conversion
Catalytic reforming processes provide high-
Iso-pentane Distillation 27 81
octane constituents in the heavier gasoline frac-
Conversion
tion, but the n-paraffin components of the lighter
Isomerization
gasoline fraction, especially butane (C4) to hex- Alkylate Alkylation 40–150 105–300
ane (C6), have poor octane ratings. The conver- Isomerate Isomerization 40–70 105–160
sion of these n-paraffins to their isomers Naphtha Distillation 30–100 85–212
(isomerization) yields gasoline components of Hydrocracking 40–200 105–390
high octane rating in this lower boiling range. Hydrocrackate
Conversion is obtained in the presence of a cata- (ii) Olefins
lyst (aluminum chloride activated with hydrochlo- Catalytic Catalytic 40–200 105–390
ric acid), and it is essential to inhibit side naphtha cracking
Cracked Steam 40–200 105–390
reactions, such as cracking and olefin formation.
naphtha cracking
Thus, the process can be used to convert
Polymer Polymerization 60–200 140–390
n-butane, n-pentane, and n-hexane into their (iii) Aromatics
respective iso-paraffins of substantially higher Catalytic Catalytic 40–200 105–390
octane number. The straight-chain paraffins are reformate reforming
converted to their branched-chain counterparts
whose component atoms are the same but are
arranged in a different geometric structure. Isom- constituents from residua, heavy oils, extra
erization is important for the conversion of heavy oils, and tar sand bitumen. The deasphalted
n-butane into iso-butane, to provide additional oil (the soluble product of propane deasphalting
feedstock for alkylation units, and the conversion that is reactively low in the presence of coke-
of normal pentanes and hexanes into higher- forming constituents and metal-containing cata-
branched isomers for gasoline blending (Table 6). lyst poisons) is then sufficiently cleaned to be sent
Isomerization processes produce sour water to a hydrotreating unit or a hydrocracking unit or
and caustic wastewater. The ether manufacturing to be used as blend stock for fuel oil.
process utilizes a water wash to extract methanol In the process, liquid propane (liquefied under
or ethanol from the reactor effluent stream. After pressure) is the usual solvent of choice due to its
the alcohol is separated, this water is recycled unique solvent properties. At lower temperatures
back to the system and is not released. In those (38–60
C, 100–140
F), paraffins are very solu-
cases, where chloride catalyst activation agents ble in propane, and at higher temperatures
are added, a caustic wash is used to neutralize (approximately 93
C, 200
F), hydrocarbons are
any entrained hydrogen chloride. This process almost insoluble in propane. The propane-
generates caustic wash water that must be treated deasphalting process is analogous to solvent
before being released. extraction in that a packed or baffled extraction
tower or rotating disc contactor is used to mix the
Deasphalting and Dewaxing oil feedstocks with the solvent.
In the well-established tower method, four to
Deasphalting eight volumes of propane are fed to the bottom of
Propane deasphalting is commonly used to pre- the tower for every volume of feed flowing down
cipitate asphaltene constituents and resin from the top of the tower. The oil, which is more
Petroleum Refining and Environmental Control and Environmental Effects 113

soluble in the propane, dissolves and flows to the selectively cracked on zeolite-type catalysts,
top. The higher-molecular-weight polar asphalt con- and the lower-boiling reaction products are
stituents flow to the bottom of the tower where they separated from the dewaxed feedstock oil by
are removed in a propane mix. Propane is recovered fractionation
from the two streams through two-stage flash sys- The dewaxing process also produces heater
tems followed by steam stripping in which propane stack gas (carbon monoxide, sulfur oxides,
is condensed and removed by cooling at high pres- nitrogen oxides, and particulate matter) as
sure in the first stage and at low pressure in the well as hydrocarbon emission such as fugitive
second stage. The asphalt recovered can be blended propane and fugitive solvents [5]. Steam strip-
with other asphalt, or heavy fuel oil, or can be used ping wastewater (oil and solvents) and solvent
as feed to the coker. The propane recovery stage recovery wastewater (oil and propane) are
results in propane-contaminated water that typically also produced. The fugitive solvent emissions
is sent to the wastewater treatment plant. may be toxic (toluene, methyl ethyl ketone,
Air emissions may arise from fugitive propane methyl isobutyl ketone).
emissions and process vents. These include Finally, the so-called finishing processes
heater stack gas (carbon monoxide, sulfur (Table 7) include a variety of individual pro-
oxides, nitrogen oxides, and particulate matter) cess units that are used to produce the finished
as well as hydrocarbon emission such as fugitive product, i.e., a product that will be ready for
propane and fugitive solvents. Steam stripping sales or ready for sales after blending.
wastewater (oil and solvents) and solvent recov-
ery wastewater (oil and propane) are also
produced. Environmental Regulations and
Pollutants
Dewaxing
The solvent dewaxing process is used to remove An environmental regulation is a legal mecha-
wax (n-paraffin hydrocarbons, straight-chain par- nism that determines how the policy directives
affin hydrocarbons) from deasphalted lubricating of an environmental law are to be carried out.
oil base stocks. The main process steps include An environmental policy is a requirement that
mixing the feedstock with the solvent, chilling the specifies operating procedures that must be
mixture to crystallize wax, and recovering the followed. An environmental guidance is a docu-
solvent. Commonly used solvents include toluene ment developed by a governmental agency that
and methyl ethyl ketone (MEK) – methyl isobutyl outlines a position on a topic or which gives
ketone is also used in a wax deoiling process to instructions on how a procedure must be carried
prepare food-grade wax. Although some of the out. It explains how to do something and provides
old methods are still in commercial use, there are governmental interpretations on a governmental
three main methods used in modern refinery act or policy. Environmental issues range from the
technology: effects of pollutants on the population at large to
effects on the lives of workers in various occupa-
i. Processes in which the feedstock is mixed tions where sickness or disability can result from
with one or more solvents - the feedstock is exposure to chemical agents.
cooled to allow the formation of wax crystals, There are a variety of regulations (Table 8) that
and the solid phase is separated from the liq- apply to crude oil refining [5]. The most popular
uid phase by filtration is the series of regulations known as the Clean Air
ii. The urea dewaxing processes in which urea Act that first was introduced in 1967 and was
straight-chain paraffin adducts are formed and subsequently amended in 1970 and most recently
then separated by filtration from the dewaxed oil in 1990. The most recent amendments provide
iii. Catalytic dewaxing processes in which stricter regulations for the establishment and
straight-chain paraffinic hydrocarbons are enforcement of national ambient air quality
114 Petroleum Refining and Environmental Control and Environmental Effects

Petroleum Refining and Environmental Control and Environmental Effects, Table 7 Finishing processes
Process Action Method Purpose Feedstock(s) Product(s)
Amine treating Treatment Absorption Remove acidic Sour gas Acid free
contaminants gases
Desalting Dehydration Absorption Remove Crude oil Desalted crude
contaminants oil
Drying and Treatment Absorption/ Remove water and Liquid Sweet and dry
sweetening thermal sulfur compounds hydrocarbons hydrocarbons
Furfural extraction Solvent Absorption Upgrade Cycle oils High-quality
extraction mid-distillate and diesel, lube oil
lubes
Hydrodesulfurization Treatment Catalytic Remove sulfur, High-sulfur Desulfurized
contaminants residuum, gas olefins
oil
Hydrotreating Hydrogenation Catalytic Remove impurities, Residuum, Cracker feed,
saturate olefins cracked distillate, lube
products
Phenol extraction Solvent Absorption/ Improve viscosity Lube oil base High-quality
extraction thermal index, color stocks lube oils
Solvent deasphalting Treatment Absorption Remove asphalt Vacuum tower Heavy lube
residual, oil, asphalt
propane
Solvent dewaxing Treatment Cool/filter Remove wax from Vacuum lube Dewaxed lube
lube stocks oil base stock
Solvent extraction Solvent Absorption/ Separate unsaturated Gas oil, High-octane
extraction precipitation oils reformate, gasoline
distillate
Sweetening Treatment Catalytic Remove hydrogen Untreated High-quality
sulfide and distillate/ distillate/
mercaptans gasoline gasoline

standards for, as an example, sulfur dioxide. not indigenous to the ecosystem, or, if the chem-
These standards do not stand alone, and there are ical occurs naturally, it becomes a pollutant when
many national standards for sulfur emissions. it is present in a concentration greater than the
Crude oil production and refining produce naturally occurring concentration thereby caus-
chemical waste [4, 5]. If this chemical waste is ing an adverse effect on the ecosystem. The pol-
not processed in a timely manner, it can become a lutant is often the product of human activity and
pollutant. Under some circumstances, chemical has a detrimental effect on the environment, in
waste is reclassified as hazardous waste, which is part or in toto. Pollutants can also be subdivided
any gaseous, liquid, or solid waste material that, into two classes: (i) primary and (ii) secondary.
if improperly managed or disposed of, may pose Thus:
hazards to human health and the environment. In
some cases, the term chemical waste is used Source ! Primary pollutant
interchangeably (often incorrectly) with the ! Secondary pollutant
term hazardous waste, but chemical waste
should be considered as hazardous unless proven A primary pollutant is a chemical that is emit-
otherwise, and the correct use of the terms must ted directly from the source. In terms of atmo-
be used. spheric pollutants from crude oil, examples
A pollutant is a chemical that is discharged into are carbon oxides (COx – carbon monoxide, CO,
a specific location (an ecosystem) that is typically and carbon dioxide, CO2), sulfur dioxide (SO2),
Petroleum Refining and Environmental Control and Environmental Effects 115

Petroleum Refining and Environmental Control and ½Scrude oil þH2 ! H2 S þ hydrocarbons
Environmental Effects, Table 8 Examples of environ- 2½Ncrude oil þ 3H2 ! 2NH3 þ hydrocarbons
mental regulations that apply to crude oil refineries
First On the other hand, a secondary pollutant is a
enacted Amended
chemical that is produced by interaction of a pri-
Clean Air Act (CAA) 1970 1977
mary pollutant with another chemical or which is
1990
Clean Water Act (Water 1948 1965a
produced by dissociation of a primary pollutant or
Pollution Control Act) 1972b
other effects within an ecosystem. Again, using
1977 the atmosphere as an example, the formation of
1987c the constituents of acid rain is an example of the
Comprehensive Environmental 1980 1986d formation of secondary pollutants:
Response, Compensation, and
Liability Act (CERCLA) SO2 þH2 O ! H2 SO3 ðsulfurous acidÞ
Hazardous Materials 1974 1990 2SO2 þO2 ! 2SO3 ðsulfur trioxideÞ
Transportation Act
SO3 þH2 O ! H2 SO4 ðsulfuric acidÞ
Occupational Safety and Health 1970 1987e
NO þ H2 O ! HNO2 ðnitrous acidÞ
Act (OSHA)
Oil Pollution Act 1973 1990f
3NO2 þ 2H2 O ! HNO3 ðnitric acidÞ
Resource Conservation and 1976 1984g
Recovery Act (RCRA) In many cases, these secondary pollutants can
Safe Drinking Water Act 1974 1986h have significant environmental effects, such as par-
Toxic Substances Control Act 1976 1984i ticipation in the formation of acid rain and smog [5].
(TSCA or TOSCA)
a
Water Quality Act
b
Water Pollution Control Act
c
Water Quality Act
Refinery Emissions
d
Also includes the Superfund Amendments and
Re-authorization Act (SARA) Amendments Crude oil, like any other raw material, can pro-
e
Several amendments during the 1980s duce chemical waste. By 1960, the crude oil refin-
f
Interactive with various water pollution acts
g
Includes the Federal Hazardous and Solid Waste
ing industry had become well established
Amendments throughout the world. Effluent water, atmospheric
h
Several amendments during the 1970s, the 1980s, and the emissions, and combustion products also became
1990s a focus of increased technical attention [4, 5].
i
Import rule enacted; as of June 22, 2016, includes the
Frank R. Lautenberg Chemical Safety for the 21st Century
Refineries produce a wide variety of products
Act, which updates the Toxic Substances Control Act from crude oil feedstocks and feedstock blends.
During crude oil refining, refineries use and gen-
erate chemical wastes, some of which are present
in air emissions, wastewater, or solid wastes
and nitrogen oxides (NOx – nitric oxide, NO, and (Table 9) [5]. Emissions are also created through
nitrogen dioxide, NO2) from fuel combustion the combustion of fuels and as by-products of
operations: chemical reactions occurring when crude oil frac-
tions are upgraded. A large source of air emissions
2½Ccrude oil þO2 ! 2CO are, generally, the process heaters and boilers that
½Ccrude oil þO2 ! CO2 produce carbon monoxide, sulfur oxides, and
½Scrude oil þO2 ! SO2 nitrogen oxides, leading to pollution and the for-
2½Ncrude oil þO2 ! 2NO mation of acid rain. Hence, there is the need for
½Ncrude oil þO2 ! NO2 gas-cleaning operations on a refinery site so that
such gases are cleaned from the gas stream prior to
Hydrogen sulfide (H2S) and ammonia (NH3) entry into the atmosphere.
are produced during refining of sulfur-containing Fugitive emissions of volatile hydrocarbons
and nitrogen-containing feedstocks: arise from leaks in valves, pumps, flanges, and
116 Petroleum Refining and Environmental Control and Environmental Effects

Petroleum Refining and Environmental Control and Petroleum Refining and Environmental Control and
Environmental Effects, Table 9 Examples of emissions Environmental Effects, Table 9 (continued)
and waste from refinery processes
Residual
Residual wastes
wastes Process Air emissions generated
Process Air emissions generated Solvent Fugitive solvents Little or no
Desalting Heater stack gasa, Crude extraction residual
fugitive emissionsb oil/desalter waste
sludge Dewaxing Fugitive solvents, Little or no
Atmospheric Heater stack gas, Little or no heaters residual
distillation fugitive emissions, residual waste
steam ejector waste Propane Heater stack gas, Little or no
emissions deasphalting fugitive propane residual
Vacuum Heater stack gas, Little or no waste
distillation fugitive emissions, residual Wastewater Fugitive emissions Various
steam ejector waste treatment (H2S, NH3, and types of
emissions hydrocarbons) sludged
Thermal Heater stack gas, Little or no a
Typically: CO, SOx, NOx, hydrocarbons, particulate
cracking fugitive emissions residual matter
waste b
Typically: hydrocarbons
Visbreaking Heater stack gas, Little or no c
Typically: CO, NOx, SOx
fugitive emissions residual d
API separator sludge, chemical precipitation sludge, bio-
waste logical sludge
Coking Heater stack gas, Coke dust
fugitive emissions,
decoking emissions
Catalytic Heater stack gas, Spent
cracking fugitive emissions catalysts, other similar sources where crude and its fractions
spent flow through the system. While individual leaks
catalyst
may be minor, the combination of fugitive emis-
fines
Catalytic Heater stack gas, Spent
sions from various sources can be substantial.
hydrocracking fugitive emissions, catalysts These emissions are controlled primarily through
catalyst fines leak detection and repair programs and occasion-
regeneration (dust) ally using special leak-resistant equipment.
Hydrotreating/ Heater stack gas, Spent The primary measure of the environmental
hydroprocessing fugitive emissions, catalyst
catalyst fines impact of refinery wastes is their toxicity to
regeneration gases exposed organisms. The toxicity of a substance
Hydrocracking Heater stack gas, Spent is most commonly reported as its concentration in
fugitive emissions, catalyst water that results in the death of half of the
catalyst fines
exposed organisms within a given length of
regeneration gases
time. Exposure times for toxicity tests are typi-
Alkylation Heater stack gas, Neutralized
fugitive emissions alkylation cally 96 h, although other times have been used.
(hydrocarbons) sludge
Isomerization Heater stack gas, Calcium Gaseous Emissions
HCl, fugitive chloride Gaseous emissions from crude oil refining create
emissions sludge
several environmental problems [5]. During com-
Polymerization Hydrogen sulfide Spent
from caustic catalyst bustion, the combination of hydrocarbons, nitro-
washing gen oxide, and sunlight results in localized low
Catalytic Heater stack gas, Spent levels of ozone or smog. This is particularly evi-
reforming fugitive emissions, catalyst dent in large urban areas and especially when air
catalyst fines
does not circulate well. Crude oil use in automo-
regeneration gasesa
biles also contributes to the problem in many
(continued)
Petroleum Refining and Environmental Control and Environmental Effects 117

areas. The primary environmental consequences of the hydrogen mixture, which results in its sep-
of air pollutants are respiratory difficulties in aration into fractions differing in carbon number,
humans and animals, damage to vegetation, and volatility, specific gravity, and other characteris-
soil acidification. Releases of hydrogen sulfide, of tics [5]. The most volatile fraction, which contains
course, can be fatal to those exposed. most of the gases which are generally dissolved in
The terms refinery gas and process gas are also the crude, is referred to as pipe still gas or pipe still
often used to include any of the gaseous products light ends and consists essentially of hydrocarbon
and by-products that emanate from a variety of gases ranging from methane to butane(s) or some-
refinery processes [4, 5]. There are also compo- times pentane(s).
nents of the gaseous products that must be The gas varies in composition and volume,
removed prior to release of the gases to the atmo- depending on crude origin and on any additions
sphere or prior to use of the gas in another part of to the crude made at the loading point. It is not
the refinery, i.e., as a fuel gas or as a process uncommon to re-inject light hydrocarbons such as
feedstock. propane and butane into the crude before dispatch
Refinery and natural gas streams may contain by tanker or pipeline. This results in a higher vapor
large amounts of acid gases, such as hydrogen pressure of the crude, but it allows one to increase
sulfide (H2S) and carbon dioxide (CO2) [3, the quantity of light products obtained at the refin-
4]. Hydrogen chloride (HCl), although not usually ery. Since light ends in most crude oil markets
considered to be a major pollutant in crude oil command a premium, while in the oil field itself
refineries, can arise during processing from the propane and butane may have to be re-injected or
presence of brine in crude oil that is incompletely flared, the practice of spiking crude oil with lique-
dried. It can also be produced from mineral matter, fied crude oil gas is becoming common.
and other inorganic contaminants are gaining In addition to the gases obtained by distillation
increasing recognition as a pollutant which of crude oil, more highly volatile products result
needs serious attention. Acid gases corrode refin- from the subsequent processing of naphtha and
ing equipment, harm catalysts, pollute the atmo- middle distillate to produce gasoline. Hydrogen
sphere, and prevent the use of hydrocarbon sulfide is produced in the desulfurization pro-
components in petrochemical manufacture. cesses involving hydrogen treatment of naphtha,
When the amount of hydrogen sulfide is large, it distillate, and residual fuel and from the coking or
may be removed from a gas stream and converted similar thermal treatments of vacuum gas oils and
to sulfur or sulfuric acid. Some natural gases residual fuels. The most common processing step
contain sufficient carbon dioxide to warrant in the production of gasoline is the catalytic
recovery as dry ice, i.e., solid carbon dioxide. reforming of hydrocarbon fractions in the heptane
And there is now a conscientious effort to mitigate (C7) to decane (C10) range.
the emission of pollutants from hydrotreating pro- In a series of processes commercialized under
cess by careful selection of process parameters the generic name reforming, paraffin and naph-
and catalysts [4, 5]. thene (cyclic nonaromatic) hydrocarbons are
Crude oil refining produces gas streams that altered structurally in the presence of hydrogen
often contain substantial amounts of acid gases and a catalyst into aromatics or isomerized to
such as hydrogen sulfide and carbon dioxide. more highly branched hydrocarbons. Catalytic
More particularly hydrogen sulfide arises from reforming processes thus not only result in the
the hydrodesulfurization of feedstocks that con- formation of a liquid product of higher octane
tain organic sulfur: number but also produce substantial quantities of
gases. The latter are rich in hydrogen but also
½Sfeedstock þH2 ! H2 S þ hydrocarbons contain hydrocarbons from methane to butanes,
with a preponderance of propane (CH3CH2CH3),
Crude oil refining involves, except for some of n-butane (CH3CH2CH2CH3), and iso-butane
the more viscous crude oils, a primary distillation [(CH3)3CH].
118 Petroleum Refining and Environmental Control and Environmental Effects

A second group of refining operations that hydrogen is to be recycled, the gases produced in
contribute to gas production is that of the catalytic this process again must be separated into lighter
cracking processes [5]. These consist of fluid-bed and heavier streams; any surplus recycle gas and
catalytic cracking in which heavy gas oils are the liquefied crude oil gas from the hydrocracking
converted into gas, liquefied crude oil gas, cata- process are both saturated.
lytic naphtha, fuel oil, and coke by contacting the Both hydrocracker gases and catalytic
heavy hydrocarbon with the hot catalyst. Both reformer gases are commonly used in catalytic
catalytic and thermal cracking processes, the latter desulfurization processes. In the latter, feedstocks
being now largely used to produce chemical raw ranging from light to vacuum gas oils are passed
materials, result in the formation of unsaturated at pressures of 500–1,000 psi with hydrogen over
hydrocarbons, particularly ethylene (CH2=CH2), a hydrofining catalyst. This results mainly in the
but also propylene (propene, CH3.CH=CH2), conversion of organic sulfur compounds to hydro-
iso-butylene [iso-butene, (CH3)2C=CH2] and the gen sulfide,
n-butenes (CH3CH2CH=CH2, and
CH3CH=CHCH3) in addition to hydrogen (H2), ½Sfeedstock þH2 ! H2 S þ hydrocarbons
methane (CH4) and smaller quantities of ethane
(CH3CH3), propane (CH3CH2CH3), and butanes The reaction also produces some light hydro-
[CH3CH2CH2CH3, (CH3)3CH]. Diolefins such as carbons by hydrocracking.
butadiene (CH2=CH.CH=CH2) are also present. Thus, refinery streams, while ostensibly being
Additional gases are produced in refineries hydrocarbon in nature, may contain large amounts
with visbreaking and/or coking facilities that are of acid gases such as hydrogen sulfide and carbon
used to process of the heaviest crude fractions. In dioxide. Most commercial plants employ hydro-
the visbreaking process, a high-boiling feedstock genation to convert organic sulfur compounds
(such as a residuum or heavy oil) is passed into hydrogen sulfide. Hydrogenation is effected
through externally fired tubes and undergoes by means of recycled hydrogen-containing gases
liquid-phase cracking reactions, which result in or external hydrogen over a nickel molybdate or
the formation of lower-boiling fuel oil compo- cobalt molybdate catalyst.
nents. Oil viscosity is thereby reduced, and some In summary, refinery process gas, in addition to
gases, mainly hydrogen, methane, and ethane, are hydrocarbons, may contain other contaminants,
formed. Substantial quantities of both gas and such as carbon oxides (COx, where x = 1 and/
carbon are also formed in coking (both delayed or 2), sulfur oxides (SOx, where x = 2 and/or 3), as
coking and fluid coking) in addition to the middle well as ammonia (NH3), mercaptans (R-SH), and
distillate and naphtha. When coking residual carbonyl sulfide (COS). The presence of these
high-boiling feedstocks, the feedstock is pre- impurities may eliminate some of the sweetening
heated and contacted with hot carbon (coke) processes, since some processes remove large
which causes extensive cracking of the feedstock amounts of acid gas but not to a sufficiently low
constituents of higher molecular weight to pro- concentration. On the other hand, there are those
duce lower-molecular-weight products ranging processes not designed to remove (or incapable of
from methane, via liquefied crude oil gas and removing) large amounts of acid gases, whereas
naphtha, to gas oil and heating oil. Products they can remove the acid gas impurities to very
from coking processes tend to be unsaturated, low levels when the acid gases are present only in
and olefin components predominate in the gases low-to-medium concentration in the gas. From an
produced during coking processes. environmental viewpoint, it is not how these gases
A further source of refinery gas is hydrocrack- can be utilized but it is the effects of these gases on
ing; a catalytic high-pressure pyrolysis process in the environment when they are introduced into the
the presence of fresh and recycled hydrogen and atmosphere.
the process is directed mainly at the production of In addition to the corrosion of equipment of
additional middle distillates and gasoline. Since acid gases, the escape into the atmosphere of
Petroleum Refining and Environmental Control and Environmental Effects 119

sulfur-containing gases can eventually lead to the such as mercaptans (RSH) and carbonyl sulfide
formation of the constituents of acid rain, i.e., the (COS). The presence of these impurities may
oxides of sulfur (SO2 and SO3). Similarly, the eliminate some of the sweetening processes
nitrogen oxides (NOx, x = 1 or 2 – nitric oxide, since some processes remove large amounts of
NO, and nitrogen dioxide, NO2) can also lead to acid gas but not to a sufficiently low concentra-
nitrous and nitric acids which are the other major tion. On the other hand, there are those processes
contributors to acid rain. The release of carbon that are not designed to remove (or are incapable
dioxide and hydrocarbons as constituents of refin- of removing) large amounts of acid gases. How-
ery effluents can also influence the behavior and ever, these processes are also capable of removing
integrity of the ozone layer. the acid gas impurities to very low levels when the
Hydrogen chloride, if produced during refin- acid gases are there in low-to-medium concentra-
ing, quickly picks up moisture in the atmosphere tions in the gas.
to form droplets of hydrochloric acid and, like On a regional level, the emission of sulfur
sulfur dioxide, is a contributor to acid rain oxides (SOx) and nitrogen oxides (NOx) can also
[5]. However, hydrogen chloride may exert severe cause the formation of acid species at high alti-
local effects because, unlike sulfur dioxide, it does tudes, which eventually precipitate in the form of
not need to participate in any further chemical acid rain, damaging plants, wildlife, and property.
reaction to become an acid, and under atmo- Most crude oil products are low in sulfur or are
spheric conditions that favor a buildup of stack desulfurized, and while natural gas sometimes
emissions in a large industrial complex or power includes sulfur as a contaminant, it is typically
plant, the amount of hydrochloric acid in rainwa- removed at the production site. At the global
ter could be quite high. level, there is concern that the increased use of
Natural gas is also capable of producing emis- hydrocarbon-based fuels will ultimately raise the
sions that are detrimental to the environment. temperature of the planet (global warming), as
While the major constituent of natural gas is meth- carbon dioxide reflects the infrared or thermal
ane, there are components such as carbon dioxide emissions from the earth, preventing them from
(CO), hydrogen sulfide (H2S), and mercaptans escaping into space (greenhouse effect). Whether
(thiols; R-SH), as well as trace amounts of sundry the potential for global warming becomes real will
other emissions. The fact that methane has a fore- depend upon how emissions into the atmosphere
seen and valuable end-use makes it a desirable are handled. There is considerable discussion about
product, but in several other situations, it is con- the merits and demerits of the global warming
sidered a pollutant, having been identified a theory, and the discussion is likely to continue for
greenhouse gas. some time [17]. Be that as it may, the atmosphere
Hydrogen chloride (HCl), although not usually can only tolerate pollutants up to a limiting value,
considered to be a major emission, is produced and that value needs to be determined. In the mean-
from mineral matter and the brines that often time, efforts must be made to curtail the use of
accompany crude oil during production and is noxious and foreign (nonindigenous) materials
gaining increasing recognition as a contributor to into the air. There are a variety of processes which
acid rain. However, hydrogen chloride may exert are designed for sulfur dioxide removal from gas
severe local effects because it does not need to streams [4, 5], but scrubbing process utilizing lime-
participate in any further chemical reaction to stone (CaCO3) or lime [Ca(OH)2] slurries has
become an acid. Under atmospheric conditions received more attention than other gas scrubbing
that favor a buildup of stack emissions in the processes. Most the gas scrubbing processes are
areas where hydrogen chloride is produced, the designed to remove sulfur dioxide from the gas
amount of hydrochloric acid in rainwater could be streams; some processes show the potential for
quite high. removal of nitrogen oxide(s).
In addition to hydrogen sulfide and carbon In summary, crude oil refining can result in
dioxide, gas may contain other contaminants, considerable gaseous emissions. It is a question
120 Petroleum Refining and Environmental Control and Environmental Effects

of degree insofar as the composition of the gas- chemical reaction. This will result in a persistent
eous emissions may vary from coal to crude oil, dark color traceable to reaction products that are
but the constituents are, in most cases, the same. redistributed as colloids. Separation may also be
difficult at low temperature because of high vis-
cosity of the stock, but this problem can be over-
Liquid Effluents come by dilution with light naphtha or with
It is convenient to divide the hydrocarbon com- propane.
ponents of crude oil into the following three clas- In addition, delayed coking also requires the
ses: (i) paraffins, which are saturated use of large volumes of water for hydraulic
hydrocarbons with straight or branched chains; cleaning of the coke drum. However, the process
(ii) naphthenes or alicyclic hydrocarbons, which water can be recycled if the oil is removed by
are saturated hydrocarbons containing one or skimming and suspended coke particles are
more rings, each of which may have one or more removed by filtration. If this water is used in a
paraffin side chains; and (iii) aromatic hydrocar- closed cycle, the impact of delayed coking on
bons, which are containing one or more aromatic water treatment facilities and the environment is
nuclei (e.g., benzene, naphthalene, and phenan- minimized. The flexicoking process offers one
threne) which may be co-joined with (substituted) alternative to direct combustion of coke for pro-
naphthene rings and/or paraffin side chains. cess fuel. The gasification section is used to pro-
Thermal processing can significantly increase cess excess coke to mixture of carbon monoxide
the concentration of polynuclear aromatic hydro- (CO), carbon dioxide (CO2), hydrogen (H2), and
carbons in the product liquid because the hydrogen sulfide (H2S) followed by treatment to
low-pressure hydrogen-deficient conditions favor remove the hydrogen sulfide. Currently, maxi-
aromatization of naphthene constituents and con- mum residue conversion with minimum coke pro-
densation of aromatics to form larger ring sys- duction is favored over gasification of coke.
tems. To the extent that more compounds like The environmental impact of hydrocarbons in
benzo(a)pyrene are produced, the liquids from water varies considerably. The toxicity of aro-
thermal processes will be more carcinogenic matic hydrocarbons is relatively high, while that
than asphalt. of straight-chain paraffins is relatively low. LC50
The sludge produced on acid treatment of values for the most common aromatic hydrocar-
crude oil distillates, even gasoline and kerosene, bons found in the crude oil industry (benzene,
is complex in nature [5]. Esters and alcohols are toluene, xylene, and ethylbenzene) are on the
present from reactions with olefins; sulfonation order of 10 ppm. Hydrocarbon concentrations of
products from reactions with aromatic com- less than 1 mg/l in water have been shown to have
pounds, naphthene compounds, and phenols; a sublethal impact on some marine organisms.
and salts from reactions with nitrogen bases. To High-molecular-weight paraffins, on the other
these, constituents must be added the various hand, are essentially nontoxic. Chronic exposures
products of oxidation-reduction reactions: coagu- of entire ecosystems to hydrocarbons, either from
lated resins, soluble hydrocarbons, water, and natural seeps or from crude oil facilities, have
free acid. shown no long- or intermediate-term impact; the
The disposal of the sludge is a comparatively ecosystems have all recovered when the source of
simple process for the sludge resulting from hydrocarbons was removed.
treating gasoline and kerosene, the so-called The toxicity of heavy metals found in the crude
light oils. The insoluble oil phase separates out oil refining industry varies widely. The toxicity of
as a mobile tar, which can be mixed and burned many heavy metals lies in their interference with
without too much difficulty. In all cases, careful the action of enzymes, which limits or stops nor-
separation of reaction products is important to the mal biochemical processes in cells. General
recovery of well-refined materials. This may not effects include damage to the liver, kidney, or
be easy if the temperature has risen because of reproductive, blood-forming, or nervous systems.
Petroleum Refining and Environmental Control and Environmental Effects 121

With some metals, these effects may also include Principal sources of releases to air from refin-
mutations or tumors. Heavy metal concentrations eries include (i) combustion plants, emitting sul-
allowed in drinking water vary for each metal but fur dioxide, oxides of nitrogen and particulate
are generally below about 0.01 mg per l. matter; (ii) refining operations, emitting sulfur
dioxide, oxides of nitrogen, carbon monoxide,
particulate matter, volatile organic compounds,
Solid Effluents
hydrogen sulfide, mercaptans and other sulfurous
Catalyst disposal is a major concern in all refiner-
compounds; and (iii) bulk storage operations and
ies. In many cases, the catalysts are regenerated at
handling of volatile organic compounds (various
the refinery for repeated use. Disposal of spent
hydrocarbons). Considering this, it is necessary to
catalysts is usually part of an agreement with the
consider the regulatory requirements, such as air
catalysts manufacturer, whereby the spent catalyst
emission permits stipulating limits for specific
is returned for treatment and remanufacture.
pollutants, and possibly health and hygiene permit
The formation of considerable quantities of
requirements as well as requirements for the mon-
coke in the coking processes is a cause for concern
itoring program and the requirements to upgrade
since it not only reduces the yield of liquid prod-
pollution abatement equipment.
ucts but also initiates the necessity for disposal of
the coke. Stockpiling to coke may be a partial
Storage and Handling of Crude Oil and Crude
answer unless the coke contains leachable mate-
Oil Products
rials that will endanger the ecosystem because of
Large quantities of environmentally sensitive
rain or snow melt. In addition, the generation and
crude oil products are stored in (i) tank farms
emission of sulfur oxides (particularly sulfur diox-
which include multiple tankage; (ii) single
ide) occurs from the combustion sulfur-containing
above-ground storage tanks, ASTs; (iii) semi-
coke as plant fuel. Sulfur dioxide (SO2) has a wide
underground, SUSTs, or underground storage
range of effects on health and on the environment.
tanks, USTs. Smaller quantities of materials may
These effects vary from bronchial irritation upon
be stored in drums and containers of assorted
short-term exposure to contributing to the acidifi-
compounds, such as lubricating oil, engine oil,
cation of lakes. Emissions of sulfur dioxide, there-
and other products for domestic supply.
fore, are regulated in many countries.
In light of this, it is also necessary to consider
(i) secondary containment of tanks and other stor-
age areas and integrity of hard standing (without
Entry into the Environment cracks, impervious surface) to prevent spills
reaching the wider environment: also secondary
It is almost impossible to transport, store, and containment of pipelines where appropriate;
refine crude oil without spills and losses. It is (ii) age, construction details, and testing program
difficult to prevent spills resulting from failure or of tanks; (iii) labeling and environmentally secure
damage on pipelines. It is also impossible to storage of drums (including waste storage);
install control devices for controlling the ecolog- (iv) accident/fire precautions and emergency pro-
ical properties of water and the soil along the cedures; and (v) disposal/recycling of waste or
length of all pipelines. The soil suffers the most “out of spec” oils and other materials.
ecological damage in the damage areas of pipe- There is a potential for significant soil and
lines. Crude oil spills from pipelines lead to irre- groundwater contamination to have arisen at crude
versible changes of the soil properties. The most oil refineries. Such contamination consists of:
affected soil properties by crude oil losses from
pipelines are filtration, physical, and mechanical i. Crude oil hydrocarbons including lower-
properties. These properties of the soil are impor- boiling, very mobile fractions (paraffins,
tant for maintaining the ecological equilibrium in cycloparaffins, and volatile aromatics such as
the damaged area. benzene, toluene, ethylbenzene, and xylenes)
122 Petroleum Refining and Environmental Control and Environmental Effects

typically associated with gasoline and similar that may draw upon the main supply of water,
boiling range distillates or water held in lagoons, or from reservoirs or
ii. Middle distillate fractions (paraffins, cyclo- neighboring water courses. Such water will be
paraffins, and some polynuclear aromatics) polluting and require containment. Thus,
associated with diesel, kerosene, and lower- refining facilities require significant volumes
boiling fuel oil, which are also of significant of water for on-site processes (e.g., coolants,
mobility blowdowns, etc.) as well as for sanitary and
iii. Higher-boiling distillates (long-chain paraf- potable use. Wastewater will derive from
fins, cycloparaffins, and polynuclear aro- these sources (process water) and from storm
matics) that are associated with lubricating water runoff. The latter could contain signifi-
oil and heavy fuel oil cant concentrations of crude oil product.
iv. Various organic compounds associated with Crude oil hydrocarbons, either dissolved,
crude oil hydrocarbons or produced during emulsified, or occurring as free-phase, will be
the refining process, e.g., phenols, amines, the key constituents although wastewater may
amides, alcohols, organic acids, nitrogen, also contain significant concentrations of phe-
and sulfur-containing compounds nols, amines, amides, alcohols, ammonia, sul-
v. Other organic additives, e.g., antifreeze fide, heavy metals, and suspended solids.
(glycols), alcohols, detergents, and various Wastewaters may be collected in separate
proprietary compounds drainage systems (for process, sanitary and
vi. Various heavy metals storm water) although industrial and storm
Key sources of such contamination at crude water systems may in some cases be com-
oil refineries are at (i) transfer and distribution bined. In addition, ballast water from bulk
points in tankage and process areas, also gen- crude tankers may be pumped to receiving
eral loading and unloading areas, (ii) land farm facilities at the refinery site prior to removal
areas, (iii) tank farms, (iv) individual above- of floating oil in an interceptor and treatment
ground storage tanks and particularly individ- as for other wastewater streams. On-site treat-
ual underground storage tanks, (v) additive ment facilities may exist for wastewater, or
compounds, and (vi) pipelines, drainage areas treatment may take place at a public wastewa-
as well as on-site waste treatment facilities, ter treatment plant. Storm water/process water
impounding basins, lagoons, especially if is generally passed to a separator or intercep-
unlined. tor prior to leaving the site which takes out
While contamination may be associated with free-phase oil (i.e., floating product) from the
specific facilities, the contaminants are rela- water prior to discharge or prior to further
tively highly mobile in nature and have the treatment (e.g., in settling lagoons). Discharge
potential to migrate significant distances from from wastewater treatment plants is usually
the source in soil and groundwater. Crude oil passed to a nearby watercourse.
hydrocarbon contamination can take several Other wastes that are typical of a refinery
forms: free-phase product, dissolved-phase, include (i) waste oils, process chemicals, and
emulsified phase, or vapor phase. Each form still resides; (ii) non-specification chemicals
will require different methods of remediation and/or products; (iii) waste alkali such as
so that cleanup may be complex and expensive. sodium hydroxide; (iv) waste oil sludge from
In addition, crude oil hydrocarbons include sev- interceptors, tanks, and lagoons; and (v) solid
eral compounds of significant toxicity, e.g., ben- wastes such as cartons, rags, catalysts,
zene and some polyaromatics are known and coke.
carcinogens. Vapor phase contamination can
be of significance in terms of odor issues. Release into the Environment
Due to the obvious risk of fire, refineries Crude oil products released into the environment
are equipped with sprinkler or spray systems undergo weathering processes with time. These
Petroleum Refining and Environmental Control and Environmental Effects 123

processes include evaporation, leaching (transfer (dispersants). These droplets may remain in the
to the aqueous phase) through solution and water column or coalesce with other droplets and
entrainment (physical transport along with the gain enough buoyancy to resurface. Dispersed oil
aqueous phase), chemical oxidation, and micro- tends to biodegrade and dissolve more rapidly
bial degradation. The rate of weathering is highly than floating slicks because of high surface area
dependent on environmental conditions. For relative to volume. Most of this process occurs
example, gasoline, a volatile product, will evapo- from about half an hour to half a day after the spill.
rate readily in a surface spill, while gasoline
released below 10 ft of clay topped with asphalt Dissolution
will tend to evaporate slowly (weathering pro- Dissolution is the loss of individual oil com-
cesses may not be detectable for years). pounds into the water. Many of the acutely toxic
An understanding of weathering processes is components of oils such as benzene, toluene, and
valuable to environmental test laboratories. xylene will readily dissolve into water. This pro-
Weathering changes product composition and cess also occurs quickly after a discharge but
may affect testing results, the ability to tends to be less important than evaporation. In a
bioremediate, and the toxicity of the spilled prod- typical marine discharge, generally less than 5%
uct. Unfortunately, the database available on the v/v of the benzene is lost to dissolution while
composition of weathered products is limited. greater than 95% v/v is lost to evaporation. For
However, biodegradation processes, which influ- alkylated polynuclear aromatic compounds, solu-
ence the presence and the analysis of crude oil bility is inversely proportional to the number of
hydrocarbons at a site, can be very complex. The rings and extent of alkylation. The dissolution
extent of biodegradation is dependent on many process is thought to be much more important in
factors including the type of microorganisms pre- rivers because natural containment may prevent
sent, environmental conditions (e.g., temperature, spreading, reducing the surface area of the slick,
oxygen levels, and moisture), and the predomi- and thus retarding evaporation. At the same time,
nant hydrocarbon types. In fact, the primary factor river turbulence increases the potential for mixing
controlling the extent of biodegradation is the and dissolution. Most of this process occurs
molecular composition of the crude oil contami- within the first hour of the spill.
nant. Multiple ring cycloalkanes are hard to Aromatics, and especially the benzene, tolu-
degrade, while polynuclear aromatic hydrocar- ene, ethyl benzene, and xylene isomers (BTEX)
bons display varying degrees of degradation. blend, tend to be the most water-soluble fraction
Straight-chain alkanes biodegrade rapidly with of crude oil. Crude oil-contaminated groundwater
branched alkanes and single-saturated ring com- tends to be enriched in aromatics relative to other
pounds degrading more slowly. crude oil constituents. Relatively insoluble hydro-
The primary processes determining the fate of carbons may be entrained in water through
crude oils and oil products after a spill are adsorption into kaolinite particles suspended in
(i) dispersion, (ii) dissolution, (iii) emulsification, the water or as an agglomeration of oil droplets
(iv) evaporation, (v) leaching, (vi) sedimentation, (microemulsion). In cases where groundwater
(vii) spreading, and (viii) wind. These processes contains only dissolved hydrocarbons, it may not
are influenced by the spill characteristics, envi- be possible to identify the original crude oil prod-
ronmental conditions, and physicochemical prop- uct because only a portion of the free product will
erties of the spilled material. be present in the dissolved phase. As whole prod-
uct floats on groundwater, the free product will
Dispersion gradually lose the water-soluble compounds.
The physical transport of oil droplets into the Groundwater containing entrained product will
water column is referred to as dispersion. This is have a gas chromatographic fingerprint that is a
often a result of water surface turbulence but also combination of the free product chromatogram
may result from the application of chemical agents plus enhanced amounts of the soluble aromatics.
124 Petroleum Refining and Environmental Control and Environmental Effects

Generally, dissolved aromatics may be found Leaching

quite far from the origin of a spill, but entrained Leaching is the loss or extraction of certain mate-
hydrocarbons may be found in water close to the rials from a carrier into a liquid (usually, but not
crude oil source. Oxygenates, such as methyl-t- always a solvent). Leaching processes introduce
butyl ether (MTBE), are even more water soluble hydrocarbon into the water phase by solubility
than aromatics and are highly mobile in the and entrainment. Leaching processes of crude oil
environment. products in soils can have a variety of potential
scenarios. Part of the aromatic fraction of a crude
oil spill in soil may partition into water that has
Emulsification
been in contact with the contamination.
Certain oils tend to form water-in-oil emulsions
(where water is incorporated into oil) or “mousse”
Sedimentation or Adsorption
as weathering occurs. This process is significant
As mentioned above, most oils are buoyant in water.
because, for example, the apparent volume of the
However, in areas with high levels of suspended
oil may increase dramatically, and the emulsifica-
sediment, crude oil constituents may be transported
tion will slow the other weathering processes,
to the river, lake, or ocean floor through the process
especially evaporation. Under certain conditions,
of sedimentation. Crude oil and crude oil constitu-
these emulsions may separate and release rela-
ents may adsorb on to sediments and sink or be
tively fresh oil. Most of this process occurs from
ingested by zooplankton and excreted in fecal pel-
about half a day to 2 days after the spill.
lets that may settle to the bottom. Crude oil stranded
on shorelines also may pick up sediments, float with
Evaporation the tide, and then sink. Most of this process occurs
Evaporative processes are very important in the from about 2 to 7 days after the spill.
weathering of volatile crude oil products and may
be the dominant weathering process for gasoline. Spreading
Automotive gasoline, aviation gasoline, and some As oil enters the environment, it begins to spread
grades of jet fuel (such as the fuel designated JP-4) immediately. The viscosity of the oil, its pour point,
contain 20–99% v/v highly volatile constituents and the ambient temperature will determine how
(i.e., constituents with less than nine carbon rapidly the oil will spread, but light oils typically
atoms). spread more rapidly than heavy oils. The rate of
Evaporative processes begin immediately after spreading and ultimate thickness of the oil slick will
oil is discharged into the environment. Some light affect the rates of the other weathering processes.
products (like 1-ring to 2-ring aromatic hydrocar- For example, discharges that occur in geographi-
bons and/or low-molecular-weight alkanes less cally contained areas (such as a pond or slow-
than n-C15) may evaporate entirely; a significant moving stream) will evaporate more slowly than
fraction of heavy refined oils also may evaporate. if the oil could spread. Most of this process occurs
For crude oils, the amount lost to evaporation can within the first week after the spill.
typically range from approximately 20 to 60%
v/v. The primary factors that control evaporation Wind
are the composition of the oil, slick thickness, Wind (aeolian) transport (relocation by wind) can
temperature and solar radiation, wind speed, and also occur and is relevant when catalyst dust and
wave height. While evaporation rates increase coke dust are considered. Dust becomes airborne
with temperature, this process is not restricted to when winds traversing arid land with little vege-
warm climates. tation cover pick up small particles such as cata-
It is not unusual for evaporative processes, lyst dust, coke dust, and other refinery debris and
however, to be working simultaneously with send them skyward. Wind transport may occur
other processes to remove the volatile aromatics through suspension, saltation, or creep of the
such as benzene and toluene. particles.
Petroleum Refining and Environmental Control and Environmental Effects 125

Toxicity 1. Ignitability-Flammability
A liquid that has a flash point of less than
With few exceptions, the constituents of crude oil, 60
C (140
F) is considered ignitable. Some
crude oil products, and the various emissions are examples are benzene, hexane, heptane, ben-
hazardous to the health. There are always excep- zene, pentane, crude oil ether (low boiling),
tions that will be cited in opposition to such a toluene, and xylene isomers.
statement, the most common exception being the 2. Corrosivity
liquid paraffin that is used medicinally to lubricate An aqueous solution that has a pH of less
the alimentary tract. The use of such medication is than or equal to 2 (high acidity), or greater than
common among miners who breathe and swallow or equal to 12.5 (high alkalinity), is considered
coal dust every day during their work shifts. corrosive. Most crude oil constituents and
Another approach is to consider crude oil con- crude oil products are not corrosive, but many
stituents in terms of transportable materials, the of the chemicals used in refineries are corro-
character of which is determined by several chem- sive. Corrosive materials also include sub-
ical and physical properties (i.e., solubility, vapor stances such as sodium hydroxide and some
pressure, and propensity to bind with soil and other acids or bases.
organic particles). These properties are the basis 3. Reactivity
of measures of leachability and volatility of indi- Chemicals that react violently with air or
vidual hydrocarbons. Thus, crude oil transport water are considered hazardous. Examples are
fractions can be considered by equivalent carbon sodium metal, potassium metal, phosphorus,
number to be grouped into 13 different fractions. etc. Reactive materials also include strong oxi-
The analytical fractions are then set to match these dizers such as perchloric acid and chemicals
transport fractions, using specific n-alkanes to capable of detonation when subjected to an
mark the analytical results for aliphatic com- initiating source, such as solid, dry <10%
pounds and selected aromatic compounds to picric acid, benzoyl peroxide, or sodium boro-
delineate hydrocarbons containing benzene rings. hydride. Solutions of certain cyanide or sul-
The range of chemicals in crude oil and crude fides that could generate toxic gases are also
oil products is so vast that summarizing the prop- classified as reactive.
erties and/or the toxicity or general hazard of The potential for finding such chemicals in
crude oil in general or even for a specific crude a refinery is subject to the function and product
oil is a difficult task. However, crude oil and some slate of the refinery and/or the petrochemical
crude oil products, because of the hydrocarbon complex.
content, are at least theoretically biodegradable 4. Hazardous Chemicals
but large-scale spills can overwhelm the ability Many chemicals have been shown in scien-
of the ecosystem to break the oil down. The tox- tific studies to have toxic, carcinogenic, muta-
icological implications from crude oil occur pri- genic, or teratogenic effects on humans or
marily from exposure to or biological metabolism other life forms and are designated either as
of aromatic structures. These implications change acutely hazardous waste or as toxic waste
as an oil spill ages or is weathered. [16]. Substances found to be fatal to humans
in low doses, or in the absence of data on
human toxicity, have been shown in studies to
Lower-Boiling Constituents have an oral LD50 toxicity (rat) of less than
Many of the gaseous and liquid constituents of the 2 mg per l, or a dermal LD50 toxicity (rabbit) of
lower-boiling fractions of crude oil and in crude less than 200 mg per kg or is otherwise capable
oil products fall into the class of chemicals which of causing or significantly contributing to an
have the one or more of the following character- increase in serious irreversible, or incapacitat-
istics are hazardous by the Environmental Protec- ing reversible illness is designated as Acute
tion Agency. Hazardous Waste. Materials containing any of
126 Petroleum Refining and Environmental Control and Environmental Effects

the toxic constituents so listed are to be con- aromatic hydrocarbon derivatives (PNAs).
sidered hazardous waste, unless, after consid- Moreover, both atmospheric and hydrospheric
ering the following factors, it might reasonably impacts must be assessed when considering
be concluded that the waste is not capable of toxic implications from a crude oil release
posing a substantial present or potential hazard containing significant quantities of these
to public health or the environment when single-ring aromatic compounds.
improperly treated, stored, transported or dis-
posed of, or otherwise managed.
The relative proportion of hazardous con- Higher-Boiling Constituents
stituents present in crude oil is typically quite Naphthalene and its homologs are less acutely
variable. Therefore, contamination will vary toxic than benzene but are more prevalent for a
from one site to another. In addition, as the longer period during oil spills. The toxicity of
volatility of the crude oil constituents different crude oils and refined oils depends on
decreases (the general progression from lower not only the total concentration of hydrocarbons
molecular weight to higher molecular weight but also the hydrocarbon composition in the
constituents), the greater the percentage of water-soluble fraction (WSF) of crude oil, water
polynuclear aromatic hydrocarbons and other solubility, concentrations of individual compo-
semi-volatile constituents or nonvolatile con- nents, and toxicity of the components. The water-
stituents (many of which are not so immedi- soluble fractions prepared from different oils will
ately toxic as the volatiles but can result in vary in these parameters. Water-soluble fractions
long-term/chronic impacts). These higher- (WSF) of refined oils – for example, No. 2 fuel
molecular-weight constituents thus need to be oil and Bunker C oil (Table 3) – are more toxic
analyzed for the semi-volatile compounds that than water-soluble fraction of crude oil to several
typically pose the greatest long-term risk. species of fish (killifish and salmon). Com-
In addition to large oil spills, crude oil hydro- pounds with either more rings or methyl substi-
carbons are released into the aquatic environ- tutions are more toxic than less substituted
ments from natural seeps as well as nonpoint- compounds but tend to be less water soluble
source urban runoffs. Acute impacts from mas- and thus less plentiful in the water-soluble
sive one-time spills are obvious and substantial. fraction.
The impacts from small spills and chronic The solubility of hydrocarbon components in
releases are the subject of much speculation crude oil products is an important property when
and continued research. Clearly, these inputs assessing toxicity. The water solubility of a sub-
of crude oil hydrocarbons have the potential stance determines the routes of exposure that are
for significant environmental impacts, but the possible. Solubility is approximately inversely
effects of chronic low-level discharges can be proportional to molecular weight; lighter hydro-
minimized by the net assimilative capacities of carbons are more soluble in water than higher-
many ecosystems, resulting in little detectable molecular-weight compounds. Lower-molecular-
environmental harm. weight hydrocarbons (C4–C8, including the aro-
Crude oil is naturally weathered according matic compounds) are relatively soluble, up to
to its physical and chemical properties, but about 2,000 ppm, while the higher-molecular-
during this process, living species within the weight hydrocarbons are nearly insoluble. Usu-
local environment may be affected via one or ally, the most soluble components are also the
more routes of exposure, including ingestion, most toxic.
inhalation, dermal contact, and, to a much Finally, the toxicity of crude oil may be
lesser extent, bioconcentration through the affected by factors such as “weathering” time or
food chain. Aromatic compounds of concern the addition of oil dispersants. Weathered crude
include alkylbenzene derivatives, toluene, oil and fresh crude oil may have different toxic-
naphthalene derivatives, and polynuclear ities, depending on oil type and weathering time.
Petroleum Refining and Environmental Control and Environmental Effects 127

Total Petroleum Hydrocarbons measured, gross quantity without identification of its

The term total petroleum hydrocarbons (TPH) is constituents, the total petroleum hydrocarbons data
used environmentally to describe the family of still represent a mixture. Thus, the data derived from
several hundred chemical compounds that origi- measurement of the crude oil hydrocarbons in an
nally come from crude oil composed of crude oil ecosystem is not a direct indicator of risk to humans
constituents and products that fall into the range or to the environment. The data may even be the
from C1 to beyond C35 [18, 19]. For example, results from one of several analytical methods, some
refined products from crude oil include of which have been used for decades and others
(i) gasoline that characteristically has total petro- developed in the past several years.
leum hydrocarbons composed of hydrocarbons Crude oil products, themselves, are the source
between the size range of normal hexane (n-C6) of the many components but do not adequately
and normal dodecane (n-C12) and (ii) diesel fuel define total petroleum hydrocarbons. However,
that characteristically is composed of hydrocar- the composition of crude oil products assist in
bons found between the size range of normal understanding the hydrocarbons that become envi-
undecane (n-C11) and normal pentacosane ronmental contaminants, but any ultimate exposure
(n-C25). However, crude oil and the higher-boiling is determined also by how the product changes
crude oil products (such as the non-distillable with use, by the nature of the release, and by the
residual fuel oil and asphalt) are not limited to the hydrocarbon’s environmental fate. When crude oil
carbon ranges shown and will contain many higher products are released into the environment,
carbon number and higher-molecular-weight con- changes occur that significantly affect their poten-
stituents. Thus, chemicals that may be included in tial effects. Physical, chemical, and biological pro-
the total petroleum hydrocarbons are hexane, hep- cesses change the location and concentration of
tane (and higher-molecular-weight homologs), hydrocarbons at any site.
benzene, toluene, xylenes (and the higher molecu- Analysis for total petroleum hydrocarbons
lar weight homologs), and naphthalene as well as [20] provides a one number value of the crude
the constituents of other crude oil products such as oil hydrocarbons in an ecosystem. It does not,
gasoline and diesel fuel. It is likely that samples of however, provide information on the composi-
the total petroleum hydrocarbons collected at a tion (i.e., individual constituents) of the hydro-
specific site will contain only some, or a mixture, carbon mixture. The amount of hydrocarbon
of these chemicals. contaminants measured by this method depends
Crude oil hydrocarbons may enter the environ- on the ability of the solvent used to extract the
ment through accidents, from industrial releases, hydrocarbon from the environmental media and
or as by-products from commercial or private uses the absorption of infrared light (infrared spec-
such as direct release into water through spills or troscopy) by the hydrocarbons in the solvent
leaks. When release into water occurs, some of the extract. The method is not specific to hydrocar-
hydrocarbons float on the water and form surface bons and does not always indicate crude oil con-
films, while others may sink and form bottom tamination since humic acid, a non-crude oil
sediments. Bacteria and microorganisms in the material and a constituent of many soils, can be
water have the potential to break down some of detected by this method. Another analytical
the hydrocarbons over varying periods of time method commonly used for total petroleum
that are dependent on the ambient conditions. On hydrocarbons [20] gives the concentration of
the other hand, hydrocarbons that are spilled on to purgeable and extractable hydrocarbons. These
the soil may remain for a long time. are sometimes referred to as gasoline range
In addition, the amount of total petroleum hydro- organics (GRO) and diesel range organics
carbons is the measurable of crude oil-based hydro- (DRO) because the boiling point ranges of the
carbon in an environmental medium, whether it is hydrocarbon in each roughly correspond to those
air, water, or land. It is, thus, dependent on analysis of gasoline (C6 to C10–12) and diesel fuel (C8–12
of the medium in which it is found, and since it is a to C24–26), respectively.
128 Petroleum Refining and Environmental Control and Environmental Effects

An important feature of the analytical methods into place over the past four decades [5]. This
for the total petroleum hydrocarbons is the use of includes all pollutants such as hazardous and
an equivalent carbon number index (EC). This non-hazardous wastes, regulated and unregulated
index represents equivalent boiling points for chemicals from all sources.
hydrocarbons and is the physical characteristic Pollution prevention can be accomplished by
that is the basis for separating crude oil (and reducing the generation of wastes at their source
other) components in chemical analysis. (source reduction) or by using, reusing, or
reclaiming wastes once they are generated
Wastewater (environmentally sound recycling). However, envi-
Several wastewater issues face the refining indus- ronmental analysis plays a major role in determining
try. These issues include chemicals in waste pro- if emissions-effluents (air, liquid or solid) fall within
cess waters. However, efforts by the industry are the parameters of the relevant legislation. For exam-
being continued to eliminate any water contami- ple, issues to be addressed are the constituents of
nation that may occur, whether it be from inad- gaseous emissions, the sulfur content of liquid fuels,
vertent leakage of crude oil or crude oil products and the potential for leaching contaminants (through
or leakage of contaminated water from one or normal rainfall or through the agency of acid rain)
more processes. In addition to monitoring from solid products such as coke.
organics in the water, metal concentration must
be continually monitored since heavy metals tend Waste Management
to concentrate in the body tissues of fish and The most effective way to minimize environmen-
animals and increase in concentration as they go tal impact from drilling and production activities
up the food chain. General sewage problems face is to develop and implement an effective waste
every municipal sewage treatment facility regard- management plan. Waste management plans iden-
less of size. tify the materials and wastes in an ecosystem and
Primary treatment (solid settling and removal) list the best way to manage, treat, and dispose of
is required, and secondary treatment (use of bac- those wastes. A waste management plan should
teria and aeration to enhance organic degradation) also include an environmental audit to determine
is becoming more routine, and tertiary treatment whether existing activities are in compliance with
(filtration through activated carbon, applications relevant regulations.
of ozone, and chlorination) has been, or is being, The effective management of each waste con-
implemented by all refineries. sists of a hierarchy of preferred steps. The first and
Wastewater pretreaters that discharge water usually most important step is to minimize the
into sewer systems have new requirements. Pol- amount and/or toxicity of the waste that must be
lutant standards for sewage sludge have been set. handled. This is done by maintaining careful con-
Toxics in the water must be identified, and plans trol on chemical inventories, changing operations
must be developed to alleviate any problems. In to minimize losses and leaks, modifying or
addition, regulators have established, and con- replacing equipment to generate less waste, and
tinue to establish, water quality standards for pri- changing the processes used to reduce or elimi-
ority toxic pollutants. nate the generation of toxic wastes.
The next step in effective waste management is
to reuse or recycle wastes. If wastes contain valu-
Managing Wastes and Future Directions able components, those components can be seg-
regated or separated from the remainder of the
Pollution prevention is, simply, reduction or elim- waste stream and recovered for use. Wastes that
ination of discharges or emissions to the environ- cannot be reused or recycled must then be treated
ment. The limits of pollutants emitted to the and disposed of. A written waste management
atmosphere, the land, and water are defined by plan that completely describes the acceptable
various pieces of legislation that have been put options for handling every waste generated at
Petroleum Refining and Environmental Control and Environmental Effects 129

every site must be developed and effectively com- Several disposal methods are available for
municated to every employee involved with the crude oil refining wastes. The method used
wastes. depends on the type, composition, and regulatory
Prudent management practices focus available status of the waste. The primary disposal method
resources on the activities that pose the greatest for aqueous wastes is to inject them into Class II
risk to both the economic health of the company wells. If the quality of wastewater meets or
and the environment. The risks associated with exceeds regulatory limits, permits to discharge it
various waste management practices can be into surface waters may be obtained in some areas.
quantified and ranked through risk assessment The primary disposal methods for solid wastes are
studies. When properly managed, the risks and to bury them or to spread them over the land
hazards of refining operations can be reduced to surface. All free liquids normally must be
low levels. removed prior to disposal, either by mechanical
separation, evaporation, or the addition of solidi-
fying agents. Land treatment of wastes may be
Waste Treatment and Disposal prohibited if volatile and leachable fractions are
Most wastes require some type of treatment present in the wastes. Disposal can occur either
before they can be sent to disposal. Waste treat- on- or off-site. Underground injection of slurries
ment may include reducing the total volume of the has also been used for solids disposal in some
waste, lessening its toxicity, and/or altering its areas.
ability to migrate away from its disposal site.
One of the most important steps in waste treat- Waste Cleanup
ment is to segregate or separate the wastes into The most appropriate cleanup method will depend
their constituents, e.g., solid, aqueous, and hydro- on the contaminant and on the site characteristics.
carbon wastes. This isolates the most toxic com- The most common contaminated sites are those
ponent of the waste stream in a smaller volume that have spilled hydrocarbons in the soil and
and allows the less toxic components to be dis- those containing old drilling fluids. Several
posed of in less costly ways. Primary separation methods can be used to clean up sites. Mobile
occurs with properly selected and operated equip- hydrocarbons can be removed by drilling wells
ment, e.g., separation tanks, and heater treaters. or digging trenches and pumping the hydrocar-
Separation can be improved by using bons to the surface with groundwater for treat-
hydrocyclones, filter presses, gas flotation sys- ment. Volatile hydrocarbons can be removed by
tems, or decanting centrifuges. In arid areas, evap- injecting air and/or pulling a vacuum to vaporize
oration and/or percolation can be used to de-water those components. The use of heat, surfactants,
some wastes. and bioremediation to remove subsurface hydro-
Several methods are available for treating carbons is being studied. Dissolved hydrocarbons
hydrocarbon-contaminated solids like drill cut- in water and volatilized hydrocarbons in air can be
tings, produced solids, or soil. Solids can be removed by filtration or by absorption with acti-
washed by agitation in a jet of high-velocity vated carbon. In some cases, however, the con-
water, perhaps with an added surfactant. Solids taminated material may need to be completely
can also be mixed with an oil-wet material such as removed for off-site treatment and disposal.
coal or activated carbon that absorbs the hydro-
carbons and can be separated from the denser
solids by subsequent flotation. An emerging and Future Directions
promising technology for hydrocarbon removal Over the past four decades, the refining industry
from contaminated solids is bioremediation. has experienced significant changes in oil market
Other treatment methods include distillation, sol- dynamics, resource availability, and technological
vent extraction, incineration, and critical/super- advancements. Advancements made in explora-
critical fluid extraction. tion, production, and refining technologies allow
130 Petroleum Refining and Environmental Control and Environmental Effects

utilization of resources such as heavy oil and tar coupled with the associated infrastructure
sand bitumen that were considered economically required to implement them leaves traditional
and technically unsuitable in the middle decades refining of crude oil hydrocarbons for transporta-
of the past century. Along with the many chal- tion fuels as the modus operandi for the foresee-
lenges, it is imperative for refiners to raise their able future, which for the purposes of this text is
operations to new levels of performance. Merely seen to be 50 years. The near future challenge for
extending today’s performance incrementally will refiners will be how to harness new technologies
fail to meet most company’s performance goals. to remain alive in a changing global marketplace.
Crude oil refining in the twenty-first century
will continue to be shaped by the factors such as
consolidation of oil companies, dramatic changes Bibliography
in market demand, customization of products, and
a decrease in the API gravity and sulfur content of 1. Ray DL, Guzzo L (1990) Trashing the planet: how
the crude oil feedstocks. In fact, in addition to a science can help us deal with acid rain, depletion of the
ozone, and nuclear waste (among other things).
(hopeful but unlikely) plentiful supply of crude
Regnery Gateway, Washington
oil, the future of the refining industry will base on 2. Majumdar SB (1993) Regulatory requirements for
the following factors such as (i) increased operat- hazardous materials. McGraw-Hill, New York
ing costs or investments due to stringent environ- 3. Mokhatab S, Poe WA, Speight JG (2006) Handbook of
natural gas transmission and processing. Elsevier,
mental requirements for facilities and products
Amsterdam
and (ii) accelerating globalization resulting in 4. Speight JG (1996) Environmental technology hand-
stronger international crude oil price scenarios. book. Taylor & Francis Publishers, Washington
The effect of these factors is likely to reduce 5. Speight JG (2005) Environmental analysis and tech-
nology for the refining industry. Wiley, Hoboken
refinery profit margins further, and crude oil com-
6. Stark JL Falkler T (2008) Method for improving liquid
panies worldwide will need to make significant yield during thermal cracking of hydrocarbons. United
changes in their operation and structure to be States Patent 7,425,259 September 16
competitive on global basis. 7. Stark JL, Falkler T, Weers JJ, Zetlmeisl MJ
(2008) Method for improving liquid yield during ther-
As global crude oil consumption increases and
mal cracking of hydrocarbons. United States Patent
resources are depleted, the production of fuels and 7,416,654 August 26
petrochemicals from residua, heavy oil, and tar 8. Schucker RC (2003) Heavy oil upgrading process.
sand bitumen will increase significantly. Thus, United States Patent 6,524,469. February 25
9. Radovanović L, Speight JG (2011) Visbreaking: a
there must be an accompanying change in refinery
technology of the future. Proceedings. first interna-
technology because of the variability of the feed- tional conference – Process Technology and Environ-
stocks. At the same time, more stringent antipol- mental Protection (PTEP 2011). University of Novi
lution regulations are forcing greater restrictions Sad, Technical Faculty “Mihajlo Pupin,” Zrenjanin,
Republic of Serbia. December 7, 2011. pp 335–338
on fuel specifications. There are fundamental lim-
10. Speight JG (2012) Visbreaking: a technology of the
itations on how far current processes can go in past and the future. Sci Iran 19(3):569–5732012C
achieving proper control over feedstock behavior. 11. Shidhaye H, Kukade S, Kumar P, Rao PVC, Choudary
This means a movement from conventional means NV (2015) Improve FCC margins by processing more
vacuum resid in feed. Hydrocarb Process 94(12):35–38
of refining heavy feedstocks by using (typically) a
12. Scherzer J, Gruia AJ (1996) Hydrocracking science
coking process to more development and use of and technology. Marcel Dekker Inc., New York
more innovative processes that will produce the 13. Dolbear GE (1998) Chapter 7, Hydrocracking: reac-
maximum yields of liquid fuels (or other desired tions, catalysts, and processes. In: Speight JG
(ed) Petroleum chemistry and refining. Taylor &
products) from the feedstock.
Francis, Washington
Technological advances are on the horizon for 14. Rana MS, Sámano V, Ancheyta J, Diaz JAI
alternate sources of transportation fuels. For (2007) A review of recent advances on process technol-
example, gas-to-liquids and biomass-to-liquids ogies for upgrading of heavy oils and residua. Fuel
86:1216–1231
are just two of the concepts currently undergoing
15. EPA (2013) 2013 TRI National Analysis: Introduction.
considerable development and application. How- Toxics Release Inventory (TRI) Program. United States
ever, the state of many of these technologies Environmental Protection Agency, Washington, DC.
Petroleum Refining and Environmental Control and Environmental Effects 131

https://www.epa.gov/toxics-release-inventory-tri-pro 30. Khan MR, Patmore DJ (1997) In: Speight JG

gram/2013-tri-national-analysis-introduction. Accessed (ed) Chapter 6, Petroleum chemistry and refining.
28 Nov 2013 Taylor & Francis, Washington
16. EPA (2016) Toxics Release Inventory (TRI) Program. 31. Lee S, Speight JG, Loyalka S (2014) Handbook of
United States Environmetnal Protection Agency, alternative fuel technologies, 2nd edn. CRC Press,
Washington, DC. https://www.epa.gov/toxics-release- Taylor & Francis Group, Boca Raton
inventory-tri-program. Accessed 28 Nov, 2016 32. Lipton S, Lynch J (1994) Handbook of health hazard
17. Speight JG, Islam MR (2016) Peak energy – myth or control in the chemical process industry. Wiley,
reality. Scrivener Publishing, Beverly New York
18. Irwin RJ, VanMouwerik M, Stevens L, Seese MD, 33. McKetta JJ (ed) (1992) Petroleum processing hand-
Basham W (1997) Environmental contaminants ency- book. Marcel Dekker Inc., New York
clopedia. National Park Service, Water Resources 34. Occelli ML, Chianelli R (1996) Hydrotreating tech-
Division, Fort Collins nology for pollution control. Marcel Dekker Inc.,
19. Weisman W (1998) Analysis of petroleum hydrocarbons New York
in environmental media. Total Petroleum Hydrocarbons 35. Reis JC (1996) Environmental control in petroleum
Criteria Working Group Series. volume 1. Amherst Sci- engineering. Gulf Publishing Company, Houston
entific Publishers, Amherst, MA. (See also: volume 2: 36. Renzoni A, Fossi MC, Lari L, Mattei N (1994) Con-
Composition of Petroleum Mixtures, volume 2, 1998; taminants in the environment: a multidisciplinary
volume 3: Selection of Representation Total Petroleum assessment of risks to man and other organisms.
Hydrocarbons Fractions Based on Fate and Transport CRC Press, Boca Raton
Considerations, 1997; volume 4: Development of 37. Shih SS, Oballa MC (1991) Tar sand upgrading tech-
Fraction-Specific Reference Does and Reference con- nology. Symposium Series No. 282. American Insti-
centrations for Total Petroleum Hydrocarbons, 1997; tute for Chemical Engineers. New York
and volume 5: human Health Risk-Based Evaluation of 38. Speight JG (2000) The desulfurization of heavy oils
Petroleum Oil Contaminated Sites – Implementation of and residua, 2nd edn. Marcel Dekker, New York
the Working Group Approach, 1999) 39. Speight JG, Ozum B (2002) Petroleum refining pro-
20. EPA (2003) Collection of methods. United States cesses. Marcel Dekker, New York
Environmetnal Protection Agency, Washington, 40. Speight JG (2005) Environmental analysis and tech-
DC. https://www.epa.gov/measurements/collection- nology for the refining industry. Wiley, Hoboken
methods. Accessed 28 Nov 2016 41. Speight JG (2007) Natural gas: a basic handbook.
GPC Books, Gulf Publishing Company, Houston
42. Speight JG (2008) Handbook of synthetic fuels: properties,
Other Information Sources processes, and performance. McGraw-Hill, New York
21. Ali MF, El Ali BM, Speight JG (2005) Handbook of 43. Speight JG (2011) An introduction to petroleum tech-
industrial chemistry: organic chemicals. McGraw- nology, economics, and politics. Scrivener Publishing,
Hill, New York Beverly
22. Boyce A (1997) Introduction to environmental tech- 44. Speight JG (2011) The refinery of the future. Gulf
nology. Van Nostrand Reinhold, New York Professional Publishing, Elsevier, Oxford
23. Carson, Mumford (1995) The safe handling of 45. Speight JG, Arjoon KK (2012) Bioremediation of
chemicals in industry. Wiley, New York petroleum and petroleum poducts. Scrivener Publish-
24. Edwards JD (1995) Industrial wastewater treatment: a ing, Beverly
guidebook. CRC Press Inc., Boca Raton 46. Speight JG (2013) Heavy and extra heavy oil
25. El-Gendy NS, Speight JG (2015) Handbook of refin- upgrading technologies. Gulf Professional Publishing,
ery desulfurization. CRC Press, Taylor & Francis Elsevier, Oxford
Group, Boca Raton 47. Speight JG (2014) The chemistry and technology of
26. Forbes RJ (1958) A history of technology. V. Oxford petroleum, 5th edn. CRC Press, Taylor & Francis
University Press, Oxford Group, Boca Raton
27. Gary JH, Handwerk GE (2001) Petroleum refining: 48. Speight JG (2014) Oil and gas corrosion prevention.
technology and economics, 4th edn. Marcel Dekker Gulf Professional Publishing, Elsevier, Oxford
Inc., New York 49. Speight JG, Islam MR (2016) Peak energy – myth or
28. Hoiberg AJ (1960) Bituminous materials: asphalts, reality. Scrivener Publishing, Beverly
tars and pitches, I & II. Interscience, New York 50. Speight JG (2017) Handbook of petroleum refining.
29. Hsu CS, Robinson PR (2006) Practical advances in CRC Press, Taylor & Francis Group, Boca Raton
petroleum processing, vol 1 and 2. Springer, 51. Thibodeaux LJ (1995) Environmental chemodynamics.
New York Wiley, New York
 Reactants Species that are to be involved in a
Internal Combustion Engines, chemical reaction (in the context of this entry,
Developments in species that are to be involved in a combustion
reaction).
Timothy J. Jacobs Rich Air-fuel mixture is such that there is less air
Department of Mechanical Engineering, Texas than is chemically necessary to oxidize the
A&M University, College Station, TX, USA available fuel.
Stoichiometric Air-fuel mixture is chemically
Article Outline balanced such that there is the correct amount
of air to fully oxidize the available fuel.
Glossary
Definition of the Subject
Introduction Definition of the Subject
The Basics of Internal Combustion Engines
A Case Study: Diesel Engines Versus Gasoline An internal combustion (IC) engine is a thermody-
Engines namic work conversion device that converts chem-
Future Directions of Internal Combustion Engines ical energy (typically delivered to the engine in the
Bibliography form of a liquid or gaseous fuel) to work energy
(typically in the form of shaft work issuing from a
rotating crankshaft). The internal combustion
List of Abbreviations engine markedly distinguishes itself from other
types of power-producing equipment – most nota-
BDC Bottom dead center bly, the heat engine (e.g., steam engine or steam
EGR Exhaust gas recirculation cycle plant). In the former, chemical energy is
HCCI Homogeneous charge compression released, via combustion (i.e., rapid oxidation)
ignition mechanism, internal the same device that converts
IC Internal combustion the released energy to work energy. In the latter,
LTC Low-temperature combustion thermal energy is passed into the device via heat
TDC Top dead center transfer; the device thereby converts the thermal
energy to work energy. Because of this subtle
difference, the thermodynamic limits of maximum
Glossary efficiency of the internal combustion engine are
constrained differently than the thermodynamic
Combustion Rapid oxidation of a fuel-air mix- limit of maximum efficiency of the heat engine.
ture (reactants) converting reactants to products There are several types of internal combustion
and in the process releasing thermal energy. engines; the two most common being the piston/
Lean Air-fuel mixture is such that there is more cylinder reciprocating engine and the gas turbine
air than is chemically necessary to oxidize the engine. This entry constrains its discussion to just
available fuel. piston/cylinder reciprocating engines. As this
Products Species formed as the result of a chem- entry is meant to be brief, readers seeking addi-
ical reaction (in the context of this entry, spe- tional and thorough information are referred to the
cies formed as the result of a combustion associated cited entries and the authorities listed in
reaction). “Books and Reviews.”

© Springer Science+Business Media, LLC, part of Springer Nature 2020 133

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_430
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2018, https://doi.org/10.1007/978-1-4939-2493-6_430-3
134 Internal Combustion Engines, Developments in

Introduction The Basics of Internal Combustion

Engines
Internal combustion engines are pervasive to our
daily activities. They are the primary powerhouse Basic Operating Cycle
of the transportation industry. They serve as As this entry concentrates its discussions of IC
neighborhood and small municipality backup engines on those of the piston/cylinder recipro-
power generators or primary power stations. cating type, Fig. 1 [2] shows the basic geomet-
They provide convenience by powering lawn rical considerations of the piston/cylinder/
mowers, leaf blowers, and weed trimmers. They crankshaft arrangement, which kinematically is
deliver excitement and entertainment in pleasure described as a crank-slider mechanism. The
boats, race cars, and motor bikes. Their scale of chemical energy to work energy conversion
usability nearly spans the scales of classical phys- occurs inside the cylinder, usually bound on
ics, from as small as “micro-engines” that fit in the the sides by the cylinder walls, on the top by
palm of your hand to as large as marine engines
that scale three stories and use human-sized door-
ways for entry into the cylinder block.
The story of the internal combustion engine dates
back around 150 years ago, when J.J.E. Lenoir built
an engine that combusted coal-air mixtures in a
cylinder outfitted with a piston with a two-stroke
type fashion (without compression). Soon there-
after, Nicolaus A. Otto and his colleague Eugen
Langen expanded on the Lenoir concept, creating
an engine that had 11% efficiency compared to
Lenoir’s 5%. Determined to improve efficiencies
of internal combustion engines (which at 11%
were not much better than the steam engine),
Otto built the first engine operating on the four-
stroke principle which today serves as the primary
operating cycle of engines [1]. It is here noted that,
although Otto built the first working four-stroke
engine, Alphonse Beau de Rochas described in
theory the principles of the four-stroke cycle.
Further, Beau de Rochas outlined several condi-
tions to achieve maximum efficiency of the inter-
nal combustion engine [1].
From this point, and with the aid of several
important developers including Rudolf Diesel, Sir
Harry Ricardo, Robert Bosch, and Charles Ketter-
ing, the internal combustion engine has become one
of the most highly efficient, power dense, cost-
effective, easily maintained, and versatile power
machineries available to consumers. This entry
will provide some of the important basic informa-
tion about internal combustion engines and indicate
Internal Combustion Engines, Developments in,
some of the more recent developments that continue
Fig. 1 Illustration of a piston/cylinder arrangement, as
to make internal combustion engines competitive as often employed in a reciprocating piston internal combus-
the preferred power-producing technology. tion engine (Used with permission from [2])
Internal Combustion Engines, Developments in 135

the cylinder head (which typically houses the V max

rc ¼ (1)
gas exchange valves, such as the intake and V min
exhaust valves, and other important hardware
such as spark plugs and/or fuel injectors), and
on the bottom by the piston which reciprocates The compression ratio, as will be described in
within the cylinder. A rigid connecting rod fas- section on “Thermodynamic Analysis of Internal
tens the piston to an eccentric location on the Combustion Engines,” is a fundamentally critical
rotating crankshaft. The eccentric placement of parameter for controlling the efficiency (i.e., the
the connecting rod converts the reciprocating ratio of work energy out to chemical energy in) of
motion of the piston (i.e., boundary motion an internal combustion piston engine.
work) to the rotating motion of the crankshaft There are two major cycles used to exploit the
(i.e., shaft work). The eccentric placement of the piston engine’s conversion of chemical energy to
connecting rod to the crankshaft also dictates the work energy: a “two-stoke” cycle and a “four-
important geometrical parameter of the piston stroke” cycle. The earliest prototype engines were
engine called the “stroke.” The stroke and of the two-stroke variety (e.g., Lenoir and Otto/
“bore,” or diameter of the cylinder, create the Langen engines) [1]. In pursuit of achieving higher
displaced volume, Vd, of the piston engine. The efficiency, Otto (for whom the thermodynamic ideal
maximum volume of the cylinder, Vmax, is “Otto cycle” is named) built the four-stroke version
attained when the piston is at its bottommost of his engine [1]. Today, four-stroke cycle engines
position, a position referred to as “bottom dead are the dominant form; thus, most of the entry will
center” or BDC. The minimum volume, or clear- center on the details of the four-stroke cycle.
ance volume, Vc, is attained when the piston is at Four-stroke cycle engines require four strokes
its topmost position, a position referred to as of the piston to complete one power-producing
“top dead center” or TDC. The ratio between cycle, as shown in Fig. 2 [3]; the reader is also
Vmax and Vmin is called the compression ratio, referred to Fig. 4a to aid the discussion. Consider
rc, and is given as Eq. 1: first the cycle starting with the piston at TDC and

Internal Combustion Engines, Developments in, Fig. 2 Illustration of the four-stroke operating principle (Used with
permission from [3])
136 Internal Combustion Engines, Developments in

the intake valve open. The piston moves from cases, each cylinder undergoes the same processes
TDC to BDC, inducting fresh mixture (conven- but usually out of phase. For example, in an
tionally, fuel and air in a gasoline engine and only in-line four-cylinder engine (i.e., an engine that
air in a diesel engine) through the open intake has four cylinders oriented sequentially in a
valve in what is called the “intake stroke.” At single-line bank, an example of which is shown
some location near BDC, the intake valve closes, in Fig. 3), the cylinder processes are typically out
and the piston reverses its motion at BDC. Once of phase by 180 (720 /4). The firing order is
the valve closes, the piston/cylinder arrangement usually not linear, however, in order to ensure
creates a closed system. As the piston moves from smooth and continuous operation. For example,
BDC to TDC, the trapped mixture is compressed a four-cylinder firing order may go 1-3-2-4; that
in what is called the “compression stroke,” is, as the engine crankshaft rotates, cylinder 1 pro-
increasing the mixture’s temperature and pressure duces a power stroke first, followed by cylinder 3,
and decreasing its specific volume (i.e., increasing then by cylinder 2, and finally by cylinder 4
its density). At a point near TDC, combustion is (which is then followed again by cylinder 1).
expected to commence. In the case of a spark Finally, when describing an engine’s displace-
ignition engine (e.g., a conventional gasoline ment, it refers to a summation of each cylinder’s
engine), combustion is initiated by the release of displacement; thus, each cylinder displacement is
spark at a point near (usually advanced of) TDC. on average the total engine displacement divided
In the case of a compression ignition engine (e.g., by the number of cylinders.
a conventional diesel engine), combustion is ini-
tiated by injecting liquid fuel directly into the
How an Engine Makes Power
cylinder; the compressed air at elevated tempera-
The reciprocating motion of the piston in a cylin-
ture and pressure atomizes, vaporizes, and mixes
der is the means by which the chemical energy
with the injected fuel. After a period of time, the
released during combustion is converted into use-
high-temperature environment causes chemical
ful work. Work is transferred when a force acts
reaction and start of combustion. Around start of
through a displacement; in the case of the piston/
combustion, the piston reaches TDC, reverses
direction, and expands the cylinder volume as
combustion converts chemical energy into work
energy. This stroke takes on many names, includ-
ing “power stroke,” “combustion stroke,” and
“expansion stroke.” As the piston approaches
BDC, the exhaust valve opens, allowing the prod-
ucts of combustion to escape the cylinder. At
BDC, the piston reverses direction and motions
toward TDC with the exhaust valve open, in what
is called the “exhaust stroke.” Depending on the
engine’s crankshaft rotational speed – which can
vary from as low as 100 rev/min for large marine
application engines to as high as 15,000 rev/min
for race car engines – the four-stroke cycle
requires as much as about 1.2 s to as little as
8 ms to complete.
Up to this point, the discussion has centered on
the processes occurring in a single cylinder. Most
Internal Combustion Engines, Developments in,
engines, however, are composed of many cylin-
Fig. 3 Illustration of the Audi 2.0-L TSI, an example of
ders and take on various forms (e.g., in-line four- an in-line four-cylinder engine (Used with permission from
cylinder, “V6,” “V8,” and “W” engine). In such [4])
Internal Combustion Engines, Developments in 137

cylinder engine, the in-cylinder pressure, P, is The torque of an engine is routinely measured
interpreted as the force, and the changing cylinder with a dynamometer; in this way, the torque is con-
volume, dV, during piston strokes is interpreted as sidered “brake torque” or the amount of resistance
the displacement. Thus, the thermodynamic work torque the dynamometer must apply to “brake” the
of an engine cycle is given by Eq. 2: engine to a certain speed condition. Via application
of Eq. 3, the “brake work,” Wb, is determined. Only
þ
determining brake work, however, reveals no insight
W ¼ PdV ½kJ=cycle (2)
into the in-cylinder work processes. The in-cylinder
work processes can be calculated using in-cylinder
The work of the cycle given by Eq. 2 is con- pressure measurement that is precisely coupled to the
sidered boundary work, since it results from the in-cylinder volume via a crankshaft encoder and
changing boundary of the control system (in this detailed knowledge of the cylinder’s and crank-
context, the control system is that enclosed by slider’s geometries. An example “pressure-
the piston/cylinder arrangement, and the moving crankangle” diagram is shown Fig. 4a, where
boundary is manifested by the moving piston). crankangle is reported in degrees after top dead
Conventional engines convert this boundary work center ( ATDC) relative to “combustion TDC.”
of the piston to shaft work through the piston Also shown in Fig. 4a are the four strokes of the
connecting rod’s eccentrically located connection four-stroke cycle as described in section “Basic
to the crankshaft. The shaft work issuing from the Operating Cycle.” In-cylinder pressure is typically
crankshaft is often best interpreted as torque, t; measured with a piezoelectric pressure transducer
the work given by Eq. 2 is related to the torque of which is able to provide a fast response indication
the crankshaft via Eq. 3: during the engine cycle [5]. It should be noted that
measuring in-cylinder pressure is not a trivial task
W ¼ 2pnR t ½kJ=cycle (3) and great care must be taken to do it properly
[6–9]. When using digital equipment (i.e., an
where nR is the number of crankshaft revolutions analog-digital converter) to electronically record
per power cycle (i.e., nR = 2 for a four-stroke in-cylinder pressure, it becomes necessary to deter-
cycle and nR = 1 for a two-stroke cycle). The mine the sample rate, which is usually determined by
units for t in Eq. 3 are kN-m. the crankshaft encoder. Varying crankangle

a 120 Compression Expansion b 120 5

Intake Stroke Exhaust Stroke
Stroke Stroke
4.5

100 100 4
Intake / Exhaust Work
Pressure (bar)

3.5 “Pump Work”

3
80 80
Pressure (bar)

Pressure (bar)

2.5

2
60 60
1.5

1
0.5 1.0
40 40 Volume (L)

Compression / Expansion Work

20 “Gross Work”
20

0 0
–360 –270 –180 –90 0 90 180 270 360 0.5 1.0
Crankangle (°ATDC) Volume (L)

Internal Combustion Engines, Developments in, pressure as a function of cylinder volume, also illustrating the
Fig. 4 (a) Pressure as a function of crankshaft rotation, or areas of the P-v plane that render gross work and pump work
crankangle, in degrees after top dead center ( ATDC) and (b) (Data from author’s laboratory, Texas A&M University)
138 Internal Combustion Engines, Developments in

resolutions can be used, depending on the level of “pump,” and “net” only have relevance from indi-
precision needed of the analysis. For calculating cated data (i.e., in-cylinder pressure data), it is
in-cylinder work (described next), a crankangle res- often dropped as a designator on the work terms.
olution of up to 10 [6] can be used; for detailed Pump work, which will be nonzero when
combustion analysis, much finer resolution must be intake pressure is different from exhaust pressure
used (e.g., about 1 for gasoline engine combustion (nearly all situations), often decreases net work
and 0.25 for diesel engine combustion). The data relative to gross work (i.e., the gas exchange pro-
shown in Fig. 4 is recorded every 0.2 . In addition to cess requires the piston to do work on the gas, or
the crankangle resolution, the engine speed also pump work is negative). There are few situations,
determines the needed sample rate of the data acqui- with the use of a turbocharger or supercharger, for
sition system. Finally, it is equally important to know example, when intake pressure is greater than
the cylinder volume at each record of pressure when exhaust pressure and pump work is positive. In
calculating in-cylinder work. This requires precise such situations, net work will be greater than
phasing between the piston’s location and the crank- gross work.
shaft encoder; it also requires knowing the precise The difference between net work and brake
geometries of the cylinder and crank-slide mecha- work, as shown in Eq. 5, is the friction work, Wf,
nism. Specifically, the minimum volume (i.e., clear- of the engine. Friction, of course, always requires
ance volume), the piston stroke, and the cylinder work from the system; thus, brake work will
bore must be precisely known. It is often best to always be less than net work. In Eq. 5, Wf captures
determine these using precise measuring instru- all forms of mechanical friction of the engine,
ments, rather than relying on manufacturer specifica- including friction among crank-slider compo-
tions (which, although have tight tolerances, are nents, in bearings, in valve springs, and in various
nominal values). An example “pressure-volume,” accessories (e.g., water and oil pumps):
or “P-V,” diagram is shown in Fig. 4b.
Once a precise P-V diagram is determined, the W f ¼ W in  W b (5)
in-cylinder work associated with each process can
be determined. The area between the P-V curves, Engine researchers typically quantify all of the
as suggested by Eq. 2, represents the in-cylinder above-described work parameters on volume-
work or “indicated” work (named for the anti- normalized parameters, which in general repre-
quated use of a mechanical stylus-indicator device sent a “mean effective pressure.” The general
to record pressure [10]). There are two portions of definition for mean effective pressure, MEP, is
the typical four-stroke engine cycle, as shown in given as Eq. 6:
Fig. 4: (1) compression and expansion strokes
which in combination result in the “gross work,” W
MEP ¼ (6)
Wig, and (2) intake and exhaust strokes which in Vd
combination result in the “pump work,” Wip. It is
important to note that gross work and pump work The various mean effective pressures and their
correspond to the respective strokes of the piston, definitions are summarized in Table 1.
not necessarily the valve events (i.e., not neces- One of the major benefits of describing the
sarily the closed portion of the cycle versus open work of an engine in terms of mean effective
portion of the cycle). In combination, i.e., the sum pressure is that the “size” of the cylinder is
of gross work and pump work result in the “net removed from consideration. In other words, it is
work,” as given by Eq. 4: possible to produce more work from an engine
that has a larger displaced volume; however, the
W in ¼ W ig þ W ip (4) mean effective pressure may be lower (relative to
a lower displaced volume engine) due to a number
Note that the subscript “i” on the terms in Eq. 4 of other possible influencing parameters that
represents “indicated”; since terms “gross,” affect the engine’s ability to make power (i.e.,
Internal Combustion Engines, Developments in 139

Internal Combustion Engines, Developments in, volumetric efficiency of the engine depends
Table 1 Summary of various mean effective pressures on the chosen reference point. Thus, if the
describing the various work transfers defined for a recip-
rocating piston internal combustion engine volumetric efficiency of the whole breathing
stream is desired, the reference point is taken as
Name Definition
upstream of the intake air filter, for example. If,
Gross indicated mean effective IMEPg = Wig/
pressure (gross IMEP) Vd however, volumetric efficiency of just the
Net indicated mean effective pressure IMEPn = Win/ intake ports and through the valves is desired,
(net IMEP) Vd then the reference point is taken as the intake
Pump mean effective pressure (PMEP) PMEP = Wip/ manifold.
Vd • ra, i, as defined above, is the density of the
Friction mean effective pressure FMEP = Wf/ intake air at some reference point (kg/m3).
(FMEP) Vd
• F/A is the mass-based fuel-air ratio of the
Brake mean effective pressure BMEP = Wb/
(BMEP) Vd mixture.
• QHV is the heating value of the fuel (typically,
lower heating value is used since the products
leave the piston/cylinder system with water as
fuel conversion and volumetric efficiencies, fuel- a vapor) (kJ/kg).
air ratio, inlet air density, and fuel heating value). • N is the engine speed (rev/s).
To make these types of assessments, refer to Eq. 7,
which is developed from the basic definition of To reveal the form of mean effective pressure,
power, P (i.e., work per unit time), and respective Eq. 8 is given as a modified form of Eq. 7:
definitions of other involved parameters as shown
by Heywood [11]. An example of the use of Eq. 7 MEP ¼ f v ra, i ðF=AÞQHV (8)
to assess factors affecting power in a technology
comparison is provided in [12]: Thus, it is apparent from Eq. 8 how an engine
with a relatively small displacement might have a
f v ra, i ðF=AÞQHV V d N higher mean effective pressure than an engine
P¼ (7)
nR with a large displacement, even though the larger
displaced engine may produce more power. It is
where also clear from Eqs. 7 and 8 how the performance
(i.e., power) of an engine may be improved
• f is the fuel conversion efficiency (described beyond the “easy” action of increasing displaced
in section on “Thermodynamic Analysis of volume. One parameter, for example, that benefits
Internal Combustion Engines”). the power and mean effective pressure of the
• v is the volumetric efficiency or the engine’s engine is the fuel conversion efficiency. This
effectiveness at “breathing” air. It is defined as very important parameter is discussed in the
v ¼ r maV d , where ma is the actual mass of air next section.
a, i

inducted per cycle (kg) and ra,i is the density of Finally, this section concludes by illustrating a
the intake air at some reference point (kg/m3) typical power/torque/speed curve . Apparent from
(usually the intake manifold but also may be Eq. 3 and the definition of power, there is a rela-
atmospheric air upstream of the engine air fil- tionship among power, torque, and speed of an
ter). Note that Vd has units of m3 in Eq. 7. It is engine. This relationship for a typical internal
important to note that volumetric efficiency combustion engine is shown in Fig. 5. There are
only quantifies the engine’s ability to breath few interesting features to point out in this figure.
air, i.e., not a fuel-air mixture. Thus, for exam- The first, a practical feature, is the identification of
ple, premixing fuel with air prior to induction “rated torque” and “rated power.” Engines are
tends to lower the engine’s volumetric effi- usually “sized” based on the speed at which they
ciency. It is also important to note that develop maximum torque (also known as rated
140 Internal Combustion Engines, Developments in

Internal Combustion
Engines,
Developments in,
Fig. 5 Torque and power
as functions of engine speed
for a typical internal
combustion engine. Data is
collected from 4.5-L-
medium-duty diesel engine
with advanced technology
such as turbocharging and
exhaust gas recirculation
(Data from author’s
laboratory, Texas A&M
University)

torque) and the speed at which they develop max- droops at low engine speeds as heat transfer
imum power (also known as rated power). The diminishes the energy available for in-cylinder
second feature, which is apparent from the knowl- work production. Friction, too, is a speed-
edge that rated torque exists, is the seemingly dependent function due to the mechanical behav-
dependent relationship of torque on engine ior of the engine’s interacting components. At
speed. It is clear from the definition of power high engine speeds, there is increased friction to
that it should have a functional relationship on be overcome on a per cycle basis; thus, torque
the speed of the engine, which dictates the amount droops at high engine speeds as increased work
of work per unit time the engine can deliver. But, energy is required to overcome increased friction.
based on assessment of Eq. 3, there is not a direct Relating this discussion to Eq. 7, both heat
relationship between in-cylinder work per cycle transfer and friction effects tend to decrease f.
and the speed of the engine. In fact, in an ideal There is, however, another factor of Eq. 7 being
sense, the torque of the engine should be constant influenced by heat transfer and friction and thus
with engine speed (and, if constraining to an ideal serving as a major contributor to the behavior of
engine, higher than the maximum attained torque the torque curve shown in Fig. 5; this factor is the
of the real engine), similar to how torque of an volumetric efficiency. The maximum amount of
electric motor is nearly constant with motor speed. work that can be developed per cycle is firstly
Thus, the behavior seen in Fig. 5 suggests that dependent on the amount of air (or oxidizer) the
there exist factors in a real engine that depend on engine can breathe; the amount of air the engine
engine speed and also affect the work produced can breathe ultimately dictates the amount of fuel
per cycle (thereby affecting the torque of the that can be delivered, which of course serves as
engine). These factors predominantly consist of the energy carrier to be converted in the cylinder.
heat transfer and friction; both of which have A representative volumetric efficiency curve as
dependencies on the speed at which the engine a function of mean piston speed (which is corre-
operates. Heat transfer of course is a time- lated to engine speed) is shown in Fig. 6. There
dependent phenomenon. At low engine speeds, is much important detail in this figure, and the
there is more time per cycle for thermal energy following will describe this in detail; after this
to be transferred from the cylinder; thus, torque discussion, a return to the effect of volumetric
Internal Combustion Engines, Developments in 141

Internal Combustion
Engines,
Developments in,
Fig. 6 Example of
volumetric efficiency,
illustrating the many factors
that impact the engine’s
ability to effectively breathe
air (Used with permission
from [13]). Improvements
to the engine’s volumetric
efficiency can yield
substantial improvements to
its ability to make power,
evident from Eq. 7

efficiency on the torque curve of Fig. 5 will be curve is called “charge heating” and is a heat transfer
made. First, notice that under ideal situations, the effect . During normal engine operation, the breath-
volumetric efficiency curve would be 100% and ing system of the engine reaches a steady-state
independent of engine speed (like work, trapped temperature that is higher than the ambient temper-
mass per cycle ideally has no functional depen- ature. Thus, the elevated temperature of the intake
dency on the number of cycles per unit time or port results in heat transfer to the intake air, increas-
speed of the engine). It is noted that, depending ing the air temperature, decreasing its density, and
on the reference point chosen for quantifying it, decreasing the amount of mass per unit volume. As
volumetric efficiency could be greater than 100% described above, heat transfer is a time-dependent
if the engine is, for example, boosted. In such a phenomenon; therefore, its presence is most notable
situation, the chosen reference point is upstream at low engine speeds (low mean piston speeds);
of the boosting device (e.g., turbocharger com- notice thirdly in Fig. 6 the effect of charge heating
pressor inlet). Although such a chosen reference on the volumetric efficiency curve, which in com-
point is useful to indicating the quantity of trapped bination with quasi-static effects lowers it to the
mass in the cylinder, it prevents the quantification curve labeled “B.” Similar to how friction affects
of the engine’s breathing system outside of the work per cycle, friction also affects air flow and has
boosting device and masks opportunities for fur- a dependency on mass flow rate (i.e., has a depen-
ther improvement to the engine’s design. dency on engine speed). The fourth factor to notice
Returning to the volumetric efficiency curve of in Fig. 6 is the effect of flow friction; its effect in
Fig. 6 and for the specific case of premixed charge concert with quasi-static and charge heating factors
engines such as conventional gasoline spark-ignited results in the curve labeled “C.” Notice at very high
engines, the premixing of fuel with air prior to speeds, the flow friction effect seems to level off and
induction diminishes the amount of trapped mass play a speed-independent role as engine speed
of air in the cycle. Such a factor is considered a increases; this results from the flow attaining
“quasi-static” effect since it will be present regard- choke conditions where, regardless of the pressure
less of the speed of the engine. Other quasi-static drop across the intake system, the mass flow rate
effects include factors such as, residual fraction due remains constant as its velocity reaches the speed of
to manifold pressure differences (between exhaust sound. Thus, the fifth factor to notice in Fig. 6 is the
manifold and intake manifold) and exhaust gas effect of choking on the volumetric efficiency curve.
recirculation. Thus, notice secondly that the The net result of all combined effects including
diminishing effect of quasi-static factors is choke is the curve labeled “D.” Interestingly, some
represented as curve “A” in Fig. 6. The next phe- phenomena help to increase the engine’s volumetric
nomenon to be captured in the volumetric efficiency efficiency; ram effect is one such example. Ram
142 Internal Combustion Engines, Developments in

effect occurs at high flow rates where fluid momen- In fact, it is interesting to note the similarities
tum results in continued charging of the cylinder between the final volumetric efficiency curve
even as the in-cylinder motion of the piston no (curve “G”) of Fig. 6 and the torque curve of
longer provides the pumping action (i.e., volume Fig. 5. Such a similarity is not a coincidence, as
increase during intake stroke). The sixth factor to the maximum amount of work attainable per cycle
note in Fig. 6 is the increase in volumetric efficiency is a strong function of how well the engine
due to ram effect . The combined effects, including breathes air per cycle. Correspondingly, based
ram effect, result in the curve labeled “E.” A similar on the assessment of Eq. 7, significant improve-
but diminishing gas exchange phenomenon occurs ments to an engine’s ability to make power can be
at low engine speeds, known as backflow. Backflow realized through creative improvements to the
occurs during the valve overlap period (the period engine’s volumetric efficiency.
during which both exhaust and intake valves are
open as exhaust valves close and intake valves
Thermodynamic Analysis of Internal
open) where in-cylinder motion becomes quiescent
Combustion Engines
as the piston reaches top dead center and exhaust
At its core, an internal combustion engine is a ther-
products of combustion backflow into the intake
modynamic device. That is, it takes one form of
port. The seventh factor to notice in Fig. 6 is the
energy (i.e., chemical energy) and converts it into
effect of backflow on the engine, which tends to
another form of energy (i.e., mechanical “shaft
pervade at low engine speeds where exhaust
work” energy). There are several considerations
momentum is low. The combined effects, including
that should be given to the thermodynamics of the
backflow, on volumetric efficiency result in the
internal combustion engine, in the pursuit of
curve labeled “F.” Finally, to end on a positive
improving its ability to make power and, more
note, the eighth and last factor to observe in Fig. 6
importantly, to make power efficiently. Efficiency is
is that of tuning. Tuning is the effort to make use of
an important thermodynamic concept which quan-
established resonating sound waves in the intake
tifies the ability of a work conversion device to
and exhaust systems to either aid in charging the
convert one form of energy (e.g., chemical energy)
cylinder (intake tuning) or discharging the cylinder
into work energy (i.e., mechanical shaft work). In
(exhaust tuning). Because these charging or
the context of internal combustion engines, fuel
discharging benefits rely upon the timing of com-
conversion efficiency is used, as given by Eq. 9:
pression and rarefaction waves being positioned
at precise locations in the intake or exhaust sys-
W
tems, fixed geometry intake and exhaust systems f ¼ (9)
mf QHV
can only be tuned at a narrow speed range. The
benefit of tuning is shown in Fig. 6, where for this
where mf is the mass of fuel per cycle. Typical
particular application tuning is designed for the
maximum fuel conversion efficiencies for internal
mid-speed range of the engine. Variable tuning
combustion engines vary from ca. 20% to 45%,
systems are becoming available in production,
depending on engine design and type of combus-
which, through the use of creative flow channel-
tion system. Fuel conversion efficiency, like
ing and blocking, can change the “effective” run-
power, can be designated as brake or indicated
ner length of the flow passage and allow for tuning
fuel conversion efficiency. The relationship
across broader speed ranges than allowed by fixed
between brake and indicated fuel conversion effi-
geometry runners. With all factors considered, the
ciency is given as Eq. 10:
final volumetric efficiency curve is the solid curve
labeled “G.”
Now that the details of the volumetric effi-  f , b ¼ f , i m (10)
ciency curve as a function of engine speed, iden-
tified as curve “G” in Fig. 6, have been discussed, where f, b is the brake fuel conversion efficiency,
a return to its effect on engine torque will be made. f, i is net indicated fuel conversion efficiency, and
Internal Combustion Engines, Developments in 143

Wb
m is mechanical efficiency and given by m ¼ W in
thermal energy, via heat transfer mechanism, from
Pb
¼ Pin ¼ IMEPn. In words, fuel conversion efficiency
BMEP a high-temperature reservoir and converts the avail-
is the ratio of work output to fuel energy deliv- able portion of this thermal energy to useful work.
ered. There is another efficiency definition that The portion of thermal energy that is not available
can be considered, called thermal efficiency, for conversion to useful work (i.e., the entropy of the
which is the ratio of work output to fuel energy thermal energy) is rejected to a low-temperature sink
released. In other words, since the combustion via heat transfer mechanism. A practical example of
process converts the fuel’s chemical energy to a heat engine is the steam engine. It is the steam
thermal energy, the thermal efficiency indicates engine that caused N.L. Sadi Carnot to establish his
the conversion of thermal energy to work energy. two postulates on the maximal conditions of the heat
It is not possible to precisely measure the energy engine [18], thus laying the foundation for what
released by the fuel during the combustion pro- would become the second law of thermodynamics
cess; it is, however, possible to precisely deter- and the mathematical axioms for the property
mine the extent of incomplete combustion or entropy. The fully reversible heat engine cycle, i.e.,
combustion inefficiency. Thus, combustion effi- the Carnot cycle, is the most efficient cycle for
ciency is introduced to allow for the determination converting thermal energy into useful work
of the engine’s thermal efficiency. Combustion (which, because of its reversible nature, is the most
inefficiency is defined as the energy of unreacted efficient cycle for converting work energy into ther-
fuel in the exhaust to fuel energy delivered to the mal energy). These details are provided because a
engine. Combustion efficiency, c, is thusly given common misconception is that the maximum effi-
as Eq. 11 [14]: ciency of an internal combustion engine is limited
by a so-called Carnot efficiency; this really has no
P technical basis because the internal combustion
yi QHV, i engine does not operate on the same principle (i.e.,
c ¼ 1  h i (11)
mf• =mf• þ ma• QHV, f the heat engine principle) around which the Carnot
cycle is created.
Perhaps much of the misdiagnosis of consider-
where yi is an unreacted fuel specie in the exhaust ing an internal combustion engine as a heat engine
(e.g., carbon monoxide), QHV,i is the heating value comes from the use of ideal heat engine cycles to
of specie i, mf• and ma• are the fuel and air mass “model” an internal combustion engine, namely,
flow rates, respectively, and QHV,f is the fuel’s the Otto and Diesel cycles . These cycles do serve
heating value. Fuel conversion efficiency, com- an important purpose in that certain fundamental
bustion efficiency, and thermal efficiency, th, parameters affecting the efficiency of the ideal heat
are all related by Eq. 12: engine cycles (i.e., Otto and Diesel cycles) tran-
scend to also affecting the efficiency of the internal
f ¼ c th (12) combustion engine. An example of such is com-
pression ratio. The thermodynamic analysis of the
Note that f of Eq. 12 could be either brake or Otto cycle, the four processes of which are shown
indicated, in which case the corresponding th will in Fig. 7, on (a) pressure-volume and
have the same basis. (b) temperature-entropy diagrams reveals that effi-
At this point, it is instructive to question what the ciency of the cycle, th,Otto, is given by Eq. 13:
maximum possible fuel conversion or thermal effi-
ciency of an internal combustion engine could 1
th, Otto ¼ 1  (13)
be. To begin to answer this question, it should r g1
c
be established that, contrary to what is some-
times simply conveyed, an internal combustion where g is the ratio of specific heats for the work-
engine is not, by rigid definition, a heat engine ing fluid (e.g., gair = 1.4 at 300 K). This result
[15–17]. A heat engine is a device that collects reveals the asymptotic relationship of increasing
144 Internal Combustion Engines, Developments in

Internal Combustion
Engines,
Developments in,
Fig. 7 (a) Pressure-volume
and (b) temperature-
entropy diagrams of the air-
standard Otto cycle (Used
with permission from [19])

Internal Combustion
Engines,
Developments in,
Fig. 8 Otto cycle thermal
efficiency as a function of
compression ratio for
several g values

thermal efficiency with compression ratio, as corollary to this is more work energy is transferred
shown in Fig. 8. A similar behavior is observed out of a mixture with higher g as the mixture is
in internal combustion engines (i.e., increasing expanded. In internal combustion engines, leaner
efficiency with increasing compression ratio) as fuel-air mixtures (i.e., those with less than stoi-
the capability to expand the working fluid (i.e., chiometric, or chemically complete, concentra-
combustion products), and thus the ability to con- tions of fuel in the mixture) have higher g values
vert thermal energy to work energy, increases with and thus realize an efficiency improvement.
increasing compression ratio. Another fundamen- Although the examples of using the Otto cycle
tal behavior is apparent in Fig. 8 related to the to assess the effects of compression ratio and g
changing value of g. Notice that thermal efficiency value on thermal efficiency are helpful, the Otto
tends to increase with an increase in g; similar to cycle efficiency cannot be used to predict maxi-
the increase in compression ratio, a mixture with mum attainable efficiency limits of an internal
a higher g suggests it requires more energy to combustion engine. There are several inaccura-
increase its temperature during a work-transfer cies in applying the Otto cycle (or any ideal heat
(i.e., constant pressure) process compared to a engine cycle) to predict internal combustion
nonwork (i.e., constant volume) process. The engine performance and efficiency. Specifically,
Internal Combustion Engines, Developments in 145

the assumptions that go into the development of

Eq. 13 are as follows (refer to Fig. 7):

1. Processes 1-2 and 3-4 are reversible and adia-

batic (i.e., isentropic or constant entropy)
processes.
2. Process 2-3 is a constant volume heat addition
process, where thermal energy of the system
increases due to heat transfer with the sur-
roundings at a temperature of no less than T3.
3. Process 4-1 is a constant volume heat rejection
process, where thermal energy of the system
decreases due to heat transfer with the sur-
Internal Combustion Engines, Developments in,
roundings at a temperature no greater than T1.
Fig. 9 Pressure-volume diagram illustrating the processes
4. Properties of the working fluid are constant assumed in fuel-air cycle analysis of an internal combus-
throughout the cycle. tion engine. Note that the shown cycle is representative of
an engine under “throttled” conditions, where intake pres-
sure is less than atmospheric (Used with permission from
Of the above four assumptions, all four are
[21])
impractical (meaning, an actual engine’s effi-
ciency will be less than the ideal limit because of
real effects, which will be discussed in more detail be transferred into the system via mass flow (i.e.,
below). Further, assumptions 2–4 are theoretically via the intake process) and allow properties of the
improper to impose on an internal combustion mixture to change with species, temperature, and
engine. Assumptions 2–4 imply a heat engine pressure. The resulting analysis, referred to as
cycle, which has already been established that an fuel-air cycle analysis [20], employs the follow-
internal combustion engine is not. Instead, an ing processes (refer to Fig. 9) and assumptions in
internal combustion engine operates as a collec- modeling an internal combustion engine cylinder
tion of processes, all which have ideal limits dic- (i.e., the control volume):
tating, in aggregate, the ideal maximum efficiency
of the engine. Energy is transferred to the control • Process 6-7-1: Ideal intake process of fuel and
system (i.e., the cylinder of the engine) through air and adiabatic mixing with residual gas from
mass flow rate, as opposed to heat transfer mech- the preceding cycle. Residual gas, or often
anism (as in the heat engine apparatus). This referred to as “residual fraction,” is leftover
energy, initially in the form of chemical energy, products of combustion not fully exhausted
is converted to thermal energy during the com- from the cylinder during the previous cycle’s
bustion process, and the expansion of the piston exhaust displacement process.
thereby converts the thermal energy to work • Process 1-2: Reversible and adiabatic com-
energy. Additionally, because of the change from pression of the fuel-air/residual mixture (i.e.,
reactants to products during the combustion pro- the reactants). Chemical species are frozen.
cess and the dependence of species’ properties on • Process 2-3: Complete, adiabatic combustion
temperature and pressure, it is not appropriate of reactants to products (the latter of which are
to assume properties of the working substance assumed to exist in equilibrium concentra-
remain constant during the engine’s operation tions). Combustion, in ideal sense, is modeled
(i.e., assumption no. 4 is not valid). as constant volume, constant pressure, or lim-
Instead, maximum ideal efficiency limits of ited pressure. Limited pressure is a combina-
internal combustion engines could be assessed tion of constant volume and constant pressure
by improving the inappropriate assumptions of processes. Constant volume combustion occurs
the Otto heat engine cycle, i.e., allow energy to until a “limit pressure” is reached, when the
146 Internal Combustion Engines, Developments in

remainder of combustion occurs at constant (and equal to P6 = P5), and hi is the enthalpy of
pressure. the fresh intake mixture. Residual fraction, as
• Process 3-4: Reversible and adiabatic expan- explained above, is the residual mass, mr (which
sion of products, which remain in chemical is equal to m6, the mass at state 6), divided by
equilibrium throughout the process. the total cylinder mass, mtotal (which is equal to
• Process 4-5: Reversible and adiabatic “blow- m1 = m2 = m3 = m4, the mass of the system at
down” of products to exhaust pressure of prod- states 1, 2, 3, and 4, respectively), and is related to
ucts remaining in the cylinder. Products remain rc, Pe, Te, and the pressure and temperature at state
in fixed composition based on concentrations 4, P4, and T4, respectively, by the following:
at state 4.
mr m6 m6 1 Pe T 4
• Process 5-6: Ideal, constant pressure, and f ¼ ¼ ¼ ¼
mtotal m1 m4 r c P4 T e
adiabatic exhaust displacement of products,
remaining in fixed composition. It is clear from the equations for h1 and f that a
priori knowledge of the cycle is needed in order to
Ideal gas behavior of the mixture and con- fix state 1 (note that rc, Pi, Pe, and hi are chosen
served mass are assumed throughout the analysis. based on desired cycle compression ratio, mani-
Further, mass transfers only occur during intake fold conditions, inlet composition, and inlet tem-
and exhaust processes (i.e., no leakage, blow-by, perature, Ti, respectively). Thus, initial values for
or crevice flow during the cycle). By employing state 1 are typically assumed for the first iteration;
first and second laws and various assumptions to state 4 temperature, pressure, and composition
each of the processes, respective states can be from the initial iteration are then used to estimate
thermodynamically fixed, thus allowing determi- f and h1 for the next iteration. This iterative pro-
nation of work transfers. An item to note about cess continues until some convergence criterion is
fuel-air cycle analysis, because it accommodates met (e.g., close match between ffinal and ffinal  1).
changing species, is that three independent prop- Rather than assuming values of f and h1, it is
erties must be known to fix the states (i.e., two perhaps more intuitive to assume initial values of
thermodynamic properties, such as temperature f and temperature at state 1, T1; the following
and pressure, and species composition). Process expressions [22] are approximations for estab-
6-7-1 is assumed adiabatic. Further, it is assumed lishing initial iteration state 1 properties:
that during the stroke from TDC to BDC, the ( "     #)1
cylinder pressure is constant and equal to the Te Pi Pi ðg1Þ=g
f ¼ 1þ rc 
intake pressure, Pi: Ti Pe Pe
 
P7 ¼ P1 ¼ Pi Pi
T 1 ¼ T e rc f
Pe
6 Q1 ¼0
It is also noted that composition of the residual
With these two statements, and the use of the gas must be assumed to fully fix state 1; it is
first law, the enthalpy at state 1, h1, is given as suggested that ideal products of stoichiometric
    combustion of the fuel-air mixture be used as the
Pi residual gas species.
h1 ¼ f he þ R e T e  1  hi þ hi
Pe Work analysis of process 6-7-1 gives the intake
work quantity:
where f is the residual gas fraction, he is the
enthalpy of the exhaust (and equal to h6 = h5), 6W1 ¼ Pi ð V 1  V 6 Þ
Re is the gas constant of the exhaust (and equal to
R6 = R5), Te is the temperature of the exhaust (and Process 1-2 is assumed reversible and adiabatic
equal to T6 = T5), Pe is the pressure of exhaust (i.e., 1Q2 = 0); from the second law, this gives
Internal Combustion Engines, Developments in 147

s1 ¼ s2
P3a ¼ P3 ¼ Plimit
Geometric relationship gives v1 = rc v2.
The assumption that chemical composition 2W 3 ¼ P3 ðV 3  V 2 Þ
remains fixed during compression (i.e., process
1-2) gives chemical composition at state 2; thus, At this point, it becomes necessary to
state 2 is fixed. The first law analysis of process describe a way in which product species of
1-2 gives combustion can be modeled. It is clear from
the above discussion that combustion is
1W2 ¼ U 2  U 1 ¼ mðu2  u1 Þ modeled as adiabatic and (a) constant volume,
(b) constant pressure, or (c) limited pressure. In
Process 2-3 is the adiabatic combustion pro- order to capture a representative combustion
cess, where reactants become products (products reaction, it is necessary to assume that several
assumed to be in chemical equilibrium) modeled product species may exist and will exist in
typically in three simple fashions: (1) constant equilibrium. Also, it is important to clarify
volume combustion, (2) constant pressure com- that this approach assumes initial and final
bustion, and (3) limited pressure combustion. The equilibrium states in predicting combustion;
first law analysis for each of the three cases results i.e., the thermodynamic process goes from
in the following, respectively: one equilibrium state (reactants) to a different
equilibrium state (products). This is in contrast
1. Adiabatic constant volume combustion to using a more detailed approach that models
the progression of combustion and relies upon
u3 ¼ u2
reaction kinetics to predict intermediate spe-
cies and their nonequilibrium concentrations.
v3 ¼ v2 See, for example, Foster and Myers [16] for
overview of detailed engine modeling and
Westbrook and Dryer [23] for overview of
2W3 ¼0 kinetic modeling of combustion. While per-
haps hundreds of species could be included as
2. Adiabatic constant pressure combustion products, Olikara and Borman [24] propose a
h3 ¼ h2 combustion reaction, given as Reaction (R1),
that might nearly be general for internal com-
P3 ¼ P2 bustion engine purposes:

n þ m=4  l=2
x13 Cn Hm Ol Nk þ ðO2 þ3:7274N2
2W3 ¼ P2 ðV 3  V 2 Þ ¼ P3 ðV 3  V 2 Þ ’
þ0:0444ArÞ ! x1 H þ x2 O
3. Adiabatic limited pressure combustion þx3 N þ x4 H2 þ x5 OH þ x6 CO
• Constant volume portion þx7 NO þ x8 O2 þ x9 H2 O þ x10 CO2
u3a ¼ u2 þx11 N2 þ x12 Ar
(R1)
P3a ¼ Plimit
where x1 through x12 are the mole fractions of the
v3a ¼ v2 respective product species. x13 represents the
moles of fuel per mole of products. Solution of
• Constant pressure portion the 13 unknowns of course requires 13 indepen-
dent equations; five of these equations come from
h3 ¼ h3a the species balance (i.e., balance of C, H, O, N,
148 Internal Combustion Engines, Developments in

and Ar species). The remaining eight equations At this point, it is useful to describe the attain-
come from the assumption of equilibrium among ment of various property values indicated above
various product species [24]. As a final note, it is (e.g., s1). Because fuel-air cycle analysis seeks to
recognized that NO appears as a product species; capture the effects of changing properties on the
this represents the equilibrium concentration of cycle analysis, the approximations for constant-
NO at the reaction temperature and pressure. specific heat ideal gas property equations cannot
Because both the formation and decomposition be used. Instead, it becomes necessary to employ
of NO in the post-flame gas regions are rate lim- techniques that accommodate changes to mixture
ited [25], it is necessary to take a chemical kinetic properties based on the mixture’s temperature,
approach to modeling IC engine exhaust NO con- pressure, and species composition. One such tech-
centrations (see, e.g., [26–29]). nique is to use NASA’s Chemical Equilibrium
Process 3-4 is reversible and adiabatic expan- with Applications program [30–34] coupled with
sion with the mixture in chemical equilibrium the JANAF Thermochemical Tables. By doing
(resulting in different species at state 4 relative to so, species at any given temperature and pressure
state 3), resulting in the following thermodynamic can be predicted (using NASA’s Chemical
state (along with geometrical constraint imposed): Equilibrium with Applications program) and
the corresponding mixture’s properties (using
s4 ¼ s3 JANAF Thermochemical Tables) can be deter-
mined. This type of technique, for example, can
v4 ¼ v1 be programmed into a computer routine.
Based on the first law analysis and constant
The first law analysis gives the following expres- pressure assumption, state 6 is fixed as follows:
sion, which is the expansion work of the cycle:
T6 ¼ T5 ¼ Tr
3 W 4 ¼ U 4  U 3 ¼ m ð u4  u3 Þ
P6 ¼ P5 ¼ Pe
Process 4-5 is the ideal blowdown process (i.e.,
mixture that remains in the cylinder expands isen- The exhaust work is given by the following:
tropically, filling the cylinder volume voided by
the products irreversibly escaping past the open 5W6 ¼ P e ðV 6  V 5 Þ
exhaust valve). Species are fixed based on state
4 concentrations. Thus, state 5 is fixed with the Based on the above analysis, the net work and
following relationships: gross work of the cycle are given by:

s5 ¼ s4 W net ¼6 W 1 þ1 W 2 þ2 W 3 þ3 W 4 þ5 W 6

P5 ¼ Pe W gross ¼1 W 2 þ2 W 3 þ3 W 4

Although specific volumes change during the Pump work is as defined by Eq. 4.
blowdown process (due to mass transfer), total Further, various efficiencies can be calculated
volume is constant; thus, work transfer is 0: using Eqs. 9, 11, and 12. Note that although fuel-
air cycle analysis is ideal, there may still be
4W5 ¼0 incomplete combustion due to the assumption of
species existing in chemical equilibrium. Thus,
Finally, process 5-6 is the ideal exhaust dis- combustion inefficiency will occur at fuel-air mix-
placement process, i.e., cylinder contents are tures approaching stoichiometric and rich condi-
transferred out of the adiabatic control volume tions, as dissociation creates concentrations of
via piston displacement at constant pressure. CO, H2, and other partially oxidized species.
Internal Combustion Engines, Developments in 149

Upon execution of the above analysis in a com- increase as ’ goes rich (i.e., ’ >1). Further, th
puter routine, for example, the theoretical limits of increases as EGR level increases. To understand
maximum efficiency of an internal combustion this behavior, it is necessary to understand how
engine can be determined. Examples of such mixture properties are affected by the (a) effect of
ideal efficiencies for constant volume combustion varying ’ on species, (b) effect of varying EGR
of isooctane/air mixtures at two different EGR level on species, and (c) effects of ’ and EGR
levels are given in Fig. 10. level on temperatures and pressures of the cycle.
Figure 10 reveals interesting behavior of two Ultimately, as fundamentally revealed by Eq. 13, a
important internal combustion engine parameters change to mixture properties that results in an
on the various efficiencies, i.e., fuel-air equiva- increase in g will cause an increase in efficiency.
lence ratio (often designated as ’) and EGR level. Figure 11 [35] summarizes how burned gas mix-
First, notice that as ’ approaches stoichiometric ture properties of an isooctane/fuel mixture are
(i.e., ’ = 1), th decreases and then begins to affected by various parameters such as ’ and

Internal Combustion Engines, Developments in, analysis of an internal combustion engine assuming con-
Fig. 10 (a) Thermal efficiency, (b) combustion efficiency, stant volume combustion of isooctane (C8H18) with
and (c) fuel conversion efficiency of an ideal fuel-air cycle Ti = 300 K, Pi = Pe = 100 kPa, TEGR = 300 K, and rc = 10
150 Internal Combustion Engines, Developments in

Internal Combustion Engines, Developments in, functions of equivalence ratio of isooctane/air mixtures at
Fig. 11 Burned mass (a) molecular weight, (b) constant various burned mass temperatures and pressure of 30 atm
pressure specific heat, and (c) ratio of specific heats as (Used with permission from [35])

temperature. The behaviors of the parameters increase th. Also contributing to gb behavior,
shown in Fig. 11 can be related to g by recognizing however, is Cp,b, which is shown in Fig. 11b.
that g can be written as given by Eq. 14: Notice that, up to about ’ = 1, Cp,b increases
with ’. A portion of this increase is due to the
1 increasing concentration of species with higher
gb ¼ (14)
1
R constant pressure specific  heats  as ’ increases
Mb Cp, b (e.g.,
 at
 1,750 K, C p =R H2 O
 5:94
 , and
Cp =R CO2  7:15 , whereas Cp =R N2  4:28
 
where gb is g for the burned gas mixture, R is and Cp =R O2  4:46 , as shown in Fig. 12b);
the universal gas constant (i.e., R  8:314 kJ=kg another portion, however, comes from the
K), Mb is the molecular weight of the burned gas increase in burned gas temperature which causes
mixture, and Cp,b is the constant pressure specific an increase in constant pressure specific heat for
heat of the burned gas mixture. As an example of most species. Specifically, as ’ increases, adia-
the type of analysis that might be done on Fig. 11, batic flame temperature and thus burned mixture
notice that Mb (Fig. 11a) decreases as ’ increases temperature increase up to a peak that occurs
(for any given temperature of the burned mixture); slightly rich of stoichiometric (for isooctane, loca-
although concentrations of CO2 ( MCO2  44 g= tion of peak adiabatic flame temperature will vary
mol) increase as ’ increases, so too do concentra- depending on the type of fuel), as shown in
tions of H2O ( MH2 O  18 g=mol ). Since isooc- Fig. 12a for isooctane/air-fuel cycle analysis
tane, being a paraffin, produces more H2O with 0% EGR (note that adiabatic flame tempera-
than CO2 in the products, the lighter, higher con- ture, or TCV,adiabatic, is equal to state 3 temperature
centration of H2O dominates the net effect of or T3 of fuel-air cycle analysis and that T3
decreasing Mb with increasing ’. As the mixture is partially influenced by the changing state 2
approaches ’ = 1 and becomes rich (’ >1), Mb temperature, or T2, as ’ varies). The effects of
decreases substantially as concentrations of par- increasing burned gas temperature on the various
tially oxidized species (e.g., CO and H2) with species’ constant pressure specific heats are
relatively lighter molecular weights increase. shown in Fig. 12b. Thus as ’ increases, Cp,b
Based on the analysis of Eq. 14, a decrease in increases as species concentrations change (i.e.,
Mb will tend to increase gb, which tends to increased concentrations of CO2 and H2O) and
Internal Combustion Engines, Developments in 151

Internal Combustion a 3000

Engines,
Developments in,
Fig. 12 (a) Constant 2800
volume adiabatic flame
temperature as function of
fuel-air equivalence ratio of 2600

Tcv, adiabatic(K)
isooctane with initial
temperature and pressure as
predicted by fuel-air cycle 2400
analysis of isooctane/air
mixture and 0% EGR and
(b) CP/R for various species 2200
typically found in burned
Fuel-Air Cycle Analysis
mixtures as function of Constant Volume Combustion of Iso-Octane / Air Mixture
2000
temperature (Data adapted Tcv, adiabatic =T3, fuel-air cycle analysis
from JANAF N.B.,T2 varies with fuel-air equivalence ratio.

Thermodynamic Tables 1800

[36]) 0.6 0.8 1 1.2 1.4
Fuel - Air Equivalence Ratio

b 8
CO2

7
H2 O
Cp / R

5
O2
N2

4
1800 2000 2200 2400 2600 2800
Temperature (K)

burned gas temperature increases. An increase in due to the abundantly increasing concentrations of
Cp,b, like Mb, will cause gb to decrease. Thus, for ’ CO, H2, and other dissociated species which have
less than stoichiometric, there are two competing lower molecular weights than the fully associated
effects on gb: a decreasing Mb and an increasing (i.e., CO2 and H2O) species. The dramatic decrease
Cp,b as ’ increases. The net result, as shown in in Mb generally dominates any potential increase in
Fig. 11c, is a decrease in gb which correspond- Cp,b, resulting in a net increase in gb. Thus, as ’
ingly explains the decrease in th as ’ increases up becomes larger than 1, th increases. Further, at
to stoichiometric (see Fig. 10a). certain temperatures (e.g., 2,750 K) Cp,b decreases
The behaviors of Mb and Cp,b change as ’ as ’ becomes greater than 1; the higher tempera-
increases beyond ’ = 1 (i.e., the mixture becomes ture of the burned mixture causes higher dissocia-
rich), causing a corresponding change in gb tion, which causes increased concentrations of
and ultimately in th. As explained above, Mb dissociated species (e.g., CO and H2), which
decreases more dramatically rich of stoichiometric have lower constant pressure specific heats than
152 Internal Combustion Engines, Developments in

the
 fully  associated species (e.g.,
 at 2,750 K, (b) introducing real effects on the engine analysis
Cp =R H2 O  6:6 , whereas Cp =R H2  4:4 ). that include such factors as heat transfer, friction,
For such temperatures where Cp,b simultaneously and other energy “losses” – but not necessarily
decreases along with Mb as ’ becomes larger than exergy losses (friction is, of course, an exergy
1, there is a dramatic increase in gb and a loss). Thus, a brief discussion will follow describ-
corresponding dramatic increase in th. ing the major benefit of conducting an exergy
Although th may tend to increase as ’ analysis of an engine system. The idea of avail-
becomes larger than 1, c tends to decrease able energy – or that portion of energy which is
(again, as concentrations of dissociated species available to do useful work – is introduced by
such as CO and H2 increase as ’ becomes larger Gibbs [37] and, in its basic form, suggests that a
than 1). The result is a general decrease in f. system with a given amount of energy can transfer
Similar analysis can be conducted to understand that energy out of the system as useful work until
the effect of EGR on the various efficiencies. It the point of equilibrium between the system and
becomes necessary at this point, however, to its environment (i.e., its surroundings). Using a
assess the extent of the utility of such “first law simplified example, a piston/cylinder arrangement
analysis” – i.e., an analysis centered only on that contains a gas at 2 atm will, if allowed to
energy transfer – in pursuing future developments interact with its surroundings at 1 atm, expand and
of internal combustion engines. For example, an transfer work energy until the system pressure is
engineer may “dream” of a situation where an in equilibrium with the surrounding’s pressure
engine operates rich at a fuel-air equivalence (i.e., 1 atm.). In a more theoretical sense, it cap-
ratio of 1.4 and 30% EGR, yielding a th of nearly tures the notion that only an orderly flow of
70%; this certainly outperforms an engine operat- energy (e.g., like that established because of a
ing lean at a fuel-air equivalence ratio of 0.5 and temperature gradient between a system and its
30% EGR (see Fig. 10a). If only an energy anal- surroundings) can be converted into useful work;
ysis is conducted to assess this “dream,” the idea disorderly energy cannot be converted to useful
will immediately be discounted because the work. Entropy is the idea of disorder within a
corresponding f for the engine running at 1.4 system; thus, exergy analysis combines the effect
fuel-air equivalence ratio and 30% EGR is only of a system’s entropy on its ability to convert its
ca. 32% (compared to the lean operation which is energy into useful work. Because real processes in
close to 54%). But the products of the rich com- net increase entropy (i.e., real processes are irre-
bustion still have useful energy, that is, energy that versible and result in entropy generation), systems
can be converted to useful work. While it is true undergoing real processes destroy exergy; that is,
the cycle cannot convert such available energy to the opportunity to convert energy into useful work
useful work, it is more than a “dream” that some is destroyed via some disordering of what had
other device could convert it (e.g., a fuel cell or a been an orderly flow of energy.
thermoelectric generator). Thus, it might be con- Examples of real effects, or irreversibilities,
ceivable to design an engine that allows it to either in a general thermodynamic system or
operate at maximal th and couple it to another between a general thermodynamic system and its
device that, while not having as high of a th, has a environment (all of which happen to be present,
high enough th to suitably convert the available also, in internal combustion engines) are heat trans-
chemical energy to useful work, so that the net fer through a finite temperature difference, mixing,
efficiency of both devices is greater than the f of unrestrained expansion, combustion, and friction.
the internal combustion engine operating at the The general premise of an irreversibility is one
lean condition (for example). which renders a system/environment combination
Analysis of such “dreams” can only be com- changed as the system undergoes a cycle and returns
pleted by (a) introducing the concept of available to its initial state. For example, if a warm body
energy, or modernly called exergy, by employing exchanges thermal energy with a cold body (i.e.,
second law considerations on the analysis and heat transfer through a finite temperature difference),
Internal Combustion Engines, Developments in 153

there is heat transfer, but not work transfer, until the what is indicated by the pressure-volume rela-
two bodies are in thermal equilibrium. To return the tionship), then a sense of friction must be
two bodies to their initial states (i.e., transfer thermal included in the analysis.
energy from the cold body to the warm body), work The inclusion of “real effects” in engine cycle
energy must be transferred into the system to effect analysis is well established. For example, compar-
the heat transfer. In the forward process (i.e., the isons between actual engine cycles and fuel-air
bodies attaining thermal equilibrium), heat transfer analysis cycles are made by Edson and Taylor
occurs without work transfer; in the reverse process [38]. Inclusion of heat transfer, finite combustion
(i.e., the bodies returning to their initial states), heat rates, and mechanical losses is demonstrated by
transfer is manifested by a necessary work transfer. Strange [39]. The beginnings of detailed com-
A net effect has been made to the surroundings bustion modeling, including spatial location of
(transfer of work energy) as the system is returned start of combustion and flame propagation, are
to its initial state; thus, the process is irreversible. If, demonstrated by Patterson and Van Wylen [40].
on the other hand, a device had been positioned Such early demonstrations, and the corresponding
between the two bodies of dissimilar temperature advancement in digital computers, result in
that was able to extract the orderly sense of energy advanced models of heat transfer (e.g., [41–43]),
motion into useful work (i.e., a heat engine), the friction [44, 45], and combustion; the latter of
irreversibility of the heat transfer through finite tem- which are described in the context of doing a
perature difference is diminished. Further, if the heat complete engine cycle simulation (e.g., [44,
engine is conceived to be ideal itself, then all the 46–51]). The influences of “real effects” on both
orderly sense of energy motion is converted into energy and exergy perspectives of internal com-
useful work, hence, the concept of the Carnot heat bustion engines are also well established (see,
engine or fully reversible heat engine. e.g., the reviews by Caton [52] and Rakopoulos
On the one hand, the heat transfer is viewed and Giakoumis [53]). For example, Fig. 13 illus-
as a loss; on the other hand, it becomes apparent trates both the energy (Fig. 13a) and exergy
that there may still be an opportunity to extract (Fig. 13b) distributions of using 0% and 20%
useful work from the heat transfer. It is the latter adiabatic EGR in a spark ignition engine, as com-
supposition that exergy analysis centers on. To puted by a thermodynamic simulation [54]. Note
begin a discussion of energy and exergy analysis that when real factors, such as heat transfer and
of an internal combustion engine, a revisit to the real combustion, are considered, the maximum
assumptions made in the fuel-air cycle analysis indicated fuel conversion efficiency is about
is necessary. First, the assumption that the 30.7% for non-EGR case (0% EGR) and 31.5%
engine is adiabatic is not realistic; all real for 20% EGR case. Most of the balance of energy
engines have heat transfer. Thus, it becomes is transferred to the coolant (ca. 29% for non-EGR
necessary to capture an understanding of the case) due to nonadiabatic control system and out
effects of heat transfer on the engine cycle. Sec- the exhaust (ca. 40% for non-EGR case) due
ond, the assumption of how combustion is to underutilized conversion of thermal to work
modeled is inaccurate. Of course, constant vol- energy. Thus, the importance of considering real
ume combustion in an operating engine is not effects is clear. It is noted that the data shown in
physically possible due to the finite time Fig. 13 do not include the effect of friction; the
required by chemical kinetics to decompose the brake fuel conversion efficiency for the studied
fuel molecule. Even constant pressure combus- conditions of Fig. 13 is available in [54]. It is
tion is not realistic because it supposes a con- further noted that friction offers no opportunity
stant burn rate with instantaneous starts and ends for conversion to useful work (i.e., energy lost
of combustion. Thus, it becomes necessary to to friction is all destroyed exergy). Finally, it
include the effects of “real” combustion on the is noted that EGR seems to have a small to negli-
engine cycle. Third, if it is desired to know the gible effect on the calculated friction of the
actual torque from the engine (as opposed to simulation [54].
154 Internal Combustion Engines, Developments in

a Unused b Destruction Unused Destruction

(0.7%) Fuel due to Fuel due to
(4.5%) (0.7%) (4.5%)
Inlet Mixing Inlet Mixing
100 100 (1.4%) (0.8%)
Destroyed
Net Transfer by Comb
Out Due (20.5%) (19.3%)
(34.3%) 80
80 (39.9%)
ENERGY (%)

to Flows

ENERGY (%)
Net Transfer
(24.3%) Out Due (21.4%)
60 60 to Flows

(28.8%) Heat (29.7%)

Transfer Heat (23.1%)
(23.1%)
40 40 Transfer

Total 20 Total
20 Indicated Indicated
Work Work
(30.7%) (31.5%) (30.1%) (30.9%)
0 0
0% EGR 20% EGR 0% EGR 20% EGR

Internal Combustion Engines, Developments in, 20% EGR in a spark ignition engine modeled using a
Fig. 13 (a) Energy distribution and (b) exergy thermodynamic simulation [54] (Used with permission
(availability) distribution of various transfer and destruc- from Professor J. Caton, Texas A&M University)
tion (in the case of exergy) mechanisms for 0% EGR and

Also clear from Fig. 13 is the importance of case under study in Fig. 13b. The high-temperature
considering the internal combustion engine from exhaust, with ease of accessibility, offers an
an exergy perspective. If considering only the opportunity to exploit the fuel’s exergy transferred
energy transfers of Fig. 13a, it may appear that through the exhaust system. There are several
ca. 70% of the energy is “lost” or unavailable for practical technologies that offer an opportunity
conversion to useful work. While it is correct to extract the fuel’s exergy transferred out of the
to say that a fraction of the fuel’s energy is engine control system; these are discussed in more
underutilized in the engine control system’s (i.e., detail in the section “Waste Heat Recovery.”
piston/cylinder chamber) conversion to useful One of the important features of Fig. 13b
work, it is not correct to say that all this energy is the exergy destroyed due to combustion.
is lost. As explained above, exergy provides Combustion, in any application, is an irreversible
insight into the fraction of energy that is available phenomenon and thus will render entropy genera-
to do useful work. For example, Fig. 13 illustrates tion (i.e., exergy destruction). Like friction, the
that (for non-EGR cases) 28.8% of the fuel’s ca. 20% of fuel exergy destroyed due to combustion
energy is transferred out of the control system cannot be recovered for conversion to useful work.
via heat transfer, whereas only 23.1% of the The major causes for exergy destruction of a typical
fuel’s exergy is transferred out of the control sys- hydrocarbon-based combustion process include
tem via heat transfer. That is, for one unit of fuel, (1) thermal energy exchange among particles within
0.288 unit of energy is transferred via heat trans- the system, (2) diffusion among fuel/oxidizer parti-
fer, but only ca. 80% (0.231/0.288) of that energy cles, and (3) mixing among product species [55]; of
is available to do useful work; thus, 0.231 unit of these three, the dominant exergy destruction source
fuel exergy is a missed opportunity for conversion is the thermal energy exchange among particles. At
to useful work. This 0.231 unit of missed oppor- the onset of combustion, with the system containing
tunity could be exploited by configuring an ideal reactants, a metastable equilibrium exists as gradi-
heat engine between, for example, the cylinder ents exist between fuel and oxidizer molecules. As
walls and the environment (i.e., representative combustion proceeds during the process, natural
TH and TL). Such a conceptualization seems phenomena cause the system to eliminate such gra-
impractical, but exergy is not just passed into the dients and maximize entropy in the pursuit of
coolant. Exergy is also transferred to the exhaust, attaining a more stable equilibrium (i.e., the product
ca. 24.3% of the fuel’s exergy for the non-EGR state). The maximization of entropy in reducing the
Internal Combustion Engines, Developments in 155

gradients is manifested entirely through entropy compression, individual species will be collected
generation. Using the dominant source of exergy forming a mixture that is a priori in equilibrium
destruction (i.e., entropy generation) during com- (thus, there are no irreversibilities due to mixing)
bustion as an example, internal thermal gradients and allowed to expand isentropically. Because of
among particles [56] established during the combus- the increase in moles of the mixture as it expands
tion process create microscopic opportunities to do isentropically maintaining equilibrium along the
useful work (i.e., thermal gradients can be converted path, net work is extracted from the collection of
to useful work via a heat engine). Of course, practi- processes. Quantitative analysis [60] of such a
cal implementations of microscopic heat engines do concept reveals 0% exergy destruction; because
not exist; thus, the thermal gradients are reduced to of exergy retention in the exhaust products, how-
zero during the combustion process without any ever, thermal efficiency of the “reversible com-
conversion to useful work; heat transfer through a bustion engine” is around 28% (for a fuel-air
finite temperature difference, as described above, equivalence ratio of 1.0, compression pressure
generates entropy (destroys exergy). and temperature of 10 MPa and 6,000 K, respec-
The general behavior of exergy destruction due tively, and an expansion ratio of 18:1). Of course,
to combustion in internal combustion engines is such concepts are presently impractical; but the
generally well characterized, as summarized by notion of reversible combustion is theoretically
[57]; generally, an increase in combustion temper- possible.
ature decreases exergy destroyed due to com- In fact, it is briefly noted that, although not
bustion. This statement should not be confused explicitly calling his concept one of “reversible
with the situation of internal thermal gradients combustion,” a close inspection of Diesel’s orig-
which, as mentioned above, are a major source inal engine design [61] describes a method of
of combustion-based exergy destruction. While it nearly attaining reversible combustion (at least,
is tempting to seek ways to minimize (or even one which eliminates internal thermal gradients).
eliminate) combustion irreversibilities (and thus This is described in more detail in section “A Case
minimize or eliminate combustion-based exergy Study: Diesel Engines Versus Gasoline Engines.”
destruction), it should be recognized that such In closing this section on “Thermodynamic
efforts may decrease the conversion of exergy in Analysis of Internal Combustion Engine,” it is
other ways. Consider, for example, the use of a noted that assigning a “maximum possible effi-
lean mixture versus the use of a stoichiometric ciency” of an internal combustion engine is any-
mixture; the former has higher exergy destruction thing but straightforward. Fuel-air cycle analysis
than the later. While a stoichiometric mixture reveals that the efficiency of an engine cycle
reduces exergy destruction, its mixture composi- depends on several different characteristics of
tion also causes lower conversion of thermal to the control system. A more advanced and appro-
work energy during the expansion process (and, priate analysis of an engine – that which includes
coincidentally, increases exergy transfer via heat real effects of heat transfer, real combustion, fric-
transfer). Thus, overall efficiency is lower with tion, and other irreversibilities – gives realistic
a stoichiometric mixture. Again, energy- and senses of tangible factors that engineers can strive
exergy-based analyses, such as that presented in to improve. Further, exergetic-based analysis
Fig. 13, are necessary to quantitatively make such offers the important insight of what future oppor-
assessments. In the case of eliminating com- tunities might be exploited in improving the over-
bustion irreversibilities, a reversible combustion all system efficiency associated with an internal
process [58] can be conceptualized [59, 60] combustion engine. In returning to the engineer’s
where initially separated reactant species are isen- “dream” engine concept described above, such
tropically compressed to their respective partial computed numbers as shown in Fig. 13 are more
pressures and a certain temperature; concentra- realistic than the ideal fuel-air cycle analysis num-
tions of individual species, when allowed to inter- bers shown in Fig. 10. Further, the exergy-based
act with each other, will be in equilibrium. After analysis of Fig. 13 could be used to make the
156 Internal Combustion Engines, Developments in

necessary assessment of the engineer’s dream, to motion among particles composing the mixture,
quantify if there are real opportunities of “down- the flame then rapidly spreads throughout the
stream” exergy conversion when the engine is mixture in a turbulent fashion, enveloping
operated close to maximum theoretical thermal microscale eddies with flame. This rapid flame
efficiency levels. propagation is referred to as the turbulent
entrainment period. Upon flame envelopment
Spark Ignition Combustion of the turbulent eddies within the mixture, the
As discussed in section on “Thermodynamic flame then laminarly burns the microscale eddies
Analysis of Internal Combustion Engines,” real in the final stage of the process referred to as
combustion in an internal combustion engine is rapid burnup. Turbulent entrainment and rapid
not ideal in the sense that it can be modeled as burn practically occur simultaneously and
precisely constant volume, constant pressure, or together compose what is referred to as rapid-
even the combination of the two (limited pres- burning period. This described sequence of
sure). Real combustion, instead, involves several spark ignition combustion for a typical operat-
complex and interacting features that, on the one ing point (i.e., 1,400 rev/min, part-load condi-
hand, make it difficult to predict and model but on tion) is shown in Fig. 14 [62] and, for the same
the other hand create opportunities for further typical operating point, requires about 10 ms
development. Conventional engines are either from spark to finish.
spark ignited or compression ignited. Since the One of the important features of Fig. 14, along
type of ignition results in substantially different with the shown in-cylinder pressure, is the corre-
combustion features, the two are separated into sponding mass fraction burned. Using the first law
respective sections. This section describes spark of thermodynamics, it is possible to develop an
ignition combustion. expression that determines the chemical-to-thermal
Spark ignition combustion is the typical form energy conversion rate (commonly called the “heat
of combustion found in the commonly called release rate”) of the combustion process. Since the
“gasoline” engine (also commonly called either mass of fuel present in the cylinder, along with the
“petrol” engine or “otto” engine); it is more fuel’s heating value, is known, the mass fraction
descriptive to refer to such engines as spark burned rate of fuel can be calculated from the heat
ignition engines. The “spark” aspect of the release rate. Figure 15 illustrates a typical mass
term implies that combustion is initiated by fraction burned curve, as a percent of the total
introducing a high-voltage electrical arc, typi- fuel, and defines the above-described periods
cally through the use of a spark plug, into the referred to as flame-development period (shown
reactant mixture at a point during the cycle in Fig. 15 as the flame-development angle, Dyd)
where sustained combustion reaction will pro- and the rapid-burning period (shown in Fig. 15 as
ceed. Most spark ignition engines (exceptions the rapid-burning angle, Dyb). These periods are
are described in section on “Future Directions conventionally defined as the angle swept from
of Internal Combustion Engines”) induct a pre- spark release to 10% mass fraction burned and
mixed and homogeneous mixture of fuel and air the angle swept from 10% mass fraction burned
during the intake process. At a point near the end to 90% mass fraction burned, respectively.
of compression, the spark plug discharges an arc The mass fraction burned profile of spark
into the mixture. This arc, in the nearby region of ignition combustion can be modified, and the
the spark plug, forms a high-temperature plasma parameters that cause alterations to the mass
that evolves into a flame kernel. The flame ker- fraction burned profile are often coupled. Per-
nel transitions from the plug region in a laminar haps the most obvious parameter that can affect
sense establishing a flame front with initial the mass fraction burned profile is the spark
velocity close to the laminar flame speed. This timing or sometimes called the spark advance.
initial flame development is referred to as the By adjusting the time in the cycle when combus-
flame-development period. Because of turbulent tion is initiated (i.e., the spark advance), the times
Internal Combustion Engines, Developments in 157

Internal Combustion Engines, Developments in, steps of spark ignition combustion shown include the spark
Fig. 14 In-cylinder images capture the sequence of spark release (indicated by “Spark”), flame kernel development
ignition combustion with the corresponding result on (a–c), turbulent entrainment and burn up (d–e), and flame
in-cylinder pressure and mass fraction burned. The various termination (f) (Used with permission from [62])

into compression stroke) generally results in

100
combustion occurring earlier in the cycle and
in-cylinder pressures becoming higher in magni-
tude. Correspondingly, a retard in spark timing
Percent

(i.e., spark timing is moved later into the com-

pression stroke or possibly even into the expan-
sion stroke) generally results in combustion
occurring later in the cycle and in-cylinder pres-
0 sures becoming lower in magnitude. There are
Δqd Δqb
qs q10% q90% several other parameters, however, that can affect
the burn profile of spark ignition combustion.
Internal Combustion Engines, Developments in, Examples of such parameters include the mix-
Fig. 15 Mass fraction burn rate (as a percent of total ture’s fuel-air ratio (i.e., stoichiometry), the level
fuel) of spark ignition combustion in an internal combus-
tion engine, illustrating the definitions of the flame-
of EGR, the level of turbulence in the mixture,
development angle, Dyd, and the rapid-burning angle, the level of heat transfer, initial mixture pressure
Dyb (Used with permission from [63]) (i.e., load variation), initial mixture temperature,
and engine speed. Analysis of each of these
required for flame-development and rapid- parameters is outside the scope of this entry;
burning periods are altered. An advance in readers are referred to the “Books and Reviews”
spark timing (i.e., spark timing is moved earlier for additional reference text on the subject.
158 Internal Combustion Engines, Developments in

One item that is important to mention, how- the other two shown spark timings (i.e., 30 and
ever, is the effect of burn profile on engine-scale 10 BTDC). This lower pressure through the
parameters such as performance, efficiency, and expansion stroke results from the increased level
emissions. The rate of combustion, and its corre- of heat transfer manifested by the higher gas mix-
sponding effect on in-cylinder pressure, flame ture temperature caused by the faster burn rate-
temperature, and mixture gas temperature, alters induced higher cylinder pressures. Thus, spark
the performance, efficiency, and emissions of the timings advanced of MBT timing result in lower
spark ignition engine. For example, Fig. 16 illus- torque due to higher levels of compression work,
trates the effect of spark timing (spark advance) on heat transfer, and friction. Conversely, the most
in-cylinder pressure and relative torque to maxi- retarded timing (10 BTDC) renders the latest
mum brake torque (MBT) . MBT is the maximum burn profile (of the studied timings), the lowest
torque attained over a given parametric sweep, magnitude of peak pressure, and the latest location
most commonly over a spark-timing sweep. of peak pressure (occurring around 40 BTDC).
Note that there is nearly a single spark timing Because of the later-phased combustion, the
that yields the maximum torque delivered by the location of the rise in pressure misses the full
engine. Too early of a timing (advanced) or too opportunity of the expansion stroke. Thus, spark
late of a timing (retarded) renders a lower torque timings retarded of MBT result in lower torque
than MBT. Three spark timings are shown in due to expansion losses. It is clear that MBT
Fig. 16: 50 , 30 , and 10 before TDC (BTDC). occurs in the balance of minimized compression
The most advanced timing (50 BTDC) renders work and maximized expansion work, both of
the earliest burn profile (of the studied timings), which are affected by combustion phasing, heat
the highest magnitude of in-cylinder pressure, and transfer, and friction. In the example of Fig. 16,
the earliest location of peak pressure (occurring this balance occurs around 30 BTDC. Notice that
around 0 BTDC). Although it may appear from peak pressure for this timing is around 20
Fig. 16 that the spark advance of 50 BTDC yields BTDC; a general “rule of thumb” is that MBT
the maximum area under the pressure-volume occurs when peak pressure is positioned between
curve (and thus, the maximum work for the 15 and 20 BTDC. As revealed in Fig. 16,
cycle), it should be noted that its pressure is this rule of thumb extends to the mass fraction
lower through most of the expansion stroke than burned profile, where maximum brake torque is

Internal Combustion Engines, Developments in, torque (MBT) as a function of spark advance of a typical
Fig. 16 (a) In-cylinder pressure as a function of engine spark ignition engine (Used with permission from [64])
crankangle and (b) relative torque to maximum brake
Internal Combustion Engines, Developments in 159

timed when 50% of the fuel burns by 10 air mixture. Compression ignition combustion is
BTDC. In some cases, MBT timing may be the typical mode of combustion used in the com-
“knock limited,” meaning that a higher torque monly called diesel engine, an engine which more
could be attained at an earlier timing if fuel descriptively is called a compression ignition
knock was not present. Fuel knock is a combus- engine.
tion abnormality of spark ignition engines that, It is immediately recognized that one challenge
due to the combustion-generated compression of of compression ignition engines is the lack of a
the reactive mixture in the end regions of the direct trigger of ignition; the spark acts as the
cylinder, results from autoignition of the mixture direct trigger of combustion initiation in a spark
prior to its controlled burn by the propagating ignition engine. Conventional applications of
flame. Fuel knock can be very damaging because compression ignition engines (e.g., the conven-
the uncontrolled combustion tends to cause dra- tional diesel engine) overcome this challenge by
matic rises in pressure near critical mechanical using the fuel injection event as the direct trigger.
components (such as piston rings). The octane Thus, conventional compression ignition engines
rating of a fuel indicates the fuel’s resistance to typically induct an air and residual mixture (i.e.,
knock in a spark ignition engine; a fuel with a no premixing of fuel) during the intake stroke.
higher octane has a higher resistance to auto- This same unreactive mixture is compressed dur-
ignition. Additional discussion about this is pro- ing the compression stroke until near the point
vided in the sections on “A Case Study: Diesel when combustion is desired to begin. At this
Engines Versus Gasoline Engines” and “Direct point fuel is introduced into the mixture, which,
Injection, Spark Ignition Engines.” due to compression, is a high-temperature, high-
At MBT timing, since fuel flow rate is gener- pressure environment. After several complex and
ally held constant during spark-timing sweeps, coupled processes occur (described below), com-
efficiency will also be correspondingly maxi- bustion initiates and chemical-to-thermal energy
mized. In the case of emissions, however, the conversion takes place.
trends are not straightforward, and a brief discus- At this point, it is instructive to briefly describe
sion of emissions is reserved for the section on the role of “glow plugs” often used in compres-
“Emissions Formation and Exhaust Pollution.” sion ignition engines. Glow plugs should not be
Finally, it is noted that several parameters confused with spark plugs. A glow plug is a
affect the spark ignition burn profile. Thus, for resistive heating element inserted into the cylinder
each change to a given parameter (e.g., initial of a compression ignition engine to aid in the
pressure, initial temperature, fuel-air equivalence initial starting of the engine. It acts as a warming
ratio, engine speed, and level of mixture turbu- device during “cold start”; after engine warm up
lence), a potentially different spark timing will and stable operation, the glow plugs deactivate.
correspond to maximum brake torque. Again, They are not necessary on a cyclic basis. Spark
the general trend of spark timing will be such plugs, on the other hand, are integral components
that 50% mass fraction burned occurs at 10 of spark ignition engines; they provide the source
BTDC for maximum brake torque. of ignition for every combustion cycle of a spark
ignition engine.
Compression Ignition Combustion The method by which fuel is injected into the
In contrast to spark ignition combustion – which compressed mixture typically falls into one of two
uses an electrical arc, or spark, to initiate combus- categories: (1) direct injection or (2) indirect injec-
tion of a reactive mixture – compression ignition tion. Indirect injection involves injecting fuel into
combustion relies on compressive heating, or the a “prechamber” which is connected to the main
increase in temperature of a gas due to the increase chamber. Fuel injected into the prechamber
in pressure resulting from the decrease in volume, ignites and causes the mixture to issue into the
where combustion initiation is kinetically driven main chamber where main heat release and work
by exceeding the ignition temperature of the fuel- extraction occur. Indirect injection engines are
160 Internal Combustion Engines, Developments in

used primarily in applications where motion of the ignition. This period of time – i.e., the time
nonreactive mixture in the main chamber is too between start of fuel injection and ignition
quiescent for ignition; this typically occurred in (commonly called start of combustion) – is often
the early implementations of small, high-speed called the ignition delay period. As will be
automotive compression ignition engines. The described below, the ignition delay period of a
use of a prechamber, where nonreactive mixture compression ignition engine is an important
forced through orifices or nozzles connecting parameter that affects the remainder of the burn
the main chamber to the prechamber suitably profile and is affected by several other parameters.
increased mixture swirl and turbulence, generated The end of the ignition delay period after start of
sufficient fluid motion to allow ignition to injection will witness start of combustion and
occur. Technological developments in intake port noticeable heat release, as indicated in Fig. 17a
design, combustion chamber design, and fuel as the positive rate of heat release. During heat
injection systems have rendered indirect injection release, the mass fraction burned profile steadily
systems nearly obsolete and technically inferior to increases until end of combustion when all the
direct injection techniques. fuel is most nearly completely burned.
Direct injection, as the name implies, indicates An interesting and important feature of Fig. 17
that the fuel is injected directly into the cylinder at is the rate of heat release profile between start of
the point near when combustion is desired to combustion and end of combustion. Notice in
begin. Because of the compressed nature of the Fig. 17a there are two distinct components of the
nonreactive mixture into which fuel is injected, rate of heat release: a premixed component and a
fuel injection pressures must be very high; i.e., diffusion component. These two respective com-
modern fuel injection systems inject fuel ponents are nearly an exclusive feature of compres-
at pressures on the order of 100–2,000 bar sion ignition combustion that relies on fuel
(ca. 3,000–30,000 psi). The processes of fuel injection as the ignition trigger. The premixed
injection, combustion initiation, and subsequent heat release component results from the complex
combustion propagation (or burning) are very processes involved in physically preparing the ini-
complex in compression ignition engines. The tial “packets” of fuel for chemical decomposition
following discussion will briefly highlight these [65]; as described above, the injected liquid fuel
complexities. must be prepared and sufficiently mixed with the
Figure 17 illustrates the sequence of events that cylinder gases before chemical decomposition can
occur in a typical direct injection compression begin. Once ignition occurs, the parcels of fuel that
ignition engine. Two load conditions at 1,400 prepared for burning during the ignition delay
rev/min are shown, with a high-load condition period react under kinetic control, with relatively
(ca. 75% peak load which is about 11.3 bar high rates of burning for the given mixture temper-
BMEP) shown in Fig. 17a and a low-load condi- ature. Fuel continues to inject during the premixed
tion (ca. 25% peak load which is about 1.9 bar portion, and the increase in temperature due to
BMEP) shown in Fig. 17b. Both plots illustrate premixed combustion accelerates the physical
the fuel injector needle lift profile (which roughly preparation of the newly introduced parcels of
correlates to the fuel delivery rate), the rate of heat fuel. In spite of this acceleration of fuel preparation,
release profile, and the mass fraction burned pro- the burn rate becomes limited by the fuel’s mixing
file. Focusing first on the high-load condition rate with the air. Since the time scales for mixing
(Fig. 17a), notice that fuel injection occurs near are larger than the time scales for chemical kinetic
2 BTDC. Fuel in typical compression ignition decomposition [66], the burn rate becomes mixing
engines is injected as a liquid. Thus, the fuel controlled, rather than kinetically controlled. The
must undergo a series of physical processes mixing-controlled, or diffusion-controlled, compo-
(e.g., penetration, breakup, atomization, and nent of heat release is observed in Fig. 17a as the
vaporization) before it undergoes its chemical longer heat release following premixed heat release
process of bond fragmentation and eventual and is identified as “diffusion heat release.” The
Internal Combustion Engines, Developments in 161

burn rate of diffusion heat release is generally The bimodal feature of the fuel injection profile
slower than premixed heat release at any given in Fig. 17b is the result of the use of a double-
reaction temperature. Figure 17a shows diffusion injection strategy on this particular engine at this
heat release having a much higher burn rate than particular condition. Modern advanced engine
premixed heat release; this is manifested, however, fuel systems are capable of introducing fuel in
by the associated higher mixture temperature sequences, or pulses, in what is commonly called
effected by the combustion process. “multi-injection strategies.” In the case of the
The situation is slightly different for the low- engine highlighted in Fig. 17b, the first injection
load case, shown in Fig. 17b. Conventional com- is considered the “pilot” injection, while the sec-
pression ignition engines that use fuel injection as ond injection is considered the “main” injection.
the ignition trigger typically also use the amount The pilot injection introduces only a small portion
of fuel delivery as the means to control engine (e.g., ca. 10%) of the total fuel to be delivered in
load. Thus, the need to produce less power at a the cycle, creating an opportunity to prepare a
given engine speed is effected by decreasing the smaller portion of fuel for burning during the
amount of fuel delivered during the injection pro- ignition delay period. This small pilot increases
cess. This is evident in Fig. 17b where the area cylinder temperature, so that the ignition delay
under the fuel injector needle lift profile is less period of the main injection is much shorter, and
than that of the high-load case (for the moment, there correspondingly is a lower fraction of pre-
disregard the bimodal fuel injection profile, which mixed heat release during the main combustion
will be explained below). Consequently, the dis- event. Such a strategy is used, for example, to
tribution of premix versus diffusion heat release is reduce combustion-generated noise of a diesel
not apparent on the low-load heat release profile. engine. Interestingly, diesel engine noise is an
This does not imply that mixing-controlled com- “age-old” problem, serving as a primary focus of
bustion does not occur during low-load operation; development for Sir Harry Ricardo (a pioneer in
it does imply, however, that combustion is pre- the development of high-speed diesel engines) [67].
dominantly kinetically controlled at low-load con- The above-described use of multiple injections
ditions, whereas it is predominantly diffusion- hints at an ever-present relationship within com-
controlled at high-load conditions. pression ignition combustion systems using fuel

a 200 1400 rev/min

1 b 100 1400 rev/min
1
High Load Low Load
Diffusion 0.8 0.8
150 heat release
Rate of Heat Release (J/deg)

Premix
Rate of Heat Release (J/deg)

heat release 0.6

Mass Fraction Burned

0.6
Mass Fraction Burned

100
50
0.4
0.4
50
0.2
0.2

0 0
Needle Lift
Profile (a.u)
0 0

Needle Lift
Profile (a.u)

–20 –10 0 10 20 30 40 50 60 70 –20 –10 0 10 20 30 40 50 60 70

Crankangle (° ATDC) Crankangle (° ATDC)

Internal Combustion Engines, Developments in, (ca. 75% peak load = 11.3 bar BMEP) and (b) low-load
Fig. 17 Fuel injector needle lift profile, rate of heat condition (ca. 25% peak load = 1.9 bar BMEP) with the
release, and mass fraction burned for a typical diesel use of pilot injection (Data from author’s laboratory, Texas
engine operating at 1,400 rev/min (a) high-load condition A&M University)
162 Internal Combustion Engines, Developments in

injection as the direct trigger; i.e., a relationship environment which enables fuel preparation for
among ignition delay, fraction of premixed heat burning at a faster rate, and combustion com-
release, and fraction of diffusion heat release exists. mences relatively sooner (i.e., shorter ignition
This relationship is shown in Fig. 18a, which illus- delay). Because of the shorter ignition delay,
trates the rates of heat release at two different there is less fuel prepared for burning; thus, there
injection timings of a typical compression ignition is lower fraction of premixed heat release and
engine operating at 1,400 rev/min, high-load con- correspondingly larger fraction of diffusion heat
dition (11.3 bar BMEP, nominal). Notice for the release.
advanced injection timing there is a substantial Because premixed heat release rate is relatively
level of premixed heat release and a correspond- faster (at a given temperature) than diffusion heat
ingly lower fraction of diffusion heat release rela- release, the “combustion duration” – or the time
tive to the retarded injection timing. The advanced taken between start of combustion and end of
injection timing, due to the injection of fuel into a combustion – for a mostly premixed heat release
relatively cool environment when rates of physical combustion event will be shorter than that of mostly
and chemical processes are low, results in a rela- diffusion heat release combustion event. Thus, there
tively longer ignition delay. Because of the longer tends to be an inverse relationship between ignition
ignition delay, there is more time for fuel prepara- delay and combustion duration, as shown in
tion prior to ignition; once ignition occurs, there is Fig. 18b, where changes to ignition delay are
a relatively higher concentration of fuel that is manifested by alterations to injection timing. The
prepared for burning and correspondingly com- shorter ignition delay results in less premixed heat
busts in a premix and kinetically controlled fash- release, more diffusion heat release, and correspond-
ion. As more fuel is prepared during the ignition ingly a longer combustion duration.
delay period for premix burn, there consequently is Injection timing is just one parameter that can
less fuel available for diffusion heat release; thus, affect ignition delay and thus the burn profile of a
as premix burn fraction increases, diffusion burn compression ignition combustion event when fuel
fraction decreases. Likewise, at the retarded injec- injection is used as the ignition trigger. Other
tion timing, fuel is injected in a relatively hotter engine parameters such as EGR level [68], initial

250
a Engine Speed = 1400 rev/min b 34
Engine Load = 11.3 bar BMEP(nominal) Engine Speed = 1400 rev/min
θinj = –21°ATDC EGR = 0% Engine Load = 11.3 bar BMEP(nominal)
Injection timing (θinj) - Variable EGR = 0%
200 Injection timing - Variable
32
Rate of Heat Release (J/deg)

θinj = –3°ATDC
Combustion Duration (deg)

150 30

100 28

50 26

0 24
–30 0 30 60 90 3 3.5 4 4.5 5 5.5
Crankangle (°ATDC) Ignition Delay (deg)

Internal Combustion Engines, Developments in, via an injection timing sweep for a typical
Fig. 18 (a) Rate of heat release at an advanced compression ignition engine operating at 1,400
injection timing (21 BTDC) and a retarded rev/min, high-load condition (11.3 bar BMEP,
injection timing (3 BTDC) and (b) the relationship nominal) (Data from author’s laboratory, Texas A&M
between ignition delay and combustion duration University)
Internal Combustion Engines, Developments in 163

temperature and initial pressure [69], injection heavy trucks or handheld engine devices (such
pressure [70–72], and swirl and turbulence [69, as lawn mowers). Because of such variability in
73] all have certain effects on ignition delay and regulation, the time-oriented nature of the regula-
the resulting relative fractions of premixed and tions (i.e., regulations have, to this point, been in a
diffusion heat release. Like the effect of various state of flux), and variation in regulation among
parameters of spark ignition engine performance, various governing agencies around the world, fur-
efficiency, and emissions, the various effects of ther description of specific emissions will not be
compression ignition engine parameters on igni- provided. The basic issues at hand, however, can
tion delay and burn profile also have an effect on be briefly described.
engine efficiency, performance, and emissions. For As shown in Reaction (R1), there are several
example, overly advanced or retarded injection species formed during the combustion reaction of
timings will cause decreases in engine torque for a typical hydrocarbon-air mixture. Some of these
a given fuel delivery rate. Additionally, fuel quality species are generally stable and nonreactive in the
is particularly important for conventional diesel atmosphere, thus pose little to no harm to the five
combustion operation, where low volatility fuels kingdoms of nature. Other species, however, are
with high ignitability are generally used to ensure either harmful or reactive in ways that lead to harm-
ignition that occurs during the cycle. An important ful consequences. Specifically, there is little scien-
fuel parameter – i.e., its cetane number – is used to tific debate about the harmful nature of certain
identify the quality of fuels appropriate for use in a combustion products such as CO, NO/NO2,
diesel combustion system. A higher cetane corre- unburned hydrocarbons (HC), and particulate mat-
sponds to a fuel with a shorter ignition delay. ter (PM or the solid/liquid components of exhaust
Finally, in closing, it is important to recognize the that can be collected on a filter. It is noted that
above-described phenomena are largely phenome- historically, sulfur oxides and sulfates have been
nological observations that can be made using con- considered either separately from [76] or in combi-
ventional diagnostics with relatively straightforward nation with [77, 78] particulate matter). There is,
analysis. Considerable development (e.g., [49, 74, perhaps, continued debate about the potential con-
75]) has been made to provide substantial insight sequences of other products of combustion; in par-
into the complex fluid, heat transfer, and chemical ticular, there is current debate on the consequence of
processes that occur during compression ignition CO2 and the role it may play in the presently
combustion where fuel injection is used as the direct observed warming of the planet (i.e., so-called
ignition trigger. Description of these details is out- global warming or global climate change). Because
side the scope of this work. of the certainty of the effects of the former species,
attention will be given to them and basic informa-
Emission Formation and Exhaust Pollution tion on their formation during combustion. Because
A discussion on the basics of internal combustion of the uncertainty of the effects of the latter species,
engines, actually any combustion-based device, is readers are referred to other literature to uncover the
incomplete without a description of the associated current state of debate of CO2 and its potential
harmful species that may exist in the products of impact on global trends currently believed to occur
combustion. Such harmful species are generally (see, e.g., [79, 80]).
called “exhaust pollution,” and many governing The first such species to describe is CO, which,
agencies around the world place restrictions due to its fatal effects on human/animal life, is one
on the emission of certain pollutants from of the first combustion products to be considered a
combustion-based devices. Because the applica- pollutant [76]. It is well established [76] that all
tion of the internal combustion engine is so varied, precursor hydrocarbon decomposition reactions
emission regulations tend to be application- firstly form CO. Thus, increased concentrations
oriented. For example, in the USA, emission of CO in engine exhaust result in incomplete reac-
regulations are placed on passenger vehicles dif- tion of the principal CO oxidation step [26], given
ferently than emission regulations placed on as Reaction (R2):
164 Internal Combustion Engines, Developments in

CO þ OH $ CO2 þ H (R2) temperature T (K) of oxygen and nitrogen, respec-

tively. The several simplifying assumptions that go
In internal combustion engines, the incomplete into Eq. 15 include the following. The first assump-
oxidation of CO most typically occurs during rich tion is that the nitrogen chemistry is decoupled
engine operation [76], where available oxidants from the combustion reactions. Although combus-
for final CO oxidation are lacking [81]. Fuel-air tion reactions generally occur much faster than
mixtures close to stoichiometric or even slightly nitrogen chemistry [25], the presence of O and
lean, however, result in nonnegligible concentra- OH radicals in the thermal mechanism (which are
tions of CO. also important species in combustion reactions)
The next species to consider is NO. NO may require the chemistries to be coupled for accu-
emerged as combustion-generated pollutant due rate NO prediction [27]. By assuming the chemis-
to its observed effect of reacting with hydrocar- tries are decoupled, O, O2, OH, H, and N2 can be
bons in the presence of sunlight to produce tropo- approximated by their equilibrium concentrations
spheric ozone [82]. Its formation in a combustion at equilibrium temperature, assuming equilibrium
system is rather complex, as there are several temperature is the second assumption applied to
major pathways through which it can form. Eq. 15. The third and last assumption applied to
These major pathways include thermal (or com- Eq. 15 is that nitrogen radical (N) is in steady-state
½N
monly called Zeldovich), prompt, and fuel-based concentration (i.e., ddt ¼ 0). Finally, it is reinforced
nitrogen [26]. For reciprocating-type internal that the forward reaction rate constant of Reaction
combustion engines, the primary NO formation (R3) is used from [26] in Eq. 15; updated reaction
mechanism is the thermal mechanism where rates are available in Dean and Bozzelli [29]. It is
atmospheric air serves as the principal source of clear from Eq. 15 the strong dependency NO for-
nitrogen in the mechanism [27]. There are three mation rate has on temperature.
reactions that compose the mechanism, given as Because of the dominance of the thermal
Reactions (R3, R4, and R5): mechanism on NO formation in internal combus-
tion engines, combustion-based efforts to reduce
O þ N2 $ NO þ N (R3) NO center on reducing the reaction temperature
and O2 concentration. Such techniques include
N þ O2 $ NO þ O (R4) altering spark advance [83, 84] or injection timing
[85–87] for spark ignition or compression ignition
N þ OH $ NO þ H (R5) engines, respectively, and introducing EGR into
the mixture [87–90].
The forward and reverse reaction rate constants As described above with NO, unburned hydro-
of Reactions (R3) and (R4) are generally expo- carbons (HC) play a role in the formation
nentially dependent on temperature. To demon- of tropospheric ozone . Further, they represent
strate the substantial role temperature plays on NO unreacted fuel; if left to discharge into the atmo-
formation, several simplifying assumptions (see sphere, there is lost opportunity to convert that
below) are applied to Reactions (R3) and (R4) to chemical energy into useful work. Formation of
yield Eq. 15 [26]: HC species during combustion reaction is com-
  plex and varied depending on the type of fuel used
d½NO 6  1016 69090 ½O2, eq 12  [91]. A general cause for HC emissions, however,
¼ exp N2, eq
dt 1
T2 T is insufficient mixing between fuel and air [76],
ðmol=cm3  sÞ where most of the HC emission species are
(15) formed during low-temperature (T <1,000 K)
reactions [26]. Typically, in conventional recipro-
where [NO] is the concentration (mol/cm3) of NO cating internal combustion engines, most of
at time, t (s), and [O2,eq] and [N2,eq] represent the these species are oxidized as combustion enters
equilibrium concentrations (mol/cm3) at high-temperature mechanisms [26]. Reciprocating
Internal Combustion Engines, Developments in 165

engines, however, contain “sources” for hydrocar- function of temperature but has a stronger depen-
bon storage that, if not oxidized upon release in dency on temperature than soot formation
the gas mixture in the later portion of the cycle, [99]. The difference between soot that is formed
emit as HC emissions in the exhaust [92]. and soot that oxidizes is ultimately released in the
Although there are special considerations given engine exhaust, and is commonly called “net soot
to engines operating under cold-start conditions release.” Compression ignition engines (e.g., die-
[93], the general storage locations of HC species sel engines) are often plagued with high PM emis-
(and thus the major source of HC emissions [94]) sions, largely due to the heterogeneity induced
include cylinder head gasket crevice, spark plug into the fuel-air mixture by use of fuel injection
crevices, piston ring pack crevices, and valve seat as a direct trigger for ignition. Spark ignition
crevices. In addition to crevice HC storage, other engines, however, are recently being considered
sources of HC emissions [94] include single-wall as sources of nanoscale soot particles [100, 101].
flame quenching, oil film layers, combustion It is clear from the above discussion that
chamber deposits, exhaust valve leakage, and liq- soot and PM emission processes are complex.
uid fuel (i.e., HC species not vaporized during the From phenomenological considerations, how-
process). Because of the relative importance of ever, Khan et al. [86, 102] and Ahmad [103]
crevice storage on HC emissions, much of the provide insight into the general behavior of
combustion-based effort to decrease HC emis- PM with, in particular, diesel engine combustion.
sions has centered on reducing the volume and Specifically, it is observed that increases in diffu-
flow pattern of crevices in the piston/cylinder sion heat release generally increase the emissions
arrangement. of PM. A portion of this may be due to the overall
The final major pollutant to briefly describe is lower reaction temperatures, manifested by rela-
particulate matter or PM. PM is essentially any tively lower premixed heat release, decreasing the
exhaust specie that can be collected on a filter; it rate of soot oxidation through the diffusion flame.
typically is structured on a solid organic compo- This observation is also supported, however, by
nent (which is mostly pyrolized carbon particles Dec’s [74] observation that precursor soot forma-
or “soot”) upon which organic (e.g., unburned tion occurs in the standing premix reaction zone
HC) and nonorganic (sulfates) components within a diffusion flame sheet that exists in diesel
build. Since soot serves as the building block for combustion. Thus, increased diffusion burn cor-
PM, much of the research efforts are dedicated to respondingly results in increased soot emission.
understanding soot formation processes [95–99]. Either way, it becomes apparent when relating this
The general path for soot formation begins with discussion to the known effects of injection
the high-temperature pyrolysis, or fragmentation timing, for example, on diffusion combustion and
to hydrocarbon radical, of the hydrocarbon fuel. NO emissions that an attempt to decrease soot by
Such pyrolysis typically leads to a certain group of creating higher temperatures with, perhaps, more
hydrocarbon radicals called polyaromatic hydro- premixed heat release will correspondingly
carbons . Such polyaromatic hydrocarbons that do increase NO emissions; hence, the establishment
not oxidize during the reaction serve as the nucle- of the conventional “soot-NO x” tradeoff of diesel
ation site for soot growth. Nuclei then begin to engines.
coalesce and substantially increase surface area. As described briefly above with each species,
Now particles, the high surface area soot particles, there are in-cylinder and combustion-based
agglomerate and allow other species (i.e., liquid methods to reduce the formation of various
HC, condensed gaseous HC, and sulfates) to exhaust pollutants. Advanced development of
absorb onto the surfaces [99]. Soot formation is combustion systems continues to focus on these
strongly dependent on mixture fuel-air ratio and in-cylinder methods, as described in section on
also temperature (both tend to govern the rate of “Future Directions of Internal Combustion
pyrolysis). Formed soot can, however, undergo Engines.” Also in use to eliminate exhaust pollu-
oxidation as well. Soot oxidation is likewise a tion are exhaust aftertreatment systems, which
166 Internal Combustion Engines, Developments in

conventionally are catalyzed devices (hence, HC to yield stabilized CO2 and H2O. The effi-
their common name of “catalyst” or “catalytic ciency of the three-way catalyst – or the conver-
converter”) [104–106]. The basic idea of a cata- sion effectiveness of converting a given species
lyst is to promote a reaction that otherwise would to its more stable species, e.g., CO to CO2 – is
not proceed. In aftertreatment of engine exhaust, strongly dependent on the constituent composi-
there are several catalysts in use depending on tion on the inlet mixtures, as shown in Fig. 19
the species composition of the exhaust. For [104]. Notice that maximum conversion efficien-
example, the conventionally named “three-way cies of NOx, CO, and HC occur very near-
catalyst” is often used with conventional gaso- stoichiometric air-fuel ratios; a small departure
line spark ignition engines. Such a catalyst is from stoichiometric conditions – e.g., about
typically an exhaust flow-through device com- 0.5% increase or decrease in A/F ratio – results
posed of a ceramic monolith (the substrate) with in a nearly 20% decrease in conversion efficiency.
thru-hole passages to allow exhaust gases to flow Because of such an intolerable response of the
and a metal-oxide “washcoat” that suspends cat- catalyst, effective control of gasoline spark igni-
alytic particles on the surfaces of the monolithic tion engines requires precise control of the mix-
passages. The monoliths are usually constructed ture stoichiometry during operation.
of coerderite, and the metal-oxide washcoat is Also clear from Fig. 19 is the challenge of
often an alumina washcoat. Catalyzed particles outfitting a typical platinum-based catalyst with
are often from the precious metals group (e.g., non-gasoline spark ignition engine technology.
platinum, palladium, and rhodium), with plati- Diesel engines which typically operate fuel lean
num being the most commonly used metal. The (i.e., oxygen rich), for example, create constituent
general operation of the three-way catalyst is to exhaust species that are difficult to catalyze using
promote the reduction-oxidation reactions conventional techniques. Advanced aftertreatment
among CO and HC (the reductants) and technologies for diesel engines, or other advanced
NO/NO2 (the oxidants). In other words, the cat- combustion/engine systems, are under develop-
alyst promotes the reduction of NO or NO2 to ment and beginning to appear in production appli-
yield stabilized N2 and the oxidation of CO and cations [105, 106].

Internal Combustion
Engines,
Developments in,
Fig. 19 Catalyst efficiency
of a typical three-way
catalyst interacting with
exhaust from a typical
gasoline spark ignition
engine as a function of
mixture air-fuel ratio (A/F
ratio) (Used with
permission from [104])
Internal Combustion Engines, Developments in 167

A Case Study: Diesel Engines Versus than about 20–40% departure from stoichiometric
Gasoline Engines will typically cause misfire [107]. It is noted that
this requirement [107] is only for ignition and
At this point, with the basics of internal combus- flame propagation. The integration of a three-
tion engines having been described, it is useful to way catalyst with a conventional spark ignition
do a “state-of-the-art” technology comparison engine requires even more precise control over
between the two dominant conventional internal fuel equivalence ratio, as described in section on
combustion engines – i.e., the gasoline spark igni- “Emissions Formation and Exhaust Pollution”
tion engine and the diesel compression ignition [104]. Thus, conventional spark ignition engines
engine – to set the stage for discussions on future are not able to tolerate substantial departures from
directions in internal combustion engines. The stoichiometric conditions. Further, because of this
elementary comparison is provided in Table 2; requirement, it is necessary to avoid mixture stri-
note that this comparison is neither comprehen- ations between fuel and air. Thus, conventional
sive nor general. It is intended to highlight the spark ignition engines make use of premixing
common state of the art of the two technologies devices (e.g., carburetors or throttle body and
and create a sense for the trends of future direc- port fuel injectors) to prepare the fuel-air mixture
tions in internal combustion engine development. prior to induction. The result is a homogeneous
It is clear from Table 2 that the two engine mixture of fuel, air, and other residuals such as
types differ in virtually every way, save for their residual fraction and purposefully introduced con-
common use of the kinematic elements of the stituents such as EGR. In order to assist with the
crank-slider and piston/cylinder components. homogenization of the fuel-air mixture, light dis-
Consider first the spark ignition engine; its use tillates and aromatics with high volatility are
of spark requires a relatively tight control on the typically used as the fuel for conventional spark
fuel-air mixture equivalence ratio. The flame ignition engines. The most common, of course,
travel time is minimized when fuel-air equiva- is the mixture of light distillates and aromat-
lence ratio is near stoichiometric, and greater ics commonly called “gasoline” or “petrol.”

Internal Combustion Engines, Developments in, general; rather, it is intended to highlight the common state
Table 2 Elementary comparison between conventional of the art and create a sense for the trends of future direc-
spark ignition engines and conventional compression igni- tions in internal combustion engine development
tion engines. The comparison is neither comprehensive nor
Conventional spark ignition engine Conventional compression ignition engine
Feature (homogeneous charge spark ignition) (heterogeneous charge compression ignition)
Also known as Gasoline, petrol, Otto engine Diesel engine
Method of Spark Compression
ignition
Fuel Precisely controlled to stoichiometric Varies depending on engine load; typically
equivalence remains lean
ratio
Fuel-air mixture Carburetion, throttle body injection, port Direct injection or indirect injection
preparation injection
Degree of fuel- Homogeneous Heterogeneous
air mixing
Fuels used Gasoline, alcohols, ethanol, hydrogen, high- Diesel, oils (transesterified), hydrogen, low-
volatility hydrocarbons volatility hydrocarbons
Compression Ca. 8–11 Ca. 14–21
ratios
Method of load Throttle restriction of air or fuel-air mixture Fuel injection duration (i.e., fuel quantity)
control (depending on mixture preparation)
168 Internal Combustion Engines, Developments in

Interestingly, largely due to the reactive nature of directly delivered to the compressed air [61]; of
the homogeneous and near-stoichiometric mix- course, it is also clear that direct (or indirect)
ture [108], compression ratios of spark ignition injection of the fuel to compressed air is necessary
engines are limited to around 11. Compression so as to control start of combustion. Further, to aid
ratios higher than this promote the undesirable in the gradual and isothermal conversion from
autoignition of certain regions of the fuel-air mix- chemical to mechanical energy, Diesel stipulated
ture during the combustion process, a phenome- that there be a chemical abundance of air (i.e.,
non known as “fuel knock.” Fuel knock can be fuel lean) in the mixture [61]; practical imple-
very damaging to an engine, mostly due to its mentations of diesel engines make use of varying
occurrence in the end regions of cylinder, where fuel equivalence ratios to control the load of the
uncontrolled autoignition of the mixture results in engine. The direct (or indirect) injection of fuel
excessively high rises in pressure near susceptible into the compressed air, contrary to Diesel’s con-
components of the piston/cylinder arrangement ception, establishes a heterogeneous mixture that
(e.g., piston rings). As an aside, initially tetraethyl within itself has widely varying fuel-air ratios and
lead and eventually other lead alkyls were used reaction temperatures. Further, since ignition is
in gasoline as antiknock agents, allowing for manifested through compression, fuels with high
compression ratios to increase in spark ignition ignitability (e.g., heavy distillates and aromatics)
engines [109]. Concerns over the effect of lead on are required (correspondingly, the fuels also have
human and animal life, as well as the implica- relatively lower volatility, further establishing the
tions of lead on catalytic devices, have mostly heterogeneous nature of diesel combustion). Also
eliminated lead-based additives from fuels related to the use of compression to ignite the
[110]. Finally, and similarly related to the spark’s mixture is the need to have high compression
need to ignite a near-stoichiometric and homoge- ratios (i.e., typically varying between 14 and 21).
neous mixture, conventional spark ignition Finally, as already described, engine load is con-
engines control their load, or power output, via trolled by the direct and exclusive control of fuel;
the use of a throttle. In other words, the power of thus, engine load varies as mixture fuel-air equiv-
the engine at any given speed is minimized by alence ratio varies. Due to the heterogeneous
decreasing the efficiency by which the engine can nature of diesel combustion, and the dependency
induct fresh mixture. of soot formation on mixture stoichiometry under
Consider now the conventional compression conventional conditions, fuel equivalence ratios
ignition engine or commonly called diesel engine. rarely exceed 90% of stoichiometric to avoid
It seems that from its first inception, the diesel “smoke limitation.”
engine was to be a compressively ignited machine, In the context of the three common attributes of
as Diesel intended to create an isothermal reaction internal combustion engines (i.e., performance,
where air and fuel were (separately) compressed efficiency, and emissions), and the comparison
to the combustion temperature and expansion given in Table 2, brief comments will be made
occurred isothermally as chemical energy con- about the pros and cons of each conventional
verted directly into mechanical energy [61]. Of technology.
course, Diesel never succeeded in attaining this In terms of power, consider Eq. 7 and the
isothermal process (nor has anyone since), but the qualitative relation of each parameter for a given
foundation had been laid for a compression igni- technology at their “wide-open throttle” or full-
tion engine. By using compression ignition com- load operation, as given in Table 3. Described in
bustion, many of the constraints placed on the more detail below, diesel engines tend to have
spark ignition engine (i.e., near-stoichiometric higher fuel conversion efficiencies. Because of
fuel equivalence ratio, homogeneous mixture, not premixing the fuel with air and generally
and throttle of intake mixture) no longer need to lower engine speeds (which allows the avoidance
apply to the diesel engine. It also seems that from of flow choke), diesel engines tend to have higher
its initial inception, Diesel intended the fuel to be volumetric efficiencies (refer to the discussion
Internal Combustion Engines, Developments in 169

Internal Combustion Engines, Developments in, are qualitative assessments of typical technology and
Table 3 Comparison of the various parameters that con- should not be viewed as absolute truths of the respective
trol an engine’s ability to make power (Eq. 7) between technologies. Since typical applications of internal com-
typical applications of conventional spark ignition and bustion engines operate on the four-stroke principle, its
conventional compression ignition engines at wide-open effect on the comparison is neutral
throttle or full-load condition. It should be noted that these
Parameter’s effect on Conventional spark ignition Conventional compression ignition
Parameter powera engine engine
f " "
v " "
ra,i " "
(F/A) " "
QHV " "
Vd " "
N " "
nR # $
a
In other words, an increase in the parameter will have the listed effect on power

surrounding Fig. 6 and [111] for more detail on As described in section on “Thermodynamic
factors affecting volumetric efficiency of engines). Analysis of Internal Combustion Engines,”
Typical applications of diesel engines use engines operating with high compression ratios
turbocharging, which increases the inlet mixture and lean fuel equivalence ratios will tend to have
density to above atmospheric conditions (typical higher efficiencies than engines operating with
spark ignition engines operate naturally aspi- relatively lower compression ratios and near-
rated). As described above, spark ignition engines stoichiometric equivalence ratios. As a result, typ-
generally operate stoichiometric, whereas com- ical compression ignition engines tend to have
pression ignition engines generally operate lean higher efficiencies than typical spark ignition
even at full power. The heating value of diesel engines. In some instances, turbocharging that is
fuel is marginally higher than that of gasoline. typically found on compression ignition engines
Displaced volumes of typical compression igni- could increase efficiency if intake manifold pres-
tion engines tend to be larger than those of typical sure is boosted to higher than exhaust manifold
spark ignition engines. Correspondingly, peak pressure; the primary function of a turbocharger,
power speeds for typical spark ignition engines however, is to increase inlet mixture density to
tend to be higher than compression ignition increase the power capabilities of the engine
engines. Mostly due to the larger displaced vol- (as shown in Eq. 7). The efficiency improvement
umes, but also assisted by higher fuel conversion of compression ignition engines at part-load con-
and volumetric efficiencies, higher inlet density, ditions becomes amplified as (a) the mixture
and higher heating value of the fuel, typical com- becomes leaner for the compression ignition
pression ignition engines tend to exhibit higher engine and (b) the use of throttle to manifest part
peak powers than typical spark ignition engines. If load in the spark ignition engine introduces a
the power is normalized by displaced volume to thermodynamic loss parameter in the cycle.
render the specific power, however, spark ignition Finally, a brief comparison of emissions
engines tend to have higher specific power than between the two engines at wide-open throttle or
compression ignition engines (in spite of com- full-load condition is made. Because of their near-
pression ignition engines having high efficiencies, stoichiometric operation, engine-out emissions of
density, and heating value, the stoichiometric and HC and CO for spark ignition engines tend to be
high-speed operation of the spark ignition engine relatively higher for those of compression ignition
tends to yield higher specific power). engines (where the latter uses fuel-lean mixtures,
170 Internal Combustion Engines, Developments in

creating an oxygen-rich exhaust and high level of its potential entry into full-scale production (e.g.,
oxidation of partially oxidized species such as CO homogeneous charge compression ignition com-
and HC). Engine-out emissions of NO tend to be bustion). This section will briefly describe such
nearly the same between engine technologies. technology.
Engine-out emissions of PM are much higher for
compression ignition engines, where mixture het- Engine Downsizing
erogeneity creates numerous opportunities for A general idea that permeates much of the tech-
soot formation. Of course, it is important to note nology under development of internal combustion
that conventional spark ignition engines are typi- engines is that of engine downsizing or the effort
cally coupled with an effective catalyst, substan- to use advanced technology to enable the use
tially lowering the catalyst-out emissions of of smaller-sized engines to produce the same
the various species to levels well below the power (i.e., increase power density) [112]. The
engine-out emissions of compression ignition application-oriented benefit is that a smaller
engines. While the use of aftertreatment systems engine likely weighs less, thus could improve
with conventional compression ignition engines is application efficiency (e.g., better vehicle fuel
less straightforward, technology is becoming economy with a lighter engine). The engine itself,
available to also allow substantial reduction of however, will likely realize improved efficiency.
their engine-out emissions [106]. For example, a smaller engine that uses turbo-
At this point, it becomes clear both engine charging to maintain same power (as a larger-
technologies have features which are more favor- sized engine) will generally operate more often
able than the other for power, efficiency, and near the location of peak efficiency. Some benefit
emission considerations. For example, the high of such is realized in diesel engines, which typi-
compression ratio and lean mixture required by cally have best efficiencies near mid-speed, and
compression ignition are attractive from effi- 75% peak load conditions. More benefit, how-
ciency perspectives. The homogeneous mixture ever, is realized in gasoline engines where
of spark ignition is attractive from uniform com- throttles are used; a smaller engine with higher
bustion and emission perspectives. The lack of a power density will require less throttle and thus
throttle to control load of a compression ignition realize larger gains in efficiency improvement.
engine is attractive, but the direct ignition trigger Overall, the major purpose of engine downsizing
of a spark ignition engine is also attractive. Thus, is to improve parameters other than Vd in Eq. 7.
it is clear that future engine developments could This idea will become more apparent as specific
exploit the favorable features of engine technolo- technology is discussed below.
gies to create the next generation internal com-
bustion engine. The next section will describe Turbocharging/Supercharging/Boosting
such efforts and offer insight into the likely future Boosting an engine – i.e., increasing the inlet
direction of internal combustion engines. mixture density to increase the trapped mass per
cycle – is a very common means to increase the
power density of an engine. As evident from
Future Directions of Internal Eq. 7, an increase in the inlet air density will
Combustion Engines directly increase the power of the engine. There
are two major types of boosting technology:
There are several types of advanced technology turbocharging and supercharging. The major dif-
that exist for internal combustion engines. Some ference between the two technologies is the for-
of this technology is very prevalent on engines mer uses a centrifugal device (i.e., a turbine) to
(e.g., turbocharging on diesel engines). Some exploit available exhaust energy for conversion to
technology is beginning to appear on production shaft work, whereas the latter absorbs shaft work
models (e.g., variable valve timing). Other tech- via a direct mechanical connection to the engine.
nology is still in its development stages, awaiting In both cases, the shaft work of the device is
Internal Combustion Engines, Developments in 171

coupled to a boosting component – typically, transfer, however, usually does not leave the con-
either a centrifugal-based compressor or a trol system (i.e., the shaft work of the turbine
positive-displacement compressor – which acts is directly coupled to the pumping action of
as an “air pump” to increase density of the inlet the compressor). In some instances, efficiency
air. In most applications, a turbocharger uses a improvements can be realized if a “negative
centrifugal turbine/compressor configuration, pumping loop” is created by boosting the intake
while a supercharger uses a positive-displacement manifold to a higher pressure than the exhaust
compressor. manifold. It is also possible to improve the overall
Diesel engines are typically favorable engines engine efficiency if boosting an engine enables the
to outfit with boosting devices (further, usually use of a smaller-sized engine for a given applica-
with turbochargers as they assist over the entire tion, an effort, as described above, known as
engine operating map). Boosting compensates for downsizing. In most instances, however, overall
the diesel engine’s typical use of fuel-lean mix- system efficiency may decrease even with the use
tures to improve its power density. Further, since of a turbocharger, as the major objective of
diesel engines do not employ throttles for load increasing inlet density to increase power density
control, boosting of a diesel engine provides ben- requires additional energy transfer through the
efit over the entire operating map. For this latter exhaust than out as shaft work. This latter aspect
reason, and due to the general higher efficiency of is often realized as an increase in exhaust mani-
a turbocharger over a supercharger, turbochargers fold pressure to “drive” the turbocharger.
are the most common boosting device on a diesel An area of technology development for
engine. Gasoline engines also can be outfitted turbocharging is the use of waste-gated and vari-
with boosting devices, with additional complexity able geometry turbochargers. In non-waste-gated
to consider. First, boosting devices on gasoline or non-variable geometry turbochargers, the tur-
engines usually offer benefit only at wide-open bine has fixed geometry and thus fixed flow
throttle conditions, a point of operation for most characteristics. As such, a fixed geometry turbo-
applications of engines that is rarely used. Also, charger is restrictively designed to provide maxi-
for this reason, directly coupled superchargers are mum benefit to the engine at a narrow operating
often used where improved response to boost is range which usually centers on the peak power
provided. Second, increasing density of the inlet condition of the engine. As such, at low-speed or
mixture tends to increase the propensity to knock. low-load conditions, the turbocharger’s boosting
Thus, while boosting provides additional trapped benefits are diminished. While decreasing the tur-
mass to deliver increased power, a potential retard bocharger’s maximum benefit over a broader
in timing to avoid fuel knock likely decreases fuel regime of the engine’s operating map, such a
conversion efficiency and introduces a trade-off in constraint also typically creates a dynamic issue
how much additional power can be expected from during engine accelerations known as “turbo lag.”
the boost. Third, because of the typical use of The large turbine designed for maximum engine
superchargers when boosting is applied to gaso- flow rates requires substantial inertia to rota-
line engines, the system efficiency tends to tionally accelerate to maximum boosting benefit.
decrease as shaft work is transferred to provide Waste-gated and variable geometry turbochargers
the boosting action (whereas turbocharging uses offer opportunities to overcome such issues. In the
available energy of the exhaust). case of a waste-gated turbocharger, usually a
Oftentimes, boosting devices – in particular, smaller turbine with less inertia is used with an
turbochargers – are viewed as devices to increase exhaust “waste gate.” The smaller turbine pro-
the overall system efficiency of the engine. In vides faster response and better boosting at low
other words, it is thought that because a turbo- loads and speeds; upon approach of the engine’s
charger uses exhaust energy that would otherwise peak power condition (where either boosting
be wasted, its conversion to useful work (i.e., becomes excessive for the intake system or turbo-
boosting) should increase efficiency. This work charger speeds exceed maximum limits), a waste
172 Internal Combustion Engines, Developments in

gate opens that allows exhaust energy to bypass in-cylinder pressure is the major property neces-
the turbine. Thus, the turbocharger is supplied by sary for assessing the rate of energy release during
a fraction of the available exhaust energy provid- the combustion process; having such information
ing suitable boost at allowable rotational speeds of could allow engine control systems to change
the turbocharger at the engine’s peak power con- parameters based on a desired burn profile in the
dition. A variable geometry turbocharger uses a cylinder [120–122]. Further, along with the con-
similar concept, except that it is designed to tinuing advancements in engine model develop-
change the flow momentum of the exhaust gases ment (e.g., [46–50]), a trend toward model-based
and create multiple pressure ratios for a given engine control [123, 124] intends to decrease
exhaust flow rate [113, 114]. engine development time and improve control
Finally, it is noted that outfitting a boosting over the several parameters now present on inter-
device to an engine is not a trivial task [115]. nal combustion engines.
There is not an exclusive match between an
engine and a given boosting device; instead, the Variable Geometry Engine Designs
match of a boosting device to its engine applica- Along with efforts to effect engine downsizing,
tion is dictated by the objectives of adding and made possible with advancements in engine
the device, whether it is to exclusively increase control systems, is the notion of variable geometry
power density, efficiency, and/or emissions of the engine designs. In other words, conventional
engine system [116–118]. reciprocating internal combustion engines have
mechanically fixed geometries, i.e., constant com-
Advanced Engine Controls pression ratios and constant displaced volumes.
Much of the advanced technology that appears on A variable compression ratio is attractive, for
modern engines, or will appear on future engines, example, since it might enable high compression
is enabled by advanced engine controls. Engine ratio operation for a spark ignition engine at part-
control has always been an integral component of load conditions where there is decreased propen-
the engine’s success at delivering cost-effective sity for knock. Similarly, it could be used in diesel
and efficient power. Early engine control systems engines to avoid excessively high-peak pressures
were purely mechanical and only concerned with at full-load conditions. Variable displacement
controlling the level of power (e.g., throttle or rack engines are attractive since they enable high-
position) or holding a constant speed with variable peak powers but use less throttle at part-load con-
load (such as a generator system, using a speed ditions (thus diminishing pumping losses).
governor). Modern-day engine control systems, An early concept of variable compression
however, are virtually all electrical-based, and at ratio was proposed by J. Atkinson, for whom the
the very least sense, several aspects of the engine’s Atkinson Cycle is named. As described by [125],
operation (e.g., cam position, throttle position, Atkinson’s original conception involved mechan-
manifold temperatures and pressures, and air ical linkages to displace the piston such that the
flow) typically control most aspects of the engine engine’s compression ratio is lower than its
(e.g., spark timing, injection timing, injection expansion ratio, the main idea being that more
pressure, EGR level, and boost pressure) [119]. expansion work will yield higher efficiency. Sim-
Like the engine itself, engine controls are ilar in idea, but different in implementation, is the
becoming more sophisticated and advanced. Spe- Miller concept [125] which uses either late intake
cifically, a general trend to use in-cylinder infor- valve closing or late exhaust valve opening to
mation as a feedback signal is an example of the shorten or extend the compression or expansion
type of complexity future engine control systems strokes, respectively. Both concepts are attempted
intend to resolve. Knowledge of in-cylinder in modern-day applications, using both variable
pressure, for example, can provide immediate compression ratio techniques [112, 126, 127]
information to the engine controller about and altered valve timing techniques [112, 127,
load produced by the engine. Additionally, 128]. The latter approach, of using altered valve
Internal Combustion Engines, Developments in 173

timing techniques, is fluidly made possible having variable valve timing allows the imple-
through the use of variable valve timing [112, mentation of the Miller approach to effecting var-
127], a concept discussed in more detail in the iable compression ratio at part-load conditions but
next section. enabling conventional operation at peak power
While variable compression ratio concepts conditions [127]. Another example is the use of
attempt to increase engine efficiency by way variable valve timing to control the amount of
of increasing expansion, variable displacement residual fraction in the cylinder, which can affect
engines attempt to increase engine efficiency by emission of certain pollutants [133]; controlling
decreasing the use of throttle (thereby, decreasing residual fraction is a way to enable advanced
the pumping work associated with the gas modes of combustion, described in more detail
exchange process of a conventional spark ignition below. Finally, variable valve timing can be used
engine). With variable displacement, an engine is to replace a throttle, for example, for load control
able to deliver high power using full displace- [134]; in doing so, trapped mass can be controlled
ment; at part-load conditions, rather than use without inducing pumping work in the engine.
throttle, cylinders can be “deactivated” so that Although not widespread technology, variable
they produce no power and allow the engine valve timing is becoming more prevalent on
to deliver part-load power [129]. The deactivated modern-day engines.
cylinders typically continue to stroke and ex-
change gases; the closed portion of the cycle Waste Heat Recovery
(i.e., compression and expansion) realizes some Waste heat recovery is the effort to take advantage
loss due to heat transfer and friction, but this is of temperature gradients created between the
intended to be less than the gain realized through engine and its environment. As described in sec-
decreased pumping work. tion on “Thermodynamic Analysis of Internal
Combustion Engines,” thermal energy is trans-
Variable Valve Timing ferred out of the system through heat transfer
Briefly discussed in the above sections is the idea (e.g., through engine coolant) and exhaust flow.
of variable valve timing or the ability to change Because of the temperature gradient that exists
the valve events (i.e., intake and exhaust valve between, for example, the high-temperature
opening and closing) at any given point during exhaust and the low-temperature surroundings, an
the engine’s operation [130]. In conventional orderly flow of thermal energy tends to cause the
engine design, the valve events are “fixed” by exhaust system to attain thermal equilibrium with
mechanical positions of the lobes on the cam the environment. Conventionally, this orderly flow
shaft. The effectiveness of an engine to induct of thermal energy is wasted (i.e., the heat transfer
and exhaust mixture depends on many things completely dissipates as generated entropy). It is
including engine speed, engine load, the use of practically possible to instead intercept the orderly
EGR, spark timing, and injection timing. Thus, flow of thermal energy and convert it to useful
there are not valve events that will universally work. In the theoretical limit, this conversion of
yield peak power, efficiency, and/or emission for thermal energy to work energy is given by the
any given engine design. completely reversible cycle, often called the Carnot
The idea of variable valve timing allows the cycle. The corresponding efficiency of useful work
engineer to decouple the valve events from the converted from thermal energy is thusly the Carnot
in-cylinder processes. In other words, flexibility efficiency, as is given by Eq. 16:
of intake and exhaust is afforded with the use of
variable valve timing. This not only allows the TL
th, Carnot ¼ 1  (16)
valve events to be uniquely tuned for each oper- TH
ating point of the engine for improved perfor-
mance [131] and efficiency [132], but it also can where th, Carnot is the Carnot efficiency
enable other advanced technology. For example, (maximum possible conversion of thermal energy
174 Internal Combustion Engines, Developments in

to work energy), TL is the sink’s temperature (e.g., mechanical energy; the cycle completes with the
the environment temperature), and TH is the usual condenser and pump processes. Such a device
source’s temperature (e.g., the exhaust tempera- is reported to increase the combined engine + waste
ture). Considering that a typical automotive heat recovery efficiency by up to 10% [137]; a
exhaust temperature may be 900 K operating in potential downside is the added complexity of
a 300 K environment, ideal efficiencies could be adding four processes (as opposed to one, for exam-
on the order of 67%. Of course, real process ple, in the case of a turbomachine) to waste heat
irreversibilities diminish actual device efficiencies recovery system.
from the ideal Carnot efficiency. In spite of the Another example of major waste heat recovery
diminished efficiency from ideal, though, waste device is the thermoelectric device. The thermo-
heat recovery in an automotive application, for electric effect, first observed by Seebeck in 1821,
example, is reported to decrease vehicle fuel con- is the generation of a voltage due to a temperature
sumption by as much as 7.4% [135]. difference between two junctions of two dissimi-
There are several technologies available to make lar materials [138]; principally, the Seebeck effect
use of internal combustion engine exhaust waste describes the operation of a thermocouple mea-
heat energy; because of its spatial accessibility and surement of temperature. The practical use of the
concentrated high temperature, exhaust energy has Seebeck effect to produce electricity as a thermo-
been the focus of much of the waste heat recovery. electric device has recently emerged with semi-
For example, major technologies include conductor materials having favorable properties
(1) mechanical or electrical turbo-compounding/ to transmit electricity with little resistance heating
generating devices, (2) Rankine cycle-type devices, (Joule heating) and thermal conductivity (which
and (3) thermoelectric-type devices. Turbo- would tend to “short-circuit” the thermoelectric
compounding or turbo-generating [136] converts device). In fact, the advent of semiconductor
exhaust thermal energy to mechanical energy materials first made possible practical thermoelec-
(in the form of pressure and kinetic energy) and tric devices as refrigerators exploiting the Peltier
couples the mechanical energy either directly to effect (i.e., the opposite of the Seebeck effect,
the engine driveshaft (to deliver additional brake where an applied voltage creates a temperature
power) or to an electric generator. One disadvantage difference between two junctions of two dissimi-
of turbo-compounding (like its turbocharging com- lar metals) [138]. Now, however, thermoelectric
panion) is the method of converting thermal energy devices create promise to exploit the available
to mechanical energy; the centrifugal device thermal energy in an internal combustion engine’s
increases an engine’s exhaust pressure based on its exhaust for conversion to electricity [135, 139].
operation. The increased exhaust pressure affects
the engine’s operation by altering the pumping Direct Injection and Spark Ignition Engines
work and initial mixture composition (increased At the end of the section on “A Case Study: Diesel
exhaust species in the initial mixture); such factors Engines Versus Gasoline Engines,” it is suggested
can deteriorate the engine’s cycle efficiency. that spark ignition and compression ignition engines
Rankine cycle-type devices make use of the ther- each have favorable features for power, efficiency,
modynamic Rankine cycle – typically with an and emissions, but that each has
organic fluid designed to undergo phase change conventionally designed limitations. Thus, it
within the temperature ranges of a typical engine becomes attractive to design each ignition system’s
exhaust system – to output shaft work for either limitations out of the engine and potentially realize
direct coupling to the engine driveshaft or electrical gains in the engine’s attributes (i.e., power, effi-
generation. In such a device, the engine exhaust ciency, and emissions). Two now-common technol-
stream provides the thermal energy to boil or vapor- ogies exist to do so: (1) direct injection spark
ize the organic working fluid of the cycle. After ignition combustion and (2) homogeneous charge
becoming saturated vapor, the fluid expands through compression ignition combustion. This and the next
a turbine converting the thermal energy to section will describe these two technologies.
Internal Combustion Engines, Developments in 175

Direct injection spark ignition combustion mixtures possess higher ratios of specific heats
attempts to create a stratified fuel-air mixture (g), which translate to higher fuel conversion
“charge” around the spark plug so that, at the efficiencies. Another, less obvious, benefit of a
point of spark release, the spark ignites a near- stratified mixture is the decreased propensity to
stoichiometric mixture. It is intended that outside fuel knock (or at least decreased propensity of
of the stratification zone, there is little to no fuel, fuel knock in the regions of the cylinder able to
thus creating an overall lean fuel-air mixture. The cause harm such as near cylinder walls and piston
use of the word “stratified” to describe the fuel-air rings). The lack of a reactive mixture –
mixture is used here, as opposed to heteroge- manifested by charge stratification – in the
neous, to reinforce the notion that under ideal regions furthest from the spark plug, which are
conditions the fuel-air mixture would be homo- the last to be controllably burned by the propa-
geneous (i.e., homogeneously stoichiometric) gating flame, decreases the propensity that the
throughout the fuel-air mixture but pure air out- mixture will autoignite and burn uncontrollably.
side the stratification zone (i.e., outside the fuel- Such a feature enables the direct injection spark
air mixture). This is in contrast to the use of ignition engine to have increased compression
“heterogeneous” to describe a mixture (e.g., diesel ratios relative to conventional spark ignition
engine mixture), where it is expected that substan- engines; again, this is an attribute that promotes
tial fuel-air ratio gradients exist throughout the an increase in efficiency of the novel engine
fuel sprays. concept. Finally, the use of direct fuel injection
In order to manifest the stratified mixture con- into the cylinder enables the elimination of a
cept (see, e.g., [140–148]), fuel is injected directly throttle to control engine load. In other words,
into the cylinder. Typically, the combustion cham- engine load is controlled by the quantity of fuel
ber of a direct injection spark ignition engine is injected into the cycle, similar to the load control
specially designed to assist the stratification of the of a diesel engine. Elimination of the throttle, as
fuel-air mixture and center it on the spark plug. repeatedly described, will improve part-load effi-
The intake stroke draws air and residual mixture ciency by eliminating pumping losses effected by
into the cylinder – notably, fuel is not inducted the throttle.
during intake as is done in conventional spark Although the benefits are plentiful, the chal-
ignition operation. After intake, direct fuel injec- lenges are also present. From practical perspec-
tion into the cylinder usually occurs at some point tives, perfect attainment of a stratified charge is
during the piston’s travel from BDC to TDC dur- difficult to accomplish. As such, heterogeneities
ing the compression stroke. Spark advance is typ- within the stratified mixture emerge and can lead
ically timed at around the same point as that in a to products of incomplete combustion such as CO,
conventional spark ignition engine (i.e., at a point HC, and PM. This, of course, is amplified at full-
near the piston reaching TDC compression). The load conditions; thus, peak power attainment
remaining processes of the direct injection spark through direct injection means alone would likely
ignition engine are basically the same as the con- be limited by smoke limitations (similar to full-
ventional spark ignition engine. load limitations of a diesel engine). Further, and
There are several potential benefits of direct like the challenges faced by diesel engines, outfit-
injection spark ignition combustion. Perhaps the ting direct injection spark ignition engines with
clear benefit is the ability to use lean mixtures in a aftertreatment devices is complicated by the use of
spark ignition engine. Because the combustion overall lean mixtures (as described in section on
chamber is designed to stratify the mixture and “Emissions Formation and Exhaust Pollution”).
create a stoichiometric mixture near the spark In spite of such challenges, direct injection spark
plug, the overall equivalence ratio of the mixture ignition engines are in production; continued
filling the entire chamber can be lean. As des- development of in-cylinder flow modeling tools
cribed in section on “Thermodynamic Analysis and engine controls contributes toward the con-
of Internal Combustion Engines,” overall lean cept’s potential success.
176 Internal Combustion Engines, Developments in

Homogeneous Charge Compression Ignition concepts do not have a direct ignition trigger, as
Engines do conventional gasoline (i.e., a spark) or diesel
An attractive feature of the conventional spark (i.e., direct fuel injection) engines. Controlling
ignition engine is its use of a homogeneous mix- ignition in the HCCI concept depends on very
ture. Although this tends to promote knock precise control of the mixture’s initial state at
(as described in section “Direct Injection Spark start of compression and the compression path
Ignition Engines”), it provides the benefit of being followed up to the point of ignition. Several fac-
kinetically rate-limiting as opposed to mixing tors which are present – even in tightly controlled
rate-limiting (as in the case of diesel engines). research environments – such as heat transfer,
One issue with using a spark, or single point, to turbulence, and the history of preceding combus-
ignite a homogeneous mixture is the establish- tion events make the practical application of
ment of a flame that must propagate the mixture HCCI very challenging. The payoffs, of course,
to convert the chemical energy. In order to more are correspondingly very high.
quickly react the mixture, in a volumetric sense, Practical implementation of HCCI is first
multipoint ignition is required; i.e., ignition that reported by Onishi et al. [149] with theoretical
occurs in several locations throughout the mixture developments experimentally provided by Najt
will result in a faster burn rate. Such a voluminous and Foster [150]. Several control parameters
ignition can be effected through compression of a could be adjusted such as compression ratio
homogenous mixture. (e.g., with the use of variable valve timing), initial
This is the basic idea of homogeneous charge temperature, and quantity of residual fraction
compression ignition (HCCI): to use compression (e.g., effected either through exhaust gas
to ignite a homogeneous mixture (similar to a recirculation or variable valve timing). Identifying
diesel engine, except that diesel engines use com- the key control parameters, and the optimal way to
pression to ignite an inherently heterogeneous adjust them during real-time operation of the
mixture). The homogeneous mixture could, for engine, continues to be on-going research activi-
example, be formed through premixing of fuel ties [151–156].
and air prior to mixture induction during the
intake stroke. A faster burn rate effected by homo-
geneous charge compression ignition allows heat Advanced Compression Ignition Engines
release to occur at near constant volume condi- A technique to control combustion of an HCCI
tions, thus establishing the possibility to approach engine is to use precisely metered amounts of
theoretical limits of maximum efficiencies of residual fraction, which not only act to alter the
internal combustion engines. Further, because kinetics of combustion but also result in substan-
compression is used to ignite the mixture rather tially lower combustion temperatures. As such,
than spark, lean mixtures can be used at part-load much of HCCI combustion is characterized by
conditions which further promotes higher effi- low-temperature mechanisms commonly referred
ciencies of the HCCI concept. Similar to the direct to as low-temperature combustion (LTC) . LTC
injection spark ignition concept, HCCI engines offers a few benefits. First, efficiency improve-
could eliminate throttles as load is controlled ments in the engine can be realized (in spite of
directly by the quantity of fuel mixed with the increased exergy destruction due to low-reaction
intake mixture. Finally, the common issue of par- temperatures) due to more favorable thermody-
ticulate matter emissions faced by diesel engines namic properties (i.e., higher ratio of specific
(which also use compression ignition) are sub- heats, see Fig. 11) of the burned mixture and
stantially decreased through the avoidance of lower rates of heat transfer. Second, and typically
locally rich fuel-air mixture zones (due to the use the driver for LTC technology development,
of a homogeneous mixture in an HCCI concept). lower nitric oxide formation is discussed in sec-
Of course, immediately the obvious problem tion “Emissions Formation and Exhaust
becomes the method of ignition control . HCCI Pollution.”
Internal Combustion Engines, Developments in 177

With this in mind and reconsidering the pre- fuels are the commonly called “gasoline” and
vailing issue of HCCI implementation – i.e., con- “diesel” fuels, but generally engines have been
trol of start of combustion – it becomes plausible shown to operate on virtually any gaseous, liquid,
to consider developing an “HCCI-type” mode of and solid dust particle specie that has heating
combustion in a diesel engine. That is, rather than value (i.e., will release thermal energy in a chem-
induct a homogeneous mixture of fuel and air and ical oxidation process). Because of the wide var-
rely upon indirectly controlled parameters to con- iability of fuels available to internal combustion
trol ignition, perhaps fuel can be injected directly engines, this topic will not be expanded in this
into the cylinder allowing for better control of entry. There are, however, certain considerations
ignition. In order to manifest LTC and harvest its that should be given to the use of a fuel in an
benefits (e.g., possibly higher efficiency and sub- engine which was not intently designed for use
stantially emission), high levels of EGR and stra- with such fuel (e.g., use of ethanol in a gasoline
tegic injection timings are used to extend ignition engine or use of biodiesel in a diesel engine). First,
delay and create a nearly all-premixed combustion ignition characteristics of the fuel may not be
event. The long ignition delay, coupled with low- favorable for the particular engine design. For
temperature mechanisms, establishes the phenom- example, short-chain volatile hydrocarbons do
enological observation of two-stage ignition not generally ignite well in conventional
characterized by the presence of cool-flame reac- unmodified compression ignition engines of typ-
tions [150, 157]. Interestingly, because of attain- ical compression ratios. Likewise, long-chain
ment of LTC, soot precursor formation is nonvolatile hydrocarbons do not generally vapor-
substantially abated, and the engine is made to ize well in conventional unmodified spark ignition
operate with very low emissions of nitric oxide engines. Second, flame temperatures of the com-
and particulate matter [158–168]. The combustion bustion of the fuel may exceed material limits of
concept has become known as premixed compres- any given engine construction. Third, fuels may
sion ignition or premixed charge compression react with other support components of the engine
ignition combustion. The ability to attain LTC in system (e.g., solvency of fuels with rubber hoses).
compression ignition engines is attributed to the Finally, combustion process will likely be altered
advancement of technology now in place on such when using an unconventional fuel in a conven-
machines, such as common-rail and electronic tional unmodified engine yielding different emis-
fuel pressure systems, variable geometry turbo- sions, efficiency, and peak power capabilities.
chargers, and exhaust gas recirculation systems. Thus, although internal combustion engines have
The idea of LTC is not limited to diesel inherent fuel flexibility, their use with unconven-
engines. Similar concepts apply to the use of tional fuels is not straightforward and requires
gasoline, in what is often called gasoline compres- careful design and engineering considerations.
sion ignition. LTC is generally plagued by limited
operating conditions; thus, engines employing
LTC often need other means to operate different Nomenclature
modes of combustion. For example, Mazda
announced in 2017 the release of SkyActiv-X Greek Symbols
engines, which use gasoline compression ignition
at most operating conditions but conventional c Combustion efficiency
spark ignition when demanded. f Fuel conversion efficiency
f,b Brake fuel conversion efficiency
Alternative Fuels f,i, Indicated fuel conversion efficiency
The term “alternative fuels” for an internal com- m Mechanical efficiency
bustion engine is somewhat baseless, as an inter- th Thermal efficiency
nal combustion engine has considerable flexibility th,Carnot Thermal efficiency of the ideal Carnot
in the type of fuel it uses. Of course, conventional cycle
178 Internal Combustion Engines, Developments in

th,Otto Thermal efficiency of the ideal heat mtotal Total mass

engine Otto cycle _a
m Mass flow rate of air
v Volumetric efficiency _f
m Mass flow rate of fuel
g Ratio of specific heats MEP Mean effective pressure
gb Ratio of specific heats of the burned Mb Molecular weight of the burned
mixture mixture
’ Fuel-air equivalence ratio. For ’ <1 nR Number of revolutions per engine
mixture is lean. For ’ = 1, mixture is cycle
stoichiometric. For ’ >1, mixture is N Engine speed
rich. Note that ’ is the inverse of the R Universal gas constant
often-used air-fuel equivalence ratio R Gas constant
or l. R5 Gas constant of mixture at state 5
ra,i Inlet air density R6 Gas constant of mixture at state 6
t Engine torque Re Gas constant of exhaust mixture
P Cylinder pressure or power
Symbols P1 Pressure at state 1
P2 Pressure at state 2
BMEP Brake mean effective pressure P3 Pressure at state 3
Cp,b Constant pressure specific heat of the P3a Pressure at state 3a
burned mixture P4 Pressure at state 4
Cv,b Constant volume specific heat of the P5 Pressure at state 5
burned mixture P6 Pressure at state 6
f Residual fraction P7 Pressure at state 7
ffinal Final calculated residual fraction Pb Brake power
ffinal – 1 Previous iteration residual fraction to Pe Exhaust pressure
final calculated residual fraction Pi Inlet (initial) pressure
(F/A) Fuel-air ratio Pin Net indicated power
FMEP Friction mean effective pressure Plimit Limit pressure
h1 Specific enthalpy at state 1 PMEP Pumping mean effective pressure
h2 Specific enthalpy at state 2 1Q 2 Heat transfer of process 1-2
h3 Specific enthalpy at state 3 6Q 1 Heat transfer for process 6-1
h3a Specific enthalpy at state 3a QHV Heating value of fuel
h5 Specific enthalpy at state 5 QHV,f Heating value of fuel
h6 Specific enthalpy at state 6 QHV,i Heating value of specie i
he Specific enthalpy of exhaust mixture rc Compression ratio
hi Specific enthalpy of inlet mixture s1 Entropy at state 1
IMEPg Gross indicated mean effective s2 Entropy at state 2
pressure s3 Entropy at state 3
IMEPn Net indicated mean effective pressure s4 Entropy at state 4
m Mass s5 Entropy at state 5
m1 Mass at state 1 T1 Temperature at state 1
m2 Mass at state 2 T4 Temperature at state 4
m3 Mass at state 3 T5 Temperature at state 5
m4 Mass at state 4 T6 Temperature at state 6
m6 Mass at state 6 Tcv, Constant volume adiabatic flame
ma Mass of air adiabatic temperature
mf Mass of fuel Te Exhaust temperature
mr Residual mass TH Temperature of a source reservoir
Internal Combustion Engines, Developments in 179

Ti Inlet temperature 2. Heywood J (1988) Internal combustion engine fun-

TL Temperature of a sink reservoir damentals. McGraw-Hill, New York, p 9
3. Heywood J (1988) Internal combustion engine fun-
Tr Residual fraction temperature damentals. McGraw-Hill, New York, p 10
u1 Specific internal energy at state 1 4. Automot Eng Int (2010) 118(1):47. http://www.sae.
u2 Specific internal energy at state 2 org/automag/
u3 Specific internal energy at state 3 5. Brown W (1967) Methods for evaluating require-
ments and errors in cylinder pressure measurement.
u3a Specific internal energy at state 3a SAE Trans 76(SAE Paper No. 670008):50–71
u4 Specific internal energy at state 4 6. Lancaster D, Krieger R, Lienesch J (1975) Measure-
U1 Internal energy at state 1 ment and analysis of engine pressure data. SAE Trans
U2 Internal energy at state 2 84(SAE Paper No. 750026):155–172
7. Randolph A (1990) Methods of processing cylinder-
U3 Internal energy at state 3 pressure transducer signals to maximize data accu-
U4 Internal energy at state 4 racy. SAE Trans J Passeng Cars 99(SAE Paper
v1 Specific volume at state 1 No. 900170):191–200
v2 Specific volume at state 2 8. Kuratle R, Marki B (1992) Influencing parameters
and error sources during indication on internal com-
v3 Specific volume at state 3 bustion engines. SAE Trans J Engines 101(SAE
v3a Specific volume at state 3a Paper No. 920233):295–303
v4 Specific volume at state 4 9. Davis R, Patterson G (2006) Cylinder pressure data
V Cylinder volume quality checks and procedures to maximize data
accuracy. SAE Paper No. 2006-01-1346
V1 Volume at state 1 10. Amann C (1983) A perspective of reciprocating-
V2 Volume at state 2 engine diagnostics without lasers. Prog Energy Com-
V3 Volume at state 3 bust Sci 9:239–267
V5 Volume at state 5 11. Heywood J (1988) Internal combustion engine fun-
damentals. McGraw-Hill, New York, pp 56–57
V6 Volume at state 6 12. Tompkins B, Esquivel J, Jacobs T (2009) Perfor-
Vd Displaced volume mance parameter analysis of a biodiesel-fuelled
Vmax Maximum cylinder volume medium duty diesel engine. SAE Paper No. 2009-
Vmin Minimum cylinder volume 01-0481
13. Heywood J (1988) Internal combustion engine fun-
W Thermodynamic work damentals. McGraw-Hill, New York, p 217
1W 2 Work for process 1-2 14. Heywood J (1988) Internal combustion engine fun-
2W 3 Work for process 2-3 damentals. McGraw-Hill, New York, p 154
15. Lauck F, Uyehara O, Myers P (1963) An engineering
3W 4 Work for process 3-4
evaluation of energy conversion devices. SAE Trans
4W 5 Work for process 4-5 71(SAE Paper No. 630446):41–50
5W 6 Work for process 5-6 16. Foster D, Myers P (1984) Can paper engines stand the
6W 1 Work for process 6-1 heat? SAE Trans 93(SAE Paper No. 840911):
Wb Brake work 4.491–4.502
17. The K, Miller S, Edwards C (2008) Thermodynamic
Wf Friction work requirements for maximum internal combustion
Wgross Gross work engine cycle efficiency, Part 1: optimal combustion
Wig Gross indicated work strategy. Int J Engine Res 9:449–465
Win Net indicated work 18. Carnot NLS (1824) Reflections on the motive power
of heat (trans and ed: Thurston RH), 2nd edn (1897).
Wip Pump work Wiley, New York
Wnet Net work 19. Borgnakke C, Sonntag R (2009) Fundamentals of
xi Mole fraction of specie i thermodynamics. Wiley, New York, p 497
yi Mass fraction of specie i 20. Heywood J (1988) Internal combustion engine fun-
damentals. McGraw-Hill, New York, p 177
21. Heywood J (1988) Internal combustion engine fun-
Bibliography damentals. McGraw-Hill, New York, p 163
22. Edson M (1964) The influence of compression ratio
Primary Literature and dissociation on ideal Otto cycle engine thermal
1. Cummins C Jr (1976) Early IC and automotive engines. efficiency. Digital calculations of engine cycles.
SAE Trans 85(SAE Paper No. 760604):1960–1971 SAE, Warrendale, pp 49–64
180 Internal Combustion Engines, Developments in

23. Westbrook C, Dryer F (1984) Chemical kinetic 41. Woschni G (1967) Universally applicable equation
modeling of hydrocarbon combustion. Prog Energy for the instantaneous heat transfer coefficient in the
Combust Sci 10:1–57 internal combustion engine. SAE Trans 76(SAE
24. Olikara C, Borman G (1975) A computer program for Paper No. 670931):3065–3083
calculating properties of equilibrium combustion 42. Hohenberg G (1979) Advanced approaches for heat
products with some applications IC engines. SAE transfer calculations. SAE Trans 88(SAE Paper
Paper No. 750468 No. 790825):2788–2806
25. Lavoie G, Heywood J, Keck J (1970) Experimental 43. Borman G, Nishiwaki K (1987) Internal-combustion
and theoretical study of nitric oxide formation in engine heat transfer. Prog Energy Combust Sci
internal combustion engines. Combust Sci Technol 13:1–46
1:313–326 44. Heywood J, Higgins J, Watts P, Tabaczynski
26. Bowman C (1975) Kinetics of pollutant formation R (1979) Development and use of a cycle simulation
and destruction in combustion. Prog Energy Com- to predict SI engine efficiency and NOx emissions.
bust Sci 1:33–45 SAE Paper No. 790291
27. Miller J, Bowman C (1989) Mechanism and model- 45. Sandoval D, Heywood J (2003) An improved friction
ing of nitrogen chemistry in combustion. Prog model for spark-ignition engines. SAE Trans
Energy Combust Sci 15:287–338 J Engines 112(SAE Paper No. 2003-01-
28. Turns S (1995) Understanding NOx formation in 0725):1041–1052
nonpremixed flames: experiments and modeling. 46. Blumberg P, Lavoie G, Tabaczynski R (1979) Phe-
Prog Energy Combust Sci 21:361–385 nomenological models for reciprocating internal
29. Dean A, Bozzelli J (2000) In: Gardiner WC Jr (ed) combustion engines. Prog Energy Combust Sci
Combustion chemistry of nitrogen in gas-phase com- 5:123–167
bustion chemistry. Springer, New York, pp 125–341 47. Assanis D, Heywood J (1986) Development and use
30. McBride B, Gordon S (1992) Computer program for of a computer simulation of the turbocompounded
calculating and fitting thermodynamic functions. diesel system for engine performance and component
NASA Report No. RP-1271 heat transfer studies. SAE Trans 95(SAE Paper
31. Svehla R (1995) Transport coefficients for the NASA No. 860329):2.451–2.476
Lewis chemical equilibrium program. NASA Report 48. Filipi Z, Assanis D (1991) Quasi-dimensional com-
No. TM-4647 puter simulation of the turbocharged spark ignition
32. Gordon S, McBride B (1999) Thermodynamic data engine and its use for 2 and 4-valve engine matching
to 20000K for monatomic gases. NASA Report studies. SAE Trans J Engines 100(SAE Paper
No. TP-1999-208523 No. 910075):52–68
33. McBride B, Gordon S, Reno M (2001) Thermody- 49. Kamimoto T, Kobayashi H (1991) Combustion pro-
namic data for fifty reference elements. NASA cesses in diesel engines. Prog Energy Combust Sci
Report No. TP-3287/Rev 1 17:163–189
34. McBride B, Zehe M, Gordon S (2002) CAP: a com- 50. Reitz R, Rutland C (1995) Development and testing
puter code for generating tabular thermodynamic of diesel engine CFD models. Prog Energy Combust
functions from NASA Lewis Coefficients. NASA Sci 21:173–196
Report No. TP-2001-210959-Rev1 51. Caton J (2003) Effects of burn rate parameters on
35. Heywood J (1988) Internal combustion engine fun- nitric oxide emissions for a spark ignition engine:
damentals. McGraw-Hill, New York, pp 136–137 results from a three-zone, thermodynamic simula-
36. Stull D, Prophet H (1971) JANAF thermochemical tion. SAE Paper No. 2003-01-0720
tables, NSRDS-NBS 37. http://www.nist.gov/data/ 52. Caton J (2000) A review of investigations using the
nsrds/NSRDS-NBS37.pdf. Accessed 5 July 2010 second law of thermodynamics to study internal-
37. Keenan J (1951) Availability and irreversibility in combustion engines. SAE Trans J Engines
thermodynamics. Br J Appl Phys 2:183–192 109(SAE Paper No. 2000-01-1081):1252–1266
38. Edson M, Taylor C (1964) The limits of engine 53. Rakopoulos C, Giakoumis E (2006) Second-law ana-
performance – comparison of actual and theoretical lyses applied to internal combustion engines opera-
cycles. In: SAE digital calculations of engine cycles. tion. Prog Energy Combust Sci 32:2–47
Society of Automotive Engineers, New York, 54. Shyani R, Caton J (2009) A thermodynamic analysis
pp 65–81 of the use of exhaust gas recirculation in spark igni-
39. Strange F (1964) An analysis of the ideal Otto cycle, tion engines including the second law of thermody-
including the effects of heat transfer, finite combus- namics. Proc Inst Mech Eng Part D J Automob Eng
tion rates, chemical dissociation, and mechanical 223:131–149
losses. In: SAE digital calculations of engine cycles. 55. Dunbar W, Lior N (1994) Sources of combustion
Society of Automotive Engineers, New York, irreversibility. Combust Sci Technol 103:41–61
pp 92–105 56. Som S, Datta A (2008) Thermodynamic irreversibil-
40. Patterson D, Van Wylen G (1964) A digital computer ities and exergy balance in combustion processes.
simulation for spark-ignited engine cycles. In: SAE Prog Energy Combust Sci 34:351–376
digital calculations of engine cycles. Society of Auto- 57. Caton J (2000) On the destruction of availability
motive Engineers, New York, pp 82–91 (exergy) due to combustion processes – with specific
Internal Combustion Engines, Developments in 181

application to internal-combustion engines. Energy combustion: an integrated view combining laser

25:1097–1117 diagnostics, chemical kinetics, and empirical valida-
58. Keenan J (1941) Thermodynamics. Wiley, tion. SAE Trans J Engines 108(SAE Paper No. 1999-
New York, p 269 01-0509):587–600
59. Obert E (1970) Internal combustion engines, 3rd edn. 76. Chigier N (1975) Pollution formation and destruction
International Textbook Company, Scranton, p 459 in flames – introduction. Prog Energy Combust Sci
60. Patrawala K, Caton J (2008) Potential processes for 1:3–15
“reversible” combustion with application to recipro- 77. Beltzer M (1976) Non-sulfate particulate emissions
cating internal combustion engines. In: Proceedings from catalyst cars. SAE Trans 85(SAE Paper
of the 2008 technical meeting of the central states No. 760038):198–208
section of the combustion institute, Tuscaloosa 78. Khatri N, Johnson J, Leddy D (1978) The character-
61. Diesel R (1897) Diesel’s rational heat motor. ization of the hydrocarbon and sulfate fractions of
A lecture delivered at the general meeting of the diesel particulate matter. SAE Trans 87(SAE Paper
society at Cassell, June 16, 1897. Original published No. 780111):469–492
in Zeitschrift des Vereines Deutscher Ingenieure 79. Hoffert M, Caldeira K, Benford G, Criswell D,
(trans: Leupold R). Progressive Age, New York Green C, Herzog H, Jain A, Kheshgi H, Lackner K,
(Reprinted) Lewis J, Lightfoot H, Manheimer W, Mankins J,
62. Heywood J (1988) Internal combustion engine fun- Mauel M, Perkins L, Schlesinger M, Volk T, Wigley
damentals. McGraw-Hill, New York, p 391 T (2002) Advanced technology paths to global climate
63. Heywood J (1988) Internal combustion engine fun- stability: energy for a greenhouse planet. Science
damentals. McGraw-Hill, New York, p 390 298:981–987
64. Heywood J (1988) Internal combustion engine fun- 80. Ghoniem A (2011) Needs, resources and climate
damentals. McGraw-Hill, New York, p 374 change: clean and efficient conversion technologies.
65. Lyn W (1963) Study of burning rate and nature of Prog Energy Combust Sci 37:15–51
combustion in diesel engines. Proc Combust Inst 81. Henein N (1976) Analysis of pollutant formation and
9:1069–1082 control and fuel economy in diesel engines. Prog
66. Plee S, Ahmad T (1983) Relative roles of premixed Energy Combust Sci 1:165–207
and diffusion burning in diesel combustion. SAE 82. Haagen-Smit A, Fox M (1955) Automobile exhaust
Trans 92(SAE Paper No. 831733):4.892–4.909 and ozone formation. SAE Trans 63(SAE Paper
67. Ricardo H (1941) The high-speed internal combus- No. 550277):575–580
tion engine (Rev: Glyde HS), 3rd edn. Interscience 83. Huls T, Nickol H (1967) Influence of engine vari-
Publishers, New York ables on exhaust oxides of nitrogen concentrations
68. Ladommatos N, Abdelhalim S, Zhao H, Hu Z (1998) from a multicylinder engine. SAE Paper No. 670482
Effects of EGR on heat release in diesel combustion. 84. Starkman E, Stewart H, Zvonow V (1969) Investiga-
SAE Paper No. 980184 tion into formation and modification of exhaust emis-
69. Meguerdichian M, Watson N (1978) Prediction of sion precursors. SAE Paper No. 690020
mixture formation and heat release in diesel engines. 85. Hames R, Merrion D, Ford H (1971) Some effects of
SAE Paper No. 780225 fuel injection system parameters on diesel exhaust
70. Lyn W, Valdmanis E (1968) Effects of physical fac- emissions. SAE Paper No. 710671
tors on ignition delay. SAE Paper No. 680102 86. Khan I, Greeves G, Wang C (1973) Factors affecting
71. Kamimoto T, Aoyagi Y, Matsui Y, Matsuoka S (1981) smoke and gaseous emissions from direct injection
The effects of some engine variables on measured engines and a method of calculation. SAE Trans
rates of air entrainment and heat release in a DI diesel 82(SAE Paper No. 730169):687–709
engine. SAE Trans 89(SAE Paper No. 800253): 87. Yu R, Shahed S (1981) Effects of injection timing
1163–1174 and exhaust gas recirculation on emissions from a
72. Dent J, Mehta P, Swan J (1982) A predictive model D.I. diesel engine. SAE Trans 90(SAE Paper
for automotive DI diesel engine performance and No. 811234):3873–3883
smoke emissions. Paper presented at the international 88. Newhall H (1967) Control of nitrogen oxides by
conference on diesel engines for passenger cars and exhaust recirculation, a preliminary theoretical
light duty vehicles. Institution of Mechanical Engi- study. SAE Trans 76(SAE Paper No. 670495):
neers, London. IMECE Paper No. C126/82 1820–1836
73. Binder K, Hilburger W (1981) Influence of the rela- 89. Benson J, Stebar R (1971) Effects of charge dilution
tive motions of air and fuel vapor on the mixture on nitric oxide emission from a single-cylinder
formation processes of the direct injection diesel engine. SAE Trans 80(SAE Paper No. 710008):7–19
engine. SAE Trans 90(SAE Paper No. 810831): 90. Komiyama K, Heywood J (1973) Predicting NOx
2540–2555 emissions and effects of exhaust gas recirculation in
74. Dec J (1997) A conceptual model of DI diesel com- spark-ignition engines. SAE Trans 82(SAE Paper
bustion based on laser-sheet imaging. SAE Trans No. 730475):1458–1476
J Engines 106(SAE Paper No. 970873):1319–1348 91. McEnally C, Pfefferle L, Atakan B, Kohse-
75. Flynn P, Durrett R, Hunter G, zur Loye A, Hoinghaus K (2006) Studies of aromatic hydrocar-
Akinyemi O, Dec J, Westbrook C (1999) Diesel bon formation mechanisms in flames: progress
182 Internal Combustion Engines, Developments in

toward closing the fuel gap. Prog Energy Combust 108. Taylor C (1985) The internal combustion engine in
Sci 32:247–294 theory and practice, vol 2: Combustion, fuels, mate-
92. Cheng W, Hamrin D, Heywood J, Hochgreb S, Min K, rials, design (rev. edition). The MIT Press, Cam-
Norris M (1993) An overview of hydrocarbon emis- bridge, MA, p 50
sions mechanisms in spark-ignition engines. SAE 109. Heywood J (1988) Internal combustion engine fun-
Trans J Fuels Lubr 102(SAE Paper No. 932708): damentals. McGraw-Hill, New York, pp 4–5
1207–1220 110. Heywood J (1988) Internal combustion engine fun-
93. Henein N, Tagomori M (1999) Cold-start hydrocar- damentals. McGraw-Hill, New York, p 475
bon emissions in port-injected gasoline engines. Prog 111. Heywood J (1988) Internal combustion engine fun-
Energy Combust Sci 25:563–593 damentals. McGraw-Hill, New York, pp 217–220
94. Alkidas A (1999) Combustion-chamber crevices: the 112. Clenci A, Descombes G, Podevin P, Hara V (2007)
major source of engine-out hydrocarbon emissions Some aspects concerning the combination of down-
under fully warmed conditions. Prog Energy Com- sizing with turbocharging, variable compression
bust Sci 25:253–273 ratio, and variable intake valve lift. Proc Inst Mech
95. Haynes B, Wagner H (1981) Soot formation. Prog Eng D J Automob Eng 221:1287–1294
Energy Combust Sci 7:229–273 113. Van Nieuwstadt M, Kolmanovsky I, Morael P (2000)
96. Smith O (1981) Fundamentals of soot formation in Coordinated EGR-VGT control for diesel engines: an
flames with application to diesel engine particulate experimental comparison. SAE Trans J Engines
emissions. Prog Energy Combust Sci 7:275–291 109(SAE Paper No. 2000-01-0266):238–249
97. Kennedy I (1997) Models of soot formation and 114. Arnold S, Slupski K, Groskreutz M, Vrbas G,
oxidation. Prog Energy Combust Sci 23:95–132 Cadle R, Shahed S (2011) Advanced turbocharging
98. Richter H, Howard J (2000) Formation of polycyclic technologies for heavy-duty diesel engines. SAE
aromatic hydrocarbons and their growth to soot – a Trans J Engines 110(SAE Paper No. 2001-01-
review of chemical reaction pathways. Prog Energy 3260):2048–2055
Combust Sci 26:565–608 115. Kessel J, Schaffnit J, Schmidt M (1998) Modeling an
99. Tree D, Svensson K (2007) Soot processes in com- real-time simulation of a turbocharger with variable
pression ignition engines. Prog Energy Combust Sci turbine geometry (VGT). SAE Paper No. 980770
33:272–309 116. Hawley J, Wallace F, Pease A, Cox A, Horrocks R,
100. Hassaneen A, Samuel S, Morrey D, Gonzalez- Bird G (1997) Comparison of variable geometry
Oropeza R (2009) Influence of physical and chemical turbocharging (VGT) over conventional wastegated
parameters on characteristics of nanoscale particulate machines to achieve lower emissions. In: IMechE
in spark ignition engine. SAE Paper No. 2009-01- autotech conference, Birmingham, pp 245–259
2651 (IMechE Seminar Publication: Automotive Engines
101. Ericsson P, Samson A (2009) Characterization of and Powertrains, Paper No. C524/070/97)
particulate emissions propagating in the exhaust 117. Hawley J, Wallace F, Cox A, Horrocks R, Bird
line for spark-ignited engines. SAE Paper G (1999) Reduction of steady state NOx levels from
No. 2009-01-2654 an automotive diesel engine using optimized
102. Khan I (1969–1970) Formation and combustion of VGT/EGR schedules. SAE Trans J Engines
carbon in a diesel engine. Proc Inst Mech Eng 108(SAE Paper No. 1999-01-0835):1172–1184
184(3J):36–43 118. Tanin K, Wickman D, Montgomery D, Das S, Reitz
103. Ahmad T, Plee S, Myers J (1982) Diffusion flame R (1999) The influence of boost pressure on emis-
temperature – its influence on diesel particulate and sions and fuel consumption of a heavy-duty single-
hydrocarbon emissions. Paper presented at the inter- cylinder DI diesel engine. SAE Trans J Engines
national conference on diesel engines for passenger 108(SAE Paper No. 1999-01-0840):1198–1219
cars and light duty vehicles. Institution of Mechani- 119. Cook J, Sun J, Buckland J, Kolmanovsky I, Peng H,
cal Engineers, London. IMECE Paper No. C101/82 Grizzle J (2006) Automotive powertrain control – a
104. Kummer J (1980) Catalysts for automobile emission survey. Asian J Control 8(3):237–260
control. Prog Energy Combust Sci 6:177–199 120. Leithgoeb R, Henzinger F, Fuerhapter A,
105. Koltsakis G, Stamatelos A (1997) Catalytic automo- Gschweitl K, Zrim A (2003) Optimization of new
tive exhaust aftertreatment. Prog Energy Combust advanced combustion systems using real-time com-
Sci 23:1–39 bustion control. SAE Paper No. 2003-01-1053
106. Johnson T (2010) Diesel emission control in review. 121. Corti E, Moro D, Solieri L (2007) Real-time evalua-
SAE Int J Fuels Lubr 2(SAE Paper No. 2009-01- tion of IMEP and ROHR-related parameters. SAE
0121):1–12 Paper No. 2007-24-0068
107. Taylor C (1985) The internal combustion engine in 122. Leonhardt S, Muller N, Isermann R (1999) Methods
theory and practice, vol 2: Combustion, fuels, mate- for engine supervision and control based on cylinder
rials, design (rev. edition). The MIT Press, Cam- pressure information. IEEE/ASME Trans Mechatron
bridge, MA, pp 21–23 4(3):235–245
Internal Combustion Engines, Developments in 183

123. Yoon M, Chung N, Lee M, Sunwoo M (2009) An 138. Goldsmid H (1960) Principles of thermoelectric
engine-control-unit-in-the-loop simulator of a devices. Br J Appl Phys 11:209–217
common-rail diesel engine for cylinder-pressure- 139. Hussain Q, Brigham D, Maranville C (2010) Ther-
based control. Proc Inst Mech Eng D J Automob moelectric exhaust heat recovery for hybrid vehicles.
Eng 223:355–373 SAE Int J Engines 2:1(SAE Paper No. 2009-01-
124. Turin R, Zhang R, Chang M (2008) Systematic 1327):1132–1142
model-based engine control design. SAE Int 140. Barber E, Reynolds B, Tierney W (1951) Elimination
J Passeng Cars Electron Electr Syst 1(SAE Paper of combustion knock ~ Texaco combustion process.
No. 2008-01-0994):413–424 SAE Trans 59(SAE Paper No. 510173):26–38
125. Caton J (2008) Results from an engine cycle simula- 141. Davis C, Barber E, Mitchell E (1961) Fuel injection
tion of compression ratio and expansion ratio effects and positive ignition ~ a basis for improved effi-
on engine performance. J Eng Gas Turbines Power ciency and economy. SAE Trans 69(SAE Paper
130(5):052809-7–052809-1 No. 610012):120–131
126. Wirbeleit F, Binder K, Gwinner D (1990) Develop- 142. Mitchell E, Cobb J, Frost R (1968) Design and eval-
ment of pistons with variable compression height for uation of a stratified charge multifuel military engine.
increasing efficiency and specific power output of SAE Trans 77(SAE Paper No. 680042):118–131
combustion engines. SAE Trans J Engines 99(SAE 143. Alperstein M, Schafer G, Villforth F III (1974)
Paper No. 900229):543–557 Texaco’s stratified charge engine – multifuel, effi-
127. Sugiyama T, Hiyoshi R, Takemura S, Aoyama S cient, clean, and practical. SAE Paper No. 740563
(2007) Technology for improving engine performance 144. Pischinger F, Schmidt G (1978) Experimental and
using variable mechanisms. SAE Trans J Engines theoretical investigations of a stratified-charge
116(SAE Paper No. 2007-01-1290):803–812 engine with direct fuel injection. SAE Paper
128. Boggs D, Hilbert H, Schechter M (1995) The Otto- No. 785038
Atkinson cycle engine: fuel economy and emissions 145. Hiraki H, Rife J (1980) Performance and NOx model
results and hardware design. SAE Trans J Engines of a direct injection stratified charge engine. SAE
104(SAE Paper No. 950089):220–232 Trans 89(SAE Paper No. 800050):336–356
129. Leone T, Pozar M (2001) Fuel economy benefit of 146. Ullman T, Hare C, Baines T (1982) Emissions from
cylinder deactivation – sensitivity to vehicle applica- direct-injected heavy-duty methanol-fueled engines
tion and operating constraints. SAE Trans J Fuels (one dual injection and one spark-ignited) and a
Lubr 110(SAE Paper No. 2001-01-3591):2039–2044 comparable diesel engine. SAE Trans 91(SAE
130. Gray C (1988) A review of variable engine valve Paper No. 820966):3154–3170
timing. SAE Trans J Engines 97(SAE Paper 147. Giovanetti A, Ekchian J, Heywood J, Fort E (1983)
No. 880386):6.631–6.641 Analysis of hydrocarbon emissions mechanisms in a
131. Payri F, Desantes J, Corberaan J (1988) A study of direct injection spark-ignition engine. SAE Trans
the performance of an SI engine incorporating a 92(SAE Paper No. 830587):2.925–2.947
hydraulically controlled variable valve timing sys- 148. Kato S, Onishi S (1988) New mixture formation tech-
tem. SAE Trans J Engines 97(SAE Paper nology of direct fuel injection stratified charge SI engine
No. 880604):6.1133–6.1145 (OSKA) ~ test result with gasoline fuel. SAE Trans
132. Ma T (1988) Effect of variable engine valve timing J Engines 97(SAE Paper No. 881241):6.1497–6.1504
on fuel economy. SAE Trans J Engines 97(SAE 149. Onishi S, Jo S, Shoda K, Jo P, Kato S (1979) Active
Paper No. 880390):6.665–6.672 thermo-atmosphere combustion (ATAC) ~ a new
133. Meacham G (1970) Variable cam timing as an emis- combustion process for internal combustion engines.
sion control tool. SAE Trans 79(SAE Paper SAE Trans 88(SAE Paper No. 790501):1851–1860
No. 700673):2127–2144 150. Najt P, Foster D (1983) Compression-ignited homo-
134. Tuttle J (1980) Controlling engine load by means of geneous charge combustion. SAE Trans 92(SAE
late intake-valve closing. SAE Trans 89(SAE Paper Paper No. 830264):1.964–1.979
No. 800794):2429–2441 151. Martinez-Frias J, Aceves S, Flowers D, Smith J, Dib-
135. Stobart R, Wijewardane A, Allen C (2010) The ble R (2000) HCCI engine control by thermal man-
potential for thermo-electric devices in passenger agement. SAE Trans J Fuels Lubr 109(SAE Paper
vehicle applications. SAE Paper No. 2010-01-0833 No. 2000-01-2869):2646–2655
136. Patterson A, Tett R, McGuire J (2009) Exhaust heat 152. Law D, Kemp D, Allen J, Kirkpatrick G, Copland
recovery using electro-turbogenerators. SAE Paper T (2001) Controlled combustion in an IC engine with
No. 2009-01-1604 a fully variable valve train. SAE Trans J Engines
137. Srinivasan K, Mago P, Zdaniuk G, Chamra L, Mid- 110(SAE Paper No. 2001-01-0251):192–198
kiff K (2008) Improving the efficiency of the 153. Rausen D, Stefanopoulou A, Kang J, Eng J, Kuo
advanced injection low pilot ignited natural gas T (2005) A mean-value model for control of homo-
engine using organic Rankine cycles. ASME geneous charge compression ignition (HCCI)
J Energy Resour Technol 130:022201-7–022201-1 engines. J Dyn Syst Meas Control 127:355–362
184 Internal Combustion Engines, Developments in

154. Shaver G, Gerdes J, Roelle M, Caton P, Edwards SAE Trans J Engines 112(SAE Paper No. 2003-01-
C (2005) Dynamic modeling of residual-affected 0742):1060–1069
homogeneous charge compression ignition engines 164. Hasegawa R, Yanagihara H (2003) HCCI combus-
with variable valve actuation. J Dyn Syst Meas Con- tion in DI diesel engine. SAE Trans J Engines
trol 127:374–381 112(SAE Paper 2003-01-0745):1070–1077
155. Bengtsson J, Strandh P, Johansson R, Tunestal P, 165. Okude K, Mori K, Shiino S, Moriya T (2004) Pre-
Johansson B (2006) Multi-output control of a mixed compression ignition (PCI) combustion for
heavy-duty HCCI engine using variable valve actua- simultaneous reduction of NOx and soot in diesel
tion and model predictive control. SAE Paper engines. SAE Trans J Fuels Lubr 113(SAE Paper
No. 2006-01-0873 No. 2004-01-1907):1002–1013
156. Chiang C, Stefanopoulou A (2009) Sensitivity anal- 166. Jacobs T, Bohac S, Assanis D, Szymkowicz P (2005)
ysis of combustion timing of homogeneous charge Lean and rich premixed compression ignition com-
compression ignition gasoline engines. J Dyn Syst bustion in a light-duty diesel engine. SAE Trans
Meas Contr 131:014506-1–014506-5 J Engines 114(SAE Paper No. 2005-01-0166):
157. Fish A, Read I, Affleck W, Haskell W (1969) The 382–393
controlling role of cool flames in two-stage ignition. 167. Lechner G, Jacobs T, Chryssakis C, Assanis D,
Combust Flame 13:39–49 Siewert R (2005) Evaluation of a narrow spray cone
158. Takeda Y, Keiichi N, Keiichi N (1996) Emission angle, advanced injection timing strategy to achieve
characteristics of premixed lean diesel combustion partially premixed compression ignition combustion
with extremely early staged fuel injection. SAE in a diesel engine. SAE Trans J Engines 114(SAE
Trans J Fuels Lubr 105(SAE Paper No. 961163): Paper No. 2005-01-0167):394–404
938–947 168. Jacobs T, Assanis D (2007) The attainment of pre-
159. Akagawa H, Miyamoto T, Harada A, Sasaki S, mixed compression ignition low-temperature com-
Shimazaki N, Hashizume T, Tsujimura K (1999) bustion in a compression ignition direct injection
Approaches to solve problems of the premixed lean engine. Proc Combust Inst 31:2913–2920
diesel combustion. SAE Trans J Engines 108(SAE
Paper No. 1999-01-0183):120–132
160. Iwabuchi Y, Kawai K, Shoji T, Takeda Y (1999) Trial Books and Reviews
of new concept diesel combustion system – premixed Ferguson CR, Kirkpatrick AT (2001) Internal combustion
compression-ignited combustion. SAE Trans engines: applied thermosciences, 2nd edn. Wiley,
J Engines 108(SAE Paper No. 1999-01-0185): New York
142–151 Heywood J (1988) Internal combustion engine fundamen-
161. Kimura S, Aoki O, Kitahara Y, Airoshizawa E (2001) tals. McGraw-Hill, New York
Ultra-clean combustion technology combining a Jennings BH, Obert EF (1944) Internal combustion
low-temperature and premixed combustion concept engines: analysis and practice. International Textbook
for meeting future emission standards. SAE Trans Company, Scranton
J Fuels Lubr 110(SAE Paper No. 2001-01-0200): Pulkrabek WW (2004) Engineering fundamentals of the
239–246 internal combustion engine, 2nd edn. Pearson Prentice-
162. Kaneko N, Ando H, Ogawa H, Miyamoto N (2002) Hall, Upper Saddle River
Expansion of the operating range with in-cylinder Taylor CF (1985) The internal combustion engine in theory
water injection in a premixed charge compression and practice – vol 1: Thermodynamics, fluid flow,
ignition engine. SAE Trans J Engines 111(SAE performance (rev), 2nd edn. The MIT Press, Cam-
Paper No. 2002-01-1743):2309–2315 bridge, MA
163. Shimazaki N, Tsurushima T, Nishimura T (2003) Taylor CF (1985) The internal combustion engine in theory
Dual mode combustion concept with premixed diesel and practice – vol 2: Combustion, fuels, materials,
combustion by direct injection near top dead center. design (rev). The MIT Press, Cambridge, MA
 Permafrost A layer of soil or bedrock that has
Alaska Gas Hydrate Research been continuously frozen for at least 2 years.
and Field Studies Petrophysical Physical properties of the
reservoir rock.
S. L. Patil, A. Y. Dandekar and S. Khataniar Resources Resources are those detected quanti-
Institute of Northern Engineering, University of ties of hydrocarbon that cannot be recovered
Alaska, Fairbanks, AK, USA profitably with the current technology but may
be recoverable in the future. It also includes
those quantities that geologically may be pos-
Article Outline sible but have not been found yet [2].

Glossary
Definition of the Subject Definition of the Subject
History of Gas Hydrates
Introduction Gas Hydrates
Global Distribution of Hydrate Resource Natural gas hydrates are solid, crystalline, ice-like
Gas Hydrate Resource Potential materials made up of small molecules of gases,
Review of Global Gas Hydrate Projects mainly methane confined inside cages of water
Japan molecules. These gas hydrates are solid solutions
China and are formed when water molecules linked by
South Korea hydrogen bonding create cavities (host) that
India enclose a variety of molecules (guest). These mol-
United States ecules can be CH4, C2H6, and CO2 to name a few.
Canada It is interesting to note that there is no chemical
New Zealand bonding between the host water molecule and the
History of Gas Hydrate Exploration in Alaska caged molecule. Hydrates can form under high
Alaska Gas Hydrate Production Potential pressures or low temperatures, which are gener-
The Hot Ice No. 1 ally the conditions that exist in the Arctic or under
The Mt. Elbert-01 deep ocean beds. Huge volumes of gas hydrates
North Slope Borough Methane Hydrate Project are expected to exist at various locations around
Review the globe (Fig. 3). It is known that 1 m3 of
Gas Hydrate Stability Modeling for BGF Pools hydrates upon dissociation release about 180 std.
The East Barrow Field Material Balance Study m3 of gas, making them a huge potential as a
Field Scale Reservoir Simulation Study future energy resource [3, 4].
Future Directions
Conclusions
Disclaimer History of Gas Hydrates
Bibliography
The discovery of hydrates dates way back in 1810
Glossary when Sir Humphry Davy conducted an experi-
ment and discovered that an aqueous solution of
In-place The term used to estimate gas hydrate chlorine can be converted into solid form when
resources disregarding technical or economical cooled below 9  C [5]. Faraday confirmed the
recoverability. Generally these are the largest existence of this solid compound and now it is
estimates [1]. known that there are more than 100 molecules that
© Springer Science+Business Media, LLC 2012 185
R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_434
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media LLC, 2012
https://doi.org/10.1007/978-1-4419-0851-3_434
186 Alaska Gas Hydrate Research and Field Studies

can form hydrates when combined with water. hydrate presence, an undetermined volume of
Few molecules that are very commonly found to associated free gas accumulations, and existing
form hydrates are CH4, C2H6, C3H8, CO2, and production infrastructure (Fig. 1).
H2S, with CH4 being the most common molecule Initial research on Alaskan gas hydrates was
to form gas hydrates found in existence. It was as focused on the Prudhoe Bay Unit (PBU), Kuparuk
early as the 1950s when through x-ray diffraction River Unit (KRU), and Milne Point Unit (MPU)
methods the crystal structure of hydrates was infrastructure areas of the Alaska North Slope.
known, and by late 1960 it was realized that the These areas were being studied to determine gas
earth’s crust is home to vast quantities of natural hydrate reservoir extent, stratigraphy, structure,
gas hydrates [5]. continuity, quality, variability, and geophysical
It was Powell, in 1948 at the University of and petrophysical property distribution. These ini-
Oxford, who was the first to describe clathrate tial studies characterized reservoirs and fluids,
structure and described clathrate compounds as which led to in-place and recoverable resource
those inclusion compounds in which two or estimates and commercial potential determina-
more components are associated without ordinary tion. Studies were conducted to determine best
chemical union but through complete enclosure of practices and procedures for gas hydrate drilling,
one set of molecules in a suitable structure formed data acquisition, completion, and production.
by another [5]. The clathrate compounds are This collaborative research provided practical
divided into two categories. One is aqueous clath- input to reservoir and economic models, deter-
rates in which water is the host species, commonly mined the technical feasibility of gas hydrate pro-
known as clathrate hydrates or gas hydrates, and duction, and influenced future exploration and
the other category of hydrates where water is not field extension of this potential unconventional
the host molecule are known as nonaqueous clath- gas resource.
rates [5]. It is known that water molecules are A successful drilling project to explore, con-
linked to each other through hydrogen bonding, firm, and evaluate gas hydrates was conducted in
which results in the formation of cavities having February 2007 through a shallow seismic anom-
dimensions of 780 and 920 pm [5]. The molecules aly interpreted as hydrates. In this project, over
that are smaller than the dimension of these small 400 ft. of sediments containing gas- and water-
cavities and do not react with the hydrogen bonds bearing core and well-developed gas hydrate
of the water molecule form gas hydrates under zones (85% recovery) were cored. An extensive
appropriate conditions of temperature and pres- suite of electric logs were run to identify the gas
sure, thereby providing stability to the structure. hydrate zones, and gas hydrate saturations were
computed. Open-hole pressure-transient analyses
and fluid sampling tests were conducted using
Introduction Schlumberger’s Modular Formation Dynamics
Tester (MDT™).
Gas hydrates may become an important global A second ANS gas hydrate research project
source of clean burning natural gas. Scientists was conducted for the Barrow Gas Field (BGF).
have considered the potential of gas hydrate as Barrow, Alaska, is a remote and the northernmost
an energy resource for nearly four decades. How- city in the United States. The energy demands of
ever, until 2002, this knowledge has not been the city and nearby localities are met by natural
systematically applied to determine the economic gas supply from the Barrow Gas Field consisting
practicality of gas hydrate resource development. of three different gas pools: the East Barrow
The estimates of in-place unconventional gas (EB) pool, the South Barrow (SB) pool, and the
hydrate volume are up to 100 Trillion Cubic Feet Walakpa (WAL) pool (Fig. 2).
(TCF) within the existing oil development infra- Gas hydrate resource potential ofthe three BGF
structure of the Alaska North Slope (ANS) pools was analyzed by developing gas hydrate
[6]. This region uniquely combines known gas stability models. Material balance studies were
 Alaska Gas Hydrate Research and Field Studies

Alaska Gas Hydrate Research and Field Studies, Fig. 1 Alaska North Slope oil field developments with Eileen and Tarn gas hydrate trends. (Courtesy Collett [32])
187
188 Alaska Gas Hydrate Research and Field Studies

Alaska Gas Hydrate Research and Field Studies, Fig. 2 Barrow gas fields, Alaska. (Modified from NETL Website,
www.netl.doe.gov)

performed on East Barrow (EB) gas pool to under- Global Distribution of Hydrate Resource
stand reservoir drive mechanisms and qualita-
tively estimate the hydrate resource. Field-scale Global volumes of gas hydrate estimated till now
dynamic reservoir simulation models were devel- span several orders of magnitude and there is no
oped for the EB and WAL gas pools. Production certainty in methane hydrate resource estimates
history data were matched, reservoir drive mech- across the globe. This uncertainty in estimating
anisms were confirmed, free gas and hydrate the volume of gas hydrates present is majorly
resources were quantified, hydrate dissociation because of two factors: (1) The commonly used
patterns were examined, optimum locations for seismic methods and other geophysical methods
drilling infill wells were identified, and future that are used in estimating conventional hydro-
production scenarios were simulated. carbon resources are not sensitive to the huge
The vast amount of unconventional gas volumes of hydrates present in the seafloor sec-
in-place combined with evaluation of conven- tion. (2) Drilling activities to directly confirm the
tional Alaska North Slope gas commercialization presence of hydrates have a limited applicability
creates industry-government alignment to assess due to limited spatial extent [7]. There is no direct
this potential future resource. This region evidence which can be used in estimating the
uniquely combines known gas hydrate presence, occurrence of inplace gas hydrates. There are
an undetermined volume of associated free gas 23 locations where the presence of gas hydrates
accumulations, and existing production infra- has been confirmed through the recovery of
structure. However, many technical and economic hydrate samples, 20 of them being in the oceanic
challenges must be resolved before the gas and the rest 3 in the arctic environments [8,
hydrate becomes a viable commercial hydrocar- 9]. Since the discovery of gas hydrates,
bon resource. researchers have tried to estimate the amount of
Alaska Gas Hydrate Research and Field Studies 189

Alaska Gas Hydrate Research and Field Studies, Fig. 3 World distribution of gas hydrates. (Adapted from oral
presentation at AAPG Annual Convention, San Antonio, Texas, April 20–23, 2008)

gas trapped in hydrates none being certain. tabulated in Table 1 after Johnson [10]. Figure 4
Going by even the most conservative approxima- depicts the hydrate resource pyramid after Bos-
tion of the estimation of the trapped gas, the well and Collett [11].
amount of gas trapped in hydrates is enormous The production and commercialization of gas
(Fig. 3). hydrate deposits rely heavily on two important
HEI (Hydrate Energy International) along with factors: (1) the advancement in technology to
the International Institute for Applied Systems prove and produce from these resources overcom-
Analysis (IIASA) with funding from the World ing existing obstacles and (2) a natural gas supply
Bank, United Nations Organizations, and national market accelerating the need to progress and
governments has undertaken the task of assessing improve the gas hydrates resource potential
the global gas hydrate potential. For this study, [12]. Although various research activities are on
data from every continental margin was studied way worldwide, gas hydrates are the most
and appropriate depositional models were uti- unlikely unconventional resource that will be
lized. The data was reported for 18 regions tapped in the near future. But based on the
defined by the United Nations. For the Arctic demand for energy in the future, these research
Ocean, separate resource assessment was done and exploration activities need to go on and prob-
without regard for national boundaries and for ably in a much more aggressive way to make these
the southern ocean (from the coast of Antarctica a viable energy source. Figure 5 depicts the time
north to 60 south latitude) [10]. The result of line for post 1990 global drilling efforts after
global hydrate resource estimate has been Ruppel [12].
190 Alaska Gas Hydrate Research and Field Studies

Alaska Gas Hydrate Research and Field Studies, Table 1 Gas in-place in hydrate-bearing sands. (After Johnson
[10])
Region (United Nations designation) Gas in-place range (TCF) Gas in-place median (TCF)
United States 1500–15,434 7013
Canada 533–8979 2228
Western Europe 36–14,858 1425
Central and Eastern Europe 0–105 13
Former Soviet Union 1524–10,235 3829
North Africa 6–1829 218
Eastern Africa 42–25,695 1827
Western and Central Africa 79–26,672 3181
Southern Africa 121–26,369 3139
Middle East 31–3848 573
China 10–1788 177
Other East Asia 14–2703 371
India 36–6268 933
Other South Asia 20–3497 557
Japan 71–471 212
Oceania 38–6750 811
Other Pacific Asia 64–25,946 1654
Latin America and the Caribbean 258–31,804 4940
Southern Ocean 144–45,217 3589
Arctic Ocean 178–55,524 6621
Total 4705–31,3992 43,311

E.g., 85 Tcf technically-recoverable

on Alaskan North Slope
(Collett et al., 2008) Arctic (permafrost-associated)
sand reservoirs
100,000 Tcf Marine sand reservoirs
estimated for
global E.g., 6717 Tcf gas in Non-sand marine reservoirs
resource-grade place, Northern Gulf (including fracture filling)
deposits of Mexico sands
(Boswell & (Frye, 2008)
Collett, 2011) Massive seafloor/shallow
? hydrate at seeps

Marine shales
(low permeability)

- Increasing in-place resources

Boswell & - Decreasing reservoir quality
Collett, 2006 - Decreasing confidence in resource estimates
- Increasing production challenges
- Likely decreasing fractional recovery

Alaska Gas Hydrate Research and Field Studies, Fig. 4 The hydrate resource pyramid. (Ruppel [12])
Alaska Gas Hydrate Research and Field Studies 191

COMPLETED FUTURE
MITI METI MH21 production test
Japan Pacific?
Deepwater Marine
USA (Gulf of Mexico first commercial scale
DOE/Chevron JIP) production from gas hydrates
for countries lacking
Korea (UBGH) significant domestic
India (NGHP) conventional reserves
China (GMGS)
Gumusut-Kapak

Arctic
Canada
Permafrost

(Mallik) first commercial-scale

Hot DOE/CP production to power
Ice CO2 hydrate test onsite plant for conventional
USA (Alaska oil production
North Slope) DOE/BP DOE-led long-term
Mt. Elbert production test
China (Tibetan Plateau)
Ocean Drilling

IODP Exp. 311:

Academic

ODP Leg 164:

Vancouver Margin
Blake Ridge
Cabled gas hydrate
ODP Leg 204: observatories
Hydrate Ridge (North Atlantic?
Gulf of Mexico?
Vancouver margin?)

1990 2000 2010 2020 2030 2040

Alaska Gas Hydrate Research and Field Studies, symbols denote completed activities, and open symbols
Fig. 5 Timeline of major post-1990 gas hydrate field pro- are potential or planned activities. Rectangles filled with
grams and future activities. Circles correspond to logging shaded pattern refer to longer-term production activities
and/or coring, while squares/rectangles denote activities that remain prospective. (After Ruppel [12])
that included/will include production testing. Solid

Gas Hydrate Resource Potential hydrates is probably enormous. However, the

in-place and recoverable gas estimates are highly
Gas hydrates are naturally occurring ice-like sub- speculative and also dependent upon the conven-
stances composed of water and gas, in which a tional petroleum system components of source,
solid water-lattice accommodates gas molecules charge, reservoir, trap, seal, and a highly uncertain
in a cage-like structure called a clathrate. This recovery factor, as well as occurrence within the
clathrate structure can store up to 180 unit vol- gas hydrate stability field (Fig. 6).
umes of methane gas per 1 unit volume clathrate. The successful production history of the
This large gas storage capacity makes methane Russian Messoyakha gas hydrate field [6] demon-
hydrate a very attractive concentrated potential strates that gas hydrates may provide an immedi-
unconventional resource. ate source of natural gas that can be produced by
Gas hydrates accumulate within reservoirs conventional depressurization of an adjacent free
located in permafrost regions and within subsea gas. Gas hydrates also represent a significant dril-
sediments of outer continental margins. While ling and production hazard. Russian, Canadian,
methane, propane, and other gases can be and American researchers have described numer-
included in the hydrate structure, methane ous problems associated with gas hydrates,
hydrates appear to be the most common in nature. including well control problems and casing
The amount of methane sequestered in gas failures.
192 Alaska Gas Hydrate Research and Field Studies

Alaska Gas Hydrate Research and Field Studies, Fig. 6 Zone of potential gas hydrate stability, Alaska, North Slope.
(Courtesy Collett [32])

Review of Global Gas Hydrate Projects (GMGS) conducted and expedition GMGS-1 in
spring 2007 to perform offshore drilling in Shenhu
Several countries have active gas hydrate field area of the South China Sea. During the expedition
programs (Fig. 3). The following gives an over- eight holes were drilled. Wireline logging was
view/status of these research activities. performed on each site. Five wells were cored
[14]. Future expeditions to the South China Sea
are being considered. Besides, hydrate exploratory
Japan drilling takes place in Qinghai-Tibet Plateau.

The first remarkable gas hydrate research program

in Japan was established in 1995 by the Govern-
ment of Japan along with the Ministry of Econ- South Korea
omy, Trade, and Industry (METI). Since then the
METI was very proactive in gas hydrate research Gas Hydrate Research and Development Organiza-
field. It contributed to Canada’s Mallik 2002 Gas tion (GHDO-K), supported by the Ministry of Com-
Hydrate Production Research Well Program. In merce, Industry, and Energy, is a South Korean gas
2001, METI established Methane Hydrate R&D hydrate program started in 2000. Its main objectives
Program, which is currently in its Phase are to confirm gas hydrate resources in the East Sea
2 (2009–2015). Within the Phase 2 production and to commercially produce gas hydrates by 2015
tests on the eastern Nankai Trough are planned [14]. It started with seismic data acquisition in the
for 2012 and 2014 respectively. Phase 2 also con- East Sea with shallow coring in the Ulleung Basin in
siders environmental impact assessment and order to define its gas hydrate potential. Ulleung
development of optimal methane hydrate produc- Basin Gas Hydrate Expedition 1 (UBGH1) was
tion approaches as its objectives [13]. the first exploration and drilling expedition launched
on behalf of the Korean Institute of Geoscience and
Mineral Resources in late 2007. During the expedi-
China tion five wells were drilled with logging, and while
drilling, some of them were cored as well [14]. In
Guangzhou Center for Gas Hydrate research was July 2010, the UBGH2 expedition was launched. It
founded in 2004 to expand the research and to included vast amount of data acquired and lasted till
evaluate gas hydrate resource potential in offshore early October. Subsequently, South Korea is plan-
China. Guangzhou Marine Geological Survey ning field production tests (GHDO-K, 2010).
Alaska Gas Hydrate Research and Field Studies 193

India Tarn gas hydrate accumulation on the North

Slope of Alaska. However, no gas hydrates
National Gas Hydrate Program (NGHP) of India were encountered [14]. In 2007, the Mount
was initiated in 1997. Its primary goals are per- Elbert Gas Hydrate Stratigraphic Test Well was
formance of ocean drilling, coring, logging, and drilled by BP Exploration (Alaska) on behalf of
other activities in order to explore gas hydrate DOE. It was located on the Milne Point, Eileen
deposits and evaluation of the possibility of gas hydrate accumulation. It revealed gas
their economical production. The NGHP Expe- hydrates and even cores were recovered.
dition 1 was successfully conducted during mid-
2006. During the expedition, 39 wells were
drilled in various sites which are:
Canada
Kerala–Konkan, Krishna–Godavari, Mahanadi,
and Andaman. Special attention was paid to
Canada does not have any official gas hydrate
Krishna–Godavari gas hydrate site, where
program. However, extensive study was
more than 10 wells were drilled [14]. Extensive
performed in Canada of which the remarkable
coring and logging were performed in a number
ones are Cascadia and Mallik. Both gas hydrate
of wells and vast amounts of data were col-
occurrences are associated with permafrost. ODP
lected. NGHP Expedition 02 was also consid-
Leg 146 and IODP Expedition 311 to the offshore
ered as a future possibility.
Vancouver Island discovered gas hydrate exis-
tence in Cascadia continental margin. Numerous
data including seismic, logs, and chloride concen-
United States
tration infer gas hydrate presence. Within IODP
Expedition 311, hydrate-bearing cores were
In the United States, the Department of Energy
retrieved. Mallik-38 onshore well in the Macken-
(DOE) has been conducting extensive research
zie River Delta was believed to be drilled through
on gas hydrates for the past several decades.
100 m of hydratebearing sandstone. Subsequently
Ocean Drilling Program (ODP) Leg 164 located
in 1998, Geological Survey of Canada on behalf
on Blake Ridge, Atlantic shelf, was the first
of Japan Petroleum Exploration Company Ltd.
special attempt to identify gas hydrates [15]. It
and Japan National Oil Corporation drilled Mallik
recovered hydrate cores and documented gas
2 L-38 gas hydrate research well. Coring and
hydrate presence within the sediment
logging both revealed gas hydrates within the
[14]. ODP Leg 204 was another deep-sea expe-
sandstone. As a next step, within Mallik 2002
dition which discovered gas hydrates in the
program the partners drilled three more wells:
Hydrate Ridge, located on the Cascadia conti-
Mallik 3 L-38, Mallik 4 L-38, and Mallik 5 L-
nental margin. Gulf of Mexico Gas-hydrate JIP
38 [14].
consortium was formed in 2001 by industry
participants and government agencies in part-
nership with DOE and led by Chevron. Its goals
were studying hydrates both as drilling hazards New Zealand
and also as a potential methane resource. In
2005, it conducted scientific drilling and coring In 2010, Gas Hydrate Resources (GHR) program
in the gulf [14]. On the North Slope of Alaska, was funded for 2 years by the New Zealand Foun-
gas hydrates encountered onshore are associ- dation for Research, Science, and Technology
ated with permafrost. Despite an apparent like- (FRST). It is mostly focused on Hikurangi margin
lihood of offshore gas hydrate existence in the located to the east of the islands. The program
Arctic Ocean, there is no proof so far. In plans are: analysis of existing 2D and 3D seismic
2002–2003 Anadarko Petroleum Corporation data acquisition, seafloor sampling in spring 2011,
drilled the Hot Ice gas hydrate research well in and production modeling [16].
194 Alaska Gas Hydrate Research and Field Studies

History of Gas Hydrate Exploration in research programs by the industry, led by the US
Alaska Department of Energy. In order to produce gas
from gas hydrate, the in situ pressure-temperature
The North Slope of Alaska contains two large gas regime must change to dissociate gas. Decreasing
hydrate accumulations within and near the Prudhoe the reservoir pressure, increasing the reservoir tem-
Bay and the Tarn oil field developments (Fig. 1). perature, and/or use of chemicals can allow gas to
This confirms the possibility that gas hydrates may dissociate from the solid gas hydrate structure.
represent an important future energy resource. The Among the various techniques for production of
occurrence of gas hydrates on the North Slope of natural gas from gas hydrates, the most economi-
Alaska was confirmed in 1972 with data from the cally promising method is considered to be depres-
Northwest Eileen State-2 well, the first dedicated surization of an adjacent free gas.
gas hydrate core and test well, located near the A growing body of evidence suggests that a
Prudhoe Bay Oil Field. Most of the gas hydrates huge volume of natural gas is stored as gas hydrates
near this Eileen trend discovery well occur in a series in northern Alaska and that production of natural
of laterally continuous sandstone reservoirs and are gas from gas hydrates may be technically feasible.
geographically restricted to the area overlying the However, numerous technical and economic chal-
western part of the Prudhoe Bay Oil Field and the lenges must be resolved before this potential
eastern part of the Kuparuk River Oil Field. Three- resource can be developed. Many wells have pen-
dimensional seismic surveys in the Prudhoe Bay Oil etrated gas hydrate during oil production operations
Field indicate the presence of an undetermined on the Alaska North Slope. During this time, gas
amount of free gas trapped down deep below some hydrates were known primarily as a drilling hazard.
of the gas-hydrate-bearing units. The volume of gas Industry has considered the resource potential of
within the gas hydrates within the Eileen trend is conventional Alaska North Slope gas during indus-
estimated to be about 37–44 trillion cubic feet. try and government efforts in working toward an
Recently, data from wells along the western mar- Alaska North Slope gas pipeline. Consideration of
gin of the Kuparuk River Field have shown presence the resource potential of this conventional Alaska
of the large Tarn trend gas hydrate accumulation North Slope gas created the industry–government
overlying the Tarn oil field. The Cirque-1 well, alignment necessary to also consider the resource
located about 4 miles southwest of the Tarn oil potential of the potentially huge (40–100 TCF
field, experienced severe well control problems in in-place) unconventional gas hydrate accumula-
1992 after drilling through what appeared to be a free tions beneath existing production infrastructure.
gas interval possibly trapped below the gas hydrate Two of the wells recently drilled as a part of
stability zone. The gas-hydrate-bearing stratigraphic the industry–government research programs to
interval within the Tarn trend appears to be the explore and evaluate gas hydrate production poten-
up-dip equivalent of the West Sak sands. These tial on the ANS were the Hot Ice #1 and the
sands are estimated to contain more than 20 billion Mt. Elbert-01.
barrels of in-place heavy oil and are the focus of
recent development activity. Preliminary analyses
of other recently completed wells along the western
margin of the Kuparuk River oil field suggest that the The Hot Ice No. 1
volume of in-place gas hydrates within the Tarn trend
may exceed that of the Eileen trend. The Hot Ice No. 1 well was drilled on the
Anadarko Petroleum leases, southwest of the
main Kuparuk development, just outside the
Alaska Gas Hydrate Production Potential Kuparuk River Unit boundary. Drilling began in
2003 and ended in 2004. The well was drilled,
The production potential of the Eileen or Tarn gas cored, and logged to a depth of 2300 ft. (700 m),
hydrate accumulations has not been adequately and ultimately abandoned as no hydrates were
tested. This area is a focus of the gas hydrate found [17]. However, the lessons learned from
Alaska Gas Hydrate Research and Field Studies 195

drilling this well were useful in designing future B-Pad. The location of the Mt. Elbert-01 relative
hydrate exploration wells such as the Mt. Elbert. to the Milne Point infrastructure is shown in
Hot Ice No.1 well also used water-based poly- Fig. 7. The location of the Mt. Elbert-01 Hydrate
mer drilling mud with KCl (salt) for freezing- Well was chosen so as to drill in an area with
point depression and a mud chiller for circulating “known” hydrate accumulations with isolation
chilled mud. This was problematic, with freeze- (stand-off) from the thermal effects of any adja-
ups and mechanical problems impeding progress. cent wells. This would allow the best chance to
This knowledge influenced decisions with regard find and test undisturbed hydrates.
to coring fluids and mud chilling on the Mt. Elbert- Based on well log data from nearby offset wells,
01 Gas Hydrate Stratigraphic Test Well. it was estimated that the Base of the Ice-Bearing
Permafrost (BIBPF) was at approximately 1850 ft.
TVDss and the Base of the Gas Hydrate Stability
The Mt. Elbert-01 Zone (BGHSZ) was at about 2950 ft. TVDss. Cor-
ing operations were planned from three potential
The drilling, coring, and testing of Mt. Elbert-01 gas-hydrate-bearing zones (zones D, C, and B)
Gas Hydrate Stratigraphic Test Well was part of within this depth range. A total of 23 coring runs
the DOE/NETL-funded “Alaska North Slope Gas were made using the wireline retrievable coring
Hydrate Reservoir Characterization Project,” and method, which prevented the cored formations
was a cooperative effort between industry, acade- from being exposed to pipe tripping effects. This
mia, and government. was the first application of wireline retrievable cor-
The Mt. Elbert-01 Hydrate Test well was ing on the Alaska North Slope. Figure 8 shows a
located on a temporary ice pad in the Milne summary of the coring runs. The overall core recov-
Point Unit on the ANS, south of the B-Pad and ery was 85%. The use of a chilled mineral-oil-based
east of the E-Pad, with ice road access from the coring fluid and wireline retrievable coring method

Alaska Gas Hydrate Research and Field Studies, Fig. 7 Section Map – Mt. Elbert-01 location relative to Milne Point
infrastructure. (Courtesy Hanson [18])
196 Alaska Gas Hydrate Research and Field Studies

Summary of Mt. Elbert-01 Core Results

Mt.Elbert-01 Core Cut vs Recovered

25

20

15

Feet
Zone D
10

Zone C Cut
5 Recovered

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Core Run

Zone B
Core recovery Core Samples: Gas Samples:
highest in hydrate Rock Properties Cuttings
bearing intervals Geochemistry Isotubes
Microbiology Headspace Gas
Hydrate Core Pore Water Chem QD Hydrates
1990’ MD – 2494’ MD Hydrates at Pressure
Top
Staines 504’ Cut/428’ Recovered Frozen Hydrate Samples
Tongue 85% Recovery Thermal Properties
60 hours / 2.5 hr roundtrip

Alaska Gas Hydrate Research and Field Studies, Fig. 8 Summary of core recovery by run relative to the log through
the hydrate interval. (Courtesy Hanson [18])

are believed to be mainly responsible for such high the economic and administrative center of the
core recovery. The main advantage of a chilled oil- North Slope Borough (NSB). The energy
based coring fluid over a chilled water-based coring demands of the city and nearby localities are met
fluid is that the oil-based fluid would not freeze. by natural gas supply from the Barrow Gas Fields
Although additives were used, freezing of water- (BGF) consisting of three different gas pools: the
based coring fluids was a major problem in the East Barrow (EB) pool, the South Barrow
previous hydrate wells Mallik and Hot Ice #1 (SB) pool, and the Walakpa (WAL) pool
[18]. Details of the drilling plan and coring opera- (Fig. 2). The Barrow Gas Fields have been pro-
tions for Mt. Elbert-01 are given by Hanson [18]. ducing natural gas for the last 25 years. It has been
The logging suite for Mt. Elbert-01 consisted of estimated that the proven gas reserves in these
three electrical resistivity tools and an extended fields exceed over 250 billion standard cubic
MDT testing. Figure 9 shows a sample plot of the foot (BCF) [19, 20]. Based on the current
well logs, clearly delineating the upper D hydrate demands for natural gas in the Barrow region
zone. Full analysis of Mt. Elbert-01 well logs showed and to enhance natural gas production and expand
no evidence of significant “contacts” between free production facilities in order to provide cost effec-
gas sands and hydrate-bearing sediments [18]. Mt. tive and assured source of energy to Barrow and
Elbert-01 is the world’s first successful application of nearby areas, the Barrow gas hydrate project was
the MDT™ tool in open-hole in gas hydrates. proposed by the NSB to the US Department of
Energy (DOE). In 2006, DOE and NSB jointly
North Slope Borough Methane Hydrate funded a research program to characterize and
Project Review quantify methane hydrate resource potential asso-
ciated with the BGF, with the ultimate goal of
Barrow, Alaska, is a remote and the northernmost producing gas from the hydrate formations asso-
city in the United States. The city serves both as ciated with individual gas pools of BGF.
Alaska Gas Hydrate Research and Field Studies 197

Upper “D” Hydrate Zone

10in Induction
2 ( ohm.m ) 200
20in Induction Neutron
2 Clay
( ohm.m ) 200 0.6( ft3/ft3 ) 0
30in Induction Density OBMI_CAL_DSI_081PUP OBMI .E
Q-F-M
2 ( ohm.m ) 200 0.6( ft3/ft3 ) 0 Horizontal Scale: 1 : 9.995
60in Induction Sonic Por Orientation North
Carbonate 0 120 240 360
2 ( ohm.m ) 200 0.6( ft3/ft3 ) 0
MD 90in Induction CMR Poros Conductive OBMI Image Resistive
1 : 240 Coal
ft 2 ( ohm.m ) 200 0.6( ft3/ft3 ) 0

2000

2050

Alaska Gas Hydrate Research and Field Studies, Fig. 9 Composite RT Scanner, Neutron, Density, Sonic, CMR and
OBMI Log showing the upper “D” – Hydrate interval. (Courtesy Hanson [18])

Hydrate Stability Zone for East Barrow Gas Reservoir

Gas Hydrate Stability Modeling for BGF Temperature (F)
Pools 8 12 16 20 24 28 32 36 40 44 48 52 56
0

An engineering approach was first adopted to 400 Stable Hydrate

Zone within EB
develop methane hydrate stability models for the 800 Reservoir
Depths ft

three gas pools. This was performed on the basis

1200
of gas analysis, formation water salinity, and pres-
sure and temperature gradient data available from 1600

various sources referenced in [21]. 2000 Middle Barrow Sand - 2000 to 2150

The studied literature indicated a gas compo- 2400

sition of practically pure methane (99%) and — 32 Deg Isotherm — P-T Curve (CSM-HYD)
water salinities in the range of 0–4%
w/w. Details of the calculation procedures can Alaska Gas Hydrate Research and Field Studies,
be found in [21]. The hydrate stability models Fig. 10 Hydrate stability model for EB pool
for EB and WAL gas pools are provided in
Figs. 10 and 11. Results from the EB pool calcu- to EB, the hydrate stability model for WAL pool
lations indicate the existence of hydrate stability also showed promising results in that stable
zones within the pay zone of 2000–2150 ft., hydrate zones were not found within deeper
which suggests that the in situ hydrates are pos- zones of the WAL sand where existing gas
sibly associated with free gas reservoir. Similar wells were completed within the 2000–2600 ft.
198 Alaska Gas Hydrate Research and Field Studies

interval but showed stable envelope existing in The East Barrow Field Material Balance
up-dip locations of the reservoir between 1700 Study
and 2000 ft. To confirm these results, detailed
geological and geophysical analysis of The objective of the material balance study was to
unexplored regions of the pools are required to qualitatively understand the behavior of the EB
determine the association of gas hydrates with pool under different drive mechanisms, and to
the prospective gas pools. Singh [21] conducted confirm if hydrate dissociation was one of them.
a material balance study to obtain additional con- Stokes and Walsh [22] performed pressure
firmation of the presence of hydrates, especially buildup tests on two EB wells and later [23] com-
associated with the EB pools. pared the static reservoir pressure data with his-
torical pressure data reported by Allen and Crouch
Hydrate Stability Zone for Walakpa Gas Reservoir
[24]. Figure 12 represents the historical perfor-
Temperature (F) mance plot of the EB gas pool, which shows an
8 16 24 32 40 48 56 64 72 initial drop in the static reservoir pressure with gas
0
production, indicating a conventional response of
300
a volumetric gas reservoir producing by gas
600 Stable Hydrate
Zone In Up-Dip expansion. However, long-term production data
900 Location
Depths ft

1200
showed gradual pressure stabilization, which indi-
1500 cated the possibility of water influx from a com-
1800 Currently Mapped Walakpa Sand Top: - 1700 , Base - 2000
municating aquifer. But the possibility of a
2100 communicating aquifer was ruled out because
Walakpa Gas Well Completion Top: - 2000 , Base - 2600
2400 (1) there was negligible water production (~1600
— 32 Deg Isotherm — P-T Curve (CSM-HYD)
bbls up to September 2007), (2) cumulative gas
production of ~8 BSCF by 2007 was much higher
Alaska Gas Hydrate Research and Field Studies, than Allen and Crouch’s [24] estimate of ~6
Fig. 11 Hydrate stability model for WAL pool. BSCF, and (3) none of the EB wells had watered
(Courtesy Singh [21])
out. Thus, the EB field performance data of

1500
1996 A&C
1350 Pressure EB#14
1988 A&C Est of ultimate
recoverable = 5.7 BCF
1200

1050 PRA 2006

Expected
900
p/z

750
A&C 1988 Report
600

450

300

150

0
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 11,000 12,000
CUMULATIVE PRODUCTION, MMCF

Alaska Gas Hydrate Research and Field Studies, Fig. 12 EB gas pool performance. (Stokes and Walsh [23])
Alaska Gas Hydrate Research and Field Studies 199

Reservoir Under Aquifer & Hydrate Supports

1200

Production Data Aquifer Thickness 100 ft

1000 HYD 22 ft
Reservoir Pressure (psi)

HYD 22 ft
800 HYD 10 ft

HYD 1 ft
600
HYD 10 ft
400
Hydrate Thickness H=0 ft HYD 1 ft
No Aquifer
200
Aquifer Thickness 33 ft

0
0 2000 4000 6000 8000 10000
Time (Days)

Production Data CMG AQ (33 ft) HYD (1 ft)

CMG AQ (33 ft) HYD (10 ft) CMG AQ (33 ft) HYD (22 ft)
CMG AQ (100 ft) HYD (1 ft) CMG AQ (100 ft) HYD (10 ft)
CMG AQ (100 ft) HYD (22 ft) Modified Hydrate Model

Alaska Gas Hydrate Research and Field Studies, Fig. 13 Hydrate- and aquifer-supported gas reservoir performance
prediction for EB pool. (Courtesy Singh [21])

Fig. 13 suggested possible contribution from an layer. Thus, a combination of aquifer support and
unknown drive mechanism, which could be gas hydrate dissociation is the most likely drive mech-
recharge from dissociating gas hydrates, or water anism in the EB pool.
influx from a new aquifer.
In the EB material balance study, the overall
performance of a tank type, static reservoir model Field Scale Reservoir Simulation Study
representing the EB gas pool was compared
against the available production data. The follow- Using the reservoir drive mechanism established
ing drive mechanisms were studied with the tank through the material balance study on the
model to see which drive mechanism would EB pool, Singh [21] conducted numerical
match actual production data: (1) volumetric gas reservoir modeling studies using a full-scale,
tank, (2) water drive, (3) dissociating-hydrate three-dimensional (3D) dynamic reservoir
drive, and (4) combination of water and hydrate model. A commercial reservoir simulator (CMG-
drive. Reservoir and fluid properties and other STARS) was used to perform the simulation
details are found in Singh [21]. studies. The static reservoir model of Panda and
The study showed that the actual production Morahan [25] was imported and built upon by
data from the EB field could not be matched by adding reservoir properties and drive mecha-
assuming volumetric behavior, water drive, or nisms. Other details can be found in Singh [21].
dissociating-hydrate drive alone. However, a The simulation study predicted field-wide per-
good match of modeled production response and formance of the EB pool for a period of 30 years
actual production data, shown in Fig. 13 by red and investigated the impacts of well completion
dotted circles, was obtained in the scenario where and orientation (vertical wells versus horizontal
a free gas reservoir was in communication with a wells). History matching of the simulation results
weak aquifer, but associated with a thick hydrate confirmed the previously estimated combination
200 Alaska Gas Hydrate Research and Field Studies

drive as the primary recovery mechanism for this the temperature of the injected solvent, the injec-
reservoir. However, gas production by expansion tion rate, or the injection time impact the produc-
dominates over hydrate dissociation mechanism tion of gas from such deposits has also been
after approximately 15–20 years of production. shown [30]. It is also proposed that a combination
The simulation study also indicated that infill of thermal and depressurization technique can be
drilling using horizontal producing wells would used to produce from gas hydrates that will help in
offer the ideal combination of lower water pro- improving the economic efficiency, paving the
duction and higher gas recovery. way for the development of NGH reservoirs [30].
Weitemeyer and Constable [7] have shown
how controlled source electromagnetics can be
Future Directions used to map gas hydrates. Further advancement
of this technique has led to the creation of a fixed
Various methods have been studied and have been offset towed sensor that was able to image gas
proved to be successful in the extraction of gas from hydrates at MC118 with successful imaging of
gas hydrate deposits. These are depressurization, hydrates in the other three sites in the Gulf of
thermal methods, and the use of inhibitors (e.g., Mexico [7]. Dai [31] has put forward a method
methanol or glycols). Gas hydrate dissociation is to locate and quantify gas hydrate deposit using
an endothermic reaction and as the hydrate deposits long-offset reflection seismic information, and
dissociate, the surrounding temperatures decrease since this method is not based on the AVO inver-
resulting in further hydrate reformation, thereby sion method it stays away from the uncertainties
reducing the amount of gas that can be produced. that are a part of AVO-based seismic studies.
Thus any method or a combination of methods used
for hydrate dissociation has to not only consider the
energy required to dissociate hydrates but also has to Conclusions
take in effect the energy that is required to counter
the drop in temperature that takes place due to the Significant improvements in the methods for dril-
endothermic nature of hydrate dissociation ling and evaluating potential gas hydrate zones in
[26]. Yutaek and SeongPil [27] have shown how ANS have taken place in recent years. The drilling
kinetic hydrate inhibitors (KHI) can be used to avoid and formation evaluation of the Mt. Elbert-01
hydrate reformation, the required concentration of stratigraphic test well has successfully field-tested
these KHI being 0.5–5 wt%. the modern geophysical prospecting methods
Recent experiments conducted by Shiang-Tai used and has enabled the selection of target
and Yen-Tien [28] illustrate how replacement of zones and field parameters for potential future
CO2 with CH4 in methane hydrates can help in the production testing. The chilled, 100% mineral-
production of methane gas from methane hydrate oil-based coring fluid provided a significant
deposits. When, liquid carbon dioxide is intro- improvement in wellbore quality as compared to
duced into the methane hydrate system, rapid the freezing-point-depressed water-based fluids
replacement of solid methane with carbon dioxide used in the previous gas hydrate exploration pro-
molecules takes place. This substitution of meth- jects. Core recovery, core quality, and electrical
ane with carbon dioxide occurs spontaneously well log quality were significantly improved by
and large amounts of carbon dioxide are observed using oil-based coring fluid, and operational
to penetrate methane hydrate molecules and expel downtime due to freeze-ups was eliminated.
methane especially near the hydrate melting point The study of a producing gas field in the ANS
[28]. It has been shown how steam injection can has demonstrated the feasibility of producing nat-
be used effectively in producing gas from gas ural gas from gas hydrates by depressurization to
hydrate deposits [29], and the use of hot-brine some extent. Diagnostic methodologies were
injection for the production of gas from natural developed to examine the production data from
gas hydrates as well as how various factors like the Barrow gas fields to test the two fields for
Alaska Gas Hydrate Research and Field Studies 201

possible hydrate accumulations and to understand associated with these two projects have contrib-
the gas recovery mechanisms. Engineering uted to the work reported here.
approach was followed to construct hydrate sta-
bility models and material balance procedures for Acknowledgment The authors are grateful to the US
identifying the reservoir drive mechanisms. Based Department of Energy (USDOE), BP Exploration
(Alaska) Inc., North Slope Borough (NSB), Petrotechnical
on these results, state-of-the-art dynamic reservoir
Resources of Alaska (PRA), and Petroleum Development
simulation models were developed using an Laboratory (PDL) at UAF for their financial support in
advanced commercial reservoir simulator to accomplishing this work. The authors would also like to
study hydrate reservoir development. Reservoir thank Dr. Tim Collet (USGS), Mr. Robert Hunter (ASRC
Energy Services), Mr. Tom Walsh (Petrotechnical
modeling confirmed the key findings of previous
Resources Alaska, Inc.), and graduate students Praveen
studies and projected the future performance of Singh, Andrew Johnson, Paul Hanson, and Vivek Peraser
the reservoir under different scenarios. Finally, for their contribution, support, and help when needed.
hydrate resource potential of EB and WAL pools
appears to be promising, and both pools have
considerable hydrate potential that can supply
large volumes of gas for future years. Bibliography

1. Folger P (2010) Gas hydrates: resource and hazard.

CRS report for congress 2010, CRS, Washington, DC
Disclaimer 2. WEC (2007) Survey of energy resources. World
Energy Council, London
This entry was prepared as an account of work 3. Islam MR (1991) A new recovery technique for gas
production from Alaskan Gas Hydrates. SPE 22924,
sponsored by an agency of the US government. Dallas
Neither the US government nor an agency thereof, 4. Garg R, Ogra K, Choudhary A, Menezes R. Chemical
nor any of their employees, makes any warranty, recovery of gas hydrates using fluorine gas and micro-
expressed or implied, or assumes any legal liabil- wave technology. SPE 113556
5. Peter E (1993) Clathrate Hydrates. Ind Eng Chem Res
ity or responsibility for the accuracy, complete- 32:1251–1274
ness, or usefulness of any information, apparatus, 6. Collett TS (1992) Geologic comparison of the
product, or process disclosed, or represents that its Prudhoe Bay- Kuparuk River (U.S.A.) and
use would not infringe privately owned rights. Messoyakha (U.S.S.R.) gas hydrate accumulations.
SPE 24469, Texas, pp 1–20
References herein to any specific commercial 7. Weitemeyer K, Constable S (2011) Mapping gas
product, process, or service by trade name, trade- hydrates with marine controlled source electromag-
mark, manufacture, or otherwise do not necessar- netics. In: Proceedings of the 7th international confer-
ily constitute or imply its endorsement, ence on gas hydrates (ICGH 2011), Edinburgh
8. George MJ, Timothy CS, Boswell R, Kurihara M,
recommendation, or favoring by the US govern- Matthew RT, Koh C, Dendy SE (2008) Toward pro-
ment or any agency thereof. The views and opin- duction from gas hydrates: current status, assessment
ions of the author expressed herein do not of resources, and simulation based evaluation of tech-
necessarily state or reflect those of the US gov- nology and potential. SPE 114163, Texas
9. Sloan ED, Koh C (2008) Clathrate hydrates of natural
ernment or any agency thereof. gases, 3rd edn. Taylor and Francis, Boca Raton
This entry is also based on research conducted 10. Johnson Arthur H (2011) Global resource potential of
by two petroleum engineering graduate students gas hydrate – a new calculation. In: Proceedings of the
(Paul Hanson and Praveen Kumar) from the Uni- 7th international conference on gas hydrates (ICGH
2011), Edinburgh
versity of Alaska Fairbanks for their M.S. degree 11. Boswell R, Collett TS (2006) The gas hydrates
in petroleum engineering. The results of the two resource pyramid. In: Fire in the ice, methane hydrate
US Department of Energy–funded gas hydrate newsletter. US Department of Energy, Office of Fossil
projects in Alaska (BP Alaska Gas Hydrate Pro- Energy, National Energy Technology Laboratory, Fall
Issue, pp 5–7
ject and the North Slope Borough Gas Hydrate 12. Ruppel C (2011) Methane hydrates and the future of
Project) are also included. Several other compa- natural gas. In: MITEI Natural gas Report, MIT, MA
nies, organizations associated, and individuals (Supplementary paper on methane hydrates)
202 Alaska Gas Hydrate Research and Field Studies

13. Nagakubo S, Arata N, Yabe I, Kobayashi H, Yama- 24. Allen WW, Crouch WJ (1988) Engineering study of
moto K (2011) Environmental impact assessment South and East Barrow fields, North Slope Alaska,
study on Japan’s methane hydrate R&D program. In: Alaska. Technical report prepared for North Slope
NETL, Methane Hydrate Newsletter, Jan 2011 Borough Gas Development Project, Barrow, pp 1–56
14. Collett TS, Johnson AH, Knapp CC, Boswell R (2009) 25. Panda MN, Morahan GT (2008) An integrated reser-
Natural gas hydrates. Rev AAPG Memoir 89:146–219 voir model description for East Barrow and Walakpa
15. Paull CK, Matsumoto R (2000) Leg 164 Overview. In: gas fields. Topical Report prepared under DoE Project
Proceedings of the ocean drilling program, Scientific No. DE-FC26-06NT42962 submitted to the USDoE,
Results, vol 164. Ocean Drilling Program Morgantown, www.netl.doe. gov, pp 1–44
16. Pecher IA, GHR Working Group (2011) Gas hydrates 26. Sandilya P, Ganguly S (2011) Conceptual design of an
in New Zealand – a large resource for a small country? experimental facility for extracting natural gas from nat-
In: NETL, Methane Hydrate Newsletter, Jan 2011 ural gas hydrate. In: Proceedings of the 7th international
17. Kadaster AG, Millheim KK, Thomson TW (2005) The conference on gas hydrates (ICGH 2011), Edinburgh
planning and drilling of hot ice #1 – gas hydrate 27. Yutaek S, Seong-Pil K (2011) Dependence of draw-
exploration well in the Alaskan Arctic. In: SPE down pressure on the hydrate reformation during
92764-MS presented at the SPE/IADC drilling confer- methane hydrate production and its inhibition with
ence, 23–25 Feb 2005, Amsterdam, pp 1–15 kinetic hydrate inhibitors. In: Proceedings of the 7th
18. Hanson P (2007) The Mt. Elbert-01 Stratigraphic test international conference on gas hydrates (ICGH
well – lessons from drilling, coring and evaluating gas 2011), Edinburgh
hydrates on the Alaska North Slope, MS project 28. Shiang-Tai L, Yen-Tien T (2011) Methane hydrate
report, University of Alaska Fairbanks, Fairbanks, growth and its recovery with carbon dioxide via
Dec 2007 molecular dynamic simulations. In: Proceedings of
19. Gruy HJ (1978) Reservoir engineering and geologic the 7th international conference on gas hydrates
study of the East Barrow Field, National Petroleum (ICGH 2011), Edinburgh
Reserve in Alaska. Technical Report submitted to 29. Lee Y, Shin C, Baek Y, Kim Y, Jang S, Lee J (2011)
Husky Oil NPR Operations, Inc., Anchorage, pp 1–62 A study on the dissociation behavior of gas hydrate
20. Darkwah SA, Allen WW (1996) Engineering study of using the steam stimulation. In: Proceedings of the 7th
the South, East and Walakpa Fields, North Slope Bor- international conference on gas hydrates (ICGH
ough, Alaska. Technical Report submitted to the North 2011), Edinburgh
Slope Borough, Barrow, pp 1–32 30. Shuxia L, Yueming C, Yongmao H (2011) Experimen-
21. Singh PK (2008) An engineering study to investigate tal study of influence factors of hot-brine stimulation
methane hydrate resource potential associated with for dissociation of NGH in porous medium. In: Pro-
barrow gas fields, Alaska. MS Thesis, University of ceedings of the 7th international conference on gas
Alaska Fairbanks, Fairbanks hydrates (ICGH 2011), Edinburgh
22. Stokes PJ, Walsh TP (2006) South and East Barrow 31. Dai J (2011) A method to locate and quantify gas
gas fields well improvements study. Technical Report hydrate using long-offset seismic. In: Proceedings of
submitted to the North Slope Borough, Barrow, the 7th international conference on gas hydrates
pp 1–70 (ICGH 2011), Edinburgh
23. Stokes PJ, Walsh TP (2007) South and East Barrow 32. Collett TS (1993) Natural gas hydrates of the Prudhoe
gas fields reserves study. Technical Report submitted Bay and Kuparuk River area, North Slope, Alaska
to the North Slope Borough, Barrow, pp 1–36 AAPG Bull 77(5): 793–812
 Fischer-Tropsch reaction The catalytic conver-
Gas to Liquid Technologies sion of synthesis gas to primarily hydrocar-
bons, the discovery being credited to Franz
Marianna Asaro1, Ronald M. Smith2 and Fischer and Hans Tropsch
Burtron H. Davis3 Gas heated reforming (GHR) Use of heat avail-
1
SRI International, Menlo Park, CA, USA able by recycling process gas (tail gas) down-
2
SRI Consulting, Menlo Park, CA, USA stream of an ATR for steam methane
3
Center for Applied Energy Research, University reforming, in a heat exchanger type reactor
of Kentucky, Lexington, KY, USA Gas to liquids (GTL) The conversion of gas to
liquid fuels and/or chemicals
Heat exchange reforming (HER) See GHR
Article Outline kWh Kilowatt hours, a measure of energy
Light distillate A distillation cut of low molec-
Glossary ular weight and low boiling range, obtained
Definition of the Subject from refining of hydrocarbonaceous feed-
Introduction stocks, used to produce liquefied petroleum
Survey of Specific Gas to Liquids Technologies gas (LPG), gasoline, and naphtha
Future Directions Liquefied natural gas (LNG) Natural gas that
Bibliography has been converted to liquid form for transport
or storage
Glossary Middle distillate A distillation cut of mid-range
boiling point, obtained from refining hydrocar-
Autothermal reforming (ATR) The reaction of bonaceous feedstocks, containing hydrocarbons
oxygen and carbon dioxide or steam with ranging from C5 through about C20 or C22. When
methane to form synthesis gas, wherein the further distilled, the portion of middle distillates
exothermic partial oxidation of methane pro- containing C5 through about C15 is often referred
vides energy for the endothermic steam to as naphtha, and the portion containing C16
reforming of methane. ATR is also used in through up to C22 is referred to as diesel. The
reference to the autothermal reformer itself. naphtha is often distilled further to produce gas-
BOE Barrel of oil equivalent oline and kerosene/jet fuel, or can be used as feed
Btu Also btu, British thermal unit, a measure of for a naphtha cracker unit to make light olefins.
energy content (Less commonly, the gasoline cut is initially
bbl Barrels (of oil) collected along with the light distillates.)
Catalytic membrane reactor (CMR) A flow- Natural gas liquids (NGL) The purified and
through reactor used to influence an condensed portion of natural gas consisting of
equilibrium-limited reaction to proceed further gaseous hydrocarbons heavier than methane,
in the forward direction via selective transport specifically ethane (C2H6), propane (C3H8),
of reactant(s) or product(s) across the n-butane (n-C4H10), and isobutane (i-C4H10)
membrane Partial oxidation (POx) The controlled oxida-
CPOx Catalytic partial oxidation tion of natural gas (primarily CH4) with oxy-
CTL Coal-to-liquids gen (O2) such that syngas is formed, rather than
DME Dimethylether forming carbon dioxide (CO2) via complete
DOE United States Department of Energy combustion

© Springer Science+Business Media, LLC, part of Springer Nature 2020 203

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_72
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_72-3
204 Gas to Liquid Technologies

Prereforming The use of an adiabatic, pre- (DME), methyl-tert-butyl ether (MTBE), and
heating zone upstream of an ATR reactor for chemicals such as ammonia (itself an important
the purpose of catalytically converting C2+ feedstock, particularly in the fertilizer industry),
hydrocarbons to a mixture of methane (CH4), light olefins, and methanol and dimethylether
hydrogen (H2), carbon monoxide (CO), and intended for chemical use.
carbon dioxide (CO2), thereby allowing oxy- For clarity the term GTL is used herein to refer
gen (O2) to be used more efficiently in the to the chemical conversion of natural gas to liq-
reformer. uids. However, a practical discussion of GTL
scf Standard cubic feet, a measure of gas volume should also include LNG, which is produced
Steam methane reforming (SMR) The high from natural gas by a phase change rather than a
temperature catalytic reaction of steam with chemical transformation. Strategists in national
methane to give synthesis gas governments and energy companies alike weigh
Synthesis gas (syngas) A mixture of primarily all of these options when considering how best to
hydrogen and carbon monoxide produced by distribute and monetize natural gas.
gasification or reforming of hydrocar- Modern GTL synthesis technology can pro-
bonaceous materials, used to synthesize fuels duce synthetic fuels (synfuels) that burn more
or chemicals cleanly than conventional fuels derived from
Water gas shift reaction (WGSR) The gas petroleum. F-T GTL technology can produce a
phase reaction of carbon monoxide (CO) with virtually sulfur-free diesel fuel, much cleaner
water to form carbon dioxide (CO2) and hydro- than conventional diesel – reducing smog and
gen (H2) acid rain – and can also produce gasoline, jet
fuel, or chemicals. GTL performed at or near the
gas well also represents an emerging option to
Definition of the Subject transport large quantities of transportation fuels
and chemical feedstocks from natural gas, using
Like oil and coal, natural gas is not what first commercially available tankers.
comes to mind when considering sustainable Production of liquid fuels and chemicals are
fuel sources. Yet as for other fossil sources, con- lower volume applications compared to use of
version of natural gas to transportation fuel is natural gas for power or as a fuel source for
currently more affordable than conversion of heating. Commercial application of GTL might
renewable resources such as wind and solar, increase in magnitude as the ease of oil recovery
which are technically far away from availability decreases and its price increases over the next few
at even a fraction of the scale required to have decades.
significant impact on meeting global demand over
the coming decades. Given that global proved
natural gas reserves are currently estimated as Introduction
capable of producing more than
1,100,000,000,000 equivalent barrels of oil (the The methane content of natural gas varies greatly
energy equivalent of 42 cubic miles of oil) [1], within the range of about 50–99%, typically about
natural gas is a key contributor when considering 95%. The remainder is a multicomponent mixture
a sustainable global fuel supply. including the heavier gaseous hydrocarbons eth-
The term gas to liquids (GTL) is frequently ane, propane, butanes, pentanes, and some higher
used in reference to the chemical transformation molecular weight hydrocarbons; the acid gases
of natural gas to liquid fuels via Fischer-Tropsch carbon dioxide, hydrogen sulfide; mercaptans
(F-T) technology. In broader usage, the term GTL such as methanethiol and ethanethiol; water, nitro-
refers to the transformation of natural gas into any gen gas; helium gas; and trace mercury. Thus the
liquid, including other fuels, such as liquefied preponderance of species in natural gas is high in
natural gas (LNG), methanol, dimethylether energy content, with cleanup of contaminants
Gas to Liquid Technologies 205

necessary but greatly simplified compared to that therefore, the type of GTL processing used,
required when using coal. Likewise the CO2 emis- including the size of the reserve, how much time
sions from combustion of natural gas [2] are con- is available to monetize the natural gas, the dis-
siderably less than from oil [3] or coal [4]: tance to market, and the gas processing require-
0.40 lb./kWh compared to 0.61 lb./kWh and ments. Environmental influences include a
0.79 lb./kWh, respectively. growing resistance to the conventional practice
The bulk of natural gas used today goes of flaring (or reinjecting) gas that otherwise
directly into power generation (32%), residential would be released in association with oil recovery
and commercial heating (27%), and industrial use operations (i.e., stranded gas that is also associ-
(22%). Only a small portion, around 5%, is used ated gas).
as a chemical feedstock, and therefore chemical Multiple opportunities exist to achieve better
uses have minimal impact on supply and pricing value for natural gas than its local fuel value. First,
decisions other than in stranded gas regions where the gas or its gaseous components can be exported
the gas is geographically remote from the main to markets having higher fuel values. Exportation
end-user markets. Thus, in the Middle East for can be performed by pipeline, as pressurized gas,
example, which has large reserves but relatively or by ocean tanker as liquefied natural gas (LNG,
small local demand, the proportion of natural gas comprised of 99% methane) or natural gas liquids
going into feedstocks is higher (around 27%). (NGL, comprised of C2–C4 alkanes). Second,
Location and transportation are key factors in GTL technologies can be used to chemically con-
defining the need for, and choice of, GTL tech- vert natural gas to higher value liquid fuels such as
nology. About 40% of known natural gas reserves naphtha or diesel, using F-T technologies; or to
are of the stranded type [5], including remote gas fuel grade methanol, DME (directly or via meth-
(located far from existing pipelines) and deeply anol), MTBE (via methanol), or gasoline (via
buried or otherwise inaccessible gas. Remote gas methanol); or to chemicals.
alone accounts for about 16% of the world’s pro- A flowchart of conceptual routes for chemical
ved natural gas reserves [6]. Related factors deter- conversion of the methane in natural gas to liquids
mine the type of transportation used, and is shown in Fig. 1. All current gas to liquids

Gas to Liquid Technologies, Fig. 1 Conceptual routes for chemical conversion of gas to liquids [7]
206 Gas to Liquid Technologies

chemical production processes require that natural effort has gone into developing catalysts for the
gas is first converted to synthesis gas (syngas), a process to produce liquid hydrocarbons. Eventu-
combination of hydrogen and carbon monoxide. ally, a precipitated Co catalyst promoted with
The syngas is subsequently converted in a synthe- ThO2 and MgO supported by Kieselguhr
sis section to product. GTL processes of today are (diatomateous earth) became known as the stan-
therefore of the indirect type, in that they proceed dard atmospheric process catalyst. In 1936,
first through the intermediate syngas instead of Fischer and Pilcher developed the medium pres-
converting methane directly to liquid product. sure (10–15 atm) Fischer-Tropsch synthesis pro-
Direct syntheses of fuels from natural gas include cess. Following this development, alkalized Fe
oxidative coupling of methane and selective oxi- catalysts were implemented into this medium
dation of natural gas to methanol. These have pressure process. Collectively, the process of
been investigated in the past but poor yields converting CO and H2 mixtures to liquid hydro-
and/or economics to date have prevented their carbons over a transition metal catalyst became
commercialization. known as the Fischer-Tropsch synthesis.
The intermediacy of synthesis gas is common Methanol is a commodity chemical, and one of
when converting various hydrocarbonaceous feed- the top ten chemicals produced globally. The
stocks such as natural gas, naphtha, residual oil, longtime interest in methanol is due to its potential
petroleum coke, coal, or biomass. The lowest cost use as a fuel and as a feedstock to the chemicals
routes for syngas production are based on natural industry. In particular, methanol can be used
gas. There are several different approaches to directly or blended with various petroleum prod-
converting natural gas to syngas, the primary ucts as a clean burning transportation fuel. Meth-
methods being catalytic steam methane reforming anol is also an important chemical intermediate
(SMR), autothermal reforming (ATR), gas-heated used to produce formaldehyde, dimethyl ether,
reforming (GHR), partial oxidation (POx), heat methyl tertiary butyl ether (MTBE), acetic acid,
exchange reforming, and variants thereof. The methylamines, and methylhalides among others.
ratio of H2/CO produced varies with the process Production of methanol began in the 1800s,
design as well as the H/C ratio in the feedstock. with the isolation of wood alcohol from the dry
A process will tend to be most economic if the distillation (pyrolysis) of wood. Research and
H2/CO ratio produced is well matched to the ratio development efforts at the beginning of the twen-
needed in the next step. Exceptions include pro- tieth century involving the reaction of syngas to
cesses configured to produce more H2 than needed, give liquid fuels and chemicals led to the discov-
because this extra H2 can be separated and used for ery of a methanol synthesis process (concurrently
power or to augment the syngas ratio of another with the development of the Fischer-Tropsch syn-
chemical process. If the H2/CO ratio is lower than thesis). Methanol synthesis is now a well-
needed, then an additional source of the expensive developed commercial catalytic process with
component H2 must be supplied. With an H/C ratio high reaction rate and selectivity (up to 99%).
of 4, methane is well suited for conversion to For economic reasons, methanol is produced
syngas when the final product slate will contain almost exclusively (over 90%) via reforming of
saturated hydrocarbons or methanol. natural gas.
The synthesis of hydrocarbons from CO
hydrogenation over transition metals was discov-
ered in 1902 when Sabatier and Sanderens pro- Survey of Specific Gas to Liquids
duced CH4 from H2 and CO mixtures passed over Technologies
Ni, Fe, and Co catalysts. In 1923, Fischer and
Tropsch reported the use of alkalized Fe catalysts This survey of GTL processes covers the sub-
to produce liquid hydrocarbons rich in oxygen- topics of liquefied natural gas, natural gas liquids,
ated compounds – termed the Synthol process. methane reforming, Fischer-Tropsch GTL, and
Succeeding these initial discoveries, considerable methanol synthesis from natural gas.
Gas to Liquid Technologies 207

Liquefied Natural Gas (LNG) into the local distribution pipeline. Independently,
Liquefaction of natural gas reduces its volume to a continuous regasification of the LNG from stor-
about 1/600 the volume of the gas as measured age tanks is performed, and the amount used for
under standard conditions. The most compelling the time period between ships must obviously
justification for the production of LNG is this match the amount unloaded. The LNG pressure
volume shrinkage that makes LNG, unlike natural is raised by pump to the pressure necessary for it
gas itself, practical to both store and transport. to be vaporized into the distribution pipeline.
Production of LNG has been commercially
practiced since 1960. The two major applications Plant Considerations
of LNG technology are: The liquefaction energy required in a LNG plant
typically has been reported as 9–12% of the heat
(a) Storage of LNG for use in peak shaving plants energy in the natural gas, and 9–10% energy
with seasonal adjustment (mostly in the USA) shrinkage is a typical number for the modern
(b) Base load LNG plants for international trade mega-tonnage capacity plants (without combined
with shipping from remote areas to developed cycle systems). Because LNG is stored and deliv-
countries via dedicated LNG ocean tankers ered at atmospheric pressure and
160  C
(
256  F), compression and very deep refrigera-
The capacities of peak shaving plants are more tion are needed, with associated large consump-
than an order of magnitude smaller than base load tion of energy. LNG projects have a very high
LNG plants. capital cost, in the range of $1.0–1.5 billion for a
Construction of pipelines depends on the rela- 3–3.3 million ton/year train on a greenfield site,
tive values for gas in the supplying and consuming with 45–60% attributed to offsites and infrastruc-
regions. Plans exist for pipeline delivery of gas ture depending particularly on requirements for
from isolated eastern Russian gas fields to consum- LNG storage, the marine system, and the heat
ing markets in Asia (Japan, Korea, and China). rejection method. The technology is relatively
Nonetheless, shipping tends to be preferred over mature, although economies of scale are being
long-range pipeline construction, because pipeline investigated to increase scale size to 4.5–5.5 mil-
costs are highly capital intensive: a 621 mile lion ton/year.
(1000 km) long pipeline would cost some $1 bil- The thermal efficiency of LNG plant is deter-
lion, depending on ground conditions. Exporting mined by two major factors:
LNG to Asia and Europe from distant production
fields has become economic as a result of improve- (a) Refrigeration cycle efficiency
ments to thermodynamic efficiencies of LNG facil- (b) Power cycle efficiency
ities. However, shipping is still expensive at a cost
of at least $15 per bbl to transport from the reser- Increasing the thermal efficiency of an LNG
voir to the consumer’s storage tanks. plant for a given turbine/driver configuration will
Recent industry trends indicate that tankers to decrease the cost of production and also minimize
hold as much as 135,000 m3 (equivalent to 3.2 both site gas consumption and greenhouse gas
billion scf of natural gas) have been specially emission of CO2 generated by combustion (0.20
built, with built capacities expected to soon ton CO2/ton of LNG at high thermal efficiency,
increase to 165,000 m3. For a shipping distance vs. 0.25–0.35 ton CO2/ton in a typical LNG plant
of 4000 or 6600 miles, the shipping requirements [8] as well as providing low NOx emission
for a 1 billion scf/day liquefaction plant can be (0.095 kg/ton)).
met by fleets of six or nine tankers, each tanker The key capital cost elements in LNG facilities
making 17 or 11 trips per year, respectively. Fol- are, in descending order,
lowing transport of the LNG by tanker from the
liquefaction site, it is off-loaded at a shore termi- 1. Gas turbines steam turbine or motor drivers
nal. Vapor generated is compressed for injection for refrigeration service
208 Gas to Liquid Technologies

2. Refrigeration compressors, typically over of the refrigeration compressors currently operated

100,000 kW by ConocoPhillips in Kenai are on the order of
3. Steam and power generation including tur- 70%, versus adiabatic efficiencies of 80–85% of
bines waste heat recovery more modern centrifugal compressors with three-
4. LNG storage, typically over 250,000 m3 dimensional blades. Even so, the plant has proved
capacity per tank (equivalent to approxi- reliable and profitable.
mately 112,500 metric tons LNG per tank) Several patents issued to Phillips (now
5. LNG loading terminal including jetty or ConocoPhillips) suggest that improvements in
causeway cycle efficiency would result from improved
6. Heat rejection system, in most cases, as refrigeration load distribution, nitrogen stripping
suggested, seawater cooling (for nitrogen rich gas), open loop methane refrig-
7. Cold box and prechilling for gas liquefaction eration, and LPG recovery [10–12]. The technol-
8. Natural gas pretreating for CO2 removal gas ogy appears driven by the desire to distribute load
drying and mercury adsorption equally among the propane, ethylene, and meth-
9. LPG and natural gasoline recovery as ane cycle. This is achieved in part by superheating
by-products the methane and ethylene refrigerant vapors fed to
10. Fuel gas cold recovery and compression the compressors, probably to about
46  C
(
50  F). Design and construction considerations
Although heat transfer for chilling (i.e., the then allow use of six identical gas turbines, such
cold box) is less important than the capital as frame 5D (nominal 30,000 kW), to be used.
investment associated with compression, the The open-loop methane refrigeration appears also
correct selection of the optimized refrigeration to use the methane refrigeration compressor as the
cycle and associated drivers affects both the fuel gas compressor.
compression and heat rejection systems and The loads of the propylene compressor are
economics. about 47.3%, the refrigeration ethylene compres-
The most common refrigeration system in sion loads are about 36.4%, and the methane
LNG plants involves prechilling with propane compressor loads are 16.3% of the total refrigera-
followed by use of the mixed refrigerant system. tion load. In practice, the combined load on the
Close to 90% of these plants are licensed by Air steam turbines may be about 27% of the total
Products Corporation, Inc. (APCI). The mixed motive power in the facility. The estimated
refrigerant liquefaction systems use a mixture of power consumption of the Phillips system, for
mostly methane and ethane, in about 1.2–2.0:1 feed gas at 38  C (100  F) and 650 psig, is
molar ratio. Depending on the feed gas composi- 371 kWh/ton, assuming very lean gas, such as
tion, up to 3 mol% nitrogen and 6–12 mol% 96 vol% methane, at a heat rejection temperature
propane may be added to optimize the LNG boil- speculated to be 38  C (100  F) [10]. This esti-
ing curve. The refrigerant cooling curve is mate probably includes 15 kWh/ton of fuel gas
adjusted to follow closely the feed gas cooling compression [13].
curve in order to achieve maximum thermody- A potential drawback of load equalization is
namic efficiency [9]. use of up to 5% additional refrigeration power
One significant exception to the propane pre- and, in case of fixed speed gas turbines, a reduc-
chilling, mixed refrigerant approach is the Phillips tion in production capacity [13]. Further, this
Kenai Peninsula plant in Alaska, which com- could result in a more complex refrigeration
menced operation in 1969. This plant produces cycle where the propane cycle, ethylene cycle,
about 1.5 million metric tons/year of LNG by a and methane cycles must be more highly
cascade refrigeration system using pure propane, heat integrated [10]. Despite potential draw-
ethylene, and methane refrigeration cycles. The backs in power consumption and capacity
efficiency of this cascade system has been reported reduction, the concept of cascade refrigeration
as at the lower end of the scale, about 88% [10]. It may be very sound [7], given the following
is reasonable to assume that the adiabatic efficiency considerations:
Gas to Liquid Technologies 209

• Even the older, closed loop methane refrigera- The choice of how syngas is produced depends
tion as used in Kenai could have an advantage on several factors, including:
by allowing higher suction pressure to the meth-
ane compressor, about 1.7 kg/cm2 (24 psia) • Match between the H2/CO ratio produced and
instead of an estimated 1.1 kg/cm2 (16 psia) the H2 demand of the overall process
for an assumed open loop compressor. • Plant and process scale
• Modern design methodology would allow use • The need for, or feasibility of, using an air
of conventional carbon steel metallurgy in the separation unit (ASU) or oxygen enrichment
compressors. facilities in the overall process
• Just one electric motor-driven fuel gas com- • Scale and logistics of capital equipment, such
pressor is used, for two trains as compressors and other gas-handling
• In case of outage of the fuel gas compressor, equipment
backup is provided by a draw from the feed gas. • Heat integration and gas recycle options
• The ethylene refrigeration cycle, aside from
superheating the suction to the first stage to Steam Methane Reforming (SMR)
about
46  C (
50  F), apparently comprises Steam methane reforming, or steam reforming for
only a single side load with a probable goal of short, is a catalytic conversion of natural gas by
obtaining the compression in a single casing – reaction with steam. The basic chemistry of steam
potentially without employing superheating of methane reforming is shown in Eqs. 2 and 3.
the suction to the ethylene compressors.
• Use of two side loads reduces refrigeration load
CH4 þ H2 O ⇆ 3H2 þ CO
by 2% and increases production capacity by (2)
2–3% (although using two casings also increases DH298 ¼ 206 kJ=mol CO
capital investment, possibly by about $5 million).
• On average about 6.2% of the total power is CH4 þ 2H2 O ⇆ 4H2 þ CO2
(3)
exported as electric power outside the boundary DH298 ¼ 165 kJ=mol CO
limits. (During times of high ambient tempera-
ture, the start-up steam turbine attached to the gas These reactions are quite endothermic, requir-
turbine increases its relative steam consumption ing input of significant transferred heat, and SMR
and electric power export drops to near zero.) is therefore carried out at high temperature with
the catalyst inside tubes within a fired furnace.
Given the pros and cons of cascade vs. conven- The amount of steam used is in excess of the
tional mixed refrigerant systems, an objective com- reaction stoichiometry requirements, as required
parative evaluation could be made only on a case- to prevent the catalyst from coking.
by-case, site-specific basis for LNG production. SMR is the dominant reforming technology
used for the production of methanol, ammonia,
Methane Reforming in F-T GTL and other petrochemical products produced from
Reforming is the means by which natural gas is methane at large scale. One reason is that the
converted to the synthesis gas used as feed for SMR process is scalable. Another reason is that
GTL processes. Syngas for use in Fischer- the high H2/CO ratio produced in SMR is stoichio-
Tropsch GTL is best characterized by the H2/CO metrically suitable for downstream formation of
ratio, which should be about 2.0, after reforming, saturated hydrocarbons and oxygenates. The theo-
as per the generic F-T stoichiometry of Eq. 1. retical ratio for H2/CO of 3:1 is not reached in
practice, the maximum practical value being about
2H2 þ CO⇆
CH2
þH2 O (1) 2.8, but in the absence of complex recycle schemes,
SMR provides the highest H2/CO ratio available in
There are many process options for reforming reforming of natural gas.
technology, used alone or in hybrid reforming Conventional steam reforming catalysts are
approaches, and these are compared in Table 1. 10–33 wt% NiO on a mineral support (alumina,
210 Gas to Liquid Technologies

Gas to Liquid Technologies, Table 1 Comparison of technologies for syngas generation from natural gas [14]
Technology Advantages Disadvantages
Steam methane reforming Most extensive industrial experience H2/CO ratio higher than typically
(SMR) O2 not required required for syngas production
Lowest process temperature Highest air emissions
(best H2/CO ratio if producing H2)
Gas-heated reforming Compact size and footprint Limited commercial experience
(GHR), heat exchange Application flexibility offers options for Usually best coupled with another
reforming (HER) incremental capacity syngas generation technology, such as
ATR
Two-step reforming (SMR Size of SMR is reduced Increased process complexity
followed by O2-blown Low methane breakthrough favors high purity Higher process temperature than
secondary reforming) syngas SMR
Methane content of syngas can be tailored by Usually requires O2
adjusting secondary reformer outlet
temperature
Autothermal reforming Typical H2/CO ratio produced is close to Higher process temperature than
(ATR) stoichiometric for F-T and methanol syntheses SMR
Lower process temperature requirement than Usually requires O2
POx
Low methane breakthrough
Methane content of syngas can be tailored by
adjusting reformer outlet temperature
Partial oxidation (POx) Feedstock desulfurization not required Low H2/CO ratio disadvantageous
Absence of catalyst permits carbon formation where ratio > 2.0 is required
and, therefore operation without steam, Very high process operating
significantly lowering syngas CO2 content temperatures
Low methane breakthrough Usually requires O2
Low H2/CO ratio advantageous where High temperature heat recovery, and
ratio < 2.0 is required, such as dimethylether soot formation and handling, add
synthesis process complexity
Low methane content of syngas not
easily modified to meet downstream
processing requirements

cement, or magnesia). Although natural gas usually Recent improvements to catalysts and the
contains only small amounts of sulfur compounds, reforming process have allowed design for oper-
generally in the form of H2S, sulfur compounds are ation of side fired reformers in SMR, at conditions
the main poisons of reforming catalysts. Use of not possible with other reforming methods. The
uranium oxide or chromium oxide as promoters most notable developments are noted below.
can impart higher tolerance to sulfur poisoning,
but even at a sulfur concentration of 0.1 ppm the • Introduction of new generations of catalysts,
catalyst can begin to deactivate and the best practice suitable for prereforming and for reforming of
is to remove sulfur from the raw feed. To maintain a heavy feedstocks
3-year catalyst lifetime, the sulfur concentration in Prereforming catalysis converts higher hydro-
the reformer feed gas should be less than 0.5 ppm. If carbons (C2+) in the natural gas feed into a
the sulfur concentration in the raw feed gas is greater mixture of methane, hydrogen, and carbon
than 1%, the sulfur must be removed by chemical or oxides.
physical scrubbing. An upfront desulfurization unit • Commercial application of advanced
is used to absorb H2S onto a ZnO bed. Any reforming at low steam/carbon (S/C) ratios
remaining organic sulfur compounds and carbonyl down to 1.5, and high outlet temperatures
sulfide are partially cracked and absorbed on the above 950  C
zinc oxide bed. Low S/C ratios improve energy efficiency.
Gas to Liquid Technologies 211

• Significant increase of heat flux development by BP, it underwent extended test

Operation at average heat flux above runs at their 300 bbl/day GTL demonstration
150,000 kcal/m3 h. has been demonstrated facility that was operated in Alaska. The reformer
in Haldor Topsøe’s Houston, Texas, process design resembles a conventional shell and tube
development unit (PDU). A high flux unit is heat exchanger, and SMR reactions occur tube
now in commercial operation. side of this reactor, with the tubes filled with
• Increase of unit capacity conventional Ni-based catalyst. Heat for the endo-
Steam reformers can now be constructed for thermic SMR reactions is provided shell side,
single train reforming units with capacity to where the tubes are heated by combustion of a
provide syngas for a methanol plant of up to fuel and air mixture. Heat transfer is claimed to
3000 ton/day production. occur more efficiently in what is described as a
• Heat exchange reforming highly countercurrent device.
Heat exchange reforming has been developed The shell side combustion zone may also be
and commercialized including with gas- under elevated pressure, leading to more efficient
heated option, where the steam reforming convective heat transfer to the tubes, and this plus
reaction is carried out in a unit heated by the considerably lower catalyst quantity required con-
product gas from an autothermal reformer. siderably reduces size relative to the conventional
SMR furnace design configuration – hence the
An obvious advantage of SMR is that it does term “compact” reforming coined by BP. The
not need an oxygen plant. However, since steam principal advantages claimed for this approach
reformers are more costly than either POx or compared to a conventional SMR are a major
autothermal reformers, there is a minimal plant reduction of capital cost along with an improve-
size above which the economy of scale of a cryo- ment in thermal efficiency. The reduced space
genic oxygen plant in combination with a POx or requirements are claimed to make this reforming
autothermal reformer is less expensive than a technology more amenable to installation on
steam reformer on its own. Use of SMR for barges or ocean-going “plantships” for the con-
Fischer-Tropsch applications does involve version of relatively small or temporary sources of
recycling of CO2 and removal of excess hydro- natural gas. A drawback of the current state of
gen, by means of pressure swing absorbers or compact reforming technology is that a large
membrane separators, to lower the H2/CO ratio number of parallel units would be required for a
to a level acceptable to the Fischer-Tropsch stoi- world scale GTL plant.
chiometry. Due to the costs involved with these
steps, it is most likely that steam reforming will Autothermal Reforming (ATR)
only be considered for F-T GTL when one or Autothermal reforming (ATR) has been recog-
more of the following conditions hold: nized since the mid-1990s as the technology of
choice for production of synthesis gas at large
• A relatively small GTL plant, with a capacity scale. Conceptually, ATR is similar to catalytic
well below 10,000 bbl/day. SMR except that ATR includes the use of oxygen,
• The additional hydrogen can be used for other which renders the reaction less endothermic on a
applications like methanol or ammonia per mole basis than SMR. The stoichiometry for
production. ATR is shown in Eq. 4.
• The natural gas feed has a high CO2 content.
• A suitable water supply can be obtained at
4CH4 þ O2 þ 2H2 O ⇆ 10H2 þ 4CO
low cost. (4)
DH298 ¼ 85:3 kJ=mol CO

Compact Reforming In practice, the product distribution is strongly

Compact reforming represents a novel mechanical affected by the choice of steam/carbon (S/C) ratio.
design approach to conventional SMR. Under At a S/C ratio of 1.3, the syngas will have a H2/CO
212 Gas to Liquid Technologies

ratio of about 2.5, which is higher than the ratio of pressure vessel. The prereformed natural gas
about 2.0 needed for Fischer-Tropsch synthesis. reacts with oxygen and steam in a sub-
The H2/CO ratio can be controlled by a combina- stoichiometric flame. In the catalyst bed, the gas
tion of lowering the S/C ratio and adding CO2 or is equilibrated with respect to the methane steam
CO2-rich gas to the ATR feed, such as by reforming and shift reactions. Product gas com-
recycling CO2 to the reformer. Decreasing the position is determined by the thermodynamic
S/C ratio in the feed gas decreases the amount of equilibrium of these reactions at the exit temper-
CO2 required when adjusting the synthesis gas ature and pressure of the reactor. The exit temper-
composition to the desired ratio for F-T synthesis. ature is determined by the adiabatic heat balance
A typical ATR process concept for the produc- based on the composition and flow of the feed,
tion of synthesis gas includes the fundamental steam, and oxygen added to the reactor. The prod-
steps of desulfurization, adiabatic prereforming, uct gas is completely free of soot and oxygen.
ATR, and heat recovery. In the desulfurization It is essential that the combined design of
section, the sulfur present in natural gas feedstock burner, catalyst, and reactor ensures that the pre-
is removed to avoid poisoning the downstream cursors are destroyed by the catalyst bed to avoid
prereforming catalyst. Steam, and optionally soot formation. The soot limit, i.e., the S/C ratio at
recycled CO2, is added to the desulfurized natural which soot formation starts, has been investigated
gas and, after further heating, the resultant mixture by experiment at many combinations of tempera-
is passed to a prereformer. In prereforming, pre- ture, pressure, so that prediction of acceptable
heating of the gas is done in a fired heater operating conditions for specific feedstocks is
upstream of the ATR. The use of a prereformer now possible.
upstream of the ATR unit reduces the oxygen Additional design parameters of the auto-
consumption per unit of syngas produced during thermal syngas generation system that influence
ATR. The exit gas from the ATR is cooled by cost and thermal efficiency include:
high-pressure steam production and boiler feed
water preheat, and CO2 is removed as needed, • The preheat temperatures of oxygen and natu-
using methyldiethanolamine (MDEA, N-methyl- ral gas.
diethanolamine) as absorbent in a CO2 capture The higher these temperatures are, the less
unit. After CO2 removal, the synthesis gas with a oxygen will be used. The maximum preheat
H2/CO ratio of 2.0 is available for Fischer- temperatures are determined by safety fac-
Tropsch synthesis. Various process variants are tors and by the need to prevent soot
possible, for example, adjustment of synthesis formation.
gas composition by recycling CO2-rich tail gas • The pressure of the steam generated in the
from the Fischer-Tropsch synthesis. waste heat reboiler.
The result is that ATR offers a reduction in the The higher the steam pressure, the more effi-
investment per barrel of product and the possibil- ciently energy can be recovered from steam,
ity of higher single-line capacity compared to but the more costly the steam and boiler
conventional steam reforming. Optimized design feed water treatment systems become. The
in combination with reduction of the S/C ratio can optimum steam pressure will be determined
be expected to result in a major increase in the by the relative cost of capital and energy.
single-line capacity, by more than 25% within the
next few years. Currently, single train methanol The feasibility of soot-free operation of the
plant units based on ATR can be designed with ATR reactor at very low S/C ratio was demon-
capacities up to about 7000 ton/day. strated in Haldor Topsøe’s Houston development
The key component of the ATR process is the unit in 1997–1999, with further testing performed
ATR reactor itself. The ATR reactor has a compact since then to explore the limits of the ATR tech-
design consisting of a burner, combustion cham- nology with respect to feed gas composition, pres-
ber, and a catalyst bed placed in a refractory lined sure, and temperatures. Operation at industrial
Gas to Liquid Technologies 213

scale at a S/C ratio of 0.6 was demonstrated in a increased complexity of needing to use reactors
South African plant in 1999 and has been in in series at larger scale, and detailed studies are
operation in Europe since the start of 2002 required to determine the most cost-effective
[15]. Two lines have been in operation at design approach.
Sasolburg since 2004. Haldor Topsøe’s ATR tech- Currently, the most attractive ATR syngas gen-
nology was also selected for the Oryx plant that eration technology appears to be oxygen blown
Qatar Petroleum and Sasol began operating in ATR at low S/C ratio. This technology has been
2006 and includes a 6000 mtd syngas generator; chosen for F-T GTL projects that are closest to
for Chevron Nigeria’s Escravos, Nigeria, GTL realization. Extensive testing has proved that the
plant that began operation in 2016; and for the technology is ready for application at a large
methanol portion of Eurochem’s gas-to-petro- industrial scale with S/C ratios down to 0.6.
chemical complex planned in Lagos State, Application of even lower S/C ratios has been
Nigeria. demonstrated at small scale, and the next-
Sasol, Syntroleum, and ExxonMobil, each generation ATR scheme is under development to
have their own proprietary versions of ATR. The cut the S/C ratio even lower, to 0.4 or possibly 0.2
most prevalent layout of the syngas production [16]. Such low S/C could decrease oxygen
section is a combination of adiabatic prereforming requirements and boost single line capacity still
and fixed bed ATR. This layout results in high further.
flexibility for variations in natural gas feed and, if Two-step reforming is a variant of ATR
needed, in tail gas recycle gas compositions, as wherein the exit gas from a fired tubular reformer
well as the ability to reduce oxygen consumption is further processed in an oxygen-blown, second-
per unit of syngas produced. ary reformer. In both steps, the ATR is carried out
ExxonMobil employs a fluidized bed catalytic in a refractory lined vessel containing a mixer/
reformer. Steam diluted oxygen is fed to the reac- burner and a catalyst bed. Conditions leading to
tor through nozzles, separately from the natural soot formation are rigorously avoided.
gas. The oxygen reacts exothermically with natu-
ral gas in a burning zone near the oxygen inlet. Gas-Heated Reforming/Heat Exchange Reforming
The balance of steam needed in the process is (GHR/HER)
admixed with natural gas and fed at a level Another technology that produces suitable syn-
below the oxygen inlet to the autothermal reaction thesis gas for GTL units is a combination of
zone. A mixture of nickel on alumina catalyst may steam reforming and ATR, performed in separate
be combined with a solids diluent (alumina) to units, in a process referred to as gas-heated
form the fluidized bed. The fluidized bed reactor reforming, GHR, or heat exchange reforming,
provides superior characteristics of heat and mass HER. The concept is to use heat available from
transfer for autothermal reforming. process gas downstream of the ATR for steam
Another ATR approach under consideration is reforming, in a heat exchanger type reactor. The
to use slurry bed F-T reactors in series within each purpose is to reduce oxygen consumption and
train, as the slurry system would facilitate appli- increase carbon efficiency by optimizing the recy-
cation of ATR technology at large scale (>10,000 cle of tail gas – beyond the range possible with
bpd). With this approach, water is removed ATR alone – and to achieve this without addi-
between stages to enhance the reactant partial tional firing that would create an undesired excess
pressures in the downstream reactor stage. The of water. The gas heated reformer is a compact
removal of other diluents (such as carbon dioxide) alternative to the fired steam reformer wherein
between stages may also be considered. This high-grade heat is recycled while the S/C ratio
allows lower recycle ratios, increased steam pro- remains below 0.6 [17].
duction from the reactor heat removal system, and Conventional reformers provide the needed
decreased overall reactor volumes. These advan- heat by combustion of fuel at atmospheric pres-
tages will not necessarily compensate for the sure in a refractory lined duct. Because the
214 Gas to Liquid Technologies

convective heat transfer coefficient of a GHR is passes through the steam reforming unit and then
considerably higher at the elevated pressures of its the ATR. The lower limit for the S/C ratio is
hot gas side, unit size reductions claimed can be typically determined by the steam reforming cat-
10–15 times smaller than conventional tubular alyst [18]. The parallel arrangement allows indi-
steam methane reformers. The absence of com- vidual optimization of the S/C ratio by feeding the
bustion flue gases results in an inherently higher two reformers independently.
thermal efficiency for a GHR, since no combus- GHR has been employed since the 1980s in
tion heat is lost with the warm flue gases some ammonia and methanol plants. In the past,
discharged to the atmosphere from the stack. Fur- highly efficient GHR was not viewed as an option
thermore, the reduced power consumption, which for high capacity syngas production, mainly
results from eliminating the need for the associ- because of concerns about metal dusting corro-
ated steam methane reformer air and flue gas sion. This reformer must operate at higher tem-
blowers, contributes to the overall energy saving peratures in the parallel arrangement than in the
of a GHR. Much of this thermal efficiency advan- series arrangement, to obtain a low methane leak-
tage may be offset however by the need to gener- age. Higher operating temperatures may lead to
ate additional steam externally in an offsites boiler higher cost and more severe conditions, and the
in order to compensate for the lost waste heat main challenge in both series and parallel arrange-
recovery potential of a conventional reformer ments is preventing metal dusting corrosion on the
system. shell side of the heat exchange reformer. Mechan-
GHR can be combined with ATR to reduce ical design, including choice of materials of con-
energy consumption and enable use of smaller struction, is critical.
air separation plants, while reducing CO2 emis- A heat exchange reformer designed by Haldor
sions possibly by as much as 50%. In addition, Topsøe for operating in tandem with an ATR at a
these GHR/ATR schemes allow shrinking of the low S/C ratio was started up at Sasol’s facilities in
required high-pressure steam system and reduc- Sasolburg, South Africa, in early 2003 [15, 19].
tion in size of steam turbines. Both series and The project at Sasolburg (also referred to as “pro-
parallel arrangements of GHR with ATR have cess gas-heated reforming” or PGR) involved
been studied and are depicted in Fig. 2 [17]. In revamping an existing ATR for an increase in
the series arrangement of HER/GHR, all gas syngas production capacity of 30% via GHR.

methane partly converted methane and steam

and steam methane oxygen
oxygen and
and steam
steam

GHR ATR GHR ATR

fully
converted fully
methane converted
methane
cooled cooled
syngas syngas

series configuration parallel configuration

Gas to Liquid Technologies, Fig. 2 Series and parallel configuration of the combination of GHR and ATR [17]
Gas to Liquid Technologies 215

The gas based 34,000 bbl/day Fischer-Tropsch combustion will be required if the energy content
plants using Sasol technology in Qatar and Nige- in the tail gas is converted to power, and a large
ria also use the Haldor Topsøe ATR technology. amount of power will be generated for export.
Engineers at Johnson Matthey have stated that This may be desirable however in remote loca-
with mechanical design and configuration tions where power may be in short supply.
improvements, GHR/ATR technology is likely to
deliver better efficiency and lower operating costs
Partial Oxidation (POx) and Catalytic Partial
than conventional ATR for F-T GTL plants of the
Oxidation (CPOx) Reforming
future. In addition, Air Products and Toyo Engi-
When the quantity of oxidant (O2) added to meth-
neering Corporation have each studied GHR
ane is very low and the reaction temperature is
technology.
raised significantly, uncatalyzed partial combus-
tion of pretreated natural gas occurs. This homo-
Air Blown Autothermal Reforming
geneous reaction is referred to as gasification or
It is possible to use air instead of purified oxygen
partial oxidation (POx). The resulting syngas is
as oxidant in autothermal reforming, and a F-T
quenched or cooled by steam production, and
synthesis process based on air blown ATR was
carbonaceous by-products such as soot are
developed by Syntroleum Corporation (now a part
removed by washing. The formation of carbona-
of Renewable Energy Group, Inc.) and demon-
ceous by-products influences carbon efficiency.
strated in their plant in Tulsa, Oklahoma, in
For partial oxidation, the syngas H2/CO ratio pro-
2002–2003. In a very similar operation, air
duced is slightly below 2.0, which is closest to the
blown secondary reforming is used extensively
optimum needed for the Fischer-Tropsch or meth-
in the production of synthesis gas in ammonia
anol syntheses. This low H2/CO ratio gas results
manufacture. The main advantage for air blown
from very little, if any, steam use in the process.
processes is that no air separation unit will be
Due to the absence of catalyst, the reformer
required, which could be a strategic advantage
operates at an exit temperature of about
for smaller scale, off shore, barge mounted F-T
1300–1400  C. This high temperature and the
GTL units. For large-scale land-based units, it
absence of catalyst have the following disadvan-
may present certain restrictions on the energy
tages as compared to an autothermal reformer.
system configuration of the plant. The savings of
eliminating the capital cost of an air separation
• Formation of soot and much higher levels of
cost will be offset by the need for compression of
ammonia and HCN appear in the syngas, com-
incoming air for example, which will result per-
pared to ATR, which necessitates the use of a
haps in as much as 50% more power required than
scrubber to clean the gas.
that required to produce oxygen.
• Higher consumption of oxygen.
In addition, the increased volumetric flow of
• Due to the absence of a water gas shift reaction
nitrogen-diluted syngas will make all downstream
with the POx catalyst, the unconverted meth-
equipment through the F-T synthesis reactors
ane as well as the methane produced by the
larger. Also, the high content of nitrogen in the
Fischer-Tropsch reaction cannot be recycled to
synthesis gas excludes the use of internal recycle
the reformer without first removing the CO2
in the F-T synthesis section and external recycle of
from the Fischer-Tropsch tail gas.
tail gas from the F-T reactor back to the syngas
section of the plant is excluded. This means that
Depending on the energy needs of the plant,
all tail gas must be used as fuel, which because of
the high nitrogen content will be of poor quality. the syngas from the POx reformer can either be
At a conversion of hydrogen and CO in the syn- cooled by means of a water quench or by the
production of steam in a heat exchanger.
thesis gas of about 70%, the heating value of the
tail gas will be so low that it will not burn unas- A quench system is the less costly of the two
sisted. Either supplemental fuel or catalytic approaches but is also less thermally efficient. In
216 Gas to Liquid Technologies

designing a POx-based GTL plant, the choice oxygen. The need to premix excludes combina-
between a quench and a waste heat boiler will tions of CPOx linked in a series arrangement, in
depend on the relative cost of capital and energy. particular with prereforming or heat exchange
As mentioned above, CO2–rich tail gas separated reforming, for safety reasons. Premixing hydro-
from the syngas may be recycled to the partial carbon feed with oxygen limits the applicable
oxidation reactor to improve carbon yield. If recy- reactor preheat temperatures to rather low values,
cle is used, the gas production in the POx units kept below the self-ignition temperature of the
must be supplemented by high hydrogen content mixture to maintain control when premixing oxy-
syngas in an auxiliary steam reformer, as in the gen and hydrocarbon. A feed gas mixture
Shell process. In Shell’s Bintulu GTL plant, each containing methane and oxygen may spontane-
POx unit originally had a capacity rating for pro- ously ignite at temperatures above the auto-
duction of 3000 bbl/day of F-T products. Shell has ignition temperature, depending on the actual
recently scaled up their gasification units to enable gas properties, and the industry does not yet
production of 10,000 bbl/day of F-T product have confidence that this issue has been ade-
equivalents. quately addressed.
When a catalyst is employed to assist the par- When consumption of natural gas and oxygen
tial oxidation, the process is referred to as catalytic are compared for CPOx and a combination of
partial oxidation (CPOx). The catalyst increases prereforming and ATR, consumption of both
the activity of natural gas conversion and, there- gases are significantly higher for CPOx than for
fore, provides options for higher throughput, the combination of prereforming and ATR. The
lower operating temperature, or both. main reason is that with CPOx operating at a
CPOx with no burner (flame) was first practiced lower inlet temperature than ATR (250–450  C
on a small scale at low pressures in the 1950s by vs. 600  C), a significant part of the hydrocarbon
Haldor Topsøe, and then later with higher pressures feed is combusted to CO2 and steam (using the
by Lurgi. Early versions of CPOx used premixing larger amount of oxygen) to generate heat to reach
of hydrocarbon/steam feed and oxygen, with an the desired syngas exit temperature of 1050  C.
ignition catalyst placed on top of a fixed bed cata- Considering that a large portion of the syngas
lyst, similar to the catalyst used in an ATR-based process section of a F-T plant is for the air sepa-
unit and operating with similar space velocities. ration unit, CPOx does not appear to be econom-
Today’s CPOx versions feature a combined cata- ical under these conditions. CPOx was being
lyst system performing both ignition and partial investigated by ConocoPhillips but have recently
oxidation. The noble metal catalyst operates at discontinued this approach.
very low residence time (very high space velocity). An alternative for CPOx could be the use of air
The hydrocarbon feed stream, oxygen, and steam as an oxidant. This alternative is being pursued by
are mixed upstream from the catalytic reactor. All many as a basis for fuel processing systems for
chemical conversion takes place in the reactor and fuel cells. For GTL and similar applications oper-
no burner is used. The exit gas composition from ating at high pressures however, the issue of safety
CPOx is similar to the exit gas from ATR, given still remains even with air, although to a lesser
like amount and properties (temperature, pressure, extent than with oxygen. Oxygen consumption
composition) of the inlet streams. may be conserved (to an amount about 10% less
The main potential advantage of catalytic par- than the amount required for the ATR reactors) if
tial oxidation is the smaller size of the reactor. the inlet temperature can be safely raised to a
There are also potential disadvantages that will higher level (about 480  C) with a lower oxygen
likely outweigh the savings in the cost of the content in the hydrocarbon/oxygen mix. Addi-
reactor, which is rather modest relative to the tional operational issues with CPOx include con-
total investment in the syngas generation section. cerns about catalyst life, carbon dusting, and
The main disadvantages of CPOx relate to pre- combustion “flash back” potential in an integrated
mixing of natural gas hydrocarbon feed and F-T GTL process plant.
Gas to Liquid Technologies 217

Shell has utilized POX at its plants in Bintula Because the production of synthesis gas from
and Qatar. The natural gas is partially oxidized in natural gas by oxygen-blown ATR accounts for
a noncatalytic reaction with 99.8% pure oxygen. about 60% of the total capital cost of the GTL
The reaction is conducted at a temperature around plant, the ATR and cryogenic ASU are the two
1350  C. Steam is generated from the waste heat most capital intensive components in the syngas
and is used to power equipment. The Shell gasifi- generation process section. ITM/ITM Syngas
cation process has been utilized in refineries since technology aims to lower capital intensity by
about 1960. combining the ASU and ATR units into a single
reactor. In the ITM/ITM Syngas process, O2 in air
Catalytic Membrane Reactors is selectively reduced and transported across the
Catalytic membrane reactors (CMR) are based on membrane to the other side, where it partially
the concept that selective transport of reactant(s) oxidizes methane to form syngas. The ITM/ITM
or product(s) can influence an equilibrium-limited Syngas concept is particularly attractive for use in
reaction to proceed further in the forward direc- locations where space for an ASU would be lim-
tion. The combination of SMR and the WGSR is ited, such as on an offshore platform. The ITM
equilibrium-limited, and H2-selective membranes Syngas Team, led by Air Products, has estimated
have been used to lower the temperature required capital cost savings of greater than 30% over a
for SMR by facilitating the selective removal of conventional ATR with ASU, as well as an
H2 as it forms. Both dense (mostly palladium) and increase in overall fuel efficiency to 61%
porous (such as silica or alumina) H2-perselective vs. 58% for ATR/ASU, with up to 40% less deck
membranes have been studied [20]. Throughput space required [21].
limitations, membrane stability, and cost com- Air Products and their partners are now in
bined with the ready availability of several other Phases 3 and 5 of their cooperative program with
reforming technologies have thus far hampered US DOE to develop a 2000 TPD oxygen plant.
commercial development of H2-permselective They presently are operating a 100 TPD and pre-
membrane reactors for reforming applications. paring the ceramic module fabrication for the
Another approach to CMR for methane 2000 TPD plant [22].
reforming allows selective passage of reactant OTM oxygen technology aims to lower capital
oxygen. Nonporous membranes prepared from intensity by replacing cryogenic ASU technology
perovskites or similar ceramic materials allow with membrane reactors that allow O2 to selec-
ionic transport of oxygen from air at low pressure, tively pass through the ceramic membrane to the
with up to 100% selectivity for O2. The selectively other side, where it partially oxidizes methane to
transported oxygen moves into a reaction mixture form syngas. The OTM Alliance team, led by
at low partial pressure of oxygen, against a large Praxair, focused intensively on materials develop-
difference in total pressure. ment [23]. This DOE program ended in 2004 and
Process concepts based on the use of does not appear to be progressing toward com-
O2-selective CMR eliminate the need for an air mercialization at this time. The OTM approach
separation unit (ASU) and associated large com- also has been studied in combination with elec-
pressors. Successful elimination of the ASU would trochemical pumping of O2 [24] and with CO2
impart significant economic potential for this tech- capture using a solid sorbent [25].
nology; however, difficulties with CMR would need The team headed by Praxair completed a 5 year
to be overcome first. One problem is in mechanical program with DOE in 2015. They developed
design, including the mechanical integrity of the OTM and seals that were robust to the operating
membranes and also the membrane-to-metal junc- pressures and temperatures. Their next step is to
tions. Significant consortia efforts formed with US operate a small-pilot scale OTM converter [26].
DOE funding led to development of two novel Although membrane materials with satisfac-
technologies, known as ITM (ion transport mem- tory oxygen permeability at the relevant condi-
brane) and OTM (oxygen transport membrane). tions do exist and progress has been made, the
218 Gas to Liquid Technologies

level of oxygen flux in all cases still requires – Availability of CO2 increases capacity
improvement. Concerns persist over mechanical where single step tubular reforming is
integrity and stability with time. Catalytic mem- attractive.
brane reactors are still under development, and it In the past, the methanol industry was
is not possible to predict at this time whether or dominated by single step, steam methane
when CMR will be successfully developed for reforming without the use of oxygen.
commercial use. The goal for further development Today, single step reforming is considered
of CMR is to avoid the cost of an oxygen plant. mainly for relatively small plants and for
cases where CO2 is present in the natural
Prereforming and Postreforming gas or is available on site or nearby from
Additional reforming may be performed upstream other sources such as an ammonia plant.
or downstream of the main reformer unit, when the • Two step reforming
main furnace unit has limited heat transfer capabil- – Steam/carbon ratio = 1.5–1.8.
ity that would otherwise limit overall conversion. – Capacity range about 1500–7000 ton/day
The prereformer reaction breaks down the heavier methanol equivalent.
hydrocarbons (mostly propane and butane) to – Heavy natural gas makes two step
methane, upstream of the reformer heater. Post- reforming less attractive compared to single
reforming uses an additional catalyst bed down- step tubular reforming.
stream of the main reformer. A down-flow reactor – Availability of CO2 is no advantage with
is placed between the outlet transfer line and the two step reforming.
steam generator that recovers excess process heat. Two step reforming is a combination of a
Adding a postreformer raises the effective resi- fired, tubular reformer and an oxygen-blown
dence time of the reacting gas and thereby boosts secondary reformer. This concept was
capacity. Pre- and, in particular, postreforming are applied by Statoil and Conoco in a 2400
often considered as retrofits, which may be com- ton/day methanol plant in Tjeldbergodden,
plicated by having insufficient space or inadequate Norway, commissioned in 1997. The facil-
metallurgy for the higher heat burden. ity has an annual capacity of about 900,000
metric tons of methanol, which corresponds
Methane Reforming for Methanol/DME to 25% of Europe’s total production capac-
Synthesis ity [27].
Synthesis gas production is best characterized by • ATR
the generalized stoichiometric ratio M (also refer- – Steam/carbon ratio = 0.6–0.8.
red to as the module), where M = (H2
CO2)/ – Capacity range about 5000–10,000 ton/day
(CO + CO2). Ideally M should be equal to 2.0 for methanol equivalent.
methanol (or for F-T). For kinetic reasons, in – Heavy natural gas makes ATR less
order to control by-products formation, a value attractive.
slightly above 2.0 is normally preferred. – Availability of CO2 is no advantage.
Because about 60% of the capital cost for a – This scheme is most attractive for the pro-
methanol plant is in the methane reforming unit, duction of fuel grade methanol.
research has aimed at improving reforming pro- Using ATR, the composition of synthesis
cesses. Over the years, three process concepts gas is raised to have adequate hydrogen
have been considered for preparing syngas content by taking advantage of the WGSR
targeted for methanol synthesis, described below. with removal of CO2.

• Single step tubular reforming The main conclusions from early studies were
– Steam/carbon ratio = 2.4. that for relatively small plants (<1500 ton/day
– Capacity range up to 2500 ton/day metha- methanol equivalent), single step tubular reform-
nol equivalent. ing is advantageous for syntheses gas production.
Gas to Liquid Technologies 219

For intermediate scale capacities (corresponding and the type of reactor used. These differences all
to present world scale capacities, about relate to differing compositions of F-T syngas feed
1500–3500 ton/day methanol), two step reform- in GTL vs. CTL. The stoichiometry of the F-T
ing is most attractive. For capacities above about reaction to make hydrocarbons is about 2.1/1 in
3500 ton/day, ATR or a combination of ATR and H2/CO as shown in Eq. 5. Syngas produced by
SMR is preferred. For example, the Lurgi com- gasification of coal has a H2/CO ratio of 0.7–1.0
bined reforming process uses a combination of and is therefore stoichiometrically deficient in
steam and autothermal reforming. Proven single- hydrogen. Use of an iron-based catalyst for the
line capacities for the oxygen plants associated F-T step of CTL is typical because Fe concurrently
with combined reforming have increased to catalyzes the water-gas shift (WGS) reaction, pro-
about 3000 ton/day, corresponding to about 7000 viding make-up hydrogen as per Eq. 6.
ton/day methanol plant equivalent using two step
reforming, or about 5000 ton/day methanol plant
CO þ H2 O ⇆ CO2 þ H2
equivalent using ATR alone. (6)
DH298 ¼
41:2 kJ=mol CO
A new process concept for the synthesis loop
for methanol has been developed and specially
adapted to syngas production by ATR at low In contrast, as shown above in Eqs. 2, 3, and 4,
steam/carbon ratios. In this scheme, CO2 removal the H2/CO ratio of syngas produced by reforming
is not applied. Instead, the gas composition is methane is at or in excess of the value of 2 required
adjusted by recovery and recycle of hydrogen by the stoichiometry of the F-T reaction. Thus iron
from the purge gas from the synthesis loop. F-T catalysts are less preferred for use in GTL
processes than cobalt F-T catalysts, which do not
catalyze the WGS reaction. Cobalt is effective in
Fischer-Tropsch Synthesis
the low temperature Fischer-Tropsch (LTFT)
The Fischer-Tropsch (F-T) reaction is a polymer-
regime, operated at about 220  C, and not for
ization or oligomerization of carbon monoxide
high temperature Fischer-Tropsch (HTFT), oper-
performed catalytically under reducing condi-
ated at about 340  C. At the higher temperatures,
tions. As shown in Eq. 5, the F-T reaction is
Co tends to make methane.
thermodynamically favorable by virtue of high
Iron F-T catalysts can be used in GTL pro-
exothermicity. Catalysis, temperature control,
cesses, performed at low temperature to lessen
and downstream processing are used to provide
the contribution of iron-catalyzed WGS reaction.
a product slate as close to that of target applica-
Operating in the LTFT regime, a slurry bed reactor
tions as possible.
(SBR) is typically preferred for several reasons.
One reason is that the slurry bed uses a much
nCO þ ð2n þ 1ÞH2 ⇆ CH3
ðCn
2 H2n
4 Þ
CH3 þ nH2 O lower catalyst loading than either the fixed bed
DH298 
159 kJ=mol CO or fluidized bed reactor technologies used in
(5) HTFT processes. A second reason is that the
slurry bed can remove reaction heat more rapidly,
Fischer-Tropsch processes were originally to avoid temperature increases that would result in
developed for use in coal-to-liquids (CTL) pro- undesired formation of high levels of methane,
cesses and are similar when applied to the syngas light hydrocarbons, and, in the extreme, catalyst
derived from natural gas. The reader is referred to deactivation due to coking, sintering, and disinte-
the CTL entry in this publication for additional gration. Deposition of carbon by the Boudouard
discussion of F-T performance variables, operat- disproportionation reaction, Eq. 7, is to be
ing regimes, and product distributions. avoided, and the heat distribution imparted by
The primary differences between F-T processes the slurry bed is advantageous. A third aspect
applied for GTL and CTL are the range of operat- favoring the slurry bed is that the resultant steam
ing temperatures used, the type of catalyst used, to carbon (S/C) ratio is superior.
220 Gas to Liquid Technologies

2CO ⇆ CðsÞ þ CO2 of what that single-carbon unit is, and how it
(7) derives from the CO or, for iron, the CO or CO2
DH298 ¼
83:4 kJ=mol CO
present. Candidates include a surface-associated
Even so, Shell does operate a fixed bed, carbon, a metal methylene, a metal hydroxycar-
Co-catalyzed F-T process in its GTL plants in bene, a metal alkyl, a metal alkoxide, or a metal
Malaysia and Qatar. Shell commissioned the carboxylate. That is a great many choices for a
Bintulu plant prior to the widespread consideration reaction whose mechanism has been under inves-
of slurry bubble reactor technology for F-T pro- tigation for about a century.
cesses and stands by the fixed bed technology. One The mechanistic complexity of F-T chemistry
advantage of fixed beds over slurry beds is the is compounded by the use of certain terms, such as
potential for a lesser degree of catalyst attrition that carbide or carbene, by different researchers to
results from impacts and shear forces in the latter. denote different intermediates. Table 2 shows the
steps of the F-T reaction according to various
mechanisms currently under consideration in the
Reaction Pathway field, presented as composites that capture the
Fischer-Tropsch chemistry has long been recog- essence of each step. The many descriptors in
nized as in effect a polymerization reaction, common use for the mechanistic chemistry and
wherein the monomeric unit is the methylene intermediates are also provided.
group,
CH2– that is generated in situ from car-
bon and hydrogen atoms of CO and H2. The basic Formation of C1 Species The initial reaction of
steps of the F-T reaction are as follows: syngas with the F-T metal involves reduction of
CO and produces a metal-bound C1 species (CH2,
1. Adsorption of reactant, CO, on the catalyst CHOH, or CH2OH). Metal carbides have also
surface been considered for the C1 intermediate, forming
2. Chain initiation, by CO association followed by splitting of the C-O bond with formation of
by hydrogenation water from the O and H2; however, the formation
3. Chain growth, by insertion of additional CO of carbides as discrete species attached with
molecules followed by hydrogenation strong, multiple bonds to the surface has been
4. Chain termination, by reductive elimination of largely discounted by isotopic tracer studies.
paraffin from catalyst surface or by b-hydride Many people do still use the term “carbide” to
elimination of olefin from catalyst surface refer to a more loosely associated metal-C1 spe-
5. Desorption of product, from the catalyst surface cies, in what has been referred to as the “modern
carbide” mechanism. Surface carbiding of iron
When tested under relatively comparable con- has been demonstrated under F-T conditions, but
ditions, Fe and Co catalysts give sufficiently sim- Co does not become carbided unless H2 is absent.
ilar carbon number distributions to suggest a Given that cobalt is a good hydrogenation cata-
common intermediate for the different F-T metals. lyst, if formed during F-T the carbide would be
The predominance of linearity in F-T products hydrogenated very quickly.
and the formation of terminal, a-olefins suggest The intermediacy of immediate C-O bond
that the intermediate responsible for chain growth cleavage and surface carbides need not be invoked
is a single-carbon unit. For at least the past if CO itself is hydrogenated prior to C-O bond
50 years, most researchers have agreed that the scission. Devising experiments that would differ-
product distribution observed in F-T is consistent entiate exactly to what degree the CO is associated
with a single intermediate and chain propagation with the metal when scission occurs would be
mechanism for all products (with caveats for C1 difficult under realistic F-T conditions. In addi-
and C2, vide infra). Research on the mechanism tion, work by Davis has shown that CO2 can
was most defining during the 1960s and 1970s, initiate chain growth in Fe catalysis (in which
and experiments and debate continue on the topic case an initial species would be formic acid) [28].
Gas to Liquid Technologies, Table 2 Elements of F-T mechanisms [7]
Formation of C1 species Chemistry Common names for C1 species
Gas to Liquid Technologies

CO dissociation and Surface-associated carbon (M—C), metal

hydrogenation carbide (M–C), methylene (CH2), carbene
(M=CH2), methylidene (M=CH2), alkylidene
(M=CH2)
Direct CO hydrogenation Enol, hydroxycarbene

CO insertion and hydrogenation Hydroxymethylene, alcohol

CO2 insertion and Oxygenate, metal alkoxide

hydrogenation

Carbene coupling Metal alkyl and methylene (CH2), carbene

(M=CH2), methylidene (M=CH2), alkylidene
(M=CH2)

Carbene coupling Metallacycle or olefin complex and methylene

(intramolecular) (CH2), carbene (M=CH2), methylidene
(M=CH2), alkylidene (M=CH2)

Carbene coupling Enol or hydroxycarbene

H OH
OH C C H2 CH3 CO, H2 CH3 OH
2M C M C +M M C + M C
H –H2O OH H
M M OH
CO insertion and hydrogenation Metal hydroxyalkyl or metal alkyl
221

(continued)
 222

Gas to Liquid Technologies, Table 2 (continued)

Formation of C1 species Chemistry Common names for C1 species
CO insertion Oxygenate, metal carboxylate

Chain propagation Chemistry Common names for propagation species

Carbene insertion into Metal alkyl and methylene (CH2), carbene
metal–alkyl chain (M=CH2), methylidene (M=CH2), alkylidene
(M=CH2)

Carbene insertion into Olefin complex and methylene (CH2), carbene

metal–olefin chain (M=CH2), methylidene (M=CH2), alkylidene
(M=CH2)

Carbene insertion into Metallacycle or olefin complex and methylene

metallacyclic chain (CH2), carbene (M=CH2), methylidene
(M=CH2), alkylidene (M=CH2)

Carbene coupling, or carbene Enol, hydroxycarbene

insertion into metal-carbene
chain
CO insertion and hydrogenation Metal alkyl

CO insertion and hydrogenation Oxygenate, metal carboxylate

Gas to Liquid Technologies
Examples of chain termination Chemistry Comments
Hydrogenation, reductive
elimination

b-elimination, olefin desorption (the reverse is olefin readsorption)

Gas to Liquid Technologies

Hydrogenation, dehydration

b-elimination, reductive
elimination, olefin desorption

Hydrogenation
223
224 Gas to Liquid Technologies

Chain Initiation and Propagation Chain initia- from the metal) allows formation of a metal-
tion occurs when the first C-C bond of the chain is hydride and a p-bound olefin, the latter then dis-
formed, by coupling of two initial C1 fragments or sociating easily from the metal center to give an
by insertion of carbon monoxide into a metal alkyl alpha-olefin. The b-elimination is usually quite
bond (or possibly a metal alkoxide bond, in the reversible, and chain growth may resume if an
postulated iron case). Propagation occurs by suc- olefin reinserts into the metal-hydride bond.
cessive insertion of carbene C1 species into the
growing metal-hydrocarbon chain, or by succes-
ASF Distribution For F-T reactions, the number
sive cycles of CO insertion and reduction at the
of moles of species with carbon number n
metal. The carbene terminology used here is that
decreases exponentially with n. The weight frac-
used by E.O. Fischer in his pioneering discovery
tion of product having carbon number, Wn, is
of species of this type and refers to any species
given by the Anderson-Schultz-Flory (ASF) dis-
M1C(A1)(A2) or M2(A1)(A2), where A1 is hydro-
tribution (Eq. 8):
gen or carbon and A2 is hydrogen, carbon, or
oxygen. For monometallic M1, the bonding is
Wn ¼ ð1
aÞ2 nan
1 (8)
formalized as M=C and for bimetallic M2 the
bonding is viewed as M–C–M. These M=CH2
The parameter a is the Schultz-Flory distribu-
or M–CH2–M carbenes are highly reactive and
tion factor and represents the ratio of the rate of
are well known to insert into metal-carbon
chain propagation to the rate of chain propagation
bonds, as required by the F-T product distribution,
plus the rate of chain termination. The parameter a
under milder conditions and more quickly than
is the chain growth probability, and 1
a is the
carbon monoxide does.
probability of chain termination. Both parameter
Early studies performed by Pettit with 13C
a and parameter a are assumed to be independent
labeling strongly supported the intermediacy of
of chain length. Either value can be obtained
M=CH2 species [29]. Including 14C labeled alco-
graphically using Eq. 8 in log form, with analyt-
hols and olefins in the feed, Emmett found that
ical data for Wn as obtained for example using gas
enols of alkoxides can initiate chain growth.
chromatography. The a values for oxygenated
Much of this early mechanistic work, as well as
products appear to be about the same as those
the concept that CO2 can initiate chain growth
for hydrocarbons, which suggests that the prod-
(Fe catalysis), has been reviewed by Davis
ucts all form by the same or similar chain propa-
[28]. The propagation mechanism also can be
gation steps and with similar rates of chain
considered a carbene mechanism, in that the initial
termination, albeit with different chain termina-
product of CO reaction with a metal alkyl is M–C
tion chemistries for the oxygenates, olefins, and
(=O)R, which has a tautomeric carbene form,
paraffins.
M+=C(O
)R. Formation of oxygenates could
The parameter a is used most often to charac-
occur by insertion of CO, to give M–C(=O)alkyl
terize product distribution. Values of a > 0.9 are
intermediates which are then reduced by H2.
representative of processes producing high
Alternatively the alcohols could form by reduc-
amounts of wax, a values of about 0.8 maximize
tive elimination of metal surface-bound OH and
diesel cuts, and a values of about 0.7 are charac-
the surface bound chain.
teristic of naphtha production.
Chain Termination The typical chain termina-
tion reaction is intramolecular, either reductive F-T Catalysts
elimination or b-hydride elimination. Hydrogena- Early studies reported that the level of reaction
tion of a metal on which a chain is actively grow- activities of metal catalysts for the Fischer-
ing allows for reductive elimination to give the Tropsch reaction decline in the order Ru, Ni, Co,
saturated alkane. Loss of an H atom from the beta- and Fe. Later studies indicated that activities
position of the chain (the second carbon away declined in the order Ru, Fe, Ni, and Co with
Gas to Liquid Technologies 225

alumina as support, and in the order Co, Fe, Ru, precipitated iron catalysts, used for LTFT, catalyst
and Ni with silica as support. In practice, Ru gives activity decreases when the alkali content is raised
rapid production of high molecular weight wax beyond a threshold level.
but this metal is cost-preclusive and supply- Cobalt catalysts produce considerably less
limited at plant scale. Nickel gives predominantly CO2 during F-T synthesis than do iron catalysts.
methane as product. The partial pressures of water and CO2 affect
Commercially, Co appears to be more active selectivity, with the pressure of CO2 being more
than Fe in that the turnover rates on Co catalysts relevant for iron than cobalt. Cobalt is a better
are about three times higher than on Fe catalysts. hydrogenation catalyst than iron, and the products
In general, the observed rates depend on the dif- of its F-T reaction tend to have a lower olefin/
fusion rates of reactants and products migrating in paraffin ratio than those of iron.
and out of the porous catalyst particles, as well as
on the intrinsic rate of the F-T reaction at the
catalyst surface. The diffusion rates in turn depend Catalyst Beds and Reactors for F-T Synthesis
on the catalyst’s porosity, pore sizes, and particle Axial temperature gradients, poor heat distribu-
size, and on whether or not liquid wax is present tion with carbon deposition, and scale-up limita-
within the catalyst particles. tion mean that multitubular reactors are usually
The product distribution can be significantly not suitable for commercial LTFT synthesis. The
influenced through use of promoters. The ability slurry bed reactor (SBR) is an attractive alterna-
of numerous promoters to increase activity with tive to fixed bed reactors. When using precipitated
Co catalysts is well documented but their ability to catalysts of identical composition, a SBR system
affect selectivity is less clear. In general, any pro- is as active as a fixed bed system despite the much
moter metal that can enhance the even distribution lower catalyst loading of the SBR. Slurry bed
of reduced cobalt on the internal and external reactors have become the technology of choice
surfaces of the catalyst particles could increase for companies planning new GTL plants, with
the number of active sites available for chain the exception of Shell as noted above. Both
growth and thereby increase the selectivity for Sasol and Exxon have operated large demonstra-
wax. Ruthenium appears to do this best, and lan- tion slurry phase F-T units with supported cobalt
thanum also helps. Rhenium is sometimes used to catalysts, and Sasol is currently operating
promote chain growth with cobalt although 17,000 bbl/day reactors in Qatar and in Nigeria
reports of its efficacy vary. with Co catalysts. A drawback of SBR technology
The traditional industrial support materials to date is that additional equipment is required to
for cobalt are alumina, silica, and titania achieve near complete separation of the finely
[30]. The choice of support material appears to divided catalyst from liquid wax in slurry systems.
have little effect on selectivity; however, cata- However, Sasol has developed catalysts and/or
lysts using mesoporous supports such as MCM- techniques to overcome this problem [32].
41 have provided somewhat increased C5+ Multitubular fixed beds do have advantages as
selectivity [31]. research reactors, because they are easy to oper-
The performance of iron catalysts is very ate. No equipment is required to separate the
dependent on the addition of strong alkali pro- heavy wax product from the catalyst, since the
moters. Oxides of the alkali metals, particularly liquid wax simply trickles down the bed and is
potassium, promote formation of wax in LTFT collected in a downstream knock-out pot. Alter-
catalysis by iron. Use of K2O also helps keep the natively, laboratory studies may be performed in
iron reduced in carbided form, which may pro- high velocity, high recycle stirred tank reactors.
mote the growth of longer chains. Not only is By varying the fresh feed gas flow to these units,
selectivity markedly dependent on the amount of the partial pressures and the kinetics at different
alkali present, but alkali promoters also have a conversion levels (tantamount to different bed
direct effect on the overall rate of reaction. For depths) can be determined.
226 Gas to Liquid Technologies

Kinetics in both fixed and fluidized bed reactors to study

Various kinetic equations based on research with iron-based catalyst activity in commercial HTFT
laboratory reactors have been generated for the and LTFT processes. The key findings are as
F-T reaction. Eqs. 9, 10, 11, 12, 13, and 14 are follows:
representative examples of kinetics proposed for
Co systems. • Reaction rate is strongly dependent on hydro-
gen partial pressure. At low conversion levels,
r ¼ kPH2 2 =PCO (9) reaction rate is solely dependent on hydrogen
partial pressure.
r ¼ aPCO PH2 =ð1 þ bPCO Þ2 (10) • Reaction rate increases with the partial pres-
sure of CO.

2 • The partial pressure of CO2 does not appear to
r ¼ aPCO PH2O 0:5 = 1 þ bPCO þ cPH2O 0:5 (11)
have a direct effect on reaction rate. It however

can affect the partial pressures of other compo-
r ¼ aPCO PH2 2 = 1 þ bPCO PH2 2 (12) nents via the water gas shift reaction and thus

may have an indirect effect.
r ¼ c PH2O 0:5 =PCO2 =ð1 þ 0:93PH2 =PH2O Þ • Reaction rate is markedly depressed by the
(13) partial pressure of water.
• The level of hydrocarbon products in the reac-
r ¼ dðPH2 PCO Þ=ðPCO þ 0:27PH2O Þ (14) tor has no apparent effect on F-T reaction rate.

Equations 15, 16, 17, and 18 are representative Based on these observations, a satisfactory rate
of kinetics proposed for Fe systems. equation for iron-based catalysts is given by
Eq. 19.
r ¼ kPH2 0:6 PCO 0:4
fr0:5 PH2O 0:5 (15)
r ¼ mPH2 PCO =ðPCO þ aPH2O Þ, where a
r ¼ mPH2 PCO =ðPCO þ nPH2O Þ (16) ¼ kH2O =kCO (19)

r ¼ aPCO PH2 2 =ðPCO PH2 þ bPH2O Þ (17) The above kinetic expression is speculative but
has been shown to estimate the overall conversion
r ¼ aPCO PH2 =ðPCO þ bPH2O þ cPCO2 Þ (18) for both HTFT and LTFT reactors. This kinetic
equation also predicts that if the total pressure of
Equation 10 for Co responds to an increase in the gas feed is increased, the rate is increased by the
pressure: as the pressure increases, the F-T reac- same factor, i.e., the residence time in the catalyst
tion rate decreases. Cobalt is much more resistant bed remains the same and the degree of conversion
to oxidation than iron, whether by oxygen or by remains unchanged. This means that the produc-
water, which could account for the absence of a tion rate is increased in proportion to the increase in
PH2O term in kinetic equations for cobalt catalysts. pressure. Pilot plant tests at Sasol for both the fixed
However, recent work indicates that water should bed LTFT and the fluidized bed HTFT operations
be included in the rate equation, especially at confirmed these kinetic predictions.
higher conversions [33]. Even so, Co crystals Given that reforming methane gives as much
smaller than a certain size (<5 mm) may be oxi- or more H2 as needed for the F-T reaction, the
dized by water vapor and therefore a more presence of water-gas shift chemistry in the F-T
detailed kinetic equation for cobalt could include reactor could easily become a liability to effi-
the partial pressure of water. ciency of a GTL process. Using the same reactor
The kinetics of Fe catalysis are more well- under the same high conversion operating condi-
studied than those for Co. Sasol performed a tions and feed gas flow, a Co catalyst system will
large number of experiments and measurements have a much higher syngas conversion than an
Gas to Liquid Technologies 227

iron catalyst. With an iron catalyst, after water has biomass, and pipeline natural gas by steam methane
been condensed out a larger portion of the tail gas reforming. This will also be the first commercial use
is recycled to the reactor. This means that more of microchannel reactors for FTS.
reactors are required and there is also a greater
need to recycle, both of which increase total fixed
F-T Product Upgrading
capital and operating costs.
Conversion of syngas through the F-T process
leads to a distribution of products essentially
Commercial Activities in Fischer-Tropsch
consisting of n-paraffins (>90%) and smaller per-
Synthesis
centages of alcohols, olefins, and methane. Typi-
In 1993, two large-scale F-T plants were built
cal hydrocarbon product cut weight fractions are
specifically for operation with natural gas
shown in Table 3 for an iron catalyst.
(as opposed to coal). Backed with funding from
A consequence of aliphatic chain growth by one-
the South African government, Mossgas (now
carbon units is the theoretical impossibility of syn-
PetroSA) began operating a 22,000 bbl/day plant
thesizing a F-T product having a narrow range of
using Sasol technology. The gas is produced from a
chain length. This wide molecular weight range of
South African offshore field. Also that year, Shell
the raw F-T product slate means that downstream
built a 10,000 bbl/day plant at Bintulu, Malaysia.
separations and refining must be used to create frac-
The $600 million plant was expanded to
tions for specific end-uses. The crude product from
12,500 bbl/day in May 2000, after a fire destroyed
the F-T reaction is most commonly separated into a
the oxygen plant. Also in 1993, a plant designed by
light, relatively low boiling condensate fraction and a
Rentech was built in Colorado Springs, Colorado
heavy, high boiling wax fraction. A typical boiling
for conversion of landfill gas to liquids. A lack of
range cut temperature would be 372  C. The con-
methane from the landfill shut down the plant.
densate fraction is a liquid that can be shipped from
A joint venture of Qatar Petroleum and Qatar
a remote area to a refinery site for upgrading.
Shell has recently built a GTL plant in Ras Laffan
Condensate containing, for example, C4–C15 prod-
Industrial City, Qatar, using Shells’ cobalt-
ucts is further separated, into cold separator liquid
catalyzed, fixed bed technology. The project is
and hot separator liquid. The cold separator liquid
referred to as Pearl and came on-stream in 2011.
contains C5+ in the boiling range of <260  C,
At $18+ billion, the cost of Pearl significantly
retains about 95% of the total oxygenates pro-
exceeded budget. This may have resulted at least
duced, and has sufficiently low olefin concentra-
in part from incorporation of an additional facility,
tion that hydrogenation is unnecessary. The hot
for producing NGL, into the design. Pearl is
separator liquid contains mostly hydrocarbons
expected to generate a total of 140,000 bbl per
and has a boiling range of 260–372  C.
day in two trains operating at full nameplate
capacity, making this plant the world’s largest
GTL project [34]. In addition, Pearl has capacity Hydrocracking Heavy Paraffins (Wax)
for 120,000 barrel-of-oil-equivalents total per day The C16+ reactor wax may be upgraded by various
of NGL from two trains. Another new GTL plant hydroconversion reactions including hydrocrack-
has been constructed recently in Qatar, by Sasol, ing, hydroisomerization, catalytic dewaxing,
and was operating at about 55% of nameplate
capacity in 2010 due to problems encountered Gas to Liquid Technologies, Table 3 Hydrocarbon
weight fractions in typical F-T product
which have been overcome and can now operate
at design capacity. Another 34,000 bbl/day plant Product cut Carbon no. Wt%
came on-line in 2015 in Nigeria. Gas C1–C4 2.5
Velocys and ENVIA Energy produced the first Naphtha C5–C9 6.7
product at their Oklahoma City plant in February Diesel C10–C19 18.5
Soft wax C20–C34 27.3
2017. The plant design is to produce about 1000 bbl/
Hard wax C35+ 45.0
day. The syngas is produced from landfill gas, waste
228 Gas to Liquid Technologies

isodewaxing, or various combinations thereof, to excellent feedstock for production of ethylene by

produce middle distillates for use as fuels, lubri- steam cracking, giving much higher selectivity for
cants, chemicals, and specialty materials such as ethylene than that obtained from steam cracking
nontoxic drilling oils, technical and medicinal of naphtha obtained from crude oil.
grade white oils, chemical raw materials monomers, Feeds derived from LTFT can be hydrocracked
polymers, emulsions, isoparaffinic solvents, and under much milder conditions than typical feeds
other specialty products. For example, wax from such as vacuum gas oils that are derived from
F-T synthesis is typically first hydrofined, to elimi- crude oil. In the hydrocracking of crude oil-
nate alkenes and oxygenated compounds, and then derived feeds, typically pressures as high as
fractionated into different grades of product waxes. 150 atm are required to prevent coking of the
The waxes can be sold directly or hydroprocessed catalyst by aromatic compounds. This is not nec-
by recycle to extinction to yield high-quality diesel essary with paraffinic feeds, wherein LTFT prod-
and kerosene fuels. ucts are hydrocracked at pressures from 35 to
Hydrocracking is the most relevant upgrading 70 atm using commercial hydrocracking catalysts.
process, including for maximum production of The relatively low levels of oxygenates, mainly
diesel fuels. Hydrocracking heavy paraffins alcohols and lesser amounts of acids or carbonyls,
serves to lower the boiling range of product from present in F-T waxes are easily and completely
that of wax to that of middle distillates, and to hydrodeoxygenated.
improve the cold flow properties of F-T diesel The refining of syncrude is often performed in
with branched molecules formed during hydro- a separate facility from the upstream GTL pro-
cracking. Because F-T products are predomi- cess. For direct on-site F-T reaction product
nantly linear, especially those from LTFT, the upgrading, however, the hot separator liquid cut
raw middle distillates have very poor cold flow may be combined with F-T wax and processed
properties that would hamper their use as trans- under much milder conditions than typical crude
portation fuel unless subjected to hydrocracking. oil derived feeds. Conventional hydrocracking of
Another property affected by hydrocracking is refinery feedstocks is performed in a fixed bed
the cetane number of diesel. As the pressure of a reactor using multiple or single Co/Mo catalyst
LTFT reactor increases, the selectivity for wax compositions under operating pressures as high as
increases, the degree of branching decreases. Sub- 2500 psi, using conventional technologies such as
sequent selective hydrocracking of crude wax those offered for license by Chevron or UOP. The
makes the largest contribution to the final diesel high operating pressure is required to prevent
fuel pool. Some chain branching occurs during coking of the catalyst by aromatic compounds
hydrocracking and thus the cetane number of the contained in the feed. Such high pressures are
diesel fuel produced is somewhat lower than that not required when hydroprocessing paraffinic
of straight run F-T diesel. The final diesel pool, feeds from LTFT synthesis, which do not contain
however, has a cetane number above 70 – signif- aromatics, and lower pressures between 500 and
icantly superior to the cetane number for refinery 1000 psi are used to hydrocrack LTFT products
naphtha, which varies from 40 to 50. with commercial hydrocracking catalysts. During
Upgrading wax by hydrocracking to middle this process, the oxygenates present in the feed are
distillates gives naphtha as about 15–20% of the easily and completely hydrodexoygenated. Typi-
final product. The naphtha fraction may be cal hydrocracking conditions for F-T products
upgraded to gasoline, via catalytic reforming, or entail operating temperatures of 320–450  C,
used as a steam cracker feedstock for olefins pro- pressures of 1000–1500 psi, and hydrogen treat
duction. Converting these two naphtha cuts into rates of 500–5000 scf/bbl.
on-specification gasoline would require a consid- The low sulfur content of syngas from natural
erable amount of further octane number gas enables consideration of hydrocracking cata-
upgrading. However, this naphtha consists almost lysts containing a noble metal component, partic-
entirely of linear alkanes and therefore makes ularly platinum, instead of the less expensive,
Gas to Liquid Technologies 229

conventional sulfur-resistant Co/Mo catalysts. hydroisomerization activity and consequently bet-

Platinum leads to higher hydroisomerization ter low temperature characteristics for the diesel
activity and less severe hydroprocessing condi- product. Most probably all C14+ material would be
tions, with better low temperature performance treated using such catalysts, due to the enhance-
of the resultant diesel fuel. For example, ment of cold flow properties resulting from simul-
ExxonMobil’s hydroisomerization process uses a taneous isomerization and hydrocracking. A high
noble metal catalyst to increase the amount of iso- blending cetane value is retained, and very little of
paraffins in the distillate fuel and to help the fuel to the diesel range material in the feed is degraded
meet pour point and cloud point specifications with- into naphtha. Catalysis using a noble metal on
out the need for additives. Hydroisomerization may amorphous silica-alumina is likely to be the pre-
be carried out with a single Pt-based catalyst in a ferred approach for maximum production of die-
fixed bed reactor. Preferred operating conditions sel blending material.
use a temperature range of 260–400  C, a pressure There are many technology licensors for
range between 500 and 1000 psi, and a hydrogen hydroconversion processes, including Chevron-
treat rate of 1000–4000 scf/bbl, and heavy wax Texaco, UOP, IFP (Axens), Criterion, and Haldor
products are recycled to extinction through the Topsøe. Catalysts can be obtained from these
hydrocracking process. companies as well as from Akzo Nobel and
The processing pressure is dependent on the Clariant, among others.
hydrogenation capacity of the catalyst. When the
hydrogen partial pressure is too low, dehydro- Hydrotreating Light Paraffins
cyclization of the paraffins starts to occur with Hydrogenation is less costly than hydrocracking.
the formation of polynuclear aromatics, which For large-scale plants, a condensate hydrotreater
eventually would lead to deactivation of the cata- should be used to process hydrocarbons that are
lyst due to coking. Lower hydrogen/wax ratios already lighter than the desired middle distillates
lead to a decrease of conversion rate of the C22+ and also for straight run middle distillates if cold
fraction and, after adjustment of reaction condi- flow properties are not an issue. A condensate
tions to achieve the same conversion levels, an fractionator may be employed if it is desired to
increase of both the isoparaffin content and the produce linear paraffin products. The condensate
selectivity to products lighter than diesel. fractionator can also serve to remove light gases in
Isoparaffin content also increases with operat- order to stabilize the condensate for intermediate
ing pressure. The hydrocracking process has to storage at atmospheric pressure. If a nearby naph-
meet the following conditions: tha cracker is available, then the condensate frac-
tionator will provide a naphtha product that need
• The chain length of the hydrocracked frag- not be hydrogenated to ensure storage stability.
ments should be predominantly in the desired Also, a kerosene fraction can be separated using
products carbon number range. the condensate fractionator.
• The components above said desired range If both hydrotreating and hydrocracking are
should be hydrocracked in preference to those used, they may be operated at the same pressure
that are already within or below the desired and use a common hydrogen loop. Alternatively,
range. the hydrotreater can be operated in a once through
• The production of less commercially attractive mode, at lower pressure, and the hydrogen tail gas
species, e.g., light hydrocarbon gases should then subsequently pressurized and fed to the
be minimized. hydrocracker loop. For a 30,000 bbl/day equiva-
lent product upgrading system comprised of
Due to the clean nature of the feed, nonsulfided hydrocracking and hydrotreating, the size of the
hydrocracking catalysts containing a noble metal unit is comparable to those used in oil refineries.
component such as platinum can be considered. There are many process configurations possible
The use of noble metal catalysts leads to higher depending on site-specific factors like plant
230 Gas to Liquid Technologies

Gas to Liquid Technologies, Table 4 Process schemes for production of LTFT naphtha
FT naphtha Production scheme
SR Straight run Fractionation of FT condensate
HT SR Hydrogenated straight run Fractionating and hydrotreating of FT condensate
HX Hydrocracked Fractionating and hydrocracking of FT wax
SPD Slurry phase distillate Blend of HT SR and HX naphthas

capacity, available land, and possible integration improve cold flow characteristics by changing the
with other facilities. boiling point, for compatibility with low temper-
While product upgrading schemes center ature operation. Blends of SPD naphtha with con-
around three fuel products (diesel, LPG, and ventional diesels could be used to meet winter
naphtha), the upgrading plant is quite amenable grade fuel specifications without decreasing the
to different configurations if market demand so production volume of diesel and other light prod-
indicates. Even with cracking and hydrotreating, ucts obtained from oil refining. Including F-T
the ASF distribution means that the F-T process naphtha in blends can assist refiners in meeting
and its associated refinery produce other transpor- low sulfur specifications in fuels since FT prod-
tation fuel and products beyond middle distillates. ucts have sulfur content near zero (below 1 ppm).
Other saleable products may include solvents, The projected threshold for economic viability
kerosene, jet fuel, illuminating paraffins (sold for of F-T GTL technology, in terms of crude oil
use in lighting and appliances in South Africa), price, is now at about $70–75 per BOE. This
and base oils (for lubricant products). Maximizing threshold is based on recycle of reformed NGL
recovery of the higher boiling jet fuel while (versus production of NGL for market sales). The
maintaining diesel production will tend to investment cost for a modest scale GTL facility
increase economic return. producing predominately diesel fuel has been esti-
mated at about $1.1 billion [7]. There are then
further opportunities to add value by producing
LTFT Naphtha and Blends
base oils and waxes.
Four LTFT naphthas can be produced using pro-
This $1 billion investment value may be taken
cess configurations summarized below in Table 4.
as a starting point when estimating actual invest-
The slurry phase distillate (SPD) naphtha is
ment costs for a first-of-a-kind plant. There is a
excellent feed for the production of lower olefins,
substantial gap between such estimate and the
particularly ethylene, by steam cracking. Such
final cost of, for example, Pearl. In the Pearl
paraffinic naphtha feed demands less feedstock
project, the initial configuration was changed to
and lower energy consumption in the cracking
one utilizing NGL for market sales and also the
process than naphthas from other sources. The
builders experienced escalating capital costs, due
optimum olefin yields and selectivities obtainable
to a shortage of major equipment, procurement
from any naphtha depend on the cracking severity
delays, shortage of skilled engineering labor, and
at commercial conditions. The mass ratio of light
an increased number of reactors from an initially
olefins, propylene/ethylene (P/E), has an inverse
conservative estimate, among other contributing
relationship to the cracking severity. High P/E
cost factors.
values correspond to low severities and also
lower coproduction of methane.
The LTFT naphthas other than SR are essen- Methanol
tially entirely paraffinic and therefore have com-
positions far from those of high-octane naphthas General Process Description
useful in gasoline blends. Even so, the conven- Currently the majority of methanol is synthesized
tional naphthas typically require that the refiner from syngas produced from natural gas, using
Gas to Liquid Technologies 231

autothermal reforming, steam reforming, or a esterification of lower alcohols with formic, ace-
combination of steam reforming and autothermal tic, and propionic acids in column bottoms).
reforming. Most existing commercial plants use Water from the bottoms of the first distillation
steam reforming. Incoming pressurized natural column is discharged to the process sewer system,
gas feed is preheated and passed through a zinc while heavier ends separated as bottoms in the
oxide catalyst bed to remove sulfur compounds second methanol recovery column are sent to the
that would otherwise poison downstream cata- reformer furnace fuel system for combustion. The
lysts. Steam is added to the natural gas stream methanol is at least 99.8% pure after this distilla-
before it enters the reformer. A typical reformer tion and is transferred from temporary storage to
is a large high temperature furnace containing a final product storage tanks where it is directly
number of vertical tubes that are filled with a pumped into ships, railcars, or trucks for delivery
nickel catalyst. Once the natural gas is reformed, to customers and distributors.
the resulting synthesis gas is cooled to about
30  C, condensed water is separated from the Reaction Pathway
gas in a process gas separator, and the syngas is Chemical reactions considered in methanol syn-
compressed using a steam turbine compressor that thesis are shown in Eqs. 6, 20, and 21.
uses steam generated in an economizer on the
syngas outlet from the reformer.
CO2 þ 3H2 ⇆ CH3 OH þ H2 O
Compressed syngas containing H2, CO, and (20)
CO2 is fed to a methanol synthesis reactor that DH ¼
49:0 kJ=mol CO2
uses a copper catalyst. The methanol reactor outlet
gas is cooled to about 40  C and then passed to a CO þ 2H2 ⇆ CH3 OH
(21)
pressure letdown tank where unreacted synthesis DH ¼
90:1 kJ=mol CO
gas is separated from the product methanol and
water. A portion of this unreacted syngas stream is Both CO and CO2 are reactive in the synthesis.
purged and burned as fuel, while the majority is Coupling the methanol synthesis with the water
compressed by a recirculating compressor and gas shift reaction (Eq. 6) allows a larger quantity
remixed with feed synthesis gas (from the of carbon to become available for most efficient
upstream syngas compressor). The unreacted syn- synthesis. Although the direct hydrogenation of
gas is thus recirculated back to the methanol reac- either CO2 or CO to methanol is thermodynami-
tor, resulting in an overall conversion efficiency of cally favorable (Eqs. 20 and 21, respectively), the
99%. activation energy for direct hydrogenation is
Crude methanol collected in the letdown tank apparently too high and instead the reaction pro-
usually contains up to 18% water plus impurities ceeds through a different mechanistic pathway.
such as methyl formate, higher alcohols, ketones, Reaction of syngas to produce methanol is
ethers, and esters. This is stored at atmospheric about 100 times slower if CO2 is not also present
pressure in an intermediate storage tank and later in the mix. Until as recently as the 1990s, the role
fed to a distillation plant. The distillation plant of CO2 in methanol synthesis was unclear. The
may be configured in a variety of ways. water gas shift activity of Cu catalysts is so high
A typical distillation plant consists of a single that is was difficult to delineate the roles of CO
distillation column that removes the light ends and CO2 in methanol synthesis. Isotopic labeling
including dissolved gases, dimethyl ether, methyl studies have now unequivocally shown that CO2
formate, and acetone that are burned as fuel. Bot- is the source of carbon in methanol. The CO
toms from the light ends column are sent to a undergoes the WGS reaction to make H2 and
system that consists of two distillation columns, CO2. The CO2 is thought to keep the catalyst
both of which deliver pure methanol overhead and in an intermediate oxidation state, thus
sequentially remove water and heavy ends such as preventing reduction of ZnO followed by forma-
higher alcohols, ketones, and esters (formed by tion of brass.
232 Gas to Liquid Technologies

A long-held mechanism for catalytic methanol Reactors and Catalysis for Methanol Synthesis
synthesis proceeds through a metal-bound for- The two main process features considered when
mate intermediate, formed by hydrogenation designing a methanol synthesis reactor are
[35, 36]. As depicted in Eqs. 20, 21, 22, 23, and (a) controlling and dissipating the large reaction
24, CO2 is adsorbed onto the catalyst surface [37], heat, and (b) overcoming the equilibrium con-
where it reacts with a partially oxidized methane straint to maximize per pass conversion effi-
intermediate to form a carbonate, which is then ciency. Methanol synthesis is a version of the
hydrogenated. This intermediate is then further classic trade-off between conversion and selectiv-
hydrogenated in the rate-limiting step. The copper ity. The maximum per pass conversion efficiency
catalyst sites have high activity for splitting the of syngas to methanol is limited to about 25%, and
first C–O bond of CO2 that helps maintain the catalyst activities generally decrease as the tem-
oxidation state of the active copper sites. At high perature is lowered. At higher temperatures, cata-
concentrations, however, CO2 actually decreases lyst activity increases but so does the chance for
catalyst activity and inhibits methanol synthesis. competing side reactions to form by-products
The feed gas composition for methanol synthesis such as dimethyl ether, methyl formate, higher
is typically adjusted to contain 4–8% CO2 for alcohols, and acetone. Catalyst lifetimes are also
maximum activity and selectivity. Although Cu reduced by continuous high temperature
has water gas shift activity, excessive water causes operation.
deactivation by blocking active sites. Site Maintaining reaction temperatures below
blocking lowers catalyst activity but improves 300  C is the trade-off strategy employed to min-
selectivity, with a 50% decrease in by-product imize catalyst sintering while maximizing conver-
formation. sion of syngas to the target product slate.
Recent modeling efforts using density func- Although the equilibrium conditions favor low
tional theory (DFT) suggest that the traditional temperatures, methanol converters must be oper-
methanol synthesis mechanism proceeding via ated at temperatures in the range 200–300  C to
CO2 accounts for only about two third of the ensure the catalysts are active and to use the heat
methanol produced [38]. The remaining methanol of reaction effectively [39].
is said to derive from the CO, mostly via direct Removing methanol as it forms is another strat-
reduction by H2. The proposed mechanism is egy used for overcoming the thermodynamic lim-
conceptually similar to that for CO2 reduction to itations and improving the per-pass conversion
methanol and is shown in Scheme 1. Overall, and process efficiency. Methanol is either physi-
methanol synthesis appears rate limited by forma- cally removed or converted to another product,
tion of methoxy (CH3O*, where * refers to the such as dimethyl ether, methyl formate, or
catalyst surface) at low CO2/(CO + CO2) ratios acetic acid.
and by CH3O* hydrogenation in CO2-rich feeds. Numerous methanol reactor designs have been
Hydrogenation of CH3O* as per Eqs. 20 and 21 is commercialized over the years and these can be
the slow step for both the CO and the CO2 meth- roughly separated into two categories, namely
anol synthesis routes. adiabatic and isothermal reactors. Adiabatic

Gas to Liquid
Technologies,
Scheme 1 Proposed
mechanism for reduction of
CO to methanol
Gas to Liquid Technologies 233

Gas to Liquid Technologies, Table 5 Methanol synthesis catalysts prepared under various aging conditions
Catalyst Metals (molar ratio) Precipitation parameters Aging parameters
Cat2-const Cu/Zn/Al (60/30/10) pH 7, 70  C pH 7const, 70  C
Cat3 Cu/Zn/Al (60/30/10) pH 8, 70  C pH 8const, 70  C
Cat2-free Cu/Zn/Al (60/30/10) pH 7, 70  C Free-aged, 70  C
CatCu/Zn Cu/Zn (70/30) pH 7, 70  C pH 7const, 70  C

reactors often contain multiple catalyst beds sep- 140

arated by gas cooling devices, using either direct
120
heat exchange or injection of cooled syngas,

commercial catalyst
[%] Productivity of
either fresh or recycled. Axial temperature profiles 100
often have a saw tooth pattern: dropping down at
80
the point of heat removal and increasing linearly
between the heat exchange sections. The isother- 60
mal reactors are designed to continuously remove 40
the heat of reaction so they can operate essentially
like a heat exchanger with an isothermal axial 20
temperature profile. 0
Currently essentially all commercial processes 0 60 100 160 200 260
produce methanol from syngas in a vapor phase time [min]
reactor with a synthesis loop. In a gas phase,
tubular or packed bed reactor, reaction heat is Gas to Liquid Technologies, Fig. 3 Methanol produc-
tivity as a function of the aging time for selected catalysts:
dissipated in situ by injecting cool, unreacted (triangle) Cat2-const; (square) Cat2-free; (circle) CatCu/Zn;
feed at various points along the reactor length, or (diamond) Cat3 [41]
by including internal cooling surfaces for heat
exchange. Temperature moderation is also A typical commercial methanol synthesis cat-
achieved by recycling large quantities of gas, alysts contain copper and zinc oxide phases with
with gas cooling in the recycle loop. Even with alumina support, i.e., Cu/ZnO/Al2O3. Recent
these heat transfer measures, CO content of the modifications to catalyst preparation at laboratory
feed is limited to about 16% as a means of con- scale include specialized co-precipitation
trolling heat content by limiting the conversion. methods [40], for a greater synergistic effect
The design challenge of removing heat while between copper and zinc, and maintaining strict
maintaining precise temperature control is signif- control of pH and of the temperatures of precipi-
icant, because excessive temperatures seriously tation and calcination [36]. As shown in Table 5
diminish the lifetime of the catalyst. A liquid pro- and Fig. 3, the productivity of commercial cata-
cess is of interest, wherein nonvolatile mineral lysts is sensitive to several factors, including con-
oils make up part of the reactor liquid phase and ditions used for preconditioning by, for example,
act as a temperature moderator and heat sink. aging the catalyst [41, 42].
A new approach investigated by Air Products
involves bubbling the syngas through a slurry
consisting of micrometer-sized methanol synthe- Activities in Methanol Synthesis
sis catalyst and mineral oil as reaction medium Catalytic methanol synthesis from syngas was clas-
that transfers heat from the catalyst surface, sically run as a high temperature, high pressure
through the slurry, to boiling water in an internal exothermic equilibrium limited synthesis reaction
tubular heat exchanger. Concentrations of catalyst operated at 320–450  C and 250–350 atm pressure.
as high as 50% have been used in the slurry BASF of Germany introduced the first commercial
system without damaging the catalyst activity. synthesis of methanol, in 1923, using a zinc/
234 Gas to Liquid Technologies

chromium oxide catalyst. New methanol synthesis newer burner system, in what is called the Mega-
technologies based on lower reaction pressures Syn process for converting oxygen and natural
(50–100 atm) were introduced in the 1960s. gas to syngas.
Whereas high pressure processes require The first plant to use the MegaMethanol ® tech-
139,000–146,000 Btu/gal of methanol produced, nology was brought on stream in 2004 at Point
all modern plants are based on low pressure tech- Lisas in Trinidad. The plant is owned by Atlas
nology and are expected to require fewer than Methanol Company, a joint venture between
100,000 Btu/gal of methanol produced, on an aver- Canada’s Methanex and BP Trinidad and Tobago
age yearly basis. Additional advantages of low- (bpTT). The 4657 metric ton/year Atlas methanol
pressure processes include lower maintenance plant was built adjacent to the existing Titan Plant,
costs and reduced downtime. which has a methanol production capacity of
Johnson Matthey (formerly ICI Chemical 850,000 ton/year. Average total methanol produc-
Industries) of the United Kingdom and Lurgi of tion from the plants is 2.5 million ton/year, which
Germany offer the two most widely used technol- makes Trinidad and Tobago the world’s leading
ogies for methanol synthesis. The ICI low- methanol exporter.
pressure process operates at pressures between The Point Lisas plant has an on-site air separa-
50 and 100 atm and temperatures of about tion plant having a production capacity of 2800
210–290  C. A highly selective copper/zinc alu- tons of pure oxygen per day. Their mega-
mina catalyst is used in a cold-shot single or methanol production process consists of oxygen-
multibed reactor. The reaction is quenched on blown autothermal reforming at a pressure of
the bed by introducing cold gas through specially about 40 atm, desulfurization and prereforming,
designed spargers. Heat recovered from the syn- a two-step Lurgi methanol synthesis in water- and
thesis loop provides part of the process steam gas-cooled reactors, and adjustment of syngas
required by the steam reforming furnace. composition via hydrogen recycle. The reformer
Lurgi’s conventional low-pressure methanol outlet temperatures range from about 950  C to
process has maximum output of 1830–2000 1050  C. The synthesis gas is then compressed
ton/day. It uses a copper-based catalyst, pressures using a single casing gas compressor. An inte-
of 50–100 atm, and temperatures of 230–265  C. grated recycler in the compressor generates the
Temperature control of the catalyst is achieved by pressure for the methanol synthesis.
indirect cooling with boiling water. The boiling The methanol reactor contains a fixed number
water in the reactor shell generates 1–1.2 kg of of tube sheets containing the copper-based cata-
steam/kg of methanol, with a pressure of up to lyst. The isothermal reactor enables partial con-
45 atm. The steam, after moderate superheating, version of the syngas to methanol, reduces the
supplies the total energy for driving the recycle by-products produced during the process and
gas compressor and is subsequently used for the increases the yields at low recycle ratios. The
reboilers in the methanol distillation section. Lurgi combined converter process employs a
A small quantity of surplus steam is often avail- gas-cooled reactor and a water-cooled reactor.
able for export. The gas containing methanol from the first reactor
Johnson Matthey, Lurgi, and Haldor Topsøe is transferred to a downstream water-cooled reac-
have been pursuing 10,000 ton/day mega- tor where cooling takes place. The purge gas is
methanol single train plant designs. Lurgi’s Mega- separated and the cooled crude methanol is pro-
Methanol ® process can accommodate a single cessed in a three-column distillation unit through
train capacity of at least 5000 ton/day and was which about 40% of the heating steam is saved.
developed for capacities larger than 1 million tons The high cost of natural gas in North American
of methanol per annum. A schematic of Lurgi’s in the 2000s caused most methanol plants to cease
process is shown in Fig. 4. Lurgi does not use operations and to move production to locations
conventional steam reforming to produce syngas, with cheaper natural gas. China greatly increased
instead employing autothermal reforming with a production of methanol during the 2000s but they
Gas to Liquid Technologies 235

Gas to Liquid Technologies, Fig. 4 The Lurgi MegaMethanol ® process [43]

used coal as the feedstock. These plants came down price to levels that would still provide ade-
on-stream in the 2010s. However, the introduction quate profitability at prices below $100 per metric
of fracking greatly increased natural gas produc- ton. This price environment creates new business
tion in North America resulting in much cheaper opportunities for methanol to compete with other
natural gas. Thus, in the mid-2010s companies are basic fuel and feedstock options. With a combus-
moving their methanol production to North Amer- tion higher heating value of 9776 Btu/lb., a $100
ica. This even includes Chinese companies such per metric ton methanol fuel has an inherent
as Yuhuang Chemical, who are constructing a energy cost of $4.6 per million Btu. This cost is
5000 tpd plant in Louisiana that will come comparable to imported LNG at the receiving/re-
on-line in 2019. gasification terminal. Since the transportation/dis-
Johnson Matthey built a 5000 tpd low pressure tribution cost for methanol as a nonviscous liquid
methanol plant in Trinidad in 2005; at that time is relatively low, methanol fuel cost at the burner
this was the largest methanol plant in the world. tip can be significantly cheaper than natural gas
Another supplier of large-scale methanol plants is derived from LNG at the burner tip. Not only can
Haldor Topsøe who provided the technology for a methanol be considered under some economic
3500 tpd methanol plant in Texas, the first Green- scenarios as a primary fuel for combustion, its
field plant in the USA in many years. very low sulfur content (even when compared to
As smaller, high cost methanol plants continue ultralow sulfur diesel) provides an opportunity for
to be replaced by mega-methanol plants, the methanol to be used as a backup fuel for utility
industry’s aggregate cash cost curve should drive and commercial power plants.
236 Gas to Liquid Technologies

Eastman Chemical and Air Products, with US Gas to Liquid Technologies, Table 6 Comparative con-
DOE support, developed and brought on-line a versions for DME, methanol, and methanol/DME
commercial scale demonstration of the liquid Conversion DME Methanol Methanol + DME
phase methanol (LPMEOH) process during the Per pass (%) 50 14 18
1990s. This plant has a capacity of about 250 met- Total (%) 95 77 85
ric tons/day and has now operated for about
20 years.
methanol synthesis. Table 6 shows the conver-
Dimethylether sions obtained for direct processes producing
High oil prices and short supplies of LNG and DME, methanol, or a mixture of DME and meth-
LPG in Asia have led to construction of many new anol. The observed optimum ratio for DME syn-
DME plants, particularly in China. DME is cur- thesis is lower than that for methanol synthesis
rently industrially important as the starting mate- and ideally should be targeted at about 1.
rial in production of the methylating agent Interestingly, use of the methanol synthesis
dimethyl sulfate. DME has the potential to be reaction 21 instead of 23 would give a net reaction
used as a diesel or cooking fuel, a refrigerant, or for direct DME synthesis having H2/CO stoichi-
a chemical feedstock. DME is also used as an ometry of 2:1, which does not agree with experi-
aerosol propellant. mental observation. Thus the direct synthesis of
DME apparently proceeds through a different
Reaction Pathway methanol-producing step than occurs in produc-
DME traditionally is formed in a two-step process tion of methanol itself.
where methanol is first synthesized and is then
dehydrated over and acid catalyst such as
g-alumina, at methanol synthesis conditions. The Commercial Activities in DME Synthesis
dehydration reaction is shown in Eq. 22: Gas phase reaction in a fixed bed was generally
used in the past for DME synthesis from metha-
nol. Development of the direct synthesis of DME
2CH3 OH ⇆ CH3 OCH3 þ H2 O
(22) in the gas phase with improved fixed bed reactors
DH ¼
24:5 kJ=mol DME is also of interest [47]. Other improvements to
direct DME production include the use of a slurry
Alternatively, DME can be synthesized reactor for liquid phase, low-pressure DME
directly from syngas – without the separate pro- synthesis.
cess step involving methanol production. Exten- Commercial production of DME originated as
sive studies have been performed to deduce the formation of by-products during high-pressure
stoichiometry of direct DME synthesis from syn- methanol production. Direct, or single-step, syn-
gas, and the topic is still under debate [44]. Work thesis of DME from synthesis gas was developed
by Air Products and the JFE Group indicates that in the 1980s, prompted by the high prices and
the stoichiometry is 1:1 in H2/CO [45, 46]. The uncertain supply of oil prevailing at the time.
net reaction is shown in Eq. 23: The objective of direct synthesis was to produce
DME as an intermediate step on route to synthetic
3H2 þ 3CO ⇆ CH3 OCH3 þ CO2 gasoline. Interest in nonoil based fuels waned
(23)
DH ¼
246 kJ=mol DME quickly as the oil supply position eased, until the
early 1990s when interest was rekindled – this
Net reaction 23 is a combination of Eqs. 20, 21, time on a much wider scale – by increasing
and 22. demand in Asia for LPG or the LPG-substitute,
Because water produced in the dehydration DME, and by growing concern to produce cleaner
step is consumed in the WGSR, the conversion fuels. The companies currently most actively
of syngas can be higher in DME synthesis than in engaged in development of an economic DME
Gas to Liquid Technologies 237

process at large scale are JFE (Japan), Haldor comprised of alumina particles of about 200 mm
Topsøe (Denmark), and Air Products & average size, with a layer of Zn/Cu oxide metha-
Chemicals (USA). nol synthesis catalyst formed around each alu-
mina particle. The solvent is a light liquid oil
Gas Phase DME Production Direct synthesis of (such as Witco-40) or a hydrocarbon such as
DME in the gas phase with fixed bed reactors has n-hexadecane and acts as both a mass transfer
been developed most extensively by Haldor medium for the reaction and a heat transfer
Topsøe [47]. For large-scale processes, the reac- medium through which heat from the exothermic
tion may be conducted in a series of three to four reaction is removed by convection [55].
reactors with interstage cooling. Each reactor is Control of the reaction temperature is even
operated adiabatically. Alternatively, the adiabatic more important in DME synthesis than in metha-
operation may be carried out at low per pass nol synthesis, because the higher equilibrium con-
conversion of CO, to keep the reaction exotherm version to DME emits more heat and a hot spot in
within limits without interstage cooling (although the reactor could damage the catalyst. In a slurry
such an option may not be economically phase reactor, the heat of reaction is quickly
attractive) [48]. absorbed by the high boiling liquid oil, which
Haldor Topsøe’s process uses natural gas has a high heat capacity. Heat is removed from
reforming for production of syngas with the ratio the reactor through immersed internal cooling
(nH2
nCO2)/(nCO + nCO2) between 2.04 and 2.1, coils, in which water is vaporized to steam. The
where n is the number of moles. The greatest CO temperature within the reactor vessel is well con-
per pass conversion (65%) and the highest selec- trolled in order to achieve higher conversion with
tivity of DME relative to that of methanol (82/18) longer catalyst life.
were observed using CO rich syngas. With no The direct synthesis technology under devel-
CO2, the observed DME selectivity and produc- opment by NKK Corporation (now a part of JFE
tivity were substantially higher. Thus in contrast Holdings, Inc. formed by the merger of NKK and
to the synthesis of methanol, described above, the Kawasaki Steel Corp. in 2002) since 1989 also
presence of CO2 in the feed syngas was found to uses a slurry phase reactor. In this process, natural
have a major negative impact on DME formation. gas is reformed in an autothermal reformer with
Current DME synthesis catalyst is formulated oxygen and steam. Carbon dioxide is recycled
to catalyze the water gas shift reaction at a mini- back from the purification section of the plant to
mum level with minimal CO2 production. The give synthesis gas with a H2/CO ratio of 1.0. The
quality of the residual syngas from the reactors is standard DME synthesis reaction conditions are
therefore good enough for recycle without the 260  C and 50 atm pressure. The condensed reac-
need for a large, dedicated CO2 removal plant. tor product is purified in two distillation columns,
Ongoing improvements to the gas-phase process with CO2 vapor from the first column returned to
include development of bifunctional catalysts to the reactor, and the bottoms from the first column
produce DME in a single step in a single distilled in the second column. DME is condensed
reactor [49]. from the second column overheads, and
by-product methanol in the column bottoms
Slurry Phase DME Production Air Products returned to the DME synthesis reactor after
did a substantial body of research toward devel- water removal. The once through CO conversion
oping a slurry-phase process for direct synthesis is 50%. DME selectivity is more than 90%, and
of DME [50–54]. The presence of CO2 in the feed the molar ratio of DME to DME + methanol is
gas was observed to have a major impact on DME 0.91. The co-production of water is very small,
formation, as expected based on the stoichiometry with the molar ratio of (DME/DME + methanol +
of Eq. 23. In the slurry reactor, syngas is bubbled water) = 0.013. With recycle the total CO conver-
upward through the solvent, which contains sion reaches 95%. A pilot demonstration plant
suspended catalyst particles. The catalyst is project of 100 ton/day DME capacity was begun
238 Gas to Liquid Technologies

by NKK at Hokkaido, Japan, in 2002 and was cost of shipping methanol long distances may be
operated for 346 days to make six runs [56]. less than the cost of shipping an equivalent quan-
Mitsubishi Gas Chemical (MGC) touts higher tity of DME due to the higher cost of the pressur-
single pass conversion for its Superconverter [52] ized refrigerated storage requirements anticipated
slurry DME reactor system, claiming 70–80% for oceangoing DME tankers.
conversion efficiency compared with 50–60% In summary, there are three processes for the
for other processes. MGC claims that DME pro- production of DME: one step, two steps, and
duced at a remote site with low cost natural gas slurry processes. For the two-step process metha-
feedstock in a 5000–7000 ton/day plant should nol is synthesized in one reactor and then
beat the delivered cost of both LNG and LPG in converted to DME in a second reactor; among
Japan. Japan DME (MGC, ITOCHU, and the companies offering and/or practicing this pro-
Mitsubishi Heavy Industries) have planned a cess are Toyo, Mitsubishi, Lurgi, and Udhe [59].
1 million ton/year capacity plant in Papua New For the single-step synthesis, the methanol and the
Guinea but this failed to be accomplished acid function is combined into one catalyst parti-
[57]. They recently announced an agreement for cle. Among companies offering this technology
the construction of a methanol and dimethyl ether are Topsøe, JFE, Air Products, and NKK. The
plant in Trinidad and Tobago that would produce slurry phase reactor is utilized by Eastman Chem-
1 million tons per year of methanol and 20,000 ical and began operation in 1997 with a produc-
tons of dimethyl ether per year. The plant is sched- tion rate of 260 tons/day (1900 bbl/day). JFE also
uled to be on-line in 2018. has a DME process that uses a slurry reactor and
One key challenge in the economic design of has operated a 100 tpd (730 bbl/day) plant.
7000 ton/day plant is the reactor design configura- While most methanol and dimethyl production is
tion and catalyst to enable dehydration using a single based on natural gas, China utilized coal as its feed-
reactor vessel, as opposed to multiple serial reactor stock. It was reported that in 2011 there were
stages or parallel reactor trains. Another important 60 DME producers in China [60] with 17 having a
consideration is the ability to achieve sufficiently capacity of 200,000 tons per year (4,000 bbl/day).
high methanol conversions to permit reaction prod- However, in late 2016 it was reported that there was
uct recovery in a single distillation train. an oversupply of DME in China and that the operat-
Another process for scales such as 7000 ton/day ing rate was only about 20% of the design capacity
is referred to as the Jumbo process and is under [61].
development by Toyo Engineering using a dehydra-
tion reactor based on radial flow design. The process Methanol to Gasoline and/or Diesel
involves integrated co-production of methanol from The methanol to gasoline (MTG) process devel-
natural gas as an intermediate to DME. DME is oped by Mobil Oil Corporation (now ExxonMobil)
produced via the vapor phase dehydration of meth- involves conversion of methanol to hydrocarbons
anol over a fixed bed of an acid catalyst, typically over a zeolite catalyst. The methanol conversion
g-alumina. Reaction temperatures are typically in process occurs in two steps with chemistry as
the range of 250–400  C, with pressures as high as shown in Eqs. 22 and 24 (thermochemistry
250 psig. Gas space velocities have been reported to shown for ethylene as product).
be in the range of 500–10,000 hr
1 [58].
Toyo’s DME process offers greater flexibility
2CH3 OH ⇆ CH3 OCH3 þ H2 O
in the methanol/DME product distribution. Since (22)
the methanol and DME plants operate indepen- DH ¼
24:5 kJ=mol DME
dently, the complex could be designed to produce
methanol or DME exclusively or in any combina- CH3 OCH3 ⇆ C2
C5 olefins þ H2 O
(24)
tion in accordance with market demand. Toyo’s DH ¼
5:5 kJ=mol DME
design approach would also permit the option of
locating their Jumbo DME plant at a different Crude methanol (17% water) is superheated to
location from that of the methanol plant. The 300  C and partially dehydrated over an alumina
Gas to Liquid Technologies 239

catalyst at 27 atm to yield an equilibrium mixture engineering services) from 1982 to 1985.
of methanol, diethyl ether, and water (75% con- Although no commercial plants have been built
version of methanol). This effluent is then mixed based on the fluid bed technology, it is considered
with heated recycled syngas and introduced into a ready for licensing and/or commercialization.
reactor containing ZSM-5 catalyst at 350–366  C More detailed process design and economics
and 19–23 atm to produce hydrocarbons (44%) information can be found in [63].
and water (56%). Because the zeolite has to be ExxonMobil also developed a methanol to gas-
regenerated frequently to burn off coke formed oline and diesel process, referred to as the MOGD
during the reaction, the MTG process is rendered process. Oligomerization, disproportionation, and
continuous by using multiple gasoline conversion aromatization of olefins are the basis for the
reactors operating in parallel with 2–6 week MOGD process. Selectivity to gasoline and distil-
cycles. Mobil’s MTG process uses different pres- late from olefins is greater than 95%. The product
sures for syngas production (15–20 atm), metha- slate from MOGD is 3 wt% paraffins, 94 wt%
nol synthesis (50–100 atm), and the fixed bed olefins, 1 wt% napthenes, and 2 wt% aromatics
MTG process step (15–25 atm). [64]. A large-scale test run was performed at a
The selectivity of ExxonMobil’s MTG process Mobil refinery in 1981. The MOGD process is not
to gasoline range hydrocarbons is about 85 + %, currently in commercial practice.
with the remainder of the product being primarily Haldor Topsøe’s Integrated Gasoline Synthesis
LPG. Nearly 40% of the gasoline is aromatic (TIGAS) process was developed to minimize cap-
hydrocarbons with product distribution of 4% ital and energy costs by integrating methanol syn-
benzene, 26% toluene, 2% ethylbenzene, 43% thesis with the MTG step into a single loop (i.e.,
xylenes, 14% trimethylbenzenes, plus 12% other without isolation of methanol as an intermediate).
aromatics. The shape selectivity of the zeolite Whereas ExxonMobil’s MTG process uses differ-
catalyst results in a relatively high concentration ent pressures for syngas production, methanol
of durene (1,2,4,5-tetramethylbenzene), represent- synthesis, and the fixed bed MTG process step,
ing about 3–5% of the gasoline produced. There- the TIGAS process involves catalysts and condi-
fore, MTG gasoline is usually distilled and the tions modified such that the system pressure levels
heavy fraction is hydroprocessed to reduce the out and separate compression steps are not
concentration of durene to below 2%. This results required. A mixture of methanol and DME is
in a high-quality gasoline with a high octane num- made prior to gasoline synthesis, which results
ber. However, the future of the MTG process in only one recycle loop from the gasoline syn-
remains uncertain because the 1990 Clean Air thesis step back to the methanol/DME synthesis
Act Amendment limits the amount of aromatics step. A 1 ton/day demonstration plant was built in
in reformulated gasoline. Houston, Texas, in 1984 and operated for 3 years.
The first commercial MTG plant came The gasoline yield for the TIGAS process (defined
on-stream in 1985 in New Zealand (Mobil’s as the amount of gasoline produced divided by the
Motunui plant), producing 4500 ton/day of meth- amount of natural gas used for feed and as fuel)
anol and 14,500 bbl/day of high octane gasoline was 56%. The TIGAS process yields a lower
from natural gas. Gasoline production was later quality gasoline with a lower octane number com-
ceased and presently this plant and the nearby pared to MTG, because of a reduced selectivity to
methanol plant at Waitara produce 2.43 million gasoline range aromatics.
ton/year of chemical grade methanol for It was announced that in 2014 there were six
export [62]. MTG plants operating in China; however, no large
A 100 bbl/day fluidized bed MTG pilot plant plants were operating outside of China [65].
was jointly designed and operated near Cologne,
Germany, by ExxonMobil (supplying process Methanol to Chemicals
technology and proprietary catalyst), Union The highest volume chemical processes using
Reinische Brankohlen Kraftstoff AG (acting as methanol as feedstock are formaldehyde, methyl
the operating agent) and Uhde GmbH (providing tertiary-butyl ether, and acetic acid. Globally,
240 Gas to Liquid Technologies

formaldehyde production is the largest consumer for the production of isobutene, and it also can be
of methanol, followed by methyl tertiary-butyl used in a number of chemical reactions including
ether (MTBE) and acetic acid. methacrolein, methacrylic acid, or isoprene pro-
Formaldehyde is produced commercially from duction. However, the use of MTBE as a gasoline
methanol by three industrial processes: additive was terminated in the USA and this
greatly slowed the growth of the manufacture
1. Partial oxidation and dehydrogenation with air in of MTBE.
the presence of silver crystals, steam, and excess Acetic acid is produced by methanol carbonyl-
methanol at a temperature of 680–720  C, other- ation in the liquid phase using Co, Rh, or Ni
wise known as the BASF process. The methanol catalysts promoted with iodine. The BASF pro-
conversion for this process is 97–98%. cess is initiated by the reaction of methanol with
2. The same as process 1 except in the presence of HI to yield methyl iodide. The active catalyst is
crystalline silver or silver gauze at a tempera- the metal carbonyl [Rh2I2(CO)2] into which
ture of 600–650  C. Then, the product is dis- methyl iodide inserts during the rate-limiting
tilled and the unreacted methanol is recycled. step. Acetic acid is formed by the hydrolysis of
The primary methanol conversion for this pro- the eliminated acetyl iodide species CH3COI that
cess is 77–87%. also regenerates HI. The process is run with over
3. Oxidation with only excess air in the presence of 99% selectivity at conditions of 180  C and
a modified iron/molybdenum/vanadium oxide 30–40 atm.
catalyst at a temperature of 250–400  C, also The Monsanto process is similar to that of
known as the Formox® process (Perstorp Spe- BASF but is less severe. All new installed capac-
cialty Chemicals, Sweden) process. The metha- ity since 1973 is based on the Monsanto process.
nol conversion for this process is 98–99%. Even so, the Rh/I catalytic system is very corro-
sive and requires expensive steels for materials of
Formaldehyde is used to make resins with phe- construction. Complete recovery of the expensive
nol, urea, or melamine for the manufacture of Rh catalyst and recycle of HI are paramount to
various construction board products. maintain favorable process economics. The high
MTBE is produced by reacting isobutene with cost of Rh has led to the search of other, lower cost
methanol in the presence of an acidic catalyst. The metals that could be used as acetic acid process
reaction temperatures and pressures are 30–100  C, catalysts with similar performance.
and 7–14 atm, respectively, so that the reaction Approximately half of the world’s production
occurs in the liquid phase. Catalysts used are solid of acetic acid comes from methanol carbonylation
acids, including zeolites (H-ZSM-5), and micro- and about one third from acetaldehyde oxidation.
porous sulfonic acid ion exchange resins such as Methanol carbonylation is the most likely tech-
Amberlyst-15. A molar excess of methanol is nology of the future. Vinyl acetate, acetic anhy-
used to increase isobutene conversion and inhibit dride, and terephthalic acid are all made from
the dimerization and oligomerization of iso- acetic acid. Latex emulsion resins for paints, adhe-
butene. At optimum reaction conditions, MTBE sives, paper coatings, and textile finishing agents
yields approaching 90% can be achieved. Cur- are made from vinyl acetate. Acetic anhydride is
rently there are over 140 MTBE plants, with a used in making cellulose acetate fibers and cellu-
total installed capacity of about 20 million tons/ losic plastics.
year, using processes developed by Snamprogetti,
Huls (now Oxeno) ARCO, IFP, CDTECH (ABB Ammonia
Lummus Crest and Chemical Research Licens- Ammonia is the basic building block of the nitro-
ing), DEA (Formerly Duetsche Texaco), Shell gen industry worldwide. Almost all ammonia is
(Netherlands), Phillips Petroleum, and Sumitomo. produced in the anhydrous form (free of water) by
Greater than 95% of the MTBE produced is a catalytic reaction of nitrogen (from air) with
used as a fuel additive in the gasoline motor pool. hydrogen from a hydrocarbon source (usually
MTBE is also used in the petrochemical industry natural gas). Ammonia is a colorless, pungent,
Gas to Liquid Technologies 241

nonflammable gas at normal pressure and temper- larger amounts of sulfur. The company ICI (now
ature and is lighter than air. For storage at atmo- Johnson Matthey) makes dirty shift conversion
spheric pressure at sea level, ammonia must be catalysts that consist of cobalt and molybdenum
cooled to
33  C and stored as liquid. Lower oxides. They operate over a temperature range of
temperatures are required at higher altitudes. In 230–500  C. The controlling factors are the ratio
the usual atmospheric temperature range of of steam to sulfur in the feed gas and the catalyst
0–40  C, the range of vapor pressure is about temperature.
4–15 atm. Approximately 10% of the ammonia The synthesis gas is then washed with a solvent,
produced never reaches the market, because of its such as monoethanolamine or other amine, potas-
volatility and losses, during conversion to other sium carbonate, or sulfinol (sulfolane and an
materials and during transportation and storage. alkanolamine) to remove most of the carbon diox-
Nitrogen fertilizer consumption accounts for more ide. The solvents typically contain an activator to
than 85% of the world ammonia market. promote mass transfer. Water scrubbing at high
pressures was employed in the past to remove CO2
General Process Description but is no longer used. Most newer steam reforming
The three main steps in producing ammonia are plants use methanation to remove residual or trace
syngas production, syngas purification, and ammo- carbon monoxide and carbon dioxide, by reaction
nia synthesis [66]. The traditional and current com- with hydrogen over a nickel catalyst to yield meth-
mercial technology for hydrogen production in ane and water.
ammonia production involves steam reforming of Scrubbing with a liquid nitrogen wash produces
natural gas to produce syngas (Eq. 2), and the water a synthesis gas with very low inert gas content.
gas shift reaction (Eq. 19), both of which occur in This additional nitrogen is supplied by using air, in
the reformer. Steam reforming is by far the least excess of the stoichiometric nitrogen requirement,
expensive and most popular method of producing in the secondary reformer or, in the case of plants
hydrogen for ammonia synthesis. Two-step steam with no reforming sections such as offgas plants,
reforming or partial oxidation also can be used to by obtaining nitrogen from an associated air sepa-
generate hydrogen-rich syngas for ammonia pro- ration unit. Older processes involved scrubbing the
duction, but these methods are most useful when synthesis gas with copper liquors, such as cuprous
the feed contains heavier hydrocarbons in addition ammonium formate, to remove carbon monoxide,
to methane. and caustic, to remove carbon dioxide. Many mod-
Raw synthesis gas must be purified before being ern processes include molecular sieve dryers
sent to the ammonia synthesis unit. A shift conver- downstream of the methanation step to remove
sion step is used to remove most of the carbon final traces of water. The synthesis gas is then
monoxide from the synthesis gas, because CO compressed and finally reaches the ammonia con-
acts as a poison to the catalyst used in ammonia verter, where the contained hydrogen and nitrogen
synthesis. Shift conversion by the WGSR (Eq. 19) chemically combine over a catalyst into ammonia.
also produces more hydrogen. The shift conversion
is performed in two stages. In the first stage, called
Reaction Pathway
the high temperature shift (300–450  C), the bulk
The chemical reaction considered in ammonia
of the carbon monoxide is oxidized to CO2 over a
synthesis is shown in Eq. 25.
chromium-promoted iron oxide catalyst. The sec-
ond stage is a low-temperature shift (180–270  C)
N2 þ 3H2 ⇆ 2NH3
in which the remaining carbon monoxide is oxi- (25)
dized to CO2 over a copper/zinc catalyst, leaving a DH ¼
92:2 kJ=mol NH3
few tenths of a percent in the syngas. Typical
catalyst lifetimes for both high temperature and Since the reaction is exothermic, maximum
low temperature shift catalysts are 3–5 years. conversion at equilibrium occurs at high pressure
Sulfur tolerant (“dirty shift”) catalysts also and low temperature. Ammonia synthesis is
have been developed. These catalysts can handle achieved in catalytic reactors at pressures ranging
242 Gas to Liquid Technologies

from 150 to 345 atm, at a minimum temperature of and favors low temperatures. Since the shift reac-
430–489  C. A maximum temperature of ~500  C tion does not approach completion in the reformer
results from the balance between increasing reac- (usually there is 10–15 volume % CO, dry basis,
tion kinetics and decreasing the equilibrium in the reformer effluent), further conversion of CO
ammonia concentration as the temperature is performed using external shift conversion
increases. Thermodynamics suggest that a low catalysts.
process temperature is favored. However, the
kinetics of the reactions dictates that high temper- Reactors and Catalysis for Ammonia Synthesis
ature is required. Nitrogen is usually fed as air, but The basic design of an ammonia synthesis reactor
in some plants (e.g., partial oxidation plants) the is a pressure vessel with sections for catalyst beds
N2 is a purified gas from an associated air separa- and heat exchangers. Over the years, there have
tion facility. In terms of the hydrogen source, been many different designs proposed and
production methods, product recovery, and over- constructed for commercial operation such that
all plant engineering, there are many process var- today, ammonia reactors are classified by flow
iants for ammonia production. type (radial, axial, or cross flow) and cooling
About 50% of the hydrogen in ammonia comes method (quench or indirect) used.
from steam. High temperatures and low pressures Axial flow reactors are essentially top-to-
favor the highly endothermic methane reforming bottom flow reactors. The design is comparatively
reaction. Higher pressures tend to lower the meth- simple; however, a fairly large pressure drop
ane conversion. In industrial reformers, the develops across the catalyst bed. A radial flow
reforming and shift reactions result in a product configuration feeds gas into an annular region
gas composition that closely approaches equilib- between the reactor wall and the outer surface of
rium. However, the following side reactions pro- the catalyst bed. Gas flows through the bed and
duce carbon in the steam reformer: exits out a central collection tube. This design
minimizes pressure drop across a shallow bed
Boudouard Coking with a large surface area. Radial flow reactors
tend to be tall vessels with relatively small diam-
eters. The cross-flow reactor configuration has a
2CO ⇆ CðsÞ þ CO2 similar principle in that gas is introduced along
(7)
DH ¼
83:3 kJ=mol CO one side of the reactor and is collected radially
across the reactor by a collector on the other side.
CO Reduction Reactant gas is preheated either by circulating
it through heat exchangers or by using water to
produce steam. However, removing the heat gen-
CO þ H2 ⇆ CðsÞ þ H2 O erated from the exothermic synthesis reaction to
(26)
DH ¼
125:5 kJ=mol C maintain control of the reaction temperature is a
challenge. Quench converters are designed to
Methane Cracking introduce cool reactant gas at various points
along the length of the catalyst bed. Interbed
CH4 ⇆ CðsÞ þ 2H2 heat exchangers also can be used to remove heat
(27) at specific intervals along the bed, effectively sep-
DH ¼ 80:6 kJ=mol C
arating the bed into multiple synthesis zones, or
continuously along the bed with cooling tubes.
The molar steam to carbon ratio in the reformer These indirectly cooled designs allow for efficient
is usually between 2 and 6, depending on the recovery of reaction heat that can be used in other
feedstock and process operating conditions. parts of the process.
Excess steam is used to prevent coking in the Early designs based on axial flow were easier
reformer tubes. The shift reaction is exothermic to build but not as efficient as the more complex
Gas to Liquid Technologies 243

radial flow reactors that have more recently been to increase the dissociative sticking probability of
designed. One example of a radial flow design, by N2 on the Fe sites and thereby increase catalytic
Haldor Topsøe, uses two radial beds with quench activity [67].
gas injection between them. A similar radial flow The most effective catalysts from a process
design uses an interbed heat exchanger in the first perspective are those that have the highest rate of
catalyst bed. Cold ammonia synthesis gas is intro- conversion at the lowest temperatures. Other cat-
duced from the bottom of the reactor through the alysts are being developed such as promoted Ru
second catalyst bed, then through the heat on high surface area graphite supports, and these
exchanger in the first catalyst bed. The cold gas recently commercialized catalysts offer the possi-
flow through the second bed also provides indirect bility of greatly decreasing synthesis temperatures
heat exchange. An additional heat exchanger is and pressures with improved activity at high
located at the bottom of the reactor to cool the ammonia concentrations. The Kellogg Advanced
reactor gases. Ammonia Process is based on a Ru catalyst that is
The Kellogg four bed axial flow reactor system claimed to be 40% more effective than Fe cata-
is a good example of an axial flow reactor design. lysts, because synthesis efficiency is increased by
This quench reactor consists of four catalyst beds lowering the process pressure from 150 to 90 atm.
held on separate grids. Quench gas is introduced With effective gas cleanup and conditioning,
in the spaces between the bed, and a heat commercial ammonia synthesis catalyst lifetimes
exchanger is located at the top of the vessel. of 5–8 years can be achieved in most cases and
Another type of Kellogg reactor is a cross-flow potentially up to 14 years. Ammonia formation
reactor design where gas flows through the cata- increases as the process pressure is increased. The
lyst bed perpendicular to the vessel axis. It is optimum H2/N2 ratio is near 2 at high space
available in both quench and indirectly cooled velocities and approaches 3 at low space veloci-
versions. ties as equilibrium becomes more dominant.
The ammonia synthesis catalyst is commonly Space velocities in commercial ammonia synthe-
iron, such as magnetite, which may be promoted sis vary from 12,000/h at 150 atm to 35,000/h at
with aluminum, potassium, and/or calcium. 790 atm.
Finely divided iron is pyrophoric, therefore Fe Per pass conversion on the order of 10–35% is
catalysts are reduced to metallic form in situ. typically achieved. Ammonia is recovered from
Porosity is developed as the catalysts are reduced. the synthesis loop by cooling to condense the
During reduction, oxygen is removed from the synthesis gas at process pressures. The liquid
magnetic crystal lattice without shrinkages. ammonia is separated from the gas effluent,
Metallic Fe is formed with essentially the same which is recycled back through to reactor. Earlier
porous structure as the magnetite precursor. The plant designs used air- or water-cooling for
production and preservation of this highly porous ammonia recovery. Modern synthesis plants use
structure during reduction of the ammonia synthe- refrigeration to condense out the ammonia. The
sis catalyst precursor leads to highly active cata- ammonia recovery is not extremely efficient so
lysts. The addition of structural promoters the recycled gas typically contains 4% ammonia
facilitates the formation of highly porous plus any inert gasses (Ar, He, CH4, etc.) that may
metallic iron. be in the process stream. Purging some of the gas
Catalyst activity is not directly associated with in the recycle loop before it is recycled minimizes
Fe surface area but is related to complex interac- inert gas concentrations, and flashing the liquid
tions of the promoters. Alkali metal promoters in ammonia in a pressure letdown step releases any
ammonia synthesis catalysts are necessary to dissolved gases.
attain high activity. Potassium is the most effec- Clearly, the ammonia synthesis process con-
tive alkali promoter, while Li and Na are poor sists of many complex unit operations apart from
promoters and are not used commercially. Potas- the actual synthesis loop. The way in which these
sium is thought to interact with the Fe crystallites process components are combined with respect to
244 Gas to Liquid Technologies

mass and energy flow has a major influence on costs has been the switch from electrically driven
efficiency and reliability. Many of the differences compression to use of steam-driven centrifugal
between various commercial ammonia processes compressors powered by waste heat. During the
lie in the way in which the above process elements past 30 years, many of the smaller, older, higher-
are integrated. cost plants in the world have been shut down.
Developments to improve energy efficiency or
overcome interruptions of natural gas supply have
Commercial Activities in Ammonia Synthesis been examined and some are finding application.
The entire ammonia industry was originally based The recovery of the hydrogen and ammonia from
on using coal as a feedstock. The industry has now the synthesis purge gas by a cryogenic unit or
moved toward natural gas as the main hydrocar- membrane is now standard and an additional 5%
bon feedstock. By 1990, only 14% of the world of ammonia can be produced from the same
ammonia capacity was based on coal or coke. amount of feedstock. The use of computer con-
Apart from a few plants operating in India and trols to improve overall plant performance is
South Africa, today the majority of coal-based increasing.
ammonia plants are found in China. More than Substantial improvements have been made
20 commercial ammonia synthesis processes have over the years in the energy efficiency of carbon
been described [68]. dioxide removal systems. The first large-scale
Commercial technology vendors for ammonia ammonia plants in the 1960s typically used mono-
include: ethanolamine as a solvent, with energy input of
over 50,000 kcal/kg-mol of carbon dioxide
• Johnson-Matthey removed, whereas today’s modern plants use
• Linde improvements such as the dual-step methyl
• Kellogg Brown and Root diethanolamine process that can reduce the energy
• Haldor Topsøe input to about 10,000 kcal/kg-mol or lower of
• Ammonia Casale carbon dioxide removed. As described above,
• Uhde improvements to efficiency also have been intro-
duced with new ammonia synthesis catalysts.
As a result of a surge of major developments in The current generation of large ammonia
ammonia process technology beginning in the late plants is much more fuel-efficient than plants
1960s, large 1000–1500 metric ton/day capacities constructed 20 or 30 years ago. A typical world-
became the industry standard for new plant con- scale plant constructed in the 1970s consumed
struction. Plants as large as 2000 metric ton/day about 42 million Btu of natural gas per ton of
are now common. The main ammonia process ammonia produced. Retrofitting such a plant to
contractors have each now developed “mega- improve fuel efficiency can reduce gas consump-
ammonia technology,” and are offering large tion to about 36 million Btu/ton; however, the
plant concepts from 3000 to 6000 metric ton/day newest generation of ammonia plants uses only
units. The world’s largest ammonia plant, in Al about 30 million Btu/ton of ammonia, are reported
Jubail, Saudi Arabia, is a collaboration between to be easier to operate, and have slightly lower
Saudi Basic Industries Corporation (SABIC) and conversion costs. Some newer plants also recover
Saudi Arabian Fertilizer Company. The plant, more than 1 million Btu/ton by generating elec-
called SAFCO IV, came onstream in 2007 and tricity, used elsewhere in the complex, from waste
has capacity of 3300 metric ton/day. This huge heat [70].
plant is the first application of a new ammonia An additional minor trend has been to con-
process from Uhde, referred to as the Dual Pres- struct an ammonia plant at a location where a
sure Process, and utilizes Johnson Matthey’s hydrogen off-gas stream is available from a
KATALCO™ catalysts [69]. A major factor nearby methanol or ethylene operation (e.g.,
resulting in much lower average production Canadian plants at Kitimat, British Columbia,
Gas to Liquid Technologies 245

and Joffre, Alberta). Gas consumption at such and the ethylene may react further to produce
production sites ranges from 25 to 27 million CO2, and the single pass combined yield of ethyl-
Btu/ton of ammonia, depending on specific cir- ene and ethane is limited to about 25%. With
cumstances. Perhaps more important, the capital better catalysts, such as SrO/La2O3 and
cost of such a plant is only about 50% of the cost Mn/Na2WO4/SiO2, a C2 selectivity of about
of a conventional plant of similar capacity 80% can be achieved, but at a low methane con-
because only the synthesis portion of the plant is version of 20%. About half of the C2 product is
required. ethane and half ethylene. The best C2 selectivities
are almost always achieved under oxygen limiting
Direct Conversion of Natural Gas to Liquids conditions. Because the reaction is exothermic, a
Substantial capital cost could be avoided if meth- zone within the catalyst bed may be 150–300  C
ane could be directly converted to methanol, or hotter than the external temperature and heat man-
other high-volume liquid fuels or chemicals, with- agement is a serious problem. This is complicated
out the intermediate production of syngas. Several by the fact that metals normally used for construc-
direct processes have been investigated over the tion of reactors catalyze the combustion of
years, including methane coupling to produce eth- methane.
ane, which can be oligomerized to give gasoline in The yield of ethylene can be improved consid-
a reactor similar to that used in the MTG process. erably by operating in a recycle mode with con-
Another consideration has been the direct produc- tinuous removal of ethylene and by applying
tion of aromatics, using an oxidative coupling various techniques for separating products from
membrane. Another option is the selective oxida- reactants including cryogenic, membrane, and
tion of natural gas to produce fuels directly. Direct adsorptive separations. For economic reasons,
methods would need to have a distinct economic either molecular sieve or membrane separation
advantage over indirect methods, but to date no systems are most likely to be employed, both of
direct processes have progressed to a commercial which have severe mass transport limitations not
stage. Product yields are generally small while suitable for scale-up. A variation of this recycle
operating in the single pass mode, which makes system is to use a zeolite catalytic reactor to con-
separations difficult and costly. vert ethylene to benzene and toluene. In principle,
aromatic yields in excess of 70% could be
Oxidative Coupling to Light Hydrocarbons achieved; however, to achieve these yields the
In the oxidative coupling reaction, methane and conversion of ethane to ethylene must be highly
oxygen react over a catalyst at elevated tempera- selective.
tures (700  C) and normal pressures to form eth-
ane as a primary product and ethylene as a Selective Oxidation of Natural Gas to Methanol
secondary product. British Petroleum (now BP) Selective alkane oxidation by transition metal
has described experiments at higher temperatures complexes in solution has been the focus of sub-
(1100  C) and short residence times, at which stantial effort since the 1970s. However, the low
conditions methane reacts with oxygen to produce activity of alkanes and the typically higher activ-
syngas as well as C2+ hydrocarbons in addition to ity of the desired products make this process a
syngas. Several catalysts including zirconia gave great challenge. In 1998, Periana et al. of
C2+ selectivities of over 50%. Catalytica Inc. reported developing an effective
Work on the lower temperature methane cou- catalyst for selective oxidation of CH4 to CH3OH
pling reaction has been done by Phillips Petro- in high yield. The initial catalyst was dichloro-
leum, Texas A&M University, Akzo Chemie, (Z2-(2,20 -bipyrimidyl)platinum(II) complex,
Amoco (now BP), the University of California at (bpym)PtCl2. In well-dried sulfuric acid (102%),
Berkeley, the University of Pittsburg, the Univer- CH4 at 3400 kPa and 220  C was converted to a
sity of Tokyo, Idmitsu Kosan, and Union Carbide mixture of CH3OSO3H (methyl bisulfate, which
among others. Unfortunately, both the methane presumably would be subsequently hydrolyzed to
246 Gas to Liquid Technologies

methanol) and CH3OH at 72% conversion with shops” for commercialization. Johnson Matthey
selectivity of 81% [71]. Although catalytic in Pt has acquired Davy Process technology and ICI,
complex, the reaction was very slow and required making the parent company a leader in reforming
over half as many moles of catalyst as moles of and methanol GTL technologies. The companies
methane converted. Another catalyst, Enichem and IFP/Axens have strong capabilities
(NH3)2PtCl2, system was more active but less particularly applicable to F-T GTL. Enichem has
stable. A complete cycle would require the regen- skills in catalysis, engineering, and the oil and gas
eration of concentrated sulfuric acid. The business. IFP/Axens has catalysis expertise, plus
Catalytica process is promising, because it poten- basic engineering, process development, licens-
tially has high yield at relatively low tempera- ing, and catalyst manufacturing [75]. The future
tures. On the other hand, it suffers from of methanol may be the status quo, because the
drawbacks including (1) the catalytic rate is too technology is already mature. Alternatively, GTL
low; (2) the separation cost is too high; (3) the methanol plants may expand in number (and pos-
solvent is too corrosive; (4) reoxidation of SO2 sibly size) given the anticipated increased avail-
with regeneration of concentrated sulfuric acid is ability of unconventional gas combined with high
too expensive; and (5) the catalyst is severely oil process.
inhibited by water.
Cold flame oxidation is another approach to
direct conversion of methane to methanol. Bibliography
A pressurized mixture of methane and oxygen is
reacted at moderate temperatures of 350–500  C. 1. Crane H, Kinderman E, Malhotra R (2010) A cubic
mile of oil. Oxford University Press, New York
The reaction mixture is very fuel rich, with meth-
2. United States Environmental Protection Agency
ane to oxygen ratios of about 20:1. The main (1998) Compilation of emission factors AP-42, v1,
chemical reaction is the oxidation of methane to 5th edn, supplement D
methanol. However, further oxidation of metha- 3. United States Environmental Protection Agency
(2010) Compilation of Emission factors AP-42, vl 1,
nol to formaldehyde often takes place simulta-
5th edn, supplement E, corrected. Calculation for low-
neously. The University of Manitoba reported sulfur No. 6 fuel oil
90% selectivity for methanol at a single pass con- 4. United States Environmental Protection Agency
version of 7.5% in an isothermal process. (1998) Compilation of emission factors AP-42, v1,
5th edn, supplement E. Calculation for medium vola-
Although this is among the best results reported
tile bituminous coal
to date, the yield of methanol from this process is 5. Rajnauth J, Ayeni K, Barrufet M (2008) Gas transpor-
too low to be of commercial interest. tation: present and future. In: CIPC/SPE gas technol-
The direct oxidation of methane to methanol ogy symposium 2008 joint conference, Calgary,
16–19 June
remains a challenge in spite of more than
6. Bellussi G, Zennaro R. New developments: energy,
100 years of effort to accomplish this. Progress transport, sustainability. In: Encyclopaedia of hydro-
continues to be made. It appears that the use of Cu carbons, vol III, Ch 2.6, pp 161–182, EniChem
containing zeolites offer an ability to accomplish 7. Smith R, Asaro M, Naqvi S (2008) Fuels of the future:
technology intelligence for coal to liquids strategies.
this [72, 73]. For example, a continuous selective
Process Economics Program, SRI Consulting, Menlo
oxidation in a flow reactor using a CuFe-ZSM-5 Park
catalyst with H2O2 shows a > 92% selectivity 8. Yost C, DiNapoli R (2003) Benchmarking study com-
although at a low 0.5% conversion [74]. pares LNG plant costs. Oil Gas J 101(15):56–59
9. Nielsen R (2001) Fundamentals of mixed refrigerant
compared to conventional refrigeration are discussed
in “Ethylene Plant Enhancement”. PEP report 29G,
Future Directions SRI Consulting, Menlo Park
10. Low WR, Andress D, Houser C (1997) Method of load
distribution in a cascaded refrigeration process. US
The immediate future of F-T and DME gas to
5611216 to Phillips Petroleum Company
liquids would seem to be in their implementation. 11. Houser C, Yao J, Andress D, Low WR (1997) Effi-
Acquisitions and ventures are creating “one stop ciency improvement of open-cycle cascaded
Gas to Liquid Technologies 247

refrigeration process. US 5669234 to Phillips Petro- reforming of methane. Chem Eng Sci
leum Company 63(11):2966–2979
12. Delong BW (1987) Method for cooling normally gas- 26. Li J (2015) Praxair’s oxygen transport membrane. In:
eous material. US 4680041 to Phillips Petroleum Technology for syngas and power applications,
Company DOE/NETL carbon capture meeting, Pittsburgh,
13. Netzer D, Nielsen R (2003) Baseload liquefied natural 26 June 2015
gas by Cascade refrigeration. PEP review 2003-15. 27. http://www.statoil.com/en/OurOperations/TerminalsRe
SRI Consulting, Menlo Park fining/Tjeldbergodden/Pages/default.aspx. Accessed
14. Huffman GP, Feeley III TJ (2000) Fuel science in the 11 Jan 2011
year 2000: where do we stand, where do we go from 28. Davis BH (2001) Fischer-Tropsch synthesis: current
here? I: power generation and related environmental mechanism and futuristic needs. Fuel Process Technol
concerns – DOE’s fine particulate and air toxics research 71:157–166
program: responding to the environmental challenges to 29. Robert C, Brady RC III, Pettit R (1981) Mechanism of
coal-based power production in the 21st century. Pre- the Fischer-Tropsch reaction. The chain propagation
prints of the Division of Fuel Chemistry of the American step. J Am Chem Soc 1981:287–1289
Chemical Society 45(1):108–112. 2000 Spring confer- 30. Oukaci R, Singleton AH, Goodwin JG Jr (1999) Com-
ence of the American Chemical Society, San Francisco parison of patented Co F–T catalysts using fixed-bed
15. Reprinted from Hydrocarbon Engineering (2006) and slurry bubble column reactors. Appl Catal A Gen
Topsoe_synthesis_g#6D6FFA1.ashx.pdf. Accessed 186:129–144
2 Apr 2011 31. Manzer L, Schwarz S (2002) Fischer-Tropsch pro-
16. Christensen TS, Østberg M, Bak Hansen J-H cesses using catalysts on mesoporous supports. US
(2001) Process demonstration of autothermal reforming 2002052289 to Conoco
at low steam-to-carbon ratios for production of synthe- 32. Godorr S (2011) GTL technology advancements. In:
sis gas. In: AIChE annual meeting, Reno, 4–9 Nov World petroleum conference
17. Wesenberg MH (2006) Gas heated steam reformer 33. Das TK, Zhan X, Li J, Jacobs G, Dry ME, Davis BH
modelling. PhD thesis (2007) Ffischer-Tropsch synthesis: kinetics and effect
18. Aasberg-Petersen K, Thomas S, Christensen TS, Char- of water for a Co-Al2O3 catalsyt, Fischer-Tropsch
lotte Stub Nielsen CS, Dybkjær I (2003) Recent devel- synthesis, catalsyts and catalysis (B.H. Davis,
opments in autothermal reforming and pre-reforming M.L. Occelli, eds.) Stud Surf Sci 163:289–314
for synthesis gas production in GTL applications. Fuel 34. http://www.zawya.com/projects/project.cfm/pid210307
Process Technol 83(1–3):253–261 061231?cc. Accessed 11 Jan 2011
19. Loock S, Ernst WS, Thomsen SG, Jensen MF 35. Spath PL, Dayton DC (2003) Preliminary screening –
(2005) Improving carbon efficiency in an auto-thermal technical and economic assessment of synthesis gas to
methane reforming plant with gas heated heat fuels and chemicals with emphasis on the potential for
exchange reforming technology. Paper no. O96-001, biomass-derived syngas. NREL/TP-510-34929
7th World congress of chemical engineering, Glasgow 36. Fisher IA, Bell AT (1998) In situ infrared study of
20. Tsuru T, Yamaguchi K, Yoshioka T, Asaeda M (2004) methanol synthesis from H2/CO over Cu/SiO2 and
Methane steam reforming by microporous catalytic Cu/ZrO2/SiO2. J Catal 178(1):153–173
membrane reactors. AICHE J 50(11):2794–2805 37. Chinchen GC, Denny PJ, Parker DG, Spencer MS,
21. Carolan MF, Chen CM, Rynders SW (2003) Develop- Whan DA (1987) Mechanism of methanol synthesis
ment of the ceramic membrane ITM syngas/ITM from CO2/CO/H2 mixtures over copper/zinc oxide/
hydrogen process. Fuel Chem Div Preprints 48(1):344 alumina catalysts: use of 14C-labelled reactants. Appl
22. Repasky J, McCarthy D, Armstrong P, Carolan Catal 30(2):333–338
M (2014) ITM technology for carbon capture on nat- 38. Grabow LC, Mavrikakis M (2011) Mechanism of
ural gas and hybrid power systems. In: Workshop on methanol synthesis on cu through CO2 and CO hydro-
technology pathways forward for carbon capture & genation. ACS Catal 1(4):365–384
storage on natural gas power systems, Washington, 39. Tijm PJA, Waller FJ, Brown DM (2001) Methanol
DC, 22 Apr 2014 technology developments for the new millennium.
23. Robinson ET (S), Sirman J, Apte P, Gui X, Bulicz T R, Appl Catal A Gen 221:275–282
Corgard D, Hemmings J (2005) Development of OTM 40. Liu J, Wei R, Zhang Y, Xu R, Li Z (2009) Preparation
syngas process and testing of syngas derived ultra- of Cu/ZnO/Al2O3 catalysts for methanol synthesis by
clean fuels in diesel engines and fuel cells. improved two-step coprecipitation method. Gongye
DE-FC26-01NT41096, final report Cuihua 17(7):22–25, C.A. 2009:1625358
24. Caro J, Wang H, Noack M, Koelsch P, Kapteijn F, 41. Baltes C, Vukojevic S, Schüth F (2008) Correlations
Kannelopolous N, Nolan J (Oct. 2007) Manufacture between synthesis, precursor, and catalyst structure
of composite membranes and their use for selective and activity of a large set of CuO/ZnO/Al2O3 catalysts
partial oxidation reactions of hydrocarbons. EP for methanol synthesis. J Catal 258(2):334–344
1847311 to Universität Hannover 42. Kaluza S, Behrens M, Schiefenhoevel N, Kniep B,
25. Dupont V, Ross AB, Knight E, Hanky I, Twigg MV Fischer R, Schloegl R, Muhler M (2011) A novel
(2008) Production of hydrogen by unmixed steam synthesis route for cu/ZnO/Al2O3 catalysts used in
248 Gas to Liquid Technologies

methanol synthesis: combining continuous consecu- 58. Pavone T (2003) Jumbo dimethyl ether production
tive precipitation with continuous aging of the precip- process via Toyo technology. PEP review 2003-9.
itate. ChemCatChem 3(1):189–199 SRI Consulting, Menlo Park
43. Lurgi brochure 0312e_MegaMethanol.pdf 59. www.oil-gasportal.com/dimethyl-ether-dme-produc
44. Smith R, Naqvi S, Asaro M (2008) Fuels of the future: tion-2/?print-fdp
technology intelligence for coal to liquids strategies. 60. Prabowo B, Yan M, Wyamiro M, Setyobudi RH,
SRI Consulting, Menlo Park Biddinika MK (2017) State of the art of global
45. Lewnard JJ, Hsuing TH, White JF, Brown DM dimethyl ether production and its potential application
(1990) Single-step synthesis of dimethyl ether in a in Indonesia. Proc Pakistan Acad Sci B Life Environ
slurry reactor. Chem Eng Sci 45(8):2735–2741 Sci 54:29–39
46. Ohno Y, Yagi H, Inou N (2007) Slurry phase DME 61. http://www.ccfgroup.com/newscenter/newsview.php?
synthesis technology: 100 ton/day demonstration Class_ID=Dooooo&Info_ID=20160311065
plant operation and scale-up study. Stud Surf Sci 62. http://www.methanex.com/ourcompany/locations_
Catal 167:403–408 newzealand.html. Accessed 4 Apr 2011
47. Haugaard J, Voss B (2001) Process for the synthesis of 63. Apanel G (1999) Liquid hydrocarbons from synthesis
a methanol/dimethyl ether mixture from synthesis gas. gas. PEP report 191A. SRI Consulting, Menlo Park
US 6191175 to Haldor Topsøe 64. Tabak SA, Yurchak S (1990) Conversion of methanol over
48. Naqvi S (2002) Dimethyl ether as alternative. Fuel ZSM-5 to fuels and chemicals. Catal Today 6(3):307–327
PEP report 245. SRI Consulting, Menlo Park 65. Methanax Corportion (2015) Annual information
49. Kang S-H, Bae JW, Kim H-S, Dhar GM, Jun K-W form, 9 March 2015
(2010) Enhanced catalytic performance for dimethyl 66. Jennings JR (2010) Catalytic ammonia synthesis: fun-
ether synthesis from syngas with the addition of Zr or damentals and practice. Barnes & Noble
Ga on a Cu
ZnO
Al2O3/g-Al2O3 bifunctional cata- 67. Spath P, Dayton D (2003) Bioproducts from syngas.
lyst. Energy Fuel 24(2):804–810 Syngas_products.pdf. Accessed 2 Apr 2011
50. Bhatt B L, Schaub E, Heydorn E (1993) Recent devel- 68. Ullmann’s encyclopedia of industrial chemistry (2002)
opments in slurry reactor technology at the LaPorte “Ammonia” published online 15 Dec 2006. https://doi.
alternative fuels development unit. In: International org/10.1002/14356007.a02_143.pub2. Accessed 4 Apr
technical conference on coal utilization & fuel sys- 2011
tems, 26–29 Apr, pp 197–208 69. Shah J (2007) SAFCO IV: catalyst start-ups in the
51. Peng X-D, Toseland B, Underwood T (1997) A novel world’s largest ammonia plant. 20th AFA international
mechanism of catalyst deactivation in liquid phase annual technical conference, Tunisia, 5_03
synthesis gas-to-DME reactions. In: Bartholomew C, John_BRIGHTLING_ Johnson Matthey
Fuentes GH (eds) Catalyst deactivation, vol 111. Catalysts_U.K.pdf. Accessed 4 Apr 2011
Elsevier, pp 175–182 70. Liu H (2013) Ammonia synthesis catalysts: innovation
52. Peng X-D (2002) Catalyst activity maintenance for the and practice. Barnes & Noble
liquid phase synthesis gas-to-dimethyl ether process. 71. Xu X, Fu G, Goddard III WA, Periana RA
Part II: development of Aluminum phosphate as the (2004) “Selective oxidation of CH4 to CH3OH using
dehydration catalyst for the single-step liquid phase the Catalytica (bpym)PtCl2 catalyst: a theoretical
syngas-to-DME process. DOE contract DE-FC22- study” in Studies in surface science and catalysis, nat-
94PC93052, final report ural gas conversion VII. In: Proceedings of the 7th
53. Peng X-D (2002) Kinetic understanding of the syngas natural gas conversion symposium, pp 147, 499–504
-to-DME reaction system and its implications to pro- 72. Kalamaras C, Palomas D, Bos R, Horton A, Crimmin M,
cess and economics. DOE contract DE-FC22-94 Hellgardt K (2016) Selective oxidation of methane to
PC93052, topical report methanol over Cu- and Fe-exchanged zeolites: the effect
54. Tijm P J (2003) Development of alternative fuels and of Si/Al molar ratio. Catal Lett 146:483–492
chemicals from synthesis gas. DOE contract number 73. Narsimhan K, Iyoki K, Dinh K, Roman-Leshkov
FC22-95PC93052, final report Y (2016) Catalytic oxidation of methane into metha-
55. Smith R (September 2009) Dimethyl ether (DME) from nol over copper-exchanged zeolites with oxygen at
coal. PEP report 245B. SRI Consulting, Menlo Park low temperature. ACS Central Sci 2:424–429
56. Ogawa T, Inoue N, Shikada T, Ohno Y (2003) Direct 74. Xu J, Armstrong RD, Shaw G, Dummer NF, Freakley
dimethyl ether synthesis. J Nat Gas Chem 12:219–227 SJ, Taylor SH, Hutchings GJ (2016) Continuous selec-
57. The Ministry of Industry, Energy and Tourism; tive oxidation of methane to methanol over Cu-and
Orkustofnun/The National Energy Authority, The Fe-modified-ZSM-5 catalysts in a flow reactor. Catal
Innovation Center Iceland; Mitsubishi Heavy Indus- Today 270:93–100
tries, Ltd.; Mitsubishi Corporation; Hekla hf.; 75. Zennaro R, Hugues F, Caprani E (2006) The Eni –
NordicBlueEnergy (2010) A feasibility study report IFP/Axens GTL technology: from R&D to a success-
for a DME project in Iceland. IDME Project Feasibil- ful scale-up. In: DGMK – SCI conference on synthesis
ity Study – 2009. Accessed 4 Apr 2011 gas chemistry, Dresden
 Combined heat and power (CHP) The term
Natural Gas Power used for fossil-fired power plants, which, in addi-
tion to their primary product, electric power
Raub W. Smith1 and S. Can Gülen2 delivered to the grid, also supply a secondary
1
GE Power, Schenectady, NY, USA product in terms of useful thermal energy.
2
Bechtel Infrastructure & Power Inc., Reston, VA, Combustor A mechanical device to facilitate
USA controlled mixing and reaction of an oxidizer
(in almost all cases, air) and a fuel (in
almost all cases, a pure hydrocarbon or a mix-
Article Outline ture thereof in gaseous or liquid phase) to
generate high-temperature gaseous product
Glossary for expansion in a turbine and useful shaft
Defining the Subject work generation.
Introduction Compressor A mechanical device that increases
A Brief History of Natural Gas the pressure of a gas by reducing its
Gas Turbine Power Plants volume. There are different types of compres-
Future Directions sors, e.g., axial, radial, and reciprocating,
Bibliography which are suitable to different types of operat-
ing regimes.
Glossary Efficiency Unless specified otherwise, the
thermal efficiency of a power generating
Brayton cycle The thermodynamic cycle system, which is the dimensionless ratio of
describing the operation of a gas turbine. In a generated kWh of electricity to the amount
combined cycle, it is the topping cycle due to of energy required to generate it. It is the
its relative position vis-à-vis Rankine cycle on inverse of the heat rate with a suitable conver-
a temperature-entropy surface. sion factor.
Carnot cycle Also known as the Carnot engine, Emissions Gases and solid particles (usually
it is the embodiment of the second law of undesirable) released into the air as byproducts
thermodynamics in the form of a theoretical of a combustion process (e.g., in the boiler of a
cycle comprising two isentropic and two iso- fossil-fired power plant, gas turbine combus-
thermal processes. No heat engine operating in tor, or other internal combustion engine) to
a thermodynamic cycle can be more efficient create electric power or propel a vehicle).
than the corresponding Carnot engine defined Firing temperature The temperature of the gas
by the constant mean-effective heat addition turbine combustor exhaust gas at the inlet to
and heat rejection temperatures. the first stage rotor, which is the starting point
Cogeneration See Combined Heat and Power of useful shaft work generation.
(CHP). Gas turbine A prime mover or internal combus-
Combined cycle power plant A fossil-fired tion engine comprising a compressor, combus-
power plant that combines two types of prime tor, and an expander connected via a common
movers, usually one or more gas turbines and shaft, through which air is compressed, burned,
one or more steam turbines, whose operation is and expanded to generate useful shaft work for
governed by their respective thermodynamic electric generation (or thrust in an aircraft jet
cycles, i.e., Brayton and Rankine. engine.)

© Springer Science+Business Media, LLC, part of Springer Nature 2020 249

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_100
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_100-3
250 Natural Gas Power

Generator A device that converts the mechani- Defining the Subject

cal shaft power generated by a prime mover
into electrical power. Natural gas is an important fossil fuel that has
Global warming The apparent increase in the played an increasingly significant role in world-
average temperature of the earth’s near-surface wide electric power generation since the 1980s.
air and oceans since the mid-twentieth century The key driver underlying the importance of nat-
and its projected continuation (per Wikipedia). ural gas as a vital enabler of modern living has
Greenhouse effect The containment of heat been its relative advantage vis-à-vis other fossil
from solar radiation striking the earth’s surface fuels in terms of emissions and pollutants. In
due to the earth’s atmospheric “greenhouse” comparison to coal, the primary fossil fuel used
gases such as carbon dioxide and methane. for electric power generation in the world on a
These gases absorb and emit radiation within constant consumption basis, natural gas emits
the thermal infrared range and are believed to nearly 45% less CO2 and 80% less nitrogen
be a primary cause of global warming. oxides (NOx) with negligible amounts of sulfur
Heat rate Amount of energy required to generate oxides, particulates, and mercury.
1 kWh of electricity. It is the inverse of the The environmental advantages of natural gas
thermal efficiency with a suitable conversion are further amplified by the significantly higher
factor. thermal efficiency of the power plants that burn it
Heat recovery steam generator (HRSG) Also for electric power generation in comparison to
known as the heat recovery boiler (HRB), other variants of oil or coal. Natural gas-burning
HRSG is a cross-flow tubular heat exchanger modern gas turbines can readily reach efficiencies
that recovers the exhaust heat from a prime of 56–57% in combined cycle configurations
mover (e.g., a gas turbine) and produces (annual average field efficiency, ISO rating effi-
steam at high pressure and temperature that is ciencies are much higher). This is well above the
used in a steam turbine for additional power average efficiency of existing coal burning steam
generation. HRSG is the key equipment that turbine plants (e.g., low 30s) as well as the most
“combines” gas and steam turbines in a com- advanced ultra-supercritical designs thereof (e.g.,
bined cycle power plant. low 40s).
Heating value The thermal energy produced by With the beginning of the twenty-first century,
completely burning a unit mass of fuel in a economic projections pointed to natural gas as the
combustor to produce carbon dioxide and fastest-growing fuel for worldwide electric power
water. If the water is in a gaseous phase, the generation, with an expected average annual rate
heating value is referred to as net or lower of growth of about 3.7% between 2005 and 2030.
heating value (LHV). If the water is in a liquid During this period natural gas will account for
phase, the heating value is referred to as gross almost one fourth of the world’s total net electric-
or higher heating value (HHV). ity generation [1]. While such long-term forecasts
Rankine cycle The thermodynamic cycle are highly susceptible to uncertainties such as the
describing the operation of a steam turbine. In impact of potential governmental policies or leg-
a combined cycle, it is the bottoming cycle due islation limiting the use of fossil fuels in favor of
to its relative position vis-à-vis Brayton cycle nuclear and renewable sources of energy (largely
on a temperature-entropy surface. driven by concerns over global warming), it is
Steam turbine A prime mover or the power gen- almost a foregone conclusion that natural gas
erating part of an external combustion engine will continue to be an important part of the world’s
comprising one or more sections connected via energy and electric power generation portfolio in
a common shaft, through which steam flows, the foreseeable future.
expands, and discharges to a condenser to gen- Given this impact of natural gas on future
erate useful shaft work for electric generation electrical production, it is important to establish
or propulsion. a basic understanding of the history, current state
Natural Gas Power 251

of the art, and future possibilities of power gener- gas that results from the decomposition of waste
ation using natural gas combustion in order to materials deposited into large landfills. Other bio-
have a complete picture of future energy gas formation mechanisms are sludge digestion in
sustainability. the tanks of sewage treatment plants (sewage gas)
and anaerobic fermentation of agricultural waste.
2016 Update Biogas formation takes place near the earth’s sur-
Six years have passed since the original draft of this face in the absence of oxygen, and the resulting
chapter. The projections made in the preceding gas usually leaks to the atmosphere. Controlled
paragraphs were proven to be largely accurate. capture of landfill gas and other biogases for uti-
For instance, in the United States, natural gas- lization in electric power generation is a promis-
fired electricity generation reached record levels ing technology. (Uncontrolled energy release
in 2016 and even surpassed coal-fired generation (explosion) from the trapped landfill gas leading
in several months. According to the US Energy to death and destruction in the poverty-stricken
Information Administration (EIA), it is ultimately areas of the world is an unfortunate event that
expected to provide 34% of US electricity genera- appears in the news from time to time.) Biogas
tion by the end of the year. According to EIA, “the utilization is also important because the main con-
natural gas share of power generation is expected to stituent methane itself is a principal greenhouse
decline for several years after 2016 as it competes gas (GHG). In fact, methane is nearly 20 times
with renewables and as natural gas prices rise [until more effective than CO2 in trapping heat emanat-
about 2020]. [Steadily climbing] over the next two ing from the earth’s surface. Therefore, combus-
decades, [n]atural gas is projected to regain the tion of methane (with resultant CO2 emissions)
largest share in the electricity mix by 2024 and that otherwise would leak into the atmosphere is a
maintain that position through 2040.” preferable trade-off from GHG reduction point
of view.
Coal mine gas is another type of natural gas
Introduction that is released during pit coal mining. Depending
on the time of capture and method, there are
Natural gas is a fossil fuel comprising hydrocar- different types. The most common types are coal
bons, primarily methane (CH4), up to 90% on a bed methane (CBM) and coal mine methane
volumetric basis. It differs from the other fossil (CMM). The former is largely pure methane, and
fuels in that it is naturally available as a gas (hence its production is independent of coal mining. The
the name) as opposed to a solid (i.e., coal) or latter is released during active mining and presents
liquid (i.e., oil). Like other fossil fuels, the basic a significant danger of violent explosion to
mechanism (thermogenic) of natural gas forma- miners. Proper ventilation of mine shafts to
tion involves the extremely prolonged high- remove and vent CMM into the atmosphere is
pressure compression of organic matter remains essential for the safety of mining operation. The
(e.g., plants, animals that lived millions of years methane content of CMM is variable (25–60%)
ago on earth). The gaseous phase is the result of and can change suddenly, which might be prob-
extremely high temperatures associated with the lematic for its use as a fuel. A third type of coal
depth of the organic matter, typically 1–2 miles mine gas is abandoned mine methane (AMM),
below the earth’s crust. At even deeper levels, which can seep from mines that are no longer
natural gas consists almost entirely of pure active. Annual coal mine methane emissions
methane. account for about 8% of total anthropogenic meth-
Another (biogenic) mechanism for forming ane emissions or about 28 billion cubic meters of
natural gas is the chemical transformation of carbon dioxide equivalent [2]. Utilization of
organic matter by microorganisms. The resulting CMM and AMM for power generation is critical
gas contains methane and carbon dioxide. A well- to the prevention of release of a GHG into the
known example of this mechanism is the landfill atmosphere.
252 Natural Gas Power

Natural Gas Power, Table 1 Composition and heating values of typical natural gases
Component (%) Coal bed methane Coal mine methane Landfill gas Biogas Average NG
O2 2.1 12.6 0–1 0–2
N2 8.2 46.8 2–5 0–10 1.6
CH4 85.9 40.0 45–60 50–75 93.1
CO2 40–65 25–50 1.0
CnH2n 3.8 4.3
LHV, kcal/nm3 (Btu/scf) 7,762 3,200 3,350–4,775 3,580–5,975 8,275
872 360 375–535 400–670 930

Energy content of the natural gas is measured combined cycle (CC) power plant. A typical CC
in terms of Btu (kJ or kcal) per lb (kg) or per power plant has a rated net efficiency of 58% on
standard cubic feet (cubic meters). In many prac- an LHV basis. In other words, for each 1 lb/s of
tical calculations, natural gas is assumed to be natural gas (about 23.5 ft3/s as 100% methane)
100% pure methane, which has an energy content that enters the combustor of the gas turbine (using
of 21,515 Btu/lb (50,044 kJ/kg or 11,953 kcal/kg) the formula 58 %  21 , 515 Btu/lb  1 lb/s
on a LHV basis or 914 Btu/ft3 (34,050 kJ/m3 or  1.05506 kW  s/Btu = 13 , 166 kW), about
8133 kcal/m3) at normal (standard) conditions 13.2 MW of electric power (net) is delivered to
(herein defined as 77
F (25
C) and 1 atmo- the grid. This is enough power to sustain nearly
sphere). The actual composition may vary 13,000 average homes for 1 h at an average power
depending on the source and treatment. Based on consumption of 1 kW.
the mean of over 6,800 samples of pipeline quality In 2005, natural gas accounted for 3.4 trillion
natural gas taken in 26 major metropolitan areas kWh of net electricity production for about 20%
of the United States [3], the methane content is of the world total (see Fig. 1). This contribution is
93.1% with a heating value of 930 Btu/ft3 projected to reach 8.4 trillion kWh in 2030 or
(8,275 kcal/m3). A sampling of composition and about 25% of the world total corresponding to a
heating value of most common natural gas vari- total natural gas consumption of 55.5 trillion
ants is given in Table 1. cubic feet (i.e., 35% of total world production of
Production and distribution companies mea- 158.6 trillion cubic feet). Using the calculation
sure natural gas in volumetric terms, i.e., in mul- above as a rough guide, this translates into an
tiples of cubic feet such as millions of cubic feet, average natural gas-fired power plant thermal effi-
trillions of cubic feet, etc. In power generation ciency of 56.4%. The natural gas reserves of the
calculations, where natural gas is the fuel burned world are estimated at about 6,186 trillion cubic
in the combustor of the gas turbine, the more feet with top 20 countries contributing nearly 90%
practical approach is to use the mass flow rate of it (see Fig. 2). When one considers the uncer-
(lb/s or kg/s) along with the LHV value or the tainty in the cited projections, it appears safe to
product of the two, which is referred to as heat say that there are ample deposits of natural gas to
consumption (MMBtu/h or kW). sustain electric power production well into the
The purpose of a power plant is to generate twenty-first century.
electric power. Therefore, the key performance As mentioned in the beginning, natural gas is a
metric of an electric power generator is the net superior fossil fuel in terms of emissions and
thermal efficiency, which is the ratio of the electric pollutants vis-à-vis other fossil fuels, i.e., coal
power measured at the generator terminals (after and oil. This is not to say that natural gas is totally
subtracting all the power that is spent to keep the harmless, especially when one considers the emis-
plant running) to the plant’s fuel energy consump- sions of CO2 – the most abundant anthropogenic
tion. The most efficient way to burn natural gas for greenhouse gas in the atmosphere and, as such,
electric power generation is in a gas turbine the main culprit in driving global warming.
Natural Gas Power 253

Natural Gas Power, Fig. 1 World net electricity generation. Renewables include hydroelectric (Source: Energy
Information Administration (EIA) [1])

Natural Gas Power, Fig. 2 World natural gas reserves, includes only those (Source: Energy Information Admin-
January 2008. Asia, Oceania, and Europe exclude coun- istration (EIA) [4])
tries that were part of the former USSR, whereas Eurasia
254 Natural Gas Power

A gas turbine combustor generates approximately context refers to the thermodynamic cycle of the
45 ft3/h of CO2 per ft3/h of natural gas burned. In compression-ignition (CI) IC engine and not the
other useful relative terms, this is approximately fuel.) While possible in theory, natural gas is
129,000 lbs (58,500 kg) of CO2 per billion Btu seldom used with external combustion technology
(1.06 million MJ) of energy input or 1,125 lbs per due to the pronounced superiority of the IC
MWh generated by a gas turbine (GT) in simple engines in efficiency, emissions, power density
cycle configuration (cf., 750 lbs in combined (i.e., favorable cost vs. size trade-off), and opera-
cycle configuration). Since not all power genera- tional flexibility. However, a fair number of for-
tion equipment performs at higher-rating effi- merly oil- or coal-fired boilers in steam turbine
ciency level, real averages would be higher, e.g., power plants run on natural gas after conversion.
1,300 and 800 for simple and combined cycles, Modern gas turbines are large air-breathing
respectively. For the projected natural gas-based turbomachines with extremely large power out-
electric power generation in 2030 cited above, at a put. For example, consider a 50 Hz (3,000 rpm)
rate of 1,300 lbs/MWh, this translates into 5 billion 300+ MW net power output unit as listed in a
metric tons of CO2 released into earth’s atmo- 2008 trade publication [6]. This machine ingests
sphere. (In order to get an idea about the magni- air at ISO conditions (15
C and 1 atmosphere at
tude of this number, consider that the estimated 60% relative humidity) at a rate of nearly
weight of carbon dioxide in the atmosphere was 1,550 lb/s (705 kg/s), compresses it to a pressure
about 2,163 billion metric tons before the indus- that is 18 times that of the ambient, and combusts
trial revolution.) This is approximately half of the it with about 35 lb/s (16 kg/s) of natural gas (100%
total 2030 projection of CO2 emission from natu- CH4) generating 312 MW net electric power for a
ral gas combustion, which itself is about 20% of net thermal efficiency of 39.3%. At the inlet to the
the 42.3 billion metric tons for all fossil fuels. expander section of the gas turbine, where they
While an important contributor to the green- produce useful shaft work, the combustion prod-
house effect, natural gas is extremely advanta- ucts are at nearly 1,500
C (2,732
F). This is well
geous from the perspective of two other above the melting point of the most advanced
significant combustion-related pollution phenom- superalloy materials that are used in the manufac-
ena, i.e., smog and acid rain. Smog (whose pri- ture of turbine expander components. In order to
mary constituent is ozone) is formed by a ensure the survival of the turbine parts under those
chemical reaction of carbon monoxide, nitrogen extreme conditions for thousands of hours of con-
oxides, volatile organic compounds, and heat tinuous operation, nickel-based superalloy com-
from sunlight [5]. Acid rain is formed when sulfur ponents are protected by thermal barrier coatings
dioxide and nitrogen oxides react with water (TBC) and internally cooled by using the “cold”
vapor and other chemicals in the presence of sun- air extracted from the turbine compressor.
light to form various acidic compounds in the air (Consider that without cooling and TBC, the
[5]. Natural gas combustion products are virtually first-stage vanes of a modern GT would survive
free of the two main culprits contributing to these barely 10 s before melting away.) Utilizing those
severe pollution problems, namely, sulfur oxides and other technologies, some of which are
(SOx) and particulate matter, and they have 80% adopted from the advanced military and civilian
less NOx than products of coal combustion. aircraft engine-related research and development,
In general (notwithstanding some future fuel land-based heavy-duty industrial gas turbines are
cell designs discussed later), the most efficient true marvels of human engineering.
means to utilize natural gas as a fossil fuel in Diesel or compression-ignition IC engines are
electric power generation is to burn it in internal no slouches themselves. In fact, the efficiency of
combustion (IC) engines. The two most widely a tri-fuel (one that can burn natural gas, light and
used IC engines in electric power generation are heavy fuel oils) unit rated at nearly 17 MW net
gas turbine and reciprocating internal combustion electric output is 47.3% on an LHV basis when
or diesel engine. (Note that the term diesel in this operating with natural gas fuel [7]. They are
Natural Gas Power 255

Natural Gas Power, Fig. 3 Dual-fuel and natural gas diesel engine and gas turbine orders, June 2005 through May 2006 [9]

eminently suitable to landfill, biogas, and coal is reflected in the annual unit order numbers
mine gas applications. They are at a disadvantage displayed in Fig. 3. In terms of the numbers of
because of their relatively low power density and units ordered in 2005–2006 [9], reciprocating IC
cost. The sample multi-fuel diesel engine cited engines dominate (nearly 2,700 units vs. 666 for
above has a power density of 0.05 kW/kg of gas turbines), but the significant bulk of the
engine weight or 215 kW/m2 of engine footprint. orders is for units rated at 3.5 MW or less. In
These numbers compare very unfavorably to the terms of the total MW rating of the orders, all
power density of a heavy-duty industrial gas dual-fuel and natural gas diesel engines add up to
turbine. For a typical advanced machine with 4.4 GW as opposed to more than 40 GW for all
the performance similar to the sample cited gas turbines.
above, the power density is about 0.8 kW/kg of At this point, it should be fairly obvious that a
engine weight or about 3,000 kW or more per m2 treatise on the subject of electric power generation
of engine footprint. In order to appreciate the using natural gas is essentially a treatise on modern
power density of a GT vis-à-vis renewable land-based or heavy-duty industrial gas turbines.
energy-based systems, consider that the power This is especially true for the past and the present
densities of typical wind and solar radiation are of the technology. The future of electric power
0.4–0.8 (depending on altitude and speed) and generation using natural gas as a fuel source might
1.4 kW/m2, respectively [8]. One would need have a place for the fuel cell technology (with or
hundreds of acres of wind turbine farms or solar without gas turbines). Ongoing research activities
collector fields to replace a single advanced include solid oxide (SOFC), molten carbonate
F-Class gas turbine. This gross disparity in the (MCFC), and polymer electrolyte (PEFC) technol-
power density of the two IC engine technologies ogies. From where one stands in 2010, the most
256 Natural Gas Power

promising use for these technologies seems to be Far East and North America, with about 30 units
found in the fields of transportation and portable each, accounted for the majority of those orders.
power. With the exception of distributed power According to the 39th Power Generation Order
generation, in which small-scale (1–500 kW) units Survey by the Diesel & Gas Turbine Worldwide
power individual residences, office buildings, etc., magazine, the number of diesel, dual-fuel, and
the prospect of fuel cell-based technologies to natural gas engines ordered in 2014 totaled around
replace internal combustion engines and specifically 32,000. The majority of orders (90%) were below
gas turbines as large-scale baseload power generat- 2 MW. The 2015 Survey revealed that 50% of the
ing units in the near term, especially in a cost- orders received were destined for continuous duty,
effective manner, is practically nil. followed by standby service at 47%. Peaking ser-
vice represented roughly 2% of the orders. Natural
gas represented roughly 9% of the orders. Top
2016 Update geographic locations for diesel, dual-fuel, and
According to the US EIA’s 2016 Outlook, the gas engines were North America (20%), the Mid-
share of natural gas in world net electricity gener- dle East (16%), Western Europe (15%), Far East
ation is expected to increase from 22% in 2012 to (13%), and Southeast Asia and Australia (12%).
28% in 2040, rising by an average of 2.7% per The US shale gas boom continued to keep North
year. In essence, similar to the case in the United America atop the geographic destination list for
States, global natural gas-fired generation will recips.
catch up with coal-fired generation, which is
expected to decline from 40% of total generation
in 2012 to 29% in 2040. A Brief History of Natural Gas
As will be covered in more detail later in the
chapter, largest gas turbines at the time of writing are Natural gas has been around for quite a while. In
pushing 350 MW in 60 Hz units, which is larger China, nearly 2,500 years ago, early innovators
than the largest 50 Hz unit considered as an example used natural gas escaping naturally from the
in the original edition of the current chapter. Simi- ground to light their temples and heat brine for
larly, largest 50 Hz unit at the time of writing is rated distillation. Bamboo sections, split lengthwise,
at more than 500 MW net output with more than were glued together and bound with twine, allo-
42% net LHV efficiency. This particular machine wing the gas to be transported. In the early seven-
has a power density of 1.2 kW/kg, which is 50% teenth century, French explorers near Lake Erie
higher than the example used above. These most reported seeing natives burning gas naturally
advanced “J class” gas turbines have a nominal seeping from the ground. The first industrial use
turbine inlet temperature of 1,600
C (2912
F). of natural gas can be traced back to England in the
In terms of worldwide engine orders, not much late eighteenth century when gas produced from
has changed from the status depicted in Fig. 3. coal was used to light houses and streets. In the
Over 90% of the engine orders above 0.5 MW are United States the first natural gas well was dug in
concentrated in the 0.5–2 MW range on a yearly 1821 in Fredonia, NY, 4 years before Colonel
basis. On a yearly basis, North America accounts Drake hit upon oil and gas at a depth of 69 ft in
20% of worldwide engine orders in that segment. Lake Erie, PA. The well, dug by William Hart,
Worldwide, orders from 2013 to 2014 were fairly was 27 ft deep in a creek bed. The extracted gas
flat overall with a slight volume shift from the was transported through hollow log pipes to
lowest 0.5–1 MW range to the 1–2 MW and Fredonia for use in street lamps. Throughout
2–3.5 MW ranges. most of the nineteenth century, natural gas was
In the 180+ MW segment, 102 gas turbine used almost exclusively for the same purpose, i.e.,
units were ordered in 2014 (nearly 28 GW capac- lighting. Following the invention of the Bunsen
ity, i.e., about 280 MW on average per unit) with burner in 1885, natural gas also presented house-
60 of those slated for continuous generation duty. holds with a more convenient means of cooking
Natural Gas Power 257

and heating. The main impediment to a wider efficiencies leveled off at the low 40% range,
utilization of natural gas was the lack of a trans- whereas then-new “F-Class” gas turbines with
portation mechanism. The primary method, still in high firing temperatures, almost exclusively in
use, is through an underground pipeline. While combined cycle configuration, opened the door
the first significantly long pipeline was for thermal efficiencies easily exceeding 50%
constructed in 1891 (120 miles long between cen- and even pushing for 60%. Recently, especially
tral Indiana and Chicago), it was only after World during the so-called energy “boom” of the early
War II when technology advances made construc- 2000s, a large fraction of new capacity addition in
tion of longer and more reliable pipelines feasible. the United States has been in natural gas-fired
Until that time, natural gas was either allowed to units, e.g., nearly 95% or 22 GW in 2000. Since
vent into the atmosphere (intentionally or then, this rate has tapered off significantly. Natural
unintentionally) or burned in flares (as a gas-fired power plant capacity addition in terms of
by-product of coal mining and oil exploration) or construction starts totaled about the same, i.e.,
just left in the ground when found alone. In fact, 22 GW, in the years 2005–2007 [10]. Scheduled
when the first astronauts looked at earth from the construction start is about 9 GW for 2010, about
orbit, the most prominent sights on the earth’s the same as in 2009 but lower than the peak year
surface were the flares from the oil fields in the of 2008 with nearly 12 GW. As pointed out earlier,
Middle East and Africa. Even now about 150 mil- natural gas is projected to be the fastest-growing
lion cubic meters of natural gas, roughly equiva- fuel for worldwide electric power generation.
lent to annual US residential consumption, is Thus, the pace of construction and commissioning
flared to the atmosphere. The contribution of the of plants burning natural gas might be expected to
flared NG to global GHG emissions is about pick up again. Construction of natural gas-fired
400 million metric tons, equivalent to nearly one power plants scheduled to start between 2011 and
third of the annual US car exhaust emissions. 2015 was estimated to be 44.5 GW [10]. Even
Nowadays natural gas is used for much more though some projects will undoubtedly be can-
than street lighting or household cooking. In fact, celled or postponed, this is a strong positive sign
residential use of natural gas accounts for only a for the future role of natural gas in electric power
fifth of the total US consumption, with more than production.
60% used by the industrial sector including elec- The driver of the ascendancy of natural gas as
tric power generation. Until the end of 1980s, the star of fossil fuels is the gas turbine, which also
high-efficiency fossil fuel (mainly coal) firing or plays a key role in its transportation and distribu-
nuclear powered steam turbine-based power tion via huge pipeline networks. Heavy-duty
plants supplied the baseload electric power gener- industrial and/or aeroderivative gas turbines
ation capacity (6,000–8,000 h per year). When drive the large centrifugal gas compressors,
newly designed high-efficiency steam plants which increase the pressure of natural gas up to
entered the service, older less-efficient units 1,500 psia (~100 bars). In most cases, these gas
were relegated to intermediate duty operation turbines burn the natural gas taken from the pipe-
(e.g., 4,000 h per year). Early generation, low- line. Additional compressor stations, placed at
efficiency gas turbines, along with old steam regular intervals along the hundreds of miles
plants and hydro pumped-storage plants, were long pipelines, maintain line pressures reduced
largely limited to short-duration peak-shaving- via friction between pipe and gas, which can
type operations. Due to changing economic reach flow velocities up to 40 km/h.
parameters, introduction of protective environ-
mental regulations, and major technological A Brief History of Gas Turbine
advances in gas turbine technology, starting in When lieutenant colonel Heinz Bär died in a civil-
the mid-1980s and the early 1990s, natural gas ian aircraft accident in 1956 in (then) West Ger-
became the fuel of choice for new power plants. many, he was the world’s leading jet fighter ace
The steady increase in coal-fired steam plant with 16 victories to his credit. Some five decades
258 Natural Gas Power

after his death and more than six decades after his turbines primarily as a military aircraft propulsion
last aerial exploits, he is still unrivaled. The device. Starting in the late 1940s and early 1950s,
machine that enabled Bär to permanently enter engineers in the United States and the United
the annals of military history was the world’s Kingdom managed to derive robust, land-based
first mass production jet fighter, the Messer- power generation and, later, marine propulsion
schmitt Me 262. It was powered, remarked engines from these aircraft designs. Claire Soares
another famous German fighter pilot, “as if essentially came to the conclusion that the
pushed by angels” – actually two 1984-lbs thrust involvement of so many people from so many
Jumo 004 gas turbine jet engines. different countries with immense pride in their
Developed under the leadership of Franz work led to selective “histories” favoring selected
Anselm, the Jumo 004 was essentially the engine milestones [16]. Her brief summary can be
that opened up the jet age. Quite naturally, consulted for alternate takes on the development
Dr. Anselm did not conceive Jumo 004 in a vac- history of gas turbines. For a history of Siemens
uum. Paraphrasing Sir Isaac Newton, he was gas turbine development, a good resource is the
standing on the shoulders of giants such as Sir paper by Leiste [17]. For a US perspective with an
Frank Whittle and Hans von Ohain, who are emphasis on GE, the reader should consult Miller
widely recognized as the pioneers of modern tur- and Nemec [18] and Brandt [19]. The paper
bojet engine. Sir Frank and Dr. Ohain themselves describing the design and development of the
almost certainly acknowledged early inventors GE MS7001F gas turbine by Brandt [20] is a
such as Barber (first GT patent in 1791), Stolze good reference laying out all design aspects of a
(first GT to be built and tried in 1900–1904), modern gas turbine and the trade-offs that are
Holzwarth, and others, who showed them the unavoidable in bringing a feasible product into
path to the ultimate gas turbine engine design the market. A recent addition to these works is
through their patents on air and steam turbines. the book by Eckardt, which gives a detailed story
It is futile to attempt to plot a straight historical of the power generation gas turbine, covering
line from Hero’s turbine (which had no purpose almost a century, from the perspective of Brown
other than entropy generation) to, say, the Boveri Company (later ABB, later Alstom and
2,600
F firing temperature class steam-cooled now GE) [21].
H-System™. Countless inventors laid down the Even though the development of the modern
building blocks that eventually led to the modern gas turbine was primarily driven by military and,
gas turbine engine. For an excellent short history to a lesser extent, civilian aircraft propulsion con-
of turbomachinery in general and gas turbines in siderations, the introduction of stationary (land-
particular, the reader should consult the relevant based) units dedicated to power production did
chapter in Wilson and Korakianitis [11], who not lag too far behind. In fact, Aurel Stodola
highlight the contributions of lesser known credited Hans Holzwarth with “having built the
(unfairly) inventors such as Aegidius Elling and first economically practical gas turbine”
George Jendrassik. Another good narrative with [22]. Holzwarth built several types of his “explo-
an emphasis on turbojets and German research sion” turbine in the first quarter of the twentieth
and development can be found in Hans von century. Holzwarth’s turbine was, strictly speak-
Ohain’s foreword to Mattingly’s book on gas tur- ing, a hybrid construction combining the spark-
bine propulsion [12]. For a detailed look at turbo- ignition (constant volume) combustion process of
jet development activities of Sir Frank Whittle and an Otto cycle with the axial expansion process of a
Hans von Ohain in their respective countries, Brayton cycle. Stodola calculated a thermal effi-
three articles by Meher-Homji can be consulted ciency of 25.6% for the test of a 1,500 rpm exper-
[13–15]. imental turbine in Mühlheim an der Ruhr in 1919
Quite obviously, wars or the prewar atmo- [22]. Burning a gas with 434 Btu/ft3 heating
sphere of the western world in the first half of value, the turbine produced about 725 kW. Air
the last century pushed the development of gas and fuel gas was compressed by steam turbine-
Natural Gas Power 259

Natural Gas Power, Combustor

Fig. 4 Schematic diagram
of Brown Boveri Co.’s
Neuch^atel gas turbine
(1939)
Starter
Generator Compressor Turbine

driven compressors and sequentially injected into this gas turbine was utilized to heat the feed water
the explosion chamber (gas first) at about 30 psia. of a conventional steam plant. In other words, the
Stodola reports an average maximum explosion first US electric utility gas turbine was in a “com-
pressure of 160 psia. The combustion products bined” cycle configuration [24]. In fact, the com-
expanded through a two-stage velocity- bined cycle concept goes back to Emmet’s
compounded turbine. Exhaust gas at about mercury-vapor process (1925) [25]. For a brief
800
F was recovered in an exhaust heat boiler. history of the CC power plants and pertinent refer-
In that sense, this unit could be considered the first ences, the reader should consult the 1994 Calvin
combined cycle. Holzwarth designed and devel- Winsor Rice Lecture by Sir John Horlock [25].
oped several different variants of his turbine One can look at the technology history of the land-
between 1907 and 1928. From 1928 on, Brown based gas turbine and combined cycle power plants
Boveri Company (BBC) took over the develop- for electric power production in four generations:
ment of Holzwarth turbine and in 1933 installed a
BFG-fired unit in a German steel mill, which was 1. Generation 1 (1949–1968)
destroyed during the World War II. (a) Smaller than 30 MW GTs (Frames 3 and 5)
It is generally accepted that the first commer- (b) Firing temperatures 1,500–1,800
F
cial stationary gas turbine for electric power gen- (c) For repowering and cogeneration
eration was erected in Neuch^atel, Switzerland, by 2. Generation 2 (1968–1990)
the former BBC in 1939, 3 years before the first (a) B and E Technology GTs (50–120 MW)
flight of Me 262 powered by Jumo 004. This fuel (b) Firing temperatures ~2,000
F
oil burning 4 MW machine, primarily used for (c) NOx emission control using GT water/
standby and peaking duties, was operational for steam injection or SCRs
nearly 70 years (see Fig. 4). The combustion (d) Non-reheat steam cycles
chamber was derived from the turbo-charged 3. Generation 3 (1990–1998)
Velox boiler, which itself resulted from the BBC (a) F Technology GTs (75–260 MW)
work done on the Holzwarth turbine. The BBC (b) Three air-cooled turbine stages
turbine was tested under the supervision of Aurel (c) Firing temperatures ~2,400
F
Stodola, who reported an overall thermal effi- (d) Performance fuel heating (365
F)
ciency of 17.4%, which is less than half of the (e) DLN combustion system for NOx control
efficiency of early F-Class machines [23]. This (f) Three-pressure, reheat (3PRH) steam cycles
machine was designated by the ASME as a His- 4. Generation 4 (1998–present)
toric Mechanical Engineering Landmark in 1988. (a) H Technology GTs (400+ MW in CC)
The first gas turbine installed in an electric (b) Closed-loop steam cooling (CL-SC) of the
utility in the United States (Oklahoma Gas & Elec- first two turbine stages (both stator and
tric, Belle Isle Station, Oklahoma) was a 3.5 MW rotor); four turbine stages
GE Frame 3 unit that entered service in 1949. In (c) Cooling of cooling air (CAC) for turbine
addition to generating power, the exhaust gas of wheel spaces and subsequent stages via a
260 Natural Gas Power

Firing
Temp
60% CC Efficiency
“H” (F)
Firing Temp
2600
Combined Cycle Efficiency

55%
“F” 2400

50% 2200

“E” 2000
45%

1800

40%
1970 1975 1980 1985 1990 1995 2000 2005
Year of Shipment

Natural Gas Power, Fig. 5 Incremental evolutions of E-, F-, and H-Class GT technology (From Ref. [26])

heat exchanger (a kettle reboiler) that gener- electric power generation systems are briefly elabo-
ates IP steam to be used in the bottoming rated upon.
cycle
(d) Active clearance control (compressor and 2016 Update
turbine) The class hierarchy of heavy-duty industrial gas
(e) Firing temperatures ~2,600
F turbines became somewhat confusing in the
(f) Performance fuel heating (400 +
F) 6 years since the original draft of the current chapter.
(g) DLN combustion system for NOx control The term H technology or H-Class is now used for
(h) Three-pressure, reheat (3PRH) steam cycle advanced F-Class (air-cooled) units with turbine
integrated with the GT Brayton cycle inlet temperatures 1,500
C (2,732
F) or higher.
These machines are thus capable of reaching firing
The historical advances in GT firing temperature temperatures of 2,600
F, which were previously
and CC efficiency are summarized in Fig. 5. The possible only with the steam-cooled H-System. In
data illustrates the significant advance in the thermal fact, a combined cycle power plant with an air-
efficiency of the GT combined cycle power plants, cooled H-Class gas turbine broke the 60% efficiency
to the tune of nearly 15 percentage points, which is barrier in 2011.
mainly driven by advances in materials, coatings, The J class (its air-cooled variant is referred to as
and cooling techniques that pushed the firing tem- JAC) is capable of 1,600
C (2,912
F) turbine inlet
peratures by almost 800
F (~450
C). In the fol- temperature (or higher). It is referred to as the HA
lowing chapters, the reader is introduced to the basic (i.e., air-cooled H) class by another original equip-
thermodynamic principles that govern the perfor- ment manufacturer (OEM). In the meantime, the
mance of natural gas-burning gas turbine power same OEM retired its “fully” steam-cooled
plants. In addition, also provided is a brief overview H technology (with six units operating in the
of state-of-the-art gas turbine power plants, key world). Note that the J class gas turbines retain
economic criteria, and operability considerations. steam cooling for the combustor transition piece
Finally, the future directions in gas turbine-based and first-stage turbine ring. Both can be considered
Natural Gas Power 261

as Generation 5 technologies. Along with high firing it forms a coherent and compact reference to be
temperatures and cycle pressure ratios (about 23), used in basic engineering analysis of gas turbine
J class gas turbines are characterized by high air combined cycles. Even then, it must be pointed out
flow rates and power ratings (almost 500 MWe for that a similar treatment can be found in the excel-
the 50 Hz units). lent brief chapter on gas turbines written by
Haselbacher cited in [27]. Another valuable short
reference incorporating the basic fundamental con-
Gas Turbine Power Plants siderations is cited in [25]. The treatment herein
focuses on land-based heavy-duty industrial gas
Gas turbine in itself or as a key component of a turbines. For a brief introduction to thermodynamic
larger power generation system is a subject that can and economic considerations pertaining to
only be covered in dedicated books. Such books do aeroderivative gas turbines and pertinent refer-
exist, and several selected treatises on the subject, ences, the reader should consult Horlock [28].
which in the opinion of the authors reflect the best
available published work in the field, are listed
among the bibliography references. The coverage Basic Thermodynamics
of the subject in the limited space allotted to the Gas turbine is a relatively simple turbomachine
current chapter aims to provide the reader with the comprising three key components: compressor,
fundamental thermodynamic considerations that combustor, and expander (Fig. 6). The last one is
lead from the Carnot cycle to the present-day commonly referred to as a turbine. The compres-
advanced gas turbine in a logically coherent man- sor and expander are connected through a com-
ner. There is a reason that the gas turbine combined mon shaft. The operation of a gas turbine is
cycle plant is the most efficient heat engine that is described by a thermodynamic cycle comprising
commercially available at the present time and, four processes: compression, heat addition or
most likely, in the foreseeable future. That reason combustion, expansion, and heat rejection. A gas
is embedded in the laws of thermodynamics. turbine can also be classified as an internal com-
Understanding this will enable someone with a bustion engine (just like a car engine), in which
basic knowledge of key engineering concepts to the compressed working fluid, i.e., air, mixes with
grasp the gas turbine design principles and carry fuel and the products of the ensuing chemical
out nearly all calculations using simple formulas reaction (i.e., combustion) expand through the
and a few technology charts. Furthermore, this will turbine. In land-based electric power generation
also enable the reader to critically evaluate new applications, part of the useful mechanical or shaft
developments and claims to future improvements work, produced by the expanding combustion
in electric power generation systems that employ products (approximately 50% of the total), is con-
turbomachines and utilize fossil fuels such as sumed by the compressor. The remainder is
natural gas. converted into electric power in a synchronous
As Descartes wrote in his Discourse on Method, alternating current (ac) generator, which is
“one cannot conceive anything so strange and so connected to the same shaft as the gas turbine. In
implausible that it has not already been said by one aircraft propulsion applications, the entire shaft
philosopher or another.” As such, in relation to the work generated in the turbine is utilized to drive
treatment of the subject in the paragraphs below, a the engine compressor and fan. The thermal
claim of pure originality would be absurd. Never- energy in the exhaust gas is converted into kinetic
theless, the material is wholly original in the sense energy in the exhaust nozzle, which generates the
that key concepts are developed in a unique manner net engine thrust.
along with all pertinent formulas, representative The thermodynamic cycle that describes the
design parameters, and technology curves so that gas turbine operation is the Brayton cycle. For
262 Natural Gas Power

Natural Gas Power, Fig. 6 Modern heavy-duty industrial gas turbine (Courtesy: GE Energy)

Natural Gas Power, Fig. 7 Temperature-entropy diagram of gas turbine Brayton cycle

simple analytical calculations, which can be found the most logical) representation of the air-standard
in elementary textbooks on thermodynamics, the Brayton cycle is on a temperature-entropy or T-s
ideal or air-standard Brayton cycle is used. The surface as illustrated in Fig. 7. The key assump-
classical (and, from an engineering perspective, tions in air-standard cycle analysis are:
Natural Gas Power 263

1. The working fluid is air, which is a calorically to the control volume) and heat rejection from the
perfect gas (i.e., cp = constant). cycle is a negative number (i.e., heat taken from
2. Isentropic (i.e., adiabatic and reversible) com- the control volume).
pression and expansion. In addition to the first law of thermodynamics
3. Constant pressure heat addition and heat for an SSSF process (or control volume analysis),
rejection. there is one more key thermodynamic relationship
that forms the heart of the GT Brayton cycle
The two key nondimensional performance calculations: pressure-temperature ratio for an
metrics for a gas turbine are specific net power isentropic process:
output and thermal efficiency:

g1
_ net T2 p2 g
W ¼ (8)
w¼ (1) T1 p1
m_  cp  T1
where g is the ratio of constant pressure and con-
_ net
W stant volume-specific heats. Combining Eqs. 1, 2,
Z¼ (2)
Q_ in 3, 4, 5, 6, and 7, in nondimensional terms and
applying the isentropic formula in Eq. 8 to com-
The net power output is the difference between pression and expansion processes, one ends up
the compressor power consumption and the with the following two expressions:
expander power production. In graphical terms,
it is exactly equal to the area encompassed by the    t3 
w ¼ pk  1  k  1 (9)
cycle 1-2-3-4-1 on the T-s diagram in Fig. 7. Each p
component’s power can be calculated by the
application of the first law of thermodynamics 1
Z¼1 (10)
for steady-state steady-flow (SSSF) successively pk
to their respective control volumes:
where p = p2/p1 is the cycle pressure ratio, t3 is
_ comp ¼ m the cycle maximum temperature
W _  cp  ðT1  T2 Þ (3)
(nondimensionalized via division by T1), and
k is the isentropic exponent, g  1/g. The two
_ turb ¼ m
W _  c p  ð T3  T4 Þ (4) simple formulas, Eqs. 9 and 10, which are strictly
valid only for an ideal air-standard cycle, illustrate
_ net ¼ W
W _ turb þ W
_ comp (5) almost all of the key facts concerning gas turbine
performance:
From Fig. 7 and Eqs. 3 and 4 one can easily see
that the compressor power consumption is a neg- 1. Gas turbine or Brayton cycle performance is
ative number (i.e., work done on the control vol- controlled by two parameters: cycle pres-
ume) and expander power generation is a positive sure ratio, p, and maximum cycle tempera-
number (i.e., work done by the control volume). ture, t3.
Application of the first law for SSSF to the heat 2. Cycle efficiency is a function of cycle pres-
addition and heat rejection processes, one gets sure ratio only; higher cycle pressure leads
to higher cycle efficiency.
Q_ in ¼ m
_  c p  ð T3  T2 Þ (6) 3. Higher cycle temperature leads to higher
cycle-specific output.
Q_ out ¼ m
_  c p  ð T1  T 4 Þ (7) 4. Higher cycle pressure ratio, beyond a cer-
tain value, is detrimental to cycle-specific
pffiffiffiffiffi
output (i.e., @@ pw2 < 0 and p ¼ 2k t3 for
2
From Fig. 7 and Eqs. 6 and 7, the heat input to
the cycle is a positive number (i.e., heat supplied @w/@p = 0).
264 Natural Gas Power

These observations are also borne out qualita- ð3

tively by a close examination of the cycle T-s h3  h2 ¼ T  ds (12)
diagram in Fig. 7. As the heat addition isobar 2
moves upward (i.e., higher cycle pressure ratio,
p), one can visually verify that the area One can imagine an (hypothetical) isothermal
encompassed by the cycle, 1-2-3-4-1, becomes process that takes place at a constant temperature,
smaller until; in the limit, it becomes zero. whose value lies between T3 and T2, and has the
Equation 10 numerically implies that, in the limit same entropy as the actual end states 2 and 3. This
when p is sufficiently high, the cycle efficiency process would have exactly the same numerical
will approach 100%, or, Z ! 100% as p ! 1. value of heat addition as the original process, i.e.,
This, of course, is impossible, because, as dictated h3 – h2, per unit mass. For that process, Eq. 12 can
by the second law of thermodynamics, the maxi- be rewritten as
mum theoretical efficiency attainable by any cycle
is given by the Carnot cycle efficiency. The Car-
h3  h2 ¼ TH  ðs3  s2 Þ (13)
not cycle is a hypothetical power cycle that is
represented by two isentropic processes, i.e., com-
Using the calorically perfect gas relationships,
pression and expansion, and two isothermal pro-
the mean-effective heat addition temperature can
cesses, i.e., heat addition and heat rejection.
be found from Eq. 13 as
Graphically, it is the rectangle enveloping the
Brayton cycle, which is bounded by T1 and T3
T3  T 2
isotherms and s1 and s3 isentropes. TH ¼
 (14)
T3
For the cycle in Fig. 7, one can easily visualize ln
the limit of p ! 1 as T3 stays constant. In the T2
limiting case, one will end up with a Brayton cycle
that is a rectangle with ~0 area (i.e., zero specific Using the same logic, the mean-effective heat
output) and heat addition and rejection at constant rejection temperature is calculated as
temperatures T3 and T1, respectively, which essen-
T4  T1
tially defines a Carnot cycle with the efficiency: TL ¼
 (15)
T4
ln
T1 T1
Z¼1 (11)
T3
These two mean-effective temperatures are
Comparing with Eq. 10, for a cycle maximum graphically depicted in Fig. 8. Thus, qualitatively
temperature of T3 = 1,500
C and g = 1.4, this any thermodynamic cycle in general and the
would correspond to a cycle pressure ratio of Brayton cycle in particular, 1-2-3-4 in Fig. 8, can
p = 577 and cycle efficiency Z = 87.4%. In other be represented by its Carnot-equivalent cycle,
words, as dictated by the second law of thermody- 10 -20 -30 -40 , with mean-effective (constant) heat
namics, the maximum theoretical cycle efficiency addition and heat rejection temperatures that are
that is attainable by a gas turbine with a “firing given by Eqs. 14 and 15. The efficiency of that
temperature” of 1,500
C is 87.4%. Clearly a cycle cycle would be given by the well-known relation-
pressure ratio of 577 is of little practical interest. For ship for the Carnot efficiency, i.e.,
the modern heavy-duty industrial gas turbines, the
typical value of p is about 18. What does the second TL
Z¼1 (16)
law say about a gas turbine for a cycle maximum TH
temperature of T3 = 1,500
C and p = 18? In order
to find that out, consider the isobaric heat addition For an ideal air-standard Brayton cycle, one can
process 2 ! 3 in Fig. 7. For a reversible and isobaric show that Eq. 16 is the same as Eq. 10. Thus, for
process, the modified Gibbs equation is any gas turbine with a cycle pressure ratio of
Natural Gas Power 265

2. The lower triangular area 1-4-4c-1, which is

equal to the rectangular area 1-10 -40 -4c-1,
and represents the irreversibility and exergy
loss via heat rejection to a heat sink

In terms of cycle temperatures, the upper trian-

gular area is quantified by the gap between the
cycle maximum temperature, T3, and the mean-
effective heat addition temperature TH . Similarly,
the lower triangular area is quantified by the gap
between the cycle minimum or ambient tempera-
ture, T1, and the mean-effective heat rejection
temperature TL . The latter mechanism of “lost
work” associated with heat rejection also points
the way to the gas and steam turbine combined
cycle power plant concept. This will be explored
Natural Gas Power, Fig. 8 Temperature-entropy dia-
grams of gas turbine Brayton cycle and its “Carnot in some more detail later in the discussion.
equivalent” The real gas turbine Brayton cycle differs from
the ideal, air-standard cycle in the following
respects:
18, the maximum theoretical efficiency is 56.2%.
Using Eqs. 8, 14, and 15, this would be the effi- 1. Non-isentropic compression and expansion
ciency of a Carnot cycle with isothermal heat (represented by component polytropic or
addition and heat rejection at 852
C and isentropic efficiencies)
219
C, respectively. It is interesting to note that 2. Pressure losses during heat addition and
changing the cycle maximum temperature from heat rejection
1,500 to 1,400
C would change these tempera- 3. Real gas effects (represented by a suitable
tures to 814
C and 203
C, respectively, but equation of state)
would not change the cycle efficiency, i.e., 4. Change in working fluid composition and
56.2%, which is only a function of p. This is in properties downstream of combustion
contrast to the performance trend of the “real” gas process
turbines, whose thermal efficiency does indeed
increase with increasing cycle maximum or firing These nonidealities can be easily incorporated
temperature. Why this is so will be understood into the simple equations above via polytropic
below. component efficiencies, pressure loss factors,
One key observation from Fig. 8 is the large and real gas equation of state. One can refer to
thermal efficiency gap between the ultimate one of the available textbooks on the subject for
Carnot efficiency given by Eq. 11 and the maxi- details, e.g., [29]. For example, Eq. 9 can be
mum theoretical entitlement level as dictated rewritten as
by the second law, Eq. 10 or Eq. 16, which is
nearly 30 percentage points. Graphically, this  ka
=

gap is represented by the two triangular areas in w ¼ 1  pc Zc þ ð1 þ f Þ  wg  t3
Fig. 8: !
1
 1  Z kg (17)
1. The upper triangular area 2-2c-3-2, which is pt t
equal to the rectangular area 20 -2c-3-30 -20 ,
and represents the heat addition (i.e., com- where pc and pt < pc are compressor and turbine
bustion) irreversibility pressure ratios, respectively, ka and kg are
266 Natural Gas Power

Natural Gas Power, Fig. 9 Real Brayton cycle effi- (in parentheses) are the ideal, air-standard cycle efficien-
ciency from Eqs. 19 and 20 as a function of cycle pressure cies from Eq. 10
ratio (p) and maximum cycle temperature (t3). Also shown

isentropic exponents for air and combustion prod- performance. Typically quantified via mechanical
ucts, respectively, and Zc and Zt are compressor and generator efficiencies, Zm and Zg, respec-
and turbine polytropic efficiencies, respectively. tively, these losses should be accounted for the
The ratio of the specific heats of combustion gas turbine generator’s net electric output. The
products and air is wg, and f is the fuel-air mass fuel mass flow rate (as a fraction of turbine air-
flow ratio. Heat addition can be found from the flow) that is requisite for a specified cycle maxi-
fuel consumption, i.e., mum temperature is given by

Q_ in ¼ m
_ f  LHV (18) wg  ðT3  Tref Þ  ðT2  Tref Þ
f¼
‘  T1 þ wf  ðTf  Tref Þ  wg  ðT3  Tref Þ
where LHV is the net or lower heating value of the
(20)
fuel at the reference temperature, Tref. The effi-
ciency of the gas turbine is the ratio of Eqs. 17 and
where the second term in the denominator is the
18, i.e.,
sensible fuel energy input and wf is the ratio of the
! ! specific heats of fuel and air. Comparing Eq. 19
ka
=Z 1 (along with Eq. 20 that sets the denominator) with
1  pc c þ ð1 þ f Þ  wg  t3  1  Z kg
pt t Eq. 10, one can easily observe that the GT thermal
Z¼ efficiency, when accounting for the “real” cycle
f ‘
(19) effects, is indeed a function of t3. While it is
difficult to discern the impact of t3 on Z just by
where ‘ is the nondimensional fuel energy con- looking at Eqs. 19 and 20, a sample calculation
tent, LHV/cp , a  T1. Note that for clarity, can illustrate that as shown in Fig. 9.
mechanical and electric losses are omitted from Typical values that are reasonably accurate for
Eqs. 17 and 19, which represent the “shaft” gas turbines burning natural gas fuel (assumed to
Natural Gas Power 267

be 100% CH4 or methane) are given below flow. One other cycle maximum temperature def-
(adopted from [29]) and used in the calculations inition, favored by European OEMs, is the ISO-
that resulted in the curves depicted in Fig. 9: TIT, which is a fictitious number as defined by
ISO-2314 (typically ~ 180
F lower than RIT).
LHV = 21,515 Btu/lb Obviously, for an uncooled turbine all these defi-
cp,a = 0.24 Btu/lb-R (1.005 kJ/kg-K) nitions are the same.
cp,g = 0.274 Btu/lb-R (1.148 kJ/kg-K) or Gas turbine efficiency as a function of PR is
wg = 1.1417 shown in Fig. 10. Specific output, w, in the range
cp,f = 0.60 Btu/lb-R (2.5125 kJ/kg-K) or wf = 2.50 covered by the turbines in Table 2 is
ga = 1.400 400–450 kJ/kg. Carbon dioxide emissions are
gg = 1.333 calculated from the airflow and fuel flow using
Zc = 87.8% chemical reaction equations for 100% CH4 gas-
Zt = 85% eous fuel assumption as 510 kg/h per MW of
Zm = 99% output at 39% net efficiency. Each one percentage
Zg = 98.9% point improvement in efficiency is worth about
ka = 0.2857 15 kg/h per MW reduction in CO2 emissions. This
kg = 0.2498 is equivalent to 27,000 metric tons for a 300 MW
Tref = 77
F (25
C) GT running at baseload for 6,000 h per year.
‘ = 172.728 Figure 10 shows that the gap between the cur-
rent SOA in gas turbine technology and the max-
Original equipment manufacturer (OEM) data imum theoretical entitlement level, as dictated by
for five 50 Hz (3,000 rpm) heavy-duty industrial the second law, is more than 15 efficiency points.
gas turbines from a recent trade publication [6] are The GT efficiency – compressor pressure ratio
given in Table 2. (PR) trend in Fig. 10 (via Eq. 19) – gives the
For an actual gas turbine, the cycle maximum impression that, for a given cycle maximum tem-
temperature T3 is equivalent to the turbine inlet perature, the efficiency can be improved indefi-
temperature (TIT) at the exit of the gas turbine nitely by increasing the PR. This is not true due to
combustor. Another value that is frequently used a key feature of modern gas turbines that is very
in the industry is the “firing temperature,” which is difficult to capture in simplified models such as
the temperature at the stage-1 blade row (rotor) Eqs. 17, 18, 19, and 20: turbine hot gas path
inlet. This temperature, also known as rotor inlet (HGP) component cooling via air extracted from
temperature (RIT), is the maximum temperature at the compressor discharge and/or inter-stage loca-
which the combustion products start producing tions. Since the early days of GT development the
useful expansion work. The difference between components exposed to the highest temperature
TIT and RIT is about 250
F for modern air-cooled environment have been cooled with air drawn
gas turbines and reflects the dilution effect of the from the compressor. The Junkers Jumo 004 jet
stage-1 nozzle and forward wheel space cooling engine of the German fighter plane Messerschmitt

Natural Gas Power, Table 2 OEM data for 50 Hz published output, efficiency, pressure ratio (PR), and
heavy-duty industrial gas turbines [6]. The maximum exhaust data with Eqs. 17, 18, 19, and 20
cycle temperature, T3, values are estimated using the
T3 Output Exhaust




Turbine C F PR PRt MW Eff. (%) kg/s C
A 1,415 2,579 17.7 15.9 285.0 39.6 690 572
B 1,443 2,629 17.0 13.5 255.6 36.9 641 602
C 1,501 2,734 18.3 14.5 279.2 37.8 655 629
D 1,435 2,616 18.2 16.3 292.0 39.8 692 577
E 1,476 2,688 18.0 15.5 312.1 39.3 720 597
268 Natural Gas Power

Natural Gas Power, Fig. 10 Gas turbine efficiency from Eq. 19 as a function of compressor PR (Data is from Ref. [6]
and summarized in Table 2)

262, which entered active service in 1944 during Fig. 11 that uses data from Cumpsty [30]). A good
the last stages of the WWII, was the first mass overview of the current SOA in advanced GT
production turbojet engine in the history with a materials and coatings can be found in [31].
turbine inlet temperature (TIT) of 1,427
F An alternative to open-loop air cooling
[13]. The hollow turbine blades (manufactured (OL-AC) has recently been commercially intro-
by folding flat sheets of 12% Cr alloy called duced at the 480 MW Baglan Bay 109H CC
Cromadur) were cooled by air extracted from power plant in Wales [32]. This machine employs
between the fourth and fifth stages of the com- closed-loop steam cooling (CL-SC) for the major-
pressor and introduced to the blade through the ity of the turbine HGP components (the first two
holes drilled into the disk. The cooling technol- stages of the four-stage turbine, to be exact).
ogy, although unchanged in basic principles, has A “lighter” version of the H-System™ is the
advanced greatly over the years with the introduc- G-Class turbine with steam cooling limited to
tion of complex serpentine cooling passages, the stationary parts upstream of the turbine
impingement, film, and transpiration cooling HGP [33].
techniques and paced by concurrent advances in A structural refinement to the basic GT
superalloys (e.g., nickel-based directionally solid- Brayton cycle is the inclusion of a second “reheat”
ified or single crystal), manufacturing techniques combustor within the turbine expansion. Reheat,
(e.g., investment casting), and oxidation/corro- or sequential combustion as it is referred to by the
sion (e.g., MCrAlY) and thermal barrier (e.g., OEM that successfully commercialized the tech-
yttria-stabilized zirconia, YSZ) coatings. This nology, increases the mean-effective cycle heat
development supported the advances in TIT, addition temperature and reduces the combustion
which nearly doubled the standard set by Jumo irreversibility [34] (the upper triangular area
004 over the course of the next 50 years (e.g., see 2-2c-3-2 in Fig. 3). From a fundamental
Natural Gas Power 269

Natural Gas Power, Fig. 11 Turbine RIT and F-Class units. The shaded region designates the melting
bucket alloy capability history [30]. One can add ~250
F point of nickel-based superalloys
to RIT to get a rough idea about TIT of the advanced

Natural Gas Power, Table 3 Comparison of CC performances based on four major GT technologies [35]: air-cooled
(OL-AC), air-cooled reheat (OL-AC+SC), steam-cooled (CL-SC), and steam-cooled reheat (CL-SC+SC)
OL-AC OL-AC+SC CL-SC CL-SC+SC
GT PR 20 30 23 30
HPT RIT,
C N/A 1,290 N/A 1,430
HPT RIT,
F N/A 2,354 N/A 2,606
RIT,
C 1,430 1,430 1,430 1,430
RIT.
F 2,606 2,606 2,606 2,606
Net CC, % (points) Base +0.81 +1.77 +2.74

thermodynamic perspective, lower combustion from GE’s 7FB, Siemens V94.3A (now SGT5-
losses give the GT cycle with reheat combustion 4000F), Alstom’s GT-26, and MHI’s M701G
a higher performance entitlement than a cycle units for calibration, they have developed very
without it. Similarly, there is no doubt that detailed cooled GT models and evaluated the CC
closed-loop (external) cooling is superior to performances. Their findings are summarized in
open-loop (internal) cooling, but the exact mag- Table 3, which shows that CL-SC has an advan-
nitude of that advantage in real machines is diffi- tage of ~1.8 points (%) in net ZCC over OL-AC.
cult to quantify. Chiesa and Macchi [35] A realistic assessment of the gas turbine
quantified the advantages of OL-AC (with and requires some means to incorporate the cooling
without reheat combustion) and CL-SC technolo- penalty into the GT calculations. Cooled turbine
gies via detailed calculations. Using OEM data expansion models, no matter how much simplified,
270 Natural Gas Power

Natural Gas Power, Fig. 12 “Real” gas turbine technology curves for different PR-RIT combinations [46]. Also shown
is data from Table 2 (As published in Ref. [6])

involve a rather cumbersome system of equations regeneration (also referred to as recuperation), a

amenable only to a numerical solution. In the open combination of both, steam injection (STIG) [47,
literature, there are numerous published works that 48], humid air turbine (HAT) [49], chemical recu-
range from Holland and Thake [36] and several peration [50], and closed cycles [51] are among
papers by Elmasri [37–41] during the 1980s to the the most well known. For a comprehensive look at
more recent works of Horlock and coworkers those variants and their relative merits, as well as
[42–45]. Young and Wilcock’s comprehensive references to the pertinent literature, one can con-
study [45] and the similar treatment in [35] incor- sult the relevant sections in Wilson and
porate detailed aerothermodynamic calculations. Korakianitis [11]. Intercooling and recuperation
On the business side, obviously not available technologies have been demonstrated successfully
for public consumption, OEMs have their in aeroderivative gas turbines [52, 53], whereas the
own in-house proprietary codes containing the combination of both (ICR) has been incorporated
field-proven knowledge (instead of theoretical into a ship-propulsion unit [54]. Nevertheless,
approximations) distilled from decades of experi- steam-cooled (G- or H-Class) and reheat
ence in manufacturing and testing gas turbines. machines are the only two commercially proven
The results from a more realistic GT calculation and successful heavy-duty industrial variants of
[46], which includes the impact of turbine cooling the basic air-cooled GT, which are suitable to
with compressor extraction air, are shown in large-scale baseload electric power generation.
Fig. 12. So far it has been demonstrated that increasing
Myriad alternate GT design variations have the compressor PR and/or TIT (i.e., maximum
been proposed and entered the literature, some to cycle temperature) increases the GT efficiency
the extent that they are even covered in elementary and specific work. The latter reaches a maximum
textbooks, e.g., compressor intercooling, before further increase in PR decreases
Natural Gas Power 271

Natural Gas Power, Fig. 13 Comparison of Brayton “bottoming” cycle partially recovers the Brayton “topping”
cycles with different PRs on a T-s diagram. Also illustrated cycle heat rejection “lost” work
is the “combined” cycle concept, in which the Rankine

it. Furthermore, increasing TIT is favorable to effective cycle heat addition and rejection tem-
CC-specific work and CC efficiency. These trends peratures, TH and TL , respectively, e.g., Eq. 16.
can be easily visualized and quantified by the ideal Increasing the cycle maximum pressure from p2
cycle diagrams shown in Fig. 13. to p20 increases the cycle efficiency by simulta-
The heat rejection “lost work” quantified by neously increasing TH and reducing TL. It is easy
the lower triangular area 1-4-4c-1 suggests that a to see that the cycle-specific work, quantified by
better approximation of the Carnot cycle can be the area 1-2-3-4, would go through a maximum
achieved by utilizing the heat source provided by before approaching zero at which point the cycle
the working fluid at the GT exhaust (i.e., state 4) in efficiency is the Carnot efficiency. The ulti-
a second power generation cycle. At this point, the mately detrimental effect of continuously
Rankine cycle presents itself as the ideal choice increasing the Brayton “topping” cycle (BTC)
due to two facts: PR on CC efficiency can also be seen from
Fig. 13. The maximum possible CC “bottoming”
1. Good match with GT exhaust (on average cycle work, limited by the second law, is equal
about 600
C for SOA F-Class machines) to the triangular area 1-4-4c-1, which can be
for steam generation at 125–165 bar shown to be exactly equal to the GT exhaust
2. Constant temperature heat rejection (i.e., in gas exergy [55]. It is easy to visually appreciate
the steam condenser) at a temperature that is the steady reduction in this maximum
much closer to T1 thanTL with increasing PR (cf. 1-4-4c-1 and
1-40 -40 c-1). The efficiency of the ideal Rankine
Earlier in the section, it was shown that the bottoming cycle (RBC) is given by the equation
Brayton cycle efficiency is given by the mean- below:
272 Natural Gas Power

T1 @ZBTC g  1 12g
ZRBC ¼ 1  (21) ¼  PR g (25)
TL @PR g
where TL is from Eq. 15. The formula in Eq. 21
clearly implies that this particular ideal Rankine From Eq. 21, once again using the ideal Brayton
cycle would be equivalent to a Carnot engine cycle relationships, the rate of decrease of the RBC
operating between the temperature reservoirs at efficiency with increasing PR is found as
TL and T1. Furthermore, the product of Eq. 21 and



the total energy content of the GT exhaust Q_ exh can T1 T4
 1
be shown to be exactly equal to the total exergy of @ZRBC g  1 TL T
the exhaust gas [55]. Exergy is a fluid property ¼   L  (26)
@PR g PR  ln T4 =T1
that can be calculated using a suitable equation of
state with two other known properties (e.g., pres- Combining Eqs. 21, 22, 23, 24, 25, and 26 with
sure and temperature) and the composition. g = 1.4 and T1 = 85
F (assuming RBC heat
By looking at the cycle T-s diagram in Fig. 13, rejection at 1.2 in. Hg or 41 mbar of steam con-
the ideal CC efficiency with ideal Brayton and denser pressure) and solving them numerically,
Rankine cycles can be deduced to depend on the the PR of the BTC for maximum CC efficiency
ratio of the mean-effective Brayton cycle heat can be plotted as a function of T3 (see Fig. 14).
addition temperature given by Eq. 14 and the Also plotted in Fig. 14 is the PR corresponding to
ambient temperature T1, i.e., pffiffiffiffiffi
the maximum BTC-specific work, p ¼ 2k t3, with
t 
3 T1 = 59
F. For comparison, representative OEM
ln k
ZCC ¼ 1  p (22) data extracted from a recent trade publication [6]
t 3  pk is also shown. It is interesting to note the excellent
Ignoring the miscellaneous topping and agreement of the PR-TIT data of actual produc-
bottoming cycle losses and minor inputs, a sim- tion machines with values corresponding to max-
plified version for the CC shaft efficiency can also imum GT-specific power output predicted using a
be written as rigorous cooled gas turbine model.
Using an uncooled quasi-ideal GT model with
ZCC ¼ ZBTC þ ð1  ZBTC Þ  ZRBC (23) reasonable assumptions outlined earlier in the
section to take into account the real turbine losses,
Taking the derivative of both sides with respect CC efficiencies corresponding to the parameters
to the PR and setting the CC efficiency derivative in Fig. 14 are calculated and plotted in Fig. 15.
to zero for the maximum, one finds that (note that The key takeaways from the plots in those figures
the RBC and BTC efficiencies are approximately are rather illuminating:
the same in magnitude)
1. Actual OEM designs are at PR values some-
@ZBTC @ZRBC what lower than those for maximum
  (24) GT-specific output in an uncooled GT but
@PR @PR
coincide extremely well with the air-cooled
This relationship states that the maximum GT-specific work line, in accordance with
Brayton-Rankine combined cycle efficiency the earlier observation made by
occurs at the point where the rate of increase of Horlock [25].
BTC efficiency with PR is the same as the rate of 2. This seems to be a reasonable choice from a
decrease of the RBC efficiency. Using the ideal cost perspective with ~0.50 percentage
Brayton cycle formulas, which can be found in points sacrifice in performance vis-à-vis
any standard thermodynamics textbook such as the maximum ZCC that has a ~50% larger
[56], the rate of increase of BTC efficiency with PR. In fact, the proximity of the two max-
increasing PR is ima is even closer in the presence of turbine
Natural Gas Power 273

Natural Gas Power, Fig. 14 PR for maximum CC effi- are from Table 2. The inset illustrates the calculation of
ciency and maximum BTC-specific work (Data points are maximum ZCC point using Eqs. 24, 25, and 26 in the text
from Ref. [6]). Cycle maximum temperature or T3 values

cooling so that this is a very reasonable in Fig. 15 (by about 10 percentage points). Spon-
trade-off indeed. sored by the DOE’s National Energy Technology
3. Sixty-five percent and 70% net CC efficien- Laboratory (NETL), the Vision 21 program has a
cies require uncooled turbines with TIT of goal to develop, by 2015, the core modules for a
2,500
F and 3,000
F, respectively. fleet of fuel-flexible, multiproduct energy plants
that boost power efficiencies to more than 60%
(75% on an LHV basis for gas-fueled plants), emit
OEM data for combined cycle performances virtually no pollutants, and with carbon seques-
based on five 50 Hz (3,000 rpm) heavy-duty tration release minimal or no carbon emissions.
industrial gas turbines from a recent trade publi- Early in the 2000s, published claims about the
cation [6] (see Table 2) are given in Table 4. The future of the CC plant performance at the end of
data is for two major combined cycle configura- the first quarter of the twenty-first century stated a
tions, which are labeled as 1  1 (i.e., one GT and goal of 65% with the possibility of over 70% with
one ST) and 2  1 (i.e., two GTs and one ST). an air-cooled GT [58]. The uncooled GT-based
Using the GT performance in Fig. 12 as a basis, CC efficiency curves and the current SOA data in
CC performance curves can be developed using Fig. 15 underline the significant hurdles to be
the exergy-based bottoming cycle calculations in overcome in terms of cooling flow reduction in
[57] (see Fig. 16). pace with increasing firing temperatures in order
2010 SOA in CC performance with advanced to achieve those goals. The gap between the
gas turbines corresponded to a range of 57% to uncooled GT TIT (i.e., same as RIT) requirement
~60% with TIT values in the range of for 65% ZCC and the current material capability is
2,600–2,800
F (first edition of this chapter). nearly 800
F (~450
C). The current trend in
This is a far cry from the theoretically possible superalloy capabilities, considering the compet-
values in the absence of turbine cooling as shown ing tendencies in strength (i.e., creep rupture and
274 Natural Gas Power

Natural Gas Power, Fig. 15 Uncooled GT-based CC calculation of maximum ZCC point using Eqs. 24, 25,
efficiencies (note that T3  TIT). The data (symbol D) is and 26 in the text. Component polytropic efficiencies are
from Table 4. The other data is from [35] for the SOA air- assumed to be 92.5% and 90% for the compressor and
cooled GT (TIT = 2,826
F). The inset illustrates the turbine, respectively

Natural Gas Power, Table 4 OEM data for combined cycle performances (1  1 and 2  1) of 50 Hz heavy-duty
industrial gas turbines [6] (Gas turbine data is in Table 2)
11 21
GT ST CC ST CC
Turbine MW MW MW  (%) MW MW  (%)
A 285.0 141.0 415.1 57.9 282.0 830.2 57.9
B 255.6 141.8 390.8 56.7 289.2 786.9 57.1
C 279.2 157.9 412.9 58.0 315.3 825.4 58.0
D 292.0 144.7 423.0 58.4 290.3 848.0 58.5
E 312.1 157.3 459.0 59.5 317.7 921.1 59.7

fatigue limits) and corrosion resistance, does not acceptable NOx and CO emissions. It is thus fairly
seem to be likely to alleviate too much of that [59]. obvious that an ZCC goal of 65% and higher in a
A combination of advanced materials including Brayton-Rankine CC power plant is a highly
ceramic matrix composites, metal foams unlikely proposition in near future. This is more
(to enable the upper limit of film cooling, i.e., or less the conclusion reached by Rao et al. [61]
effusion or transpiration cooling) [60], and exter- who investigated the plant system configurations
nal cooling (most likely via steam) accompanied that might be able to meet the performance and
by bottoming cycle advances is imperative to even emission goals of Vision 21 program. They con-
approach the stated efficiency goals with cluded that hybrid systems integrating fuel cells
Natural Gas Power 275

Natural Gas Power, Fig. 16 “Real” combined cycle technology curves for different PR-RIT combinations. They are
based on the gas turbine technology curves in Fig. 12 (Also shown is data from Table 4 (as published in Ref. [6]))

with gas turbines, e.g., SOFC/GT hybrid system and, most likely, 61+%. Finally, in early 2016, a
[62], were required to meet the program goals. J class gas turbine was reported to break the world
There is little doubt about the very attractive effi- record with 62.22% CC efficiency. Based on the
ciency of the SOFC/GT hybrid systems, e.g., publicly available information, it is not clear
65–70% for atmospheric cells and 74–76% for whether this is net or gross efficiency (or corrected
pressurized cells [63]. What is really doubtful is to ISO conditions or not). Whatever the case may
whether they will be competitive with gas turbine- be, there is little doubt that a J class gas turbine is
based systems for large-scale electric power gen- readily capable of 61% CC efficiency (net ISO) or
eration considering their low power density, e.g., even higher with favorable site conditions (i.e.,
less than 5 kW/m2 (cf. ~3,000 kW/m2 for a large access to a natural coolant source such as river,
GT), and high cost, e.g., about $1,500–3,000/kW lake, ocean, etc., for very low steam condenser
in 2010 dollars [64] (cf., $350–500/kW for a GT pressures). As predicted originally, while significant
depending on its size). progress has been made in a span of less than a
decade, there is still a long way to go to reach 65%.
2016 Update The current status of the technology with
As one would expect, thermodynamics did not advanced F- (also known as H) and J class gas
change in the 6-year period since the preceding turbines is summarized in Table 5 based on data
section was written. After 60+% net CC efficiency extracted from Gas Turbine World’s 2016 Perfor-
was first achieved by an H-Class gas turbine mance Specs (32nd Edition), a trade publication
(60.75% to be exact – with a steam condenser issued every year.
pressure of 0.6 in. of mercury), several other The evolution of gas turbine CC technology
H-Class power plants achieved that performance over the last three decades can be gleaned from the
276 Natural Gas Power

Natural Gas Power, Table 5 2016 OEM data for combined cycle performances (1  1 and 2  1) of 50 Hz heavy-duty
industrial gas turbines (SC: steam cooled)
11 21
GT ST CC ST CC
Turbine MW MW MW  (%) MW MW  (%)
F+/H 60 Hz 269.1 130.4 394 60.4 263.4 792 60.7
F+/H 50 Hz 337.8 178.7 508 61.1 357.5 1,033 61.4
J 60 Hz 334.7 180.5 509 62 364.2 1.021 62.3
J 50 Hz 502.8 280.8 774 62.7 562 1,552 62.8
J 60 Hz (SC) 322 148 470 61.5 298.9 943 61.7

Natural Gas Power, Fig. 17 Evolution of GTCC tech- refer to the rating averages from respective annually
nology in terms of net efficiency. (TMI Turbomachinery published handbooks.) Data points 1–8 refer to individual
International, GTW Gas Turbine World, data points A-D CC plants, which are in commercial operation

plot in Fig. 17, which is an updated version of the having a respectable baseload performance rating,
original published in [65]. the following operability features became imper-
Increasing adoption of renewables – with their ative for a competitive CC product offering:
intermittency and unpredictability – has signifi-
cantly changed the operating duties required of • Fast start after overnight and/or weekend shut-
combined cycle power plants. Recently, CC downs with high starting reliability and good
power plants are primarily operated cyclically, at equipment life, with low start-up emissions
varying load factors with fairly large numbers of • Ability to run at low loads within applicable
shutdowns and start-ups. Therefore, in addition to emissions regulations
Natural Gas Power 277

Natural Gas Power, Fig. 18 Typical gas turbine start-up Purge credit shaves off 12 min of that time. Eliminating GT
diagram (conventional and fast versions). Conventional hold saves another 15+ minutes. With fast start feature,
start with GT hold can take up to 50 min to reach MECL. MECL point is passed in less than 10 min

• Fast load ramping (or start from standstill) to its full load at full speed (FSFL). Final steam
upon demand and at short notice, exploiting temperature control is accomplished by terminal
unforeseen market opportunities (such as loss attemperators (desuperheaters). Until steam
of load caused by renewable units going unex- temperatures reach acceptable levels for admis-
pectedly off-line) sion into the ST, steam is bypassed via a route
including the reheat superheater so that the latter
Major OEMs developed technologies under is pressurized and “wet” (i.e., cooled by steam
different trade names to facilitate aforementioned flow obviating the need for expensive alloys).
fast start capability [66]. The prime enabler is the Conventional and fast start differences are
thermal decoupling of gas and steam turbine start illustrated by the loading profiles depicted in
processes, which hinges on elimination of direct Fig. 18 [67].
HRSG steam temperature control via GT load and
exhaust energy [67]. In particular, GT is started Today’s SOA in CC Power Plants (2010)
and rolled to full speed at no load (FSNL) at the A schematic diagram of a CC system is shown in
maximum rate dictated by the size of static starter, Fig. 19, which depicts the four major components:
shaft torque limit, and particular DLN combustor gas turbine, steam turbine, condenser, and the heat
limits among others. Following synchronization, recovery steam generator (HRSG). Not shown is
GT is loaded as fast as possible first to its mini- the cooling tower (CT), which cools the con-
mum emissions-compliant load (MECL) and then denser cooling water. Usually, the heat rejection
278 Natural Gas Power

Natural Gas Power,

Fuel
Fig. 19 Three-pressure, Fuel Heater
reheat (3PRH) combined
cycle diagram (Source: GE
Energy [68])

G Heat Recovery Steam Generator

Gas
Turbine
IP
Hot Reheat HP Cold Reheat LP

G Legend
Steam
Steam Turbine Water
Gas
Condenser
Fuel
Condensate Pump

system (including condenser, CT, and circulation that theoretical maximum, which can only be
pump) and feed pumps along with myriad smaller achieved in a Carnot engine [57]. For rapid esti-
heat exchangers (such as the GT fuel gas heater in mations, this translates into an ST power output
the upper left corner in Fig. 19) are collectively that is about 50% of the combined power output
referred to as the “balance of plant” (BOP). of all GTs in the plant.
The equipment that combines the GT (i.e., As the fundamental CC efficiency relationship,
Brayton cycle) and ST (i.e., Rankine cycle) is Eq. 23, and the graphical representation in Fig. 13
the HRSG. Hot GT exhaust gas (>1,100
F for make it very clear, the cycle’s heat rejection tem-
modern F-Class units) transfers heat to conden- perature is of prime importance. This temperature
sate and generates superheated steam, typically is dictated by the steam condenser’s operation
at multiple pressure levels, which is used in the pressure, which is significantly below the atmo-
ST for power generation. Current SOA is a three- spheric pressure. For the best performance, lowest
pressure system with reheat (i.e., heating of HP possible steam condensation pressures, around
steam turbine exhaust, referred to as “cold 1.5 in. of Hg or even lower, are imperative.
reheat,” to the same temperature as the main or Achieving those low-pressure levels requires
HP steam) with steam temperatures of 1,050
F striking a delicate balance between heat rejection
or higher. Highest steam pressure at the ST inlet equipment size, cost, type, and site ambient con-
(throttle) is around 1,800–2,500 psia, beyond ditions. The key heat rejection system equipment
which the pressures at the generation point in is the steam condenser. There are many types of
the HRSG exceed the critical pressure and one condensers that are used to handle the heat rejec-
would require a “once-through” steam genera- tion from the CC plant. Depending on the site
tion system as opposed to the existing “drum- ambient conditions, cooling medium availability,
type” system. From the fundamental thermody- environmental regulations, and customer-specific
namics discussed earlier, the theoretically maxi- economic criteria, a suitable choice can be made.
mum (net) power output that can be obtained The major options are listed below.
from the Rankine bottoming cycle of the CC
plant is exactly equal to the exergy of the GT 1. Water-cooled surface condenser
exhaust gas [55]. Modern CC plants with – Once through or open loop (utilizing a nat-
advanced F-, G-, and/or H-Class GTs and 3PRH urally available cooling water source such
steam systems are capable of achieving ~72% of as lake, river, ocean)
Natural Gas Power 279

– Closed loop with a natural (very rare for CC systems preclude the optimal condenser pres-
plants) or mechanical (forced or induced sures and optimistic performance projections
draft) cooling tower (CT) displayed in Fig. 16.
2. Dry, air-cooled (also known as the “A The key to affordable and sustainable large-
Frame”) condenser scale electric power generation is economies of
3. Air-cooled wet surface (hybrid) or “plume scale that can be achieved by ever-larger prime
abatement” condenser movers (gas and steam turbines of 300+MW) and
4. Direct contact condenser with dry CT plant blocks. Modern CC plants include one or
(Heller) system more GTs with a single ST, either in a “multi-
shaft” or “single-shaft” configuration (e.g., see
In general, water-cooled systems, in particular Figs. 20 and 21). The most common multi-shaft
open-loop systems near a relatively cold natural block is 2  1 (i.e., two GTs and one ST) with
source of cooling water (e.g., in coastal locations power ratings of 600–900 MW. A few larger
in the northern hemisphere), offer the best per- plants in 3  1 or 4  1 configuration or in
formance, i.e., the lowest condenser pressure. multiple blocks of 2  1 are also available (e.g.,
How low in condenser pressure one can go is see Fig. 20). Single-shaft (i.e., 1  1) plants with
essentially a question of plant economics. power ratings of 300–500 MW per “block” can
Recently, permitting requirements in conjunction also be combined into very large generating sta-
with environmental considerations and water tions via multiple blocks. For a representative
scarcity in many geographic locations render selection of modern CC power plant configuration
water-cooled heat rejection systems infeasible. options and performances, the reader is referred to
Significant size and cost of “dry” or air-cooled the white papers listed in [68–72].

Natural Gas Power, Fig. 20 Illustration of a 2x207FA (crossover pipe and downflow LP section are clearly visi-
(two Frame 7 60 Hz GTs with single ST) NG CC power ble). Note the elevation of the ST structure to accommodate
plant. From left to right, for each GT-HRSG “block” one the steam condenser. Partially visible to the right is the
can see the GT inlet air filter housing and duct (on top of the multicell mechanical cooling tower. In the lower right
GT enclosure), HRSG with drums on the top, and the stack. corner are two distillate oil (GT backup fuel) tanks
Two STs are located to the right of the GT-HRSG blocks (Courtesy: GE Energy)
280 Natural Gas Power

Natural Gas Power,

Fig. 21 A single-shaft
combined cycle train
(Courtesy: GE Energy)

Even better economies of scale are achieved that such a cycle would not be feasible for most
via the “reference plant” approach, in which gas turbine sizes and CC configurations
OEMs develop pre-engineered modules that can [73]. However, very large H-/J class gas turbines
be shipped and installed anywhere in the world with exhaust gas flows pushing beyond 2,000 lb/s
with minimal customization to accommodate dif- (more than 900 kg/s) in a 2  1 or 3  1 multi-
ferent site and customer requirements. This shaft configuration can make a supercritical steam
ensures that natural gas-burning GTCC power turbine feasible. Whether a several gigawatt CC
plants are not only the most efficient means of power plant with a supercritical bottoming cycle
generating electric power available to us but also will be commercially offered in near future
the most cost-effective (e.g., at a turnkey capital remains to be seen. For typical CC power plant
cost of $600–1,000/kW and cost of electricity of ratings available at the time of writing, please refer
a few cents per kWh) and fastest to build (e.g., to Table 5.
less than 24 months of construction time). On top
of all that, GTs with modern “dry low NOx”
Cogeneration
combustors are also the cleanest fossil fuel-
Discussion thus far has focused on production of
burning technology in terms of emissions. This
electricity from natural gas. When electricity is the
aspect will be explored in more detail in the
sole product from the thermal cycle, the heat
following section.
rejected to ambient is termed “waste heat”
because it is no longer economically useful for
2016 Update generation of additional power. In the larger con-
Three-pressure, reheat steam bottoming cycle is text of a modern industrial society, one finds that
still the SOA. High-pressure and hot reheat steam low temperature heat of the quality least useful for
temperatures of 1,112
F (600
C) have been power generation has many economically impor-
making inroads into commercial offerings utiliz- tant uses. It is thus common in industrial and
ing advanced H- and J class gas turbines with high urban settings to integrate a power plant with
exhaust temperatures. HRSGs with once-through another process such that heat energy for that
HP evaporator design have been deployed com- process is provided in whole or in part by the
mercially by various OEMs, especially for CC power cycle after the highest-quality fuel energy
power plants with “fast start” capability. This has produced some useful work. Such installa-
technology makes a “supercritical” steam tions are termed cogeneration systems or com-
bottoming cycle conceptually possible. It has bined heat and power (CHP) systems and can be
been shown via rigorous thermodynamic analysis very efficient from an environmental and fuel use
Natural Gas Power 281

Natural Gas Power, Table 6 A sampling of natural gas- • Large industrial gas turbines ranging from ~20
fired cogeneration opportunities [74] to 500 MW electrical output are optimized for
Industrial best combined cycle efficiency and have
Paper and board Pharmaceuticals and fine exhaust energy of about 60% of heat input
manufacture chemicals with exhaust temperatures between about
Ceramics Brewing, distilling and
1,020
F and 1,220
F (550–660
C). This
malting
Textile processing Food processing
class of machine is well suited to process heat
Oil refineries Minerals processing demands that are large in an absolute sense but
Timber processing and Horticulture and glasshouses small or intermittent in relation to electrical
paper power demand, (0 to ~30% of gas turbine
Buildings heat input). This level of heat energy supply
Hotels District heating and cooling can readily be accommodated by modifying
Airports Hospitals the power generation combined cycle system
Supermarkets and large College campuses and to include the capability to extract or divert
stores schools steam from the bottoming cycle steam turbine
Individual houses Office buildings
to the process. This is a common solution for
district heating which may be highly variable
and seasonal.
perspective. Table 6 provides a sampling of • Aeroderivative gas turbines ranging from ~5 to
cogeneration opportunities with a synergistic 60 MW electrical output are optimized for best
combination of heat and power needs. simple cycle efficiency and have exhaust
In the ideal case, the heat energy available from energy of about 55% of heat input with exhaust
the prime mover (engine or thermal cycle) will temperatures of about 850
F (455
C). These
always match the needs of the customer process. machines are large enough to support substan-
In practice, this is challenging since many process tial process heat loads and are often configured
heat demands are variable (such as building or in a system without a steam turbine to save cost
district heating). The electrical loads can also be and make a higher portion of gas turbine
out of phase with process heat demands (e.g., an exhaust energy available to process. In many
electric grid with high but unsteady wind or solar industrial cogeneration facilities, it is more
renewable power contributions). These variations important to match process heating demand
in electrical load versus process heat loads can be than to meet a specific electric load. The gas
accommodated with appropriate system design turbine load is thus varied to follow process
and operating strategies. A variety of cogeneration energy demand independent of electric load.
plant options are available between the extremes Any excess electricity generated is sold to the
of a non-cogeneration facility, e.g., fired heater or grid, and any power shortfall is purchased. If
boiler, with no power generation at all (e.g., a the design process heat demand is high in
home heating system capturing the heat content relation to electric load, or may occasionally
of the fuel but wasting its potential to do work need to be much higher than the unfired gas
[75]) and the aforementioned power generation turbine exhaust energy, the system will need to
combined cycle with no process heat demand, include a supplemental burner in the HRSG,
which has accordingly been optimized for produc- which can be used to boost process heat export
tion of electricity. capability. A burner may also be used in com-
Primary considerations in cogeneration system bination with part load operation of the gas
design are the size of the plant, the relative need turbine to maintain process heat export while
for heat versus electrical power, and the electricity reducing electricity production to match
and process heat demand variability. The follow- its load.
ing bullets summarize the range of system design • A micro gas turbine ranging from ~0.03 to
solutions. 0.3 MW electrical output may be selected if
282 Natural Gas Power

Natural Gas Power, Table 7 Typical heat-to-power ratios and efficiencies for GT-based CHP systems

Notes to the table:

1. y is heat to power ratio in kWth / kWe
2. b is heat quality, which is the ratio of export exergy to export energy
3. Zel. is electric power generation efficiency
4. Zex is gas turbine + burner exergetic efficiency. Represents the combined performance entitlement with an ideal
Carnot bottoming cycle.
5. Zc,l is first law cogeneration efficiency (energy-to-fuel input)
6. Zc,ll is second law cogeneration efficiency (exergy-to-fuel input). Shortfall vs. Zex reflects exergy losses for steam
generation and steam to power conversion in the bottoming cycle.
7. Emissions are in grams per kWh of energy (A) and exergy (B)

the heat load and/or electric load is small. controlled steam extraction. Efficiency at low
Microturbines typically employ uncooled process demand is inferior to gas turbine-based
regenerative cycles for best simple cycle effi- systems, but fuel flexibility is generally better.
ciency and have exhaust energy of about 75%
of heat input and exhaust temperature of about Table 7 shows some examples of gas turbine-
530
F (275
C). These units are too small to based natural gas-fired cogeneration systems.
economically justify a bottoming cycle steam These are presented to elucidate the behavior of
turbine so that they fall into the unfired or such systems and draw conclusions regarding the
supplementary-fired heat recovery-type system role of cogeneration in promoting sustainable use
configuration. of natural gas. Columns tabulate net electric
• In the size range between microturbines and power export, export heat energy to power ratio,
aeroderivatives, a variety of small industrial quality of export heat expressed as a ratio of its
gas turbines and other prime movers such as a availability to produce work (exergy), electric
spark ignition reciprocating engine or fuel cell power generation efficiency, gas turbine + burner
may also be considered. In the range from ~10 exergetic (second law) efficiency, cogeneration
to 150 MW direct-fired steam Rankine systems efficiency expressed on both an energy and exergy
with steam turbine extraction for process heat basis, and specific CO2 emissions also expressed
can also be an attractive configuration that is on both an energy and exergy basis. Export heat is
well suited to highly variable process heat provided as steam in all cases.
needs. Power can be varied across a wide Cogeneration efficiency Zc,I in Table 7 is the
range with modulated fuel consumption, and straight ratio of the sum of power (kWe) and heat
heat export can also be widely varied via (kWth) products to the fuel burned in the plant on
Natural Gas Power 283

an LHV basis. This is a commonly used help frame the influence of steam extraction to
(sometimes with fuel input on an HHV basis) process on Zc,I and Zc,II.
measure, which is ultimately misleading and ther- Case 1 is a microturbine-based design with no
modynamically incorrect because it does not dif- supplemental fuel addition to exhaust heat recov-
ferentiate between the “quality” of the different ery. Export energy is high in relation to power
products (steam and electricity). As seen in production since the engine is only 25% efficient,
Table 7, especially with supplementary firing, it but export heat quality is low since exhaust tem-
can reach impossibly high values over 100%. perature is a modest 530
F (275
C). The low
A more rational measure is the cogeneration quality of the microturbine exhaust energy is also
efficiency per the second law of thermodynam- reflected in the low gas turbine + burner exergetic
ics, Zc,II, which uses the exergy of the thermal efficiency of Zex = 46.8%. Low exhaust temper-
product rather than its energy (i.e., enthalpy) ature also compromises the cogeneration energy
[76]. Exergy, as discussed earlier in the present efficiency (particularly Zc,II ) in this example
article, is a direct measure of the work-generating because only a small portion of the unfired
ability of a material stream via Kelvin-Planck exhaust energy is recoverable to steam.
statement of the second law of thermodynamics. Case 2 adds a supplemental burner fired to
Exergy is a property of a given fluid, and, as 1,600
F (871
C) to boost export energy substan-
such, for two given properties (e.g., pressure tially while also improving its quality. Fuel heat
and temperature), it can be calculated using a energy is now effectively utilized to provide
suitable equation of state such as ASME steam power and process heat (50.9% cogeneration effi-
tables. The relationship between Zc,II and Zc,I is ciency, exergy basis). This is a 10.6% improve-
given as ment versus case 1, but is still modest since the
engine is inefficient. Note also that specific CO2
1þby emissions (exergy basis) improved because the
Zc, II ¼ Zc, I 
1þy incremental efficiency due to increased and
improved heat export is better than the electrical
where y is the heat-to-power ratio of the cogene- efficiency of the gas turbine.
ration plant and b is the ratio of the exergy to Cases 3 and 4 show the corresponding story
energy (enthalpy). Since b < 1 , Zc , II < Zc , I for a 40 MW aeroderivative gas turbine, while
and has typically a value, which is more in line cases 5 and 6 present a 100 MW E class industrial
with power-only combined cycle plant efficien- gas turbine scenario. These gas turbines both
cies. In passing, note that the Public Utility Reg- improve exergetic performance over the micro-
ulatory Policies Act ((PURPA), a US federal law turbine cases consistent with their higher gas tur-
enacted in 1978 intended to encourage more bine + burner exergetic efficiency Zex. The more
energy-efficient and environmentally friendly efficient aeroderivative has slightly higher system
commercial energy production) efficiency, which performance but lower process heat export capa-
is another cogeneration plant efficiency definition, bility versus the industrial gas turbine.
is equivalent to Zc,II with b = 0.5. Even Zc,II is an Case 7 shows the 100 MW industrial gas tur-
inflated value because b is the efficiency of a bine in a combined cycle configuration with steam
hypothetical Carnot engine that converts the ther- extraction to process increased from 0% (power
mal energy to useful shaft work. only CC) to 100% (noncondensing steam tur-
Orange-shaded cases in Table 7 represent cases bine), at fixed steam export conditions (b). Elec-
with no steam turbine in the cycle and all steam tric power generation efficiency falls off as more
generated sent directly to process. b for these energy is diverted to process while cogeneration
cases is consistent with steam temperatures efficiencies improve. Case 8 begins where case 7
achievable from the GT exhaust stream (higher left off with full supplemental firing to 1,600
F
for the supplementary-fired cases. Blue-shaded (871
C). This yields almost 50% increased gen-
cases are power-only combined cycles. These erator output and ~40% increased steam export
284 Natural Gas Power

from the noncondensing steam turbine with very 2. Cogeneration system configuration will be a
little change in plant efficiencies. As supplemen- function of the quality and quantity of heat
tary firing increases, the specific CO2 emissions energy required, its variability, and the rel-
(exergy basis) stay approximately constant. ative heat and power needs. A wide range of
Case 9 is a large modern 500 MW H-Class applications can be served by appropriate
industrial turbine-based design with no supple- system design.
mental fuel addition to exhaust heat recovery. 3. Gas turbine-based systems are most effi-
Export energy is reduced in relation to power cient when not supplementary fired, while
production since the engine is very efficient, and gas turbine combined cycle cogeneration
export heat quality is high since exhaust temper- systems maintain high cogeneration
ature is high at 1,190
F (643
C). The high exergetic efficiency even with supplemen-
quality of the H turbine exhaust energy is also tal firing. This is because fuel energy can
reflected in the high gas turbine + burner exergetic be recovered to power all the way down
efficiency of Zex = 69.05%. This translates into to the temperature level required by the
high power-only Zel and cogeneration Zc,II process user. Best exergetic performance
(exergy-to-fuel input) efficiency for the examples is achieved when all of the exported
in case 10. These mirror the case 7 examples with process steam has first expanded in
the lower performing E class gas turbine. Note the steam turbine unless the required
that maximum heat-to-power y is considerably process steam conditions are very high
lower than for the last case 7 example, again (high b).
reflecting the higher conversion of fuel energy to
electricity for these engines. Figure 22 shows cogeneration’s contribution to
Note that the supplementary-fired scenarios electricity production in European Union
presented here stop at 1,600
F (871
C) consis- countries.
tent with uncooled metal ductwork. Firing tem- The higher penetration seen in four of these
perature can be pushed further by application of countries is attributed to a combination of favor-
refractory or water-cooled walls until the excess able economics and supportive government
oxygen is nearly expended or increased even fur- energy policy [74]. In the United States approx-
ther if supplemental air is also added. High sup- imately 4% of electricity production currently
plementary firing may be required for some comes from cogeneration facilities of which
applications but is less desirable from a resource 71% are fueled by natural gas (e.g., see
conservation perspective than a design with little Table 8.2c in [1]).
(or no) supplementary firing. How far could cogeneration penetrate in the
The following general conclusions can be United States and what benefits would accrue?
drawn from this sampling: Just as it is possible to look at the utilization rate
of renewable energy sources such as wind, hydro,
1. Cogeneration efficiency should be and solar in relation to the size and distribution of
expressed on an exergy basis to promote these resources, the thermal power plant heat
the most sustainable use of fossil fuels. available for low temperature end users can be
Direct use of fuel for low-quality end uses quantified to estimate the potential for increased
represents a large waste of the thermody- cogeneration and hence its potential to offset other
namic potential inherent in the fuel. Cogen- fuel use and emissions.
eration efficiency as commonly expressed in The annual US thermal power generation
relation to export energy gives the false totaled 3.7 trillion kWh in 2008 [1]. Factoring in
impression that supplementary firing the mix of natural gas-fired systems, nuclear, and
improves the environmental (CO2) and direct-fired (mostly coal) plants and their respec-
resource (natural gas) utilization tive thermal efficiencies allow estimation of heat
performance. resources available for cogeneration. Table 8
Natural Gas Power 285

Natural Gas Power, Fig. 22 Cogeneration as a share of According to the same source, EU average is 15%, whereas
national power production (1999) (Source: European the global average is 9% (the United States is close to the
Association for the Promotion of Cogeneration [74]). global average). More up-to-date data can be obtained
Based on International Energy Agency’s CHP report from the IEA or other websites
from 2008, Denmark is still number one with about 52%.

Natural Gas Power, Table 8 Annual US thermal energy available supply and rejection, 2008 [1, 77]
Natural gas
Combined Simple
Coal Petroleum cycle cycle Cogeneration Nuclear Total
Electricity generation 1,986 46.2 562 122 200 806 3,721
(billion kW-Hr)
Heat rate (Btu/kW-Hr, 9,884 9,884 7,598 11,526 11,156 10,488
HHV)
Heat input (trillion Btu) 19,628 457 4,268 1,402 2,227 8,456 36,062
Approximate heat 10,889 254 1,408 847 1,159 5,451 19,730
rejection to ambient or
process (trillion Btu)
Heat rejection beta (b) 0.20 0.20 0.20 0.44 0.44 0.20

shows an estimated breakdown of heat rejection steam can be effectively distributed is limited to
from thermal power plants in the United States. about a 10 km radius, while hot water for district
While cogeneration with direct-fired or heating can be distributed out to about a 20 km
nuclear-fueled Rankine systems is certainly pos- radius. For this reason, cogeneration systems
sible, the sheer size of these plants makes it tend to be relatively small in terms of electrical
difficult to economically match them to end output, which in turn aligns well with the
users that they are large enough to benefit from NG-fired gas turbine and gas turbine combined
the full heat energy available. The distance that cycle. Natural gas accounts for 883 billion kWh
286 Natural Gas Power

or 23.7% of total electricity generation [1, 77]; 2008 [1]. Figure 24 shows the incremental retail
199.7 billion kWh or 22.6% of which comes from electric load distributed to each sector and gives a
cogeneration facilities (e.g., see Table 2.1 in breakdown between residential and commercial
[77]). Taking the data in Table 7 for cases 5 and sectors. The natural gas contribution to the resi-
6 as a typical system configuration, it is possible dential and commercial sectors is 8.2 trillion Btu
to estimate the current fuel savings from natural [1]. Assuming that its primary use is for space
gas-fired cogeneration as follows: heating, it can be seen that it far exceeds the
available waste heat from gas-fired power plants
Natural gas-fired cogeneration (GT + supplemental tabulated in Table 8. It is difficult to make the
firing) case that small (micro) cogeneration systems
Electricity generation (billion kWh, 2008) 200 could be aligned to the heat and power needs at
Combined cycle heat rate (Btu/kWh, HHV) 8,305 the individual house level, though work on fuel
GT heat input (trillion Btu, 2008) 1,658 cell systems for this purpose is active. The com-
System heat rate (Btu/kWh, HHV) 11,156 mercial sector and high-density residential (e.g.,
Burner supplemental heat input (Trillion Btu, 569 apartment complex) do offer real potential for
2008)
expansion. In either case the performance of the
System heat input (trillion Btu, 2008) 2,228
cogeneration system needs to be high enough to
Heat export to process (trillion Btu, 2008) 1,159
improve on the fuel use and environmental per-
Equivalent heat input to meet process (trillion 1,302
Btu, 2008) formance of an efficient utility scale power plant
Cogeneration natural gas savings (trillion Btu, 733 in combination with direct use of natural gas for
2008) heating. While increased use of natural gas-fired
(Line 8 minus line 5) combined cycles in cogeneration systems can
provide significant near-term environmental and
The fuel savings from application of cogene- fuel use benefits; the worldwide trend toward
ration represent ~25% reduction in fuel burn ver- increasing use of renewables to provide electric-
sus making the power with a gas turbine ity acts as a countercurrent to application of
combined cycle and supplying the heat indepen- cogeneration to support any use that be served
dently with a fired boiler. CO2 emission reduc- by green electricity.
tions follow the fuel saving directly if the fuel
saving is booked as natural gas. If booked at the
current fuel mix for US electricity production, the Combustion and Emissions
CO2 savings are higher by about 33% or higher by There are different types of GT combustors that
75% if booked as reduced coal-fired generation. have been used at one time or another by different
Looking at the heat rejection from natural gas- OEMs in their products:
fired power generation in the United States in
Table 8, in comparison to current heat export 1. Silo (obsolete now, once used by Siemens
from natural gas-fueled cogeneration, it appears and former ABB)
that the amount of cogeneration could perhaps be 2. Annular (aeroderivatives such as GE
doubled without a major shift in fuel use patterns. LM6000)
Reaching the electricity contribution levels seen 3. Can or tubular (heavy-duty industrial GTs
in some small European countries would likely by Siemens, GE)
require more distributed generation with replace-
ment of large coal-fired power plants with more In the following discussion a parameter will
small cogeneration systems (likely gas fired) appear frequently, i.e., the equivalence ratio, f,
proximate to potential end users. which is the ratio of the actual fuel-air ratio, f, to
Turning now to the potential customers for the stoichiometric fuel-air ratio. The latter is the
heat energy, Fig. 23 shows the primary energy theoretical value requisite for complete combus-
production by source in the United States for tion. Thus
Natural Gas Power 287

Natural Gas Power, Fig. 23 Primary energy production energy.” “Natural gas” excludes supplemental gaseous
by source, 2008. Total primary consumption is 99.3 qua- fuels. “Coal” includes less than 0.1 quadrillion Btu of
drillion Btu. Renewable energy sources are hydroelectric coal coke net imports (Source: Energy Information Admin-
power, geothermal, solar/photovoltaic, wind, and biomass. istration (EIA) [1]). More up-to-date data can be obtained
“Petroleum” does not include the fuel ethanol portion of from the EIA website
motor gasoline – fuel ethanol is included in “renewable

Natural Gas Power, Fig. 24 Primary and total consumption by sector, 2008. See Fig. 23 for a breakdown of sources
(Source: Energy Information Administration (EIA) [1]). More up-to-date data can be obtained from the EIA website
288 Natural Gas Power

Natural Gas Power,

Fig. 25 Single-nozzle
diffusion (SND) combustor

f ¼ f act =f st is critical because of the high speed of the air

stream, even at a fraction of its full value,
When the equivalence ratio is equal to 1.0, the which is an order of magnitude higher than
flame temperature is the highest and the chemical the turbulent flame speed at which the flame
reactions are the fastest. Thus propagates into the fuel-air mixture.
Secondary zone: This is the intermediate zone
1. If f is less than 1.0, the mixture that reacts is where a portion of the remaining combustion
referred to as “lean” or “fuel lean.” air (~30% of the total) is introduced in the mid-
2. If f is greater than 1.0, the mixture that section of the can through holes or cooling
reacts is referred to as “rich” or “fuel rich.” slots in the combustor liner. The hot combus-
tion gas from the PZ gets thermally soaked,
In either lean or reach combustion, flame tem- and the combustion is completed. The air that
perature is lower, and chemical reaction is slower flows between the liner and flow sleeve and
than their stoichiometric counterparts. enters the can forms a film to cool the
The can-type combustor can be divided into liner wall.
three zones (see Fig. 25): Dilution zone: This is where the remaining air
enters the can through the metering holes and
Primary zone: This is the combustor reaction or cools the gases to the desired temperature.
flame zone where fuel is injected and mixes A design to promote turbulence and thorough
with air (about 15–20% of the compressor mixing of hot and cold streams is essential to
discharge air is introduced). In order to cap ensure a homogeneous temperature distribu-
the combustor exit temperatures at a level, tion and prevent hot streaks.
which downstream turbine HGP components
can tolerate, the requisite value of f is typically A successful combustor design should satisfy the
0.4–0.5, which is too lean for stable and effi- following criteria:
cient burning. This is why only a portion of the
combustion air is introduced into the primary 1. Efficient and stable operation, which com-
zone. High-temperature rapid combustion prises the following characteristics
takes place in the PZ. The design of this part 2. Complete combustion at or near stoichio-
of the combustor must facilitate highly turbu- metric flame temperature
lent and recirculating flow pattern to anchor the 3. Flame stability and reliability at high
flame at the exit of the fuel injector nozzle. This turndown
Natural Gas Power 289

(a) Ignition 3. Carbon dioxide (CO2), which contributes to

(b) Cross fire (propagation of flame from fired greenhouse effect and global warming
chambers with spark plugs to the unfired 4. Carbon monoxide (CO), which is toxic and
chambers via tubes interconnecting the asphyxiating
annular combustion chambers or cans) 5. Unburned hydrocarbons (UHC), including
(c) Lean blowout volatile organic compounds (VOC), which
4. Operational stability and reliability at high contribute to urban smog
turndown 6. Particulate matter (PM), which is suspected
(a) Dynamics (noise), i.e., large amplitude to cause respiratory diseases
pressure oscillations in combustion
chamber, driven by heat release oscilla- Detailed information on GT emissions and
tions; also known as combustion “hum- control can be found in [80]. Some pollutants,
ming.” These oscillations are destructive i.e., UHC, PM, and SOx, are generally of no
to engine hardware. concern for GTs burning gaseous fuel with the
(b) Pattern factor, which is a measure of the exception of coal gasification SG, which leads to
temperature distribution at the combus- conversion of fuel-contained sulfur to SOx. (Note,
T  Tmean
tor exit defined as PF ¼ Tpeakmean  Tinlet
however, that in some cases contamination of fuel
(c) Pressure drop. in the pipeline, e.g., rust particles, can be a prob-
5. Low emissions in compliance with regula- lem for HGP components.) In any event, currently
tory laws there is no available GT combustion technology to
prevent or control SOx emissions. Similarly, CO2
Achieving these goals simultaneously is very is a direct product of the complete combustion
difficult due to the counteracting impacts of process and cannot be controlled or reduced
design parameters on each criterion. The design within the GT control volume (except, of course,
history of the GE heavy-duty industrial gas tur- increasing the thermal efficiency of the GT, which
bine combustors can be found in [78]. The design will reduce the amount of fuel consumed and CO2
evolution can be summarized in three major generated for a given power output). Pre-
designs: combustion or post-combustion capture and stor-
age of carbon dioxide, known as CCS, is a
1. Single-nozzle diffusion (SND) combustor currently hot topic of research and development.
2. Multi-nozzle, quiet combustor (MNQC) A key pollutant emitted by the GT is NOx.
3. DLN or dry low NOx combustor There are two major sources of NOx:

The design evolution was primarily driven by 1. Atmospheric N2 in post-flame gases, which
the desire to satisfy the increasingly stringent leads to formation of NO and NO2 via three
emission control requirements with improved effi- principle mechanisms
ciency via higher combustion and turbine inlet (a) Thermal NO (extended Zel’dovich
temperatures (TIT). In order to clarify this, one mechanism)
must be familiar with the chemical mechanisms of (b) Nitrous oxide mechanism
NOx production, which is one of the most critical (c) Prompt NO mechanism
pollutants present in GT exhaust gas. Principal 2. Organically bounded nitrogen in fuel or
pollutants emitted by a GT are as follows [79]: fuel-bound nitrogen (FBN)

1. Oxides of nitrogen (NOx), which are toxic, The dominant NOx formation mechanism is
contribute to chemical smog, and lead to the thermal or Zel’dovich mechanism. Although
depletion of ozone in stratosphere not a significant source of NOx for NG, FBN in
2. Oxides of sulfur (SOx), which are toxic and low-quality liquid fuels and LCV gases with hot-
corrosive gas cleanup (via ammonia) can be high enough to
290 Natural Gas Power

merit attention. NOx (and CO) formation is a

function of three primary parameters:

Increasing NOx
1. Residence time in the combustion zone
2. Chemical reaction rate
Increasing p
3. Mixing rate

These three reaction parameters can be related

to the turbine operating conditions and combustor
size [79]. In particular,
Increasing T st
p  Vc
Residence time / (27)
_  Tpz
m Natural Gas Power, Fig. 26 Effect of pressure and stoi-
chiometric flame temperature on NOx
Reaction time / pn  expðk  Tst Þ (28)
flame temperatures near the stoichiometric value
where (i.e., f ~ 1.0) and negligible at leaner mixtures
with lower flame temperatures. Consequently,
• Tst is stoichiometric flame temperature (or, effective control of the NOx production in the
equivalently, f or f or firing temperature) combustion process is a matter of reducing the
• Tpz is the average gas temperature in the reac- flame temperature. There are essentially three
tion or primary zone ways to accomplish this:
• Vc is combustion volume
• p is combustion pressure (raised to a power of n) 1. Design a primary (reaction) zone with
• m_ is the combustor mass flow rate leaner fuel-air mixture
2. Injection of a diluent, i.e., water or steam,
Therefore, NOx production rate can be into the primary zone of the combustor to
expressed in an empirical formula that can be act as a “heat sink” and reduce the flame
written as temperature
3. Staged combustion with two distinct reac-
Vc
NOx / (29) tion zones, which lead to the modern “lean-
_  Tpz 
m  expðk  Tst Þ
p1þn
premixed” designs, which are referred to as
dry low NOx (or DLN).
The relationship between NOx production and
key parameters per Eq. 29 is shown qualitatively
The dramatic impact of reducing f, i.e., lean-
in Fig. 26. In general, NOx values are reported in
ing out the PZ, on flame temperature and NOx
parts per million (ppm) on a dry basis (usually
production can be seen in Fig. 27. Leaner fuel-air
expressed as ppmvd) and corrected to 15% O2 in
mixtures lead to lower values flame and gas tem-
combustion air. It is extremely important to make
peratures and favorably impact the NOx produc-
sure that the basis of the reported number is
tion rate via shorter residence and reaction times.
clearly defined; otherwise comparisons of data
Unfortunately, things get complicated rapidly.
from different sources can be misleading. The
For starters, the full operating range of an indus-
expression for correction is as follows:
trial GT combustor requires significant turndown
½NOx   5:9 ratios, e.g., 40 to 1 in fuel flow, 30 to 1 in air flow,
½NOx 15%O2 ¼   (30) and 5 to 1 in fuel-air ratio. Obviously, the com-
20:9  O2, meas
bustor design must ensure stable and efficient
combustion across the entire operating range.
The trends in Fig. 26 show that the impact of The early single-nozzle (diffusion flame) combus-
pressure on NOx formation is stronger at high tors, with direct injection of fuel and air into the
Natural Gas Power 291

Natural Gas Power,

GT11657B
No.2 Oil, 10 ATM
Fig. 27 Thermal NOx Air Preheat 590 K (600°F)
production rate (Source: GE Temperature
Energy [83]) 2500
4000 High CO 300
Emissions

Flame Temperature
Rate of Production
2000 NOx of Thermal NOx
3000
(k) 200 d NO
dt
(°F) High Smoke
(ppmv/MS)
Emissions
1500
2000
100

1000

0.5 1.0 1.5

Equivalence Ratio
Lean Rich

PZ, were designed for combustion at or near stoi- increased maintenance frequencies. High levels of
chiometric conditions. For a combustor designed water injection excited discrete dynamic pressure
for optimal full load operation at f ~ 1, combus- tones within the combustor. Frequencies within
tion at lower loads is bound to be very lean. Thus, the range of the hardware’s natural frequency
further leaning out of the PZ for NOx reduction is resulted in combustor damage. In terms of
a very limited knob. In fact, initial efforts to dynamic pressure response, steam is a better dilu-
reduce NOx via leaning out reaction zone of a ent than water, although a larger mass flow rate is
SND combustor did not achieve more than 20% needed. Significant reduction in NOx production
reduction [81]. is achievable via steam injection, e.g., ~60% at a
When it became clear that meeting increas- steam-to-fuel ratio of unity [81, 82].
ingly stringent pollution control regulations via The other significant drawback on water or
leaner reaction zones in SND combustors was steam injection for NOx abatement is the overall
insufficient, other methods to reduce flame tem- simple or combined cycle performance penalty.
peratures were investigated. Injection of water or Even though the turbine output increases via
steam into the PZ to create a “heat sink” for increased mass flow, the associated increase in
temperature reduction proved successful. Starting fuel consumption is detrimental to the heat rate.
in 1979, when regulations required that NOx Furthermore, since the quality of injection water
emissions be limited to 75 ppmvd, SND combus- must be high, i.e., of boiler quality, to prevent
tors with water or steam injection have been suc- deposits and corrosion in HGP components, addi-
cessfully used to meet or exceed these tional water treatment system (WTS) cost for sim-
requirements successfully [78]. However, SND ple cycle applications must be considered. In
combustors that use water and steam injection combined cycle applications, where steam from
are limited in their ability to reduce NOx levels the HRSG or ST (typically cold reheat steam) is
below 42 ppmvd on gas fuel (65 ppmvd on oil extracted for injection, increased makeup flow
fuel). The main reason is the practical limit to the rates and ensuing WTS size increase and water
amount of diluent injection into the reaction zone consumption are additional concerns. Even
before the increase in dynamic pressure oscilla- disregarding the combustion noise and life issues,
tions (i.e., noise) starts imposing unacceptable it is easy to imagine that ever-increasing turbine
penalties in terms of reduced hardware life and firing temperatures for improved cycle efficiency
292 Natural Gas Power

Natural Gas Power,

NOx, ppmv
Fig. 28 Influence of PZ

CO, ppmv
temperature on CO and
NOx [79]

Optimal Range

60 15

2240 2420 2600 2780 2960 3140 3320

Primary Zone Temperature, F

would drive the diluent injection flows to such system eschewed the low-speed ignition and
levels where they would literally put the flame blowout problems associated with excessive lean-
out. This is notwithstanding the counterproduc- ing out of the PZ in SND systems at low loads (see
tive effect of reduced cycle efficiency via higher Fig. 29). Furthermore, the MNQC system also
fuel consumption, which defeats the purpose of helped with the dynamic pressure oscillation
increasing the firing temperature in the first place. (i.e., noise) problem by maintaining a flat
Another limit associated with lowering the dynamic response at increased steam injection
equivalence ratio and flame temperature for NOx levels (that is the reason for the moniker
abatement is the increase in CO emissions. The “quiet”), thus enabling the design of combustion
conflicting emission benefits and harms of lower- systems with higher firing temperatures and com-
ing or increasing the flame temperature can be pliant NOx emissions without the problems of
seen in Fig. 28. Carbon monoxide is a measure increased hardware wear and reduced mainte-
of combustion inefficiency, which becomes more nance intervals. Since 1987, MNQC systems
pronounced at leaner combustions. This is due to enabled heavy-duty GTs to meet stringent NOx
shorter residence time at leaner mixtures, which emission requirements, i.e., 25 ppmvd with gas
does not lend itself to complete CO burnout. Thus, fuels and 42 ppmvd with liquid fuels. Increase in
the allowable temperature range for emissions CO and UHC emissions and combustion stability
control is limited to a relatively narrow band of issues at high injection rates, e.g., steam to fuel
~400
F. >2, precluded further improvement with this
When SND combustors hit a practical limit for technology [84].
NOx emissions reduction, even with diluent injec- Reduction in NOx emissions to even lower
tion, designers responded by developing the levels, the need for balancing the low emissions
“multi-nozzle quiet combustion” or MNQC sys- with ever-increasing firing temperatures, and the
tem. For a detailed history of the combustor devel- need to avoid the cycle thermal efficiency penalty
opment, please consult [78]. Utilizing a multi- associated with steam and water injection lead to
nozzle configuration, i.e., six nozzles located the development of DLN combustors. DLN com-
peripherally at the combustor head cap, this bustion system is the natural next step in
Natural Gas Power 293

End Cover
Gas Inlet
Gas Manifold
Water Distributor

Oil Inlet

Oil Distribution Valve

A/A Inlet Duel Fuel

Nozzle
(6 Total)
A/A Manifold
Water Inlet Dilution Holes

Liner
Liner Cap Sleeve
Fuel Nozzle Cup Combustion Holes

Bluff Body X–Fire Tube Conn.

Natural Gas Power, Fig. 29 MNQC combustor (Source: GE Energy [83])

Natural Gas Power,

Fig. 30 Dry low NOx
(DLN) combustor (Source:
GE Energy [84])

combustor development, i.e., a “staged” combus- production at an optimal level, i.e., near the mid-
tion system with lean-premixed fuel-air mixture. dle of the allowable range shown in Fig. 27. The
The primary and secondary combustion zones in first successful DLN system that went into pro-
the staged combustor are operated at the lowest duction, DLN-I, incorporated a multi-nozzle sys-
possible f at high fuel flow conditions (see tem similar to that in MNQC for the PZ and a
Fig. 30). During start-up, acceleration, and low- conventional, large secondary fuel nozzle for the
load conditions, fuel is introduced only in the PZ SZ. At normal (50–100% load) operation, fuel and
with good ignition and blowout characteristics. At air are “premixed” in the PZ, but ignition takes
increasing load levels, fuel is introduced into the place in the SZ with the flame attached to the
SZ at a fuel split to balance the CO and NOx secondary nozzle.
294 Natural Gas Power

Natural Gas Power, Table 9 Cost and performance of nonrenewable electric power generation plants (Source:
Cambridge Energy Research Associates, MA, USA)
Economic Lead Load
life time factor Efficiency Fuel cost Fixed Variable
Capital cost % $/kW-
per kW Years % % (CHP) $/MMBtu year $/MWh
NG-fired $350–500 30 1.5 5–50 38 9–12 4
GT
NG-fired $550–900 30 2 60–85 55 4–5 10–15 2
GICC
NG-fired $600–700 30 2 50–85 40–55 79–84 10–15 2
GTCC
with CHP
US SCPC $1,300–1,800 30 4 60–85 35–42 0.75–1.50 20–25 3
US IGCC $1,700–2,100 30 4 60–85 39–42 30–35 2
Nuclear $2,000–3,000 40 6 85–92 33 0.5 60–70 1.5
Load factor specifies the percent of (equivalent) hours in a year (8,760 h) during which the plant runs at its nominal
baseload. The data reflects the 2004 prices
CHP combined heat and power (also known as cogeneration), SCPC supercritical pulverized coal, IGCC integrated (coal)
gasification combined cycle

DLN technology successfully balanced the example. Nevertheless, a dramatic shift in the eco-
requirements for low combustion noise (i.e., nomic advantages of the CC plants vis-à-vis other
dynamic stability) at part load operation and nonrenewable large central stations or renewable
optimal equivalence ratio for minimum NOx at options (i.e., wind, solar, etc.), which cost nearly
acceptable CO levels. DLN achieves this bal- ten times per kW capacity and available only in
ance without diluent injection (and associated small sizes amenable to distributed power genera-
efficiency penalty and cost) while satisfying the tion, should not be expected near term. A very brief
requirement for low NOx emissions, i.e., introduction to the key aspects of the CC plant
25 ppm or lower on gas fuels. A history of the economics, i.e., plant capital (turnkey) cost or
development of DLN-I and DLN-II (for F-Class price and the cost of generating 1 kw-h of electric
turbines with high firing temperatures of 2,400+ power, is therefore quite useful.


F) systems can be found in Ref. [78] along with Turnkey combined cycle plant price levels
a detailed description of DLN operation modes. from 2009 Gas Turbine World (GTW) Handbook
are shown in Fig. 31. Note that listed GTW prices
Basic Economics are in 2009 US dollars for a basic natural gas-fired
Natural gas-fired gas turbines and GTCC power combined cycle with gas turbine generator,
plants are the most economic choices for fossil unfired multi-pressure heat recovery boiler with
fuel-based electric power generation. This is no bypass stack, condensing multi-pressure steam
clearly represented by the data in Table 9. The turbine generator, step-up transformer, water-
data in the table reflects 2004 prices. The numbers cooled heat rejection, standard controls, starting
and the relative position of NG-fired plants should system, and plant auxiliaries. Obviously,
be expected to show a certain variation as a func- depending on the scope of equipment, site-
tion of changing economic climate (i.e., inflation, specific requirements, geographic location, and
demand for generating capacity, absolute and rela- competitive market conditions, these prices will
tive fuel prices, labor shortage, raw material prices, vary considerably. Also excluded from the prices
etc.) of which the recent worldwide depression that in GTW are the indirect costs and other commer-
started in the last quarter of 2008 is a very good cial outlays. For a more realistic price level, the
Natural Gas Power 295

Natural Gas Power, Fig. 31 Turnkey CC plant prices turnkey equipment package price, to account for items
from Gas Turbine World 2009 Handbook. An adder of not covered (see Fig. 32 and Table 10). Provided for
60% is applied to the listed price, which is budgetary reference only

Natural Gas Power,

Fig. 32 Direct cost Civil
breakdown for a typical Engineering & Site
400 MW CCPP (Source: Facilities Infrastructure
[85], pp. 16) 18% 3% Gas Turbine
Gen-Set
32%
Controls &
Instrumentation
3%

Electrical
Systems HRSG
9% 10%
Steam Turbine
Mechanical Gen-Set
Systems 8%
17%

data in Fig. 31 reflects an adder of about 60% to k ¼ 2, 890  P0:2012 (31)

the listed price, which is based on the assumption
that the prices in GTW reflect only 75% of the where k is the specific price ($/kW) and P is the
direct costs, which are 85% of the total installed net plant output in MW (with R2 = 0.994). In
cost (see Fig. 32 and Table 10). general, an exponent of 0.2–0.3 is adequate to
The data plotted in Fig. 31 can be summarized estimate the capital cost of a CC plant from a
by the power law: known data point (i.e., price and net output).
296 Natural Gas Power

Natural Gas Power, Table 10 Typical CC capital cost and its importance from a design engineer’s per-
breakdown between direct and indirect costs. From Energy spective. All design performance calculations
Issues (World Bank), No. 20 June 1999, which is based on
a 350/700 MW CC plant with a Siemens-Westinghouse should be evaluated with an eye to the final plant
V94.3A (now SGT5-4000F) GT cost. That final plant cost is the money, i.e., the
Integrated services Contribution to plant “price,” that is going to come out of the plant
total owner’s pockets or his credit sources and thus
(Indirect costs) CC cost (%) directly impact the generation cost of 1 kWh of
Project management and 4 electricity.
subcontracting Probably the most critical customer criterion
Plant and project engineering/ 2 for choosing an electric power generation prod-
software
uct is the life-cycle cost of ownership. The metric
Plant construction, commissioning, 8
and training that is used to quantify this cost is the unit (per
Transport and insurance 1 kWh) cost of generating electricity. Cost of elec-
Direct costs Contribution to tricity (COE), in its most commonly used form, is
total a simple formula that combines a power genera-
CC cost (%) tion system’s cost (capital and operating) and
Civil works 15 thermal performance into a single figure of
Gas and steam turbine generators 32 merit that enables comparison of different sys-
Boiler island 11 tems. It is given as
BOP 16
Electrical systems 7 bk Lf
Instruments and control 4 COE ¼ þ þ OM (32)
H Z

where
There are several methods and tools to calculate
CC power plant cost (from a plant owner’s per- b = Capital charge factor
spective), which are used within OEMs and A&E k = Total plant specific cost in $/kW
firms and not available for public consumption. H = Annual operating hours
Well-established cost-estimating techniques, f = First-year fuel cost in $/MM Btu(HHV)
especially those that are adopted by the CPI, are L = Levelization factor
readily available in textbooks and scholarly arti- Z = Net efficiency of the CC plant
cles. For a discussion of those and the underlying OM = Operating and maintenance costs, $/kWh
details, the reader is referred to [86]. Another
valuable source is [87]. The first term on the right-hand side of Eq. 32
Calculation of the final cost of the CC plant represents the carrying charges, the second term is
requires consideration of many elements. In the cost of fuel, and the last term is the cost of
fact, the total cost of GT and ST generators and maintaining and operating the plant. Fuel cost and
the HRSG, which are the major equipment in the operation and maintenance (O&M) costs are the
CC power plant, comprises about 50% of the variable components of the COE, whereas carry-
total plant cost. A typical distribution is shown ing charge is the fixed component. COE is typi-
in Fig. 32 from [85]. In addition to these “direct” cally calculated over the economic lifetime of the
costs, there are “indirect costs” that can contrib- plant (typically 30 years).
ute anywhere between 15% and 30% to the final Levelization factor, L, converts the fuel cost that
cost of the CC power plant. As an example usually escalates over the plant’s economic life due
consider the representative breakdown in to market conditions and inflation into average
Table 10. annual (constant) values. The formula for L can
The presented data should suffice to give an be found in any financial reference and is a func-
idea about the scope of the CC plant capital cost tion of discount, escalation, and inflation rates.
Natural Gas Power 297

The fixed charge rate, b, represents the plant’s exclude owner’s costs and interest during construc-
financial and tax situation and is a function of the tion (IDC). As mentioned above, indirect costs can
economic life of the power plant, debt-equity ratio add about 40% to that, which makes the range
used in financing the plant, debt and equity rates, $2,300–3,200.
corporate tax rate and tax credits, book deprecia- A 2014 reference [88] recommends the capital
tion, property tax, and insurance. Various methods costs for selected generation technologies as fol-
are available for calculating b, which can be found lows (representing all-in plant costs and are inclu-
in plant economic studies. Typical values of b are sive of all engineering, procurement, and
16–18% for independent power producers (IPPs) construction (EPC) costs as well as owner’s
and 12–13% for government-owned utilities, costs and IDC):
depending on the interest and taxation rates.
If the new and clean performance conditions • SCPC $3,700
are used within the calculations, the cost of gen- • IGCC with CCS (carbon capture and seques-
eration as given by the COE formula should be tration) $8,200
treated as a figure of merit rather than an accurate • Advanced GTCC $1,225–1,300
assessment of the “real” generating cost. Impor- • Nuclear $8,000
tant drivers – such as seasonal variations in ambi-
ent conditions, part load operation, recoverable Turnkey CC plant prices from Gas Turbine
and unrecoverable component deterioration, World 2015 Handbook are found to be not too
uncertainty in fuel prices, start-up and shutdown, far off from those in Fig. 31.
and unforeseen events – are not considered. Nev- An expansion of the basic COE formula in
ertheless, levelized COE is an important metric Eq. 32 enables the inclusion of operability and
that provides insight to performance-cost trade- dispatch characteristics of the power plant in addi-
offs for OEMs (system design) as well as cus- tion to the base ISO baseload performance param-
tomers (system selection). eters (output and efficiency) [89]. Calculations
The exact distribution of the cost of generating show that, with the increasing importance of flex-
1 kWh of electric energy varies depending on the ibility in support of renewable generation and
financial assumptions. Typically, about two thirds other improved operability features such as better
of the COE for a NG-fired, medium-loaded part load efficiency, the results are significantly
(~6,000 h per year) modern 3PRH CC plant with different than those obtained from the simplistic
an F-Class gas turbine is the fuel cost. About one version in Eq. 32. The reader is referred to [89, 90]
fourth of the COE is the capital cost. The distri- for an expansive coverage of salient details.
bution is markedly different (actually, nearly
exactly the opposite of NGCC) for the IGCC
plant with very high capital expenditure ($2,000 Future Directions
per kW or even higher) utilizing a cheaper fuel,
coal ($1–2). At the time of the first edition of this chapter in
2010, natural gas accounts for about 22% of US
2016 Update electricity generation. The share of natural gas in
In terms of the capital cost ranking of the technolo- electricity generation in OECD member countries
gies in Table 9, the situation has not changed. (Note in Europe is about the same. The share of natural
the vintage of the cost data in that table, 2004, gas is expected to increase in the future to the tune
12 years before the time of writing this update.) As of 25% of total world electricity generation by
far as the absolute values are concerned, consider 2030 [4]. However, recent developments suggest
year 2014 (i.e., a 10-year period). The numbers in that this projection might in fact be a low estimate.
Table 9 have to be inflated by 25%. For example, the One reason is the new drilling technologies that
SCPC cost range becomes $1,625–1,250. Note that made extraction of natural gas from shale rocks
the prices in Table 9 are “turnkey” prices, which feasible. The technique is called hydraulic
298 Natural Gas Power

fracturing and involves injection of high-pressure per 1,000 cubic feet), and they were as high as $6
water into rock formations to destroy them and per million Btu (about $6.80 per 1,000 cubic feet)
enable the trapped gas to be released to the surface at the beginning of 2010. Note that extremely low
for capture. So much so that this development price levels are not likely to continue for long if
resulted in a 35% jump in US natural gas reserves they are below the production costs. Extreme
from about 1,500 trillion cubic feet in 2006 to price volatility can be expected when demand
more than 2,000 trillion cubic feet in 2008 and supply (or capacity) go neck to neck.
[91]. Note that this number is the total estimated In this context, one should also mention the
reserves including speculative, possible, and problems associated with natural gas transporta-
probable sources and of which only 237 trillion tion via pipelines. In January 2009, the price quar-
cubic feet is considered to be proven by the rel between Russia and Ukraine ended in Russia
Energy Department. According to the same cutting off all gas flows through Ukraine for
source, shale gas constitutes nearly a third of the 13 days. This resulted in severe distress in South-
total reserves. eastern Europe, which is nearly completely reliant
Another potential source of natural gas is gas on Russian gas, and other parts Europe. While not
hydrates (also known as gas clathrates), which an issue in North America, political stability of
are “crystalline solids consisting of gas molecules, producer and transit countries is a significant con-
usually methane, each surrounded by a cage of cern for European countries that can make them
water molecules.” It looks very much like water reluctant in investing in natural gas-fired power
ice. Methane hydrate is stable in ocean floor sed- plants. Alternate routes between Russia,
iments at water depths greater than 300 m, and Azerbaijan, and other countries, such as thou-
where it occurs, it is known to cement loose sed- sands of miles long South Stream and Nabucco
iments in a surface layer several hundred meters pipelines, are actively pursued projects to provide
thick. The worldwide amount of carbon bound in answers to such concerns. The only other practical
gas hydrates is conservatively estimated to total way to transport natural gas is to move it by sea by
twice the amount of carbon to be found in all cooling it to 162
C (260
F) and converting it
known fossil fuels on Earth (Source: U.S. Geolog- to a liquid (liquefied natural gas or LNG), which
ical Survey Fact Sheet). Unfortunately, commer- reduces its volume by 600 times. Specially built,
cially feasible production of natural gas from gas double-hulled cryogenic ships are used to trans-
hydrates is not a reality in near future. port LNG from its port of origin to its port of
The dramatic increase in natural gas reserves destination, which are purpose-built facilities
and actual production via new technology that can used exclusively to export or import LNG. From
extract it from shale rocks, taken at its face value, the receiving port, LNG is shipped to its final
is a strong driver for its increasing share in elec- destination via tanker trucks, trains, or pipelines
tricity generation, at least in the United States. The (after regasification). Expensive infrastructure
advantage of natural gas vis-à-vis coal, which investment requirements, safety concerns, large
roughly accounts for half of US electricity gener- “carbon footprint” during liquefaction, transport,
ation, in terms of reduced carbon emissions, is and regasification are key concerns associated
thus fortified from lower prices and increased with LNG, which limit its share in the overall
availability. But the situation may not be as simple natural gas consumption (currently less than
as it seems. The price of natural gas hits its 7-year 10% of US imports).
low of about $3 per 1,000 cubic feet in the sum- In terms of cost, regulatory costs and incen-
mer of 2009. This huge drop from a peak of $13 in tives, e.g., those associated with GHG emission
the summer of 2008 was a direct result of the caps and cap-and-trade system requirements (e.g.,
Great Recession that led to a decrease in demand, for manufacturers and utilities to purchase pollu-
coupled with the increase in production. At the tion permits), should be considered along with
time of writing, NYMEX natural gas futures are straight dollars per Btu. Coal-fired generation
trading at about $4 per million Btu (about $4.50 can compete with natural gas in terms of GHG
Natural Gas Power 299

and other pollutant emissions only at a signifi- especially with the increase of renewable energy
cantly high capital cost and reduced efficiency. power plants (mostly solar and wind) connecting
Currently, there are two major approaches to to the grid, make them eminently suitable to cyclic
accomplish this: post-combustion carbon capture on-demand duties to compensate for the variability
and storage (CCS) technologies that can clean the of the renewable generation. Presently, all major
stack gas of the existing (via retrofit) and new OEMs provide CC plants that can reach baseload
coal-fired plants and precombustion technologies (several hundred megawatts) within 30–60 min
that are a very good fit with integrated gasifica- without sacrificing baseload efficiency, parts life,
tion CC (IGCC) power plants. Note that carbon and RAM. Aeroderivative gas turbines, such as
emission abatement technologies via stack gas GE’s intercooled LMS100, can provide 100 MW
treatment are equally applicable to natural gas- within 10 min in an emergency or for peak shaving
fired CC power plants. In terms of both efficiency in hot summer months. Especially in Europe and to
and cost, detailed studies indicate that NG-fired a certain extent in the United States, large, highly
CC technology is superior to oil or coal-fired flexible natural gas-fired CC power plants that start
plants as well as IGCC [92]. and shutdown daily for up to 250 starts and a total of
Objectively viewed, these factors are strong 4,000 h of operation per year are expected to sup-
incentives for natural gas-burning power plants plement renewable power generation. These plants
to replace aging, inefficient, and carbon-emitting are based on advanced F-Class and H-Class gas
coal-fired power plants in the United States. To a turbines that are rated at 58–59% efficiency and
certain extent this can be accomplished even with- expected to reach more than 60% in the near future.
out building new natural gas-fired CC power Clearly, then, no matter how regional politics
plants. A recent report estimated that CC power in the United States, Europe, and other parts of the
plants had an average 2007 capacity factor of 42% world play out, and what kind of legislative pack-
[93]. In comparison, in the same year, coal-fired ages eventually emerge from governing bodies to
plants had a much higher capacity factor of 75%. control, cap, or regulate carbon and other GHG
Clearly, at least on paper, the replacement poten- emissions, natural gas-fired electricity generation
tial by simply switching generation from coal- with advanced gas turbine-based power plants
fired to natural gas-fired plants is pretty big. In will be an important part of the overall generating
reality, factors such as geographical proximity of technology portfolio. In all likelihood, natural
the power plants to each other and to the locations gas-fired advanced CC power plants will be the
of highest demand, transmission system limita- bridge technology between coal-based power
tions, and dispatch issues limited this potential generation and renewable (and probably nuclear)
severely, i.e., to about 5–9% of total US coal- power generation to alleviate the lurking problem
based electricity generation [93]. While the pro- of global warming. (In fact, this is pretty much the
posed switch from coal to natural gas would also conclusion reached by a recent MIT report, which
require an increase of 5–20% in natural gas sup- was made public after the writing of the present
ply, this should not be a significant problem in the article [94].) The question is what will be the
light of recent developments discussed in the ear- primary technologies in the next, say, quarter
lier paragraphs. A recent article states that several century?
major US utilities have announced that they will Without a doubt, gas turbine-based power
replace at least some of their coal-fired plants plants, in simple cycle (both heavy-duty industrial
slated for decommissioning (total capacity of and aeroderivative variants) and combined cycle
more than 4.5 GW) [10]. configurations, will be by far the dominant tech-
Another factor that favors the natural gas-fired nology in natural gas electric power generation.
gas turbine plants over the coal-fired steam turbine Advanced air-cooled F-Class and H-Class, steam-
power plants is operational flexibility. The rapid cooled G-Class, GE’s H-System™, and J class
start-up and good turndown characteristics of gas engines with firing temperatures up to 1,600
C
turbine simple and combined cycle power plants, and PRs up to 23 (up to 35 in reheat machines) are
300 Natural Gas Power

currently capable of 58%, have recorded 59%, and high HGP temperatures. Manufacturing techniques
are slated for 60% (or even 61% as claimed by one have always been a significant driver in the imple-
OEM) in 2011–2012 [95]. These are large units mentation of aforementioned innovations, e.g.,
with 500–600 MW per block in 1  1 configura- TBC application processes (e.g., electron beam
tion designed not only for baseload duty but for physical vapor deposition) and casting and
operational flexibility as described earlier. machining technologies to produce SC airfoils. In
(In passing, note that the cited efficiencies are a similar vein, laser drilling and casting methods to
typically ISO rating numbers that reflect the achieve transpiration (effusion) cooling (the ulti-
most favorable operating assumptions (e.g., mate limit of film cooling) are related areas of
once-through open-loop cooling with water research and development to further the goal of
drawn from a natural source such as river, ocean, increasing turbine inlet temperatures. An associ-
etc.). Depending on site and ambient characteris- ated effort is made on the bottoming cycle side,
tics, performances of actual installed units will be i.e., development of HRSG heat exchanger tubes,
somewhat lower with wide variability.) steam headers and piping, and steam turbine mate-
Increasing power plant efficiency is the sim- rials (casing, rotor, and buckets), that can withstand
plest and arguably easiest way to reduce carbon increased steam conditions as high as 180 bar and
emissions without breaking the bank or worrying 600–700
C [96].
about what to do with all that CO2 or where to put Closed-looped steam cooling of the combustor
it. As mentioned earlier, ambitious targets as high transition piece and stage-1 stators (nozzles) is an
as 65% or even 70% have been in circulation for existing technique to increase the firing tempera-
some time [58–61]. The discussion in the preced- ture without raising the combustor exit (turbine
ing sections of the article hopefully makes it clear inlet) or flame temperatures [32, 33]. This enables
that the key to ever-increasing gas turbine com- performance improvement without a concomitant
bined cycle efficiencies is increasing firing tem- increase in NOx emissions. Another method to
peratures and commensurate cycle pressure ratios. facilitate extreme combustor exit and firing tem-
The factors that prevent easy realization of that peratures (up to 1,700
C) is exhaust gas
simple goal are primarily related to available com- recirculation (EGR), in which a fraction of the
ponent materials that can withstand the extreme stack gas (up to 35%) is recirculated into the
pressures and temperatures and limits imposed by compressor or combustor inlet [97]. EGR reduces
pollutant emission regulations. State-of-the-art thermal NOx production via reduced O2 in com-
nickel-based alloys used in turbine HGP compo- bustion air and peak temperatures in flame. EGR’s
nents require significant amounts of cooling air benefit becomes more pronounced at increased
supplied from extraction ports in the compressor, flame temperatures, and it can reduce NOx emis-
which is detrimental to performance. Higher sions by up to 50% [98]. EGR is also beneficial for
flame zone temperatures in DLN combustors post-combustion CCS because it increases the
result in excessive NOx and/or CO in flue gas, CO2 concentration in stack gas (from ~3–4% to
which is incompatible with increasingly stringent ~8–10% by volume), which leads to a reduction in
environmental regulations that limit the emissions the capital cost of the separation plant [98].
of those pollutants to low single-digit numbers. Lean premix or DLN combustion technology is
Single-crystal (SC) and directionally solidified the industry standard for heavy-duty industrial gas
(DS) superalloys with TBCs helped push firing turbines for large-scale electricity production. In
temperatures up to 1,500
C. With closed-loop terms of the ability to meet the stringent emissions
steam cooling and advanced film cooling tech- requirements, the technology is nearly at its limit at
niques, J class turbines are rated at 1,600
C. Cur- the current turbine inlet temperatures. EGR is one
rent research efforts focus on advances in possibility to extend the applicability of DLN or
materials, e.g., ceramic matrix composites (CMC) even diffusion combustors to 1,700
C turbine inlet
[59] or metal foams [60], which can reduce or even temperatures with low NOx emissions. Another
eliminate cooling air requirements at extremely way to mitigate the NOx emissions is selective
Natural Gas Power 301

catalytic reduction (SCR) in the HRSG. SCR is an possible. Detailed studies show that fuel cell-GT
existing post-combustion NOx reduction technol- hybrid plants are capable of thermal efficiencies
ogy. Other combustion technologies are being approaching 70%. Bhargava et al. [103] can be
investigated for deployment in IGCC and other consulted for more information on these and other
low Btu syngas and blast furnace gas applications advanced GT-based power plants such as humid
to extend the fuel flexibility of gas turbines. The air turbine (HAT) and intercooled recuperated
reader can consult the relevant sections in The Gas engine (ICR). Almost all variants have been stud-
Turbine Handbook published by NETL (US DOE, ied in detail and reached various stages of devel-
Office of Fossil Energy). Various high-hydrogen opment and commercialization in the past. They
and oxyfuel combustion technologies are being have not been able to pass the audition in terms of
actively investigated for CO2-free emissions. reliability, operability, capital cost, scalability, and
Since the subject of the present article is natural other important considerations (e.g., water con-
gas-burning technologies, the reader is referred to sumption) to present viable alternatives to simple
relevant academic journals such as ASME’s Jour- Brayton cycle GT in CC configuration. Closed-
nal of Engineering for Gas Turbines and Power, loop steam cooling and reheat (also known as
which contain a wide selection of recent research sequential combustion) have been the only suc-
papers on the subject. cessful and commercially accepted variations on
A potentially promising combustion technol- the basic cycle. At this time, to the best of the
ogy that directly attacks the single largest source authors’ knowledge, there is no reason to expect
of exergy destruction in the GT Brayton cycle is this situation to change in the foreseeable future;
pulse detonation combustion (PDC). Originally at least in terms of large-scale electric power gen-
proposed for jet engine aircraft propulsion, the eration. As always, dual-fuel diesel engines capa-
technology has been around for more than a half ble of burning gaseous fuels will continue to play
century [99]. Recently, PDC is actively consid- a role in small-scale and/or distributed generation
ered for land-based power generation [100]. The applications. The same can be true for very small
technology involves combustion of a fuel-air mix- (several kilowatts) fuel cell-based hybrid units.
ture across detonation waves in tubular devices
[101]. Detonation waves are composite waves 2016 Update
comprised of a “frozen” shock wave with pressure Long-term natural gas (NG) price forecasts are
and temperature jump followed by a relaxation difficult to make as illustrated by the chart in
zone where chemical combustion reactions take Fig. 33 [104], which superimposes outlooks by
place. The benefit of the process is temperature the US DOE (consistently predicting increasing
and pressure rise during the cycle heat addition scarcity and rising prices) and The National Petro-
process with a correspondingly higher mean- leum Council (NPC), with the latter comprising
effective temperature vis-à-vis the conventional pessimistic (reactive path) and optimistic
diffusion or DLN combustors with approximately (balanced future) scenarios. (The two scenarios
constant pressure (actually a 5–6% pressure loss). are classified in terms of public policy conflict
Detailed cycle calculations indicate that a PDC between supply and demand.)
gas turbine in CC configuration has potentially Except for the 2003–2008 period, when prices
1.5–3 percentage points advantage over standard spiked above historical levels due to a tight market
units at comparable firing temperatures and pres- caused by several factors, i.e., weak supply and
sure ratios [102]. growth in demand for peaking power in particular,
SOFC- or MCFC-based hybrid fuel cell plus long-term expectation of annual ~2% growth in
gas turbine power plants are by far the most effi- NG prices pretty much held (going back to the
cient future technologies to “burn” natural gas for Carter administration era and the Alaska Natural
electric power generation. In these machines, the Gas Transportation System (ANGTS) project).
fuel cell replaces the combustor to oxidize the Right after that peak price period, development
gaseous fuel. Different configurations are of new sources of shale gas, driven by hydraulic
302 Natural Gas Power

Natural Gas Power, Fig. 33 Evolution of US DOE natural gas price forecasts and actual prices [89]

20
16.6 16.0 16.0
14.7
15
$ per MMBtu

11.5 11.8 Japan (LNG)

10.8 10.5 10.4
10.1
8.9 9.1 8.9 8.8
8.9
10 8.3 8.9
Europe 7.3 7.0
6.5
4.4 4.0 3.7 4.4 3.6
5 2.8 3.3
2.6 2.3 2.7
USA
0
2010 2012 2014 2016 2018

Natural Gas Power, Fig. 34 Comparison of gaseous fuel prices (past and short-term outlook) in the United States,
Europe, and Japan (EIU Economic and Commodity Forecast, December 2015)

fracturing and horizontal drilling technologies, monthly capacity factors for 16 different fossil
has more than compensated for the decline in and non-fossil fuel and technology combinations.
conventional supply and has led to major The data for natural gas-fired combined cycle
increases in reserves of US natural gas. The power plants is summarized in Fig. 35. Prior to
so-called shale gas boom, although not a guaran- 2010, these plants were run at a very low capacity
tee by any means, is expected to prevent non- factor (CF), but the situation changed quite dramat-
seasonal, years-long price spikes and exorbitant ically in recent years. Clearly, shale gas “boom”
long-term growth rates in the United States. The and ensuing low natural gas prices played a signif-
situation is somewhat different in Europe and icant role in this. Even so, it is hard to envision that
Japan (see Fig. 34). the annual average CF for natural gas-fired CC
US Energy Information Administration plants will be much higher than 55–60% in the
(Electric Power Monthly, Table 6.7a) provides foreseeable future (especially with increasing
Natural Gas Power 303

Natural Gas Power, 70

Fig. 35 US EIA (Electric Monthly Variation
Power Monthly, Table 6.7a) 65
capacity factor data for
60

CAPACITY FACTOR
natural gas-fired combined
cycles 55
First 4 Months
50

45

40

35 Values for 2014 and prior years are final.

Values for 2015 and 2016 are preliminary.
30
2006 2008 2010 2012 2014 2016 2018
YEAR

Natural Gas Power,

Fig. 36 Actual field
performance of CC power
plants

renewable resource penetration). In passing, note illustrated in Fig. 36. The data used in the chart
that capacity factors in Fig. 35 are significantly is available from the EIA website. A compilation
higher than that for a typical cyclic operation, of the “Top 20 Performers” of the previous year is
which is an indication of significantly higher load annually published in the November–December
factor (e.g., more hours at full load), power aug- issue of the Electric Light and Power magazine. In
mentation (via supplementary firing and/or GT the future, up-to-date values can be obtained from
inlet conditioning), or a combination thereof. these sources.
The impact of actual field operation regime and Most recent gas turbine-enabling technologies
site ambient and loading conditions on CC effi- toward higher CC efficiencies are summarized in
ciency (vis-à-vis rating performance, which Fig. 37. Two notable emerging technologies are
should be considered as a “sticker” number additive manufacturing (AM, also known as 3D
achievable by a new and clean power plant printing) and data analytics. The AM technology
under best possible boundary conditions) is (e.g., selective laser melting adopted by a major
304 Natural Gas Power

Dry Low Nox – Dual Fuel Capable

1,600°C TIT
Can– Annular (Except Alstom)
3D Aero Airfoils Reheat/Sequential (Alstom’s EV/SEV)
Multistage Variable Guide Vanes
Steam Cooled Transition (MHPS)
“Super Finish” Technology Axial Fuel Staging (GE) H
Hydrogen Cooled
99% Efficiency Field Replaceable Blades
20+ Pressure Ratio Natural Gas Fuel H C H
300+ MWe (60Hz)
40+% Simple Cycle Efficiency H
60+% Combined Cycle Efficiency

600+°C
≤ 25 ppm NOx
≤ 10 ppm CO
No Toxic Emissions
< ½ Coal–Fired CO2

Welded Rotor (Alstom) Future Enablers Four Stage Turbine (Except Alstom)
Bolted Construction (All Others) 1700°C TIT Single Crystal Ni–Based Superalloys
Hirth Coupling (Siemens) Ceramic Matrix Composites Thermal Barrier Coating
Hydraulic Clearance Optimization (Siemens) Additive Manufacturing 3D Aero Blades and Vanes
Data Analytics Steam Cooling (H System,MHPS G/J Class)
Controls & Optimization Advanced Film Cooling
Brush Seals

Natural Gas Power, Fig. 37 Gas turbine-enabling technologies

OEM) facilitates rapid prototyping, repair, and 50 Hz version of the reheat (sequential combus-
manufacturing of complex parts such as turbine tion) gas turbine technology by another European
blades with many microscopic cooling holes. OEM, which also inherited the J class technology
Consequently, component validating and testing of the same OEM (expected to be commercially
can become an integral part of the design and offered in 2017). At the time of writing, the future
development process, while spare parts can be of the reheat gas turbine technology is uncertain.
manufactured on demand and, ultimately, on site. Furthermore, the H-System, the only fully steam-
Data analytics (also marketed under trade cooled gas turbine system, is no longer offered by
names such as digital solutions) comprises its OEM and is replaced by its air-cooled variant,
advanced sensor and monitoring and diagnostics the “HA” class, which is essentially an air-cooled
(M&D) technologies to improve diagnostic and J class (nominal 1,600
C TIT).
prognostic capabilities of power plant equipment
to improve performance (i.e., heat rate and output)
and operability, extend parts life, and reduce risk Bibliography
of failure. A large quantity of data from the equip-
ment (e.g., gas turbine) can be processed locally in 1. Energy Information Administration (EIA)
(2009) Annual Energy Review 2008. http://www.
the DCS and transmitted remotely over, say, the
eia.doe.gov/aer
Internet to a central OEM facility for archival, 2. United States Environmental Protection Agency
additional analysis, and anomaly detection utiliz- (EPA). Methane. www.epa.gov/methane
ing specially designed algorithms comprising sta- 3. Liss WH, Thrasher WR (1992) Gas Technology Insti-
tute, variability of natural gas composition in select major
tistical methods, time series analysis, and rigorous
metropolitan areas of the United States. GRI-92/013
aerothermodynamic models (“digital twins”). 4. Energy Information Administration (EIA) (2008) Inter-
Note that one major European OEM has been national Energy Outlook 2008. DOE/EIA-0484(2008).
acquired by another OEM (in the United States) in www.eia.doe.gov/oiaf/ieo/index.html
5. Natural gas and the environment, from http://www.
2015. European community regulations led to a
naturalgas.org/environment/naturalgas.asp
divestiture of some of the gas turbine assets of the 6. Turbomachinery International 49(6). 10/2008, Hand-
former. This resulted in the acquisition of the book 2009
Natural Gas Power 305

7. Wärtsilä 50DF generating set, from www.wartsila.com economic perspectives. J Eng Gas Turbines Power
8. Khan BH (2006) Non-conventional energy sources. 119:119–123
Tata McGraw Hill Publishing Co. Ltd., New Delhi 29. Cohen H, Rogers GFC, Saravanamuttoo HIH
9. McNeely M (2006) Power generation order survey. (1987) Gas turbine theory, 3rd edn. Longman Group
Diesel & Gas Turbine Worldwide, Article from Oct UK Limited, Cambridge
2006 issue 30. Cumpsty N (2003) Jet propulsion, 2nd edn. Cam-
10. Burt B, Mullins S (2010) U.S. gas-fired power devel- bridge University Press, Cambridge, UK
opment: last man standing. Power, Sept 2010, 31. Schilke PW (2004) Advanced gas turbine materials
pp 71–73. http://www.powermag.com and coatings GER-3569G. www.gepower.com
11. Wilson DG, Korakianitis T (1998) The design of high 32. Pritchard JE (2003) H-System™ technology update,
efficiency turbomachinery and gas turbines, 2nd edn. GT2003-38711, ASME turbo expo – power for land,
Prentice-Hall, Uppersaddle River sea & air, Atlanta, 16–19 June 2003
12. Von Ohain H (1996) Foreword in elements of gas 33. Koeneke C (2006) Steam cooling of large frame gas
turbine propulsion. In: Mattingly JD (ed). Tata turbines one decade in operation. VDI-Berichte
McGraw Hill Edition 2005 1965:33–42
13. Meher-Homji CB (1997) The development of the 34. Imwinkelried B (1995) Advanced cycle system gas
Junkers Jumo 004B – the world’s first production turbines GT24/GT26: the highly efficient gas tur-
turbojet. J Eng Gas Turbines Power 119:783 bines for power generation. In: Proceedings of the
14. Meher-Homji CB (1998) The development of the 21st international congress on combustion engines,
whittle turbojet. J Eng Gas Turbines Power 120:249 CIMAC 1995, Interlaken
15. Meher-Homji CB (2000) Pioneering turbojet devel- 35. Chiesa P, Macchi E (2004) A thermodynamic analy-
opments of Dr. Hans von Ohain – from the HeS 1 to sis of different options to break 60% electric effi-
the HES 011. J Eng Gas Turbines Power 122:191 ciency in CC power plants. J Eng Gas Turbines
16. Soares C (2006) Gas turbines in simple cycle and Power 126:770–785
combined cycle applications, Section 1.1 in the gas 36. Holland MJ, Thake TF (1980) Rotor blade cooling in
turbine handbook. US DOE, Office of Fossil Energy, high pressure turbines. J Aircr 17:412–418
NETL 37. Elmasri MA, Pourkey F (1986) Prediction of cooling
17. Volker L (1999) Development of the Siemens gas flow requirements for advanced utility gas turbines
turbine and technology highlights. Siemens Power part 1: analysis and scaling of the effectiveness curve,
Generation, Erlangen 86-WA/HT-43, ASME winter annual meeting, Ana-
18. Miller H, Nemec T (2006) Gas turbines, Chapter 24. heim, 7–12 Dec 1986
In: Kutz M (ed) Mechanical engineers’ handbook, 38. Elmasri MA (1986) Prediction of cooling flow
Energy and Power, 3rd edn. Wiley, Hoboken requirements for advanced utility gas turbines part
19. Brandt D (2007) A brief history of GE energy prod- 2: influence of ceramic thermal barrier coatings,
uct lines. General Electric Company, Schenectady ASME winter annual meeting, Anaheim, 7–12 Dec
20. Brandt D (1988) The design and development of an 1986
advanced heavy-duty gas turbine. J Eng Gas Tur- 39. Elmasri MA (1985) On thermodynamics of gas tur-
bines Power 110:243–250 bine cycles part 1 – second law analysis of combined
21. Eckardt D (2013) Gas turbine powerhouse. De cycles. J Eng Gas Turbines Power 107:880–889
Gruyter Oldenbourg 40. Elmasri MA (1986) On thermodynamics of gas tur-
22. Stodola A (1927) Steam & gas turbines, authorized bine cycles: part 2 – a model for expansion in cooled
translation from the 6th German edition by L. C. turbines. J Eng Gas Turbines Power 108:151–159
Löwenstein. McGraw-Hill Book Company, New York 41. Elmasri MA (1986) On thermodynamics of gas tur-
23. Langston LS (2010) World’s first gas turbine power bine cycles: part 3 – thermodynamic potential and
plant. ASME Mech Eng 132(4):51 limitations of cooled reheat gas turbine combined
24. Tomlinson LO, Lee DT. Combined cycles, Chapter 7. cycles. J Eng Gas Turbines Power 108:160–170
In: Sawyer JW, PE, Japikse D (eds) Sawyer’s gas 42. Horlock JH, Watson DT, Jones TV (2001) Limita-
turbine engineering handbook tions on gas turbine performance imposed by large
25. Horlock JH (1994) Combined cycle power plants – turbine cooling flows. J Eng Gas Turbines Power
past, present, and future. J Eng Gas Turbines Power 123:487–494
117:608–616 43. Horlock JH (2001) The basic thermodynamics of tur-
26. Gebhardt E (2000) The F technology experience bine cooling. J Eng Gas Turbines Power 123:583–591
story. GER-3950C. http://www.gepower.com/ 44. Wilcock RC, Young JB, Horlock JH (2005) The
27. Elliott TC, Editors of POWER Magazine (eds) effect of turbine blade cooling on the cycle efficiency
(1989) Standard handbook of power plant engineer- of gas turbine power cycles. J Eng Gas Turbines
ing, Chapter 2.4. In: Haselbacher H (ed) Gas turbine Power 127:109–120
fundamentals. McGraw-Hill 45. Young JB, Wilcock RC (2002) Modeling the air-
28. Horlock JH (1997) Aero-engine derivative gas tur- cooled gas turbine: parts 1 and 2. J Turbomach
bines for power generation: Thermodynamic and 124:207–222
306 Natural Gas Power

46. Gülen SC (2010) A simple mathematical model for 64. Massardo AF, Magistri L (2003) Internal reforming
cooled gas turbines, GT2010-22160, ASME turbo solid oxide fuel cell – gas turbine combine cycles
expo – power for land, sea & air, Glasgow, (IRSOFC-GT); part II: exergy and thermoeconomic
14–18 June 2010 analyses. ASME J Eng Gas Turbines Power 125:67–74
47. Rice IG (1995) Steam-injected gas turbine analysis: 65. Gülen SC (2015) Étude on gas turbine combined
steam rates. J Eng Gas Turbines Power 117:347–353 cycle power plant – next 20 years. J Eng Gas Tur-
48. Cheng DY, Nelson ALC (2002) The chronological bines Power 138:#051701
development of the Cheng cycle steam injected gas 66. Gülen SC, Jones CJ (2011) GE’s next generation
turbine during the past 25 years. ASME international – CCGT plants: operational flexibility is the key. Mod-
IGTI turbo expo 2002, GT2002-30119 ern Power Systems, June 2011, pp 16–18. www.
49. Rao A (1989) Process for producing power. US Pat- modernpowersystems.com
ent 4,289,763 67. Gülen SC (2013) Gas turbine combined cycle fast
50. Adelman ST, Hoffman MA, Baughn JW A methane- start: the physics behind the concept. Power Engi-
steam reformer for a basic chemically recuperated neering, June 2013, pp 40–49. www.power-eng.com
gas turbine. J Eng Gas Turbines Power 117:16–23 68. Chase DL, Kehoe PT (2000) GE combined-cycle prod-
51. McDonald CF, Boland CR (1981) The nuclear uct line and performance. GER-3574g, GE Energy
closed-cycle gas turbine (HTGR-GT) – dry cooled 69. Maslak CE, Tomlinson LO (1994) GE combined-cycle
commercial power plant studies. J Eng Gas Turbines experience. GER-3651. http://www.gepower.com
Power 103:89–100 70. Tomlinson LO, McCullough S (1996) Single-shaft
52. Reale MJ (2004) New high efficiency simple cycle combined -cycle power generation system. GER-
gas turbine – GE’s LMS100™, GER-4222A. www. 3767c. http://www.gepower.com
gepower.com 71. Matta RK, Mercer GD, Tuthill RS (2000) Power
53. Mercury 50, Recuperated gas turbine generator set, systems for the 21st century – H GT combined-
Solar ® Turbines. www.solarturbines.com cycles. GER-3935B, GE Energy
54. Cox JC, Hutchinson D, Oswald JI (1995) The Wes- 72. Smith RW, Polukort P, Maslak CE, Jones CM, Gar-
tinghouse/Rolls Royce WR-21 gas turbine variable diner BD (2001) Advanced technology combined
area power turbine design. ASME paper 95-GT-54, cycles. GER-3936a, GE Power Systems
international gas turbine and aeroengine congress 73. Gülen SC (2013) Performance entitlement of super-
and exposition, Houston, 5–8 June 1995 critical steam bottoming cycle. J Eng Gas Turbines
55. Hofer DC, Gülen SC (2006) Efficiency entitlement Power 135:#124501
for bottoming cycles, GT2006-91213. ASME turbo 74. European Association for the Promotion of Cogene-
expo – power for land, sea & air, Barcelona, ration (2001) A guide to cogeneration
8–11 May 2006 75. Phylipsen GJM, Blok K, Worrell E (1998) Handbook
56. Moran MJ, Shapiro HN (1988) Fundamentals of on international comparisons of energy efficiency in
engineering thermodynamics. Wiley, New York the manufacturing industry. Department of Science,
57. Gülen SC, Smith RW (2008) Second law efficiency Technology and Society, Utrecht University, The
of the Rankine bottoming cycle of a combined cycle Netherlands
power plant. ASME Paper GT2008 -51381, ASME 76. Gülen SC (2010) A proposed definition of CHP
Turbo Expo 2008, Berlin, 9–13 June 2008 efficiency, Power, June 2010, pp 58–63. http://
58. Bohn D (2006) SFB 561: aiming for 65% CC effi- www.powermag.com
ciency with an air-cooled GT. Modern power sys- 77. Energy Information Administration (EIA)
tems, pp 26–29 (2010) Electric power annual. http://www.eia.doe.
59. Mutassim Z (2008) New gas turbine materials. Tur- gov/fuelelectric.html
bomachinery international, Sept/Oct 2008 issue, 78. Davis LB, Black SH (2000) Dry low NOx combus-
pp 38–42 tion systems for GE heavy-duty gas turbines. GER-
60. Bohn D, Dilthey U, Schubert F (2004) Innovative 3568g. http://www.gepower.com
Technologien für ein GuD-Kraftwerk mit 65% 79. Lefebvre AH (1995) The role of fuel preparation in
Wirkungsgrad. VDI-Berichte 1857:13–25 low-emission combustion. J Eng Gas Turbines
61. Rao AD, Robson FL, Geisbrecht RA (2002) Power Power 117:617
plant system configurations for the 21st century, 80. Roointon P, Moore GD (2001) Gas turbine emissions
ASME turbo expo 2002, Amsterdam, 3–7 June 2002 and control. GER-4211. http://www.gepower.com
62. Lundberg WL, Veyo SE, Moeckel MD (2003) A high 81. Hilt MB, Waslo J (1984) Evolution of NOx abate-
efficiency solid oxide fuel cell hybrid power system ment techniques through combustor design for
using the Mercury 50 advanced turbine system gas heavy-duty gas turbines. J Eng Gas Turbines Power
turbine. ASME J Eng Gas Turbines Power 125:51–58 106:825
63. Massardo AF, Lubelli F (2000) Internal reforming 82. Touchton GL (1984) An experimentally verified NOx
solid oxide fuel cell – gas turbine combine cycles prediction algorithm incorporating the effects of steam
(IRSOFC-GT); part I: cell model and cycle thermo- injection. J Eng Gas Turbines Power 106:833
dynamic analysis. ASME J Eng Gas Turbines Power 83. Davi MA (1994) GE gas turbine combustion flexi-
122:27–35 bility. GER-3946, GE energy
Natural Gas Power 307

84. Miller HE (1994) Development of the GE quiet com- 103. Bhargava R, Bianchi M, Campanari S et al
bustor and other design changes to benefit quality (2010) A parametric thermodynamic evaluation of
GER-3551. http://www.gepower.com high performance gas turbine based power cycles.
85. Kehlhofer R, Warner J, Nielsen H, Bachmann R J Eng Gas Turbines Power 132:#022001
(1999) Combined cycle gas & steam turbine power 104. Lynch M (2016) The confusion about natural gas
plants, 2nd edn. PennWell Corp., Tulsa prices, Forbes Blog, 23 May 2016
86. Peters MS, Timmerhaus KD, West RE (2004) Plant
design and economics for chemical engineers,
5th edn. McGraw-Hill, Boston
87. Bejan A, Tsatsaronis G, Moran M (1996) Thermal Books and Reviews
design & optimization. Wiley, New York 105. Bejan A (2006) Advanced engineering thermody-
88. Energy and Environmental Economics (2014) Capi- namics, 3rd edn. Wiley, New Jersey
tal cost review of generation technologies. Energy 106. Boss M (1996) Steam turbines for STAG™ com-
and Environmental Economics, San Francisco. bined cycle power systems. GER-3582E. http://
www.ethree.com www.gepower.com
89. Gülen SC, Mazumder I (2013) An expanded cost of 107. Chase D (2001) Combined cycle development evolu-
electricity model for highly flexible power plants. tion and future. GER-4206. http://www.gepower.com
J Eng Gas Turbines Power 136:#011601 108. Colegrove D, Mason P, Retzlaff K, Cornell D (2001)
90. Gülen SC (2011) A more accurate way to calculate Structured steam turbines for the combined cycle
the cost of electricity. Power, June 2011, pp 62–65 market GER-4201. http://www.gepower.com
91. As reported in the press per Potential Gas Committee 109. Constant EW II (1980) The origins of the turbojet
report, Potential supply of natural gas in the United revolution. The Johns Hopkins University Press,
States (31 Dec 2008), Potential Gas Agency, Colo- Baltimore/London
rado School of Mines, Golden 80401–81887 110. Cotton KC (1998) Evaluating and improving steam
92. Gambini M, Vellini M (2003) CO2 emission abate- turbine performance, 2nd edn. Cotton Fact, Rexford
ment from fossil fuel power plants by exhaust gas 111. Denton JD (1993) Loss mechanisms in turbomachines,
treatment. J Eng Gas Turbines Power 125:365–373 the 1993 IGTI scholar lecture. J Turbomach 115:621–656
93. Wagman D (2010) Can natural gas displace coal? 112. Dunn MG (2001) Convective heat transfer and aero-
Power Eng (issue):4 dynamics in axial flow turbines. J Eng Gas Turbines
94. The future of natural gas – an interdisciplinary MIT Power 123:637–686
study, Interim report by MIT energy initiative, ISBN 113. Elmasri MA (2007) Design of gas turbine combined
(978-0-9828008-0-5), 2010, Massachusetts Institute cycle and cogeneration systems – theory, practice and
of Technology, Boston optimization. Seminar Notes, Thermoflow, Sudbury.
95. Robb D (2010) CCGT: breaking the 60 percent effi- info@thermoflow.com
ciency barrier. Power Eng Int 18(Issue 3). www. 114. Han JC, Dutta S, Ekkad SV (2000) Gas turbine heat
peimagazine.com transfer and cooling technology. Taylor & Francis,
96. Review of status of advanced materials for power New York
generation, technology status report, cleaner coal 115. Horlock JH (2001) Combined power plants: includ-
technology programme, Department of Trade and ing combined cycle gas turbine (CCGT) plants.
Industry, London Krieger Publishing Company, Malabar
97. Tukagoshi K, Muyama A, Uchida S et al (2005) Lat- 116. Kehlhofer R, Hannemann F, Stirnimann F, Rukes B
est technology for large capacity gas turbine. MHI (2009) Combined cycle gas & steam turbine power
Tech Rev 42(3):1–5 plants, 3rd edn. PennWell Corp, Tulsa
98. ElKady AM, Evulet A, Brand A (2009) Application 117. Lakshminarayana B (1996) Fluid dynamics and heat
of exhaust gas recirculation in a DLN F-class com- transfer of turbomachinery. Wiley, New York
bustion system for postcombustion carbon capture. 118. Lefebvre AH, Ballal DR (2010) Gas turbine combus-
J Eng GTs Power 131:#034505 tion: alternative fuels and emissions, 3rd edn. CRC
99. Kailasanath K (2000) Review of propulsion applica- Press, Taylor & Francis Group, Boca Raton
tions of detonation waves. AIAA J 38(9):1698–1708 119. Boyce MP (2006) Gas turbine engineering handbook,
100. Goldmeer J, Tangirala V, Dean A (2008) System-level 3rd edn. Gulf Professional Publishing, Houston
performance estimation of a pulse detonation based 120. Nag PK (2006) Power plant engineering, 2nd edn.
hybrid engine. J Eng Gas Turbines Power 130:#011201 Tata McGraw-Hill Publishing Co. Ltd., New Delhi
101. Tangirala VE, Rasheed A, Dean AJ (2007) Perfor- 121. Saravanamuttoo HIH, Rogers GFC, Cohen H,
mance of a pulse detonation combustor-based hybrid Straznicky PV (2009) Gas turbine theory, 6th edn.
engine, GT2007-28056, ASME turbo expo – power Pearson Prentice Hall
for land, sea & air, Montreal, 14–18 June 2007 122. Traupel W (1977) Thermische Turbomaschinen,
102. Gülen SC (2010) Gas turbine with constant volume Erster Band, Thermodynamisch-strömungstechnische
heat addition, ESDA2010-24817, ASME 2010 10th Berechnung, 3rd edn, neuarbeitete und erweiterte
Biennial conference on engineering systems design Auflage. Springer, Berlin/Heidelberg/New York
and analysis, Istanbul, 12–14 July 2010
 achievable heat release obtainable in a furnace at
Coal and Peat: Global constant pressure.
Resources and Future Supply Carboniferous A geologic period and system
that extends from 360 to 300 million years ago.
Mikael Höök Coal preparation Physical and mechanical pro-
Department of Earth Sciences, Uppsala cesses often applied to coal to make it suitable
University, Uppsala, Sweden for a particular use.
Coalification The process that transforms vege-
tation to peat succeeded by the transformation
Article Outline of peat to lignite and later higher coal ranks.
Cretaceous A geologic period and system span-
Glossary ning from 145 to 66 million years ago.
Definition of the Subject and Its Importance Devonian A geologic period and system span-
Introduction ning from 416 to 360 million years ago.
Coal and Peat: Global Resources Exploration The examination of an area by means
Future Directions of surface geological mapping, seismic techniques,
Bibliography geophysical methods, drilling of boreholes, and
sampling of the underlying sediments and rocks.
Glossary Jurassic A geologic period and system
extending from 200 to 145 million years ago.
Anthracite Anthracite is the highest rank of coal Macarels Microscopic organic constituents of
because it has undergone the greatest degree of coal that may differ significantly in composition
metamorphosis away from peat. It features low and properties depending on rank and geology.
volatile matter (<10%) and high carbon, giving Ordovician A geologic period and system span-
it the highest energy content of all coals. Semi- ning from 490 to 445 million years ago.
anthracite is somewhere in the middle between Permian A geologic period and system
low volatile bituminous coal and anthracite. extending from 300 to 250 million years ago.
Ash Inorganic residues remaining after combus- Reserves The recoverable volumes that are depen-
tion. It is less than the initial mineral matter dent on certain arbitrary limits in respect of thick-
content because of chemical changes during ness, depth, quality, economics, legal restrictions,
combustion, i.e., the loss of water, carbon diox- and other factors. Reserves are always less than
ide, and sulfurous compounds. resources.
Bituminous coal Bituminous coal lies between Resources The amount of coal in place before
subbituminous coal and semianthracite in exploitation; all resources may or may not be
terms of rank. This rank of coal is commonly economically or technically recoverable; see
divided into additional subgroups dependent Reserves.
upon the content of volatile material. Silurian A geologic period and system spanning
Calorific value Corresponds to the amount of heat from 445 to 416 million years ago.
per unit mass when combusted. Can be expressed Stripping ratio The ratio of the thickness of over-
as gross calorific value, which is the amount of burden to that of the total workable coal section.
heat liberated during combustion under standard- Tertiary A geologic period and system
ized conditions at constant volume so that all of extending from 66 to 2.6 million years ago.
the water in the products remains in liquid form, Triassic A geologic period and system spanning
or as net calorific value, which is the maximum from 250 to 200 million years ago.

© Springer Science+Business Media, LLC, part of Springer Nature 2020 309

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_161
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_161-3
310 Coal and Peat: Global Resources and Future Supply

Volatile matter Non-moisture component of production mechanism and the limitations that are
coal that is liberated at high temperature in imposed on it require a wide variety of approaches
the absence of air. and conceptual infrastructures.

Definition of the Subject and Its

Importance Introduction

Coal is the second most important fuel currently Natural resources are vital in supporting the con-
used by mankind, accounting for nearly 30% of tinued well-being of the world’s population.
the world’s primary energy supply. It provides Metals, fossil fuels, minerals, clean water, and
41% of global electricity supplies and is a vital many other resources are critical components for
fuel or production input for the steel, cement, and many aspects of modern society. Of all available
chemical industries. However, coal is a fossil fuel resources, few are as important as energy. In fact,
formed from organic material by geological pro- energy has been described as the ultimate resource
cesses over millions of years. Hence, coal is a by several studies [1–4]. In physics, energy is
finite resource in terms of human time scales, defined as the ability to do work. Einstein even
and its continued availability is important to the showed the equivalence between energy and mass
world economy. [5], implying that everything in the universe is
Peat is a related substance but is classified made from energy in various forms. Naturally, it
somewhere between a fossil fuel and biomass. follows that energy influences everything and
The energy sector uses peat as a fuel to generate cannot be substituted for other resources.
electricity and heat. It also has applications in Energy resources take many forms, ranging
industrial, residential, and other sectors, but from fossil fuels and uranium to biomass, human
global consumption of peat is insignificant in labor, or wind power. Currently, over 81% of all
comparison to coal. Peat shares many similarities primary energy used by mankind is derived from
with coal and is increasingly often grouped with fossil fuels [6]. Petroleum dominates, with
coal for resource estimates in reports and assess- roughly 31% of world primary energy being
ments by public agencies. derived from this particular hydrocarbon, coal at
Knowing how coal and peat are created is vital number two (~29%), followed quickly by natural
to understanding how deposits are formed and gas (~21%). The contributions from other energy
what their basic properties are. Geology provides sources range from 2% to 10% (biomass, nuclear,
models and methodologies for describing deposits hydropower) to less than 0.5% (wind, solar).
and where to find them. Exploration, drilling, and To understand the conceptual importance of
surveys provide the data necessary to map deposits coal, one can begin by studying the exceptional
and assess the resources they contain. Classifica- expansion of fossil energy in the last 200 years
tion schemes are also central to understanding how (Fig. 1). The development of modern society, pop-
the terms relate to the underlying data. ulation, and the world economy under this brief
Future production of coal and peat is essential period has been remarkable. Nearly all this growth
for the development of global energy supplies. It is has been powered by fossil energy in various
only the produced volumes that can be used in forms, originally coal but later oil and natural gas.
human activities, and a detailed appreciation of The growth soared and mankind both prospered
the production process is essential in understanding and multiplied in many places, all thanks to cheap
future supply developments. Factors such as econ- and abundant fossil fuels that provided useful
omy, technology, legal, and environmental con- energy for a multitude of activities.
straints affect the recoverable share of the The postwar boom also coincides with the
available resources, i.e., the reserves. Understand- green revolution in agriculture and the use of
ing the complexity and the greater whole of the high yield agricultural techniques based on
Coal and Peat: Global Resources and Future Supply 311

Coal and Peat: Global

Resources and Future
Supply, Fig. 1 World
production of fossil energy
since 1800 based on the
author’s compilation and
interpolation of historical
production statistics

synthetic fertilizers and high energy use. In fact, different qualities and ranks as well as different
fossil energy has been described as the principal degrees of structural complexity.
raw material of modern agriculture [7–10]. Fur- This entry provides a fundamental overview of
thermore, energy is intimately linked to economic global peat and coal resources, their distribution,
growth and development [11, 12]. To summarize, and some of the central characteristics in their
nothing is as important to society as energy, and recovery and utilization. In addition, it presents a
this should never be overlooked when dealing brief summary of resource classification schemes
with questions of sustainability. and their importance in understanding the
The rise of the industrial nations in the West- amounts that may be available for future supply.
ern world was chiefly driven by the coal that Finally, some resource and reserve statistics will
powered steam engines, transportation, and be presented with discussion. Estimates of global
industry as well as heating homes. Important resources must be dated and can naturally change
industrial complexes and cities arose near coal- as we learn more about their geology. The figures
rich regions, such as the Ruhr or Pennsylvania. In presented here should be seen as a snapshot of the
the USA, coal had replaced biomass as the dom- current state of our knowledge. Nevertheless,
inant primary energy source by the late nine- such estimates indicate where the major deposits
teenth century [13]. Similar events occurred in and mining activity are currently located and from
the UK, Germany, and other nations around the where future production might originate.
early twentieth century. Today, we are witnessing
the rise of China and India as new industrial
powers, both countries growth fueled chiefly by Coal and Peat: Global Resources
coal. However, petroleum eventually surpassed
coal as the dominant fuel globally in the 1950s, Formation of Coal and Peat
but coal has remained the second most important Coal and peat derive from vegetable remains that
energy resource ever since. have been deposited in various environments
Sedimentary layers of coal and peat beds range where it has been transformed by geological pro-
in age from the Upper Paleozoic to the recent cesses to potentially useful energy resources. The
(geologically) and are located all over the world. sedimentation of coal and peat has been exten-
The layers are a result of the accumulation of sively studied for over a century. This section
vegetable matter in a specialized deposition envi- describes the general mechanism behind the for-
ronment. Local conditions and events can produce mation of peat and coal.
312 Coal and Peat: Global Resources and Future Supply

Mires are the accepted general term for peat- during the initial stages. A compaction ratio of
forming ecosystems of all types. The cycling of 9:1 or more is not uncommon [16]. Continued
matter in most ecosystems is relatively fast and burial and the addition of heat and time will
complete. In contrast, mires are characterized by break down the complex hydrocarbon compounds
an incomplete cycling resulting in a positive carbon in the peat through a series of possible reactions.
balance. Peat-forming areas are generally wet places Methane and other gaseous alternation products
where the annual input of dead organic matter from are typically expelled from the deposit, while
the photosynthetic activity of the plant cover other elements disperse. The result is that the
exceeds annual breakdown. Over time organic deposit becomes increasingly carbon rich. The
material accumulates and becomes preserved. Peat process begins with plant debris and continues
deposits typically occur in waterlogged environ- through peat, lignite, subbituminous coal, bitumi-
ments where water prevents oxidation of the organic nous coal, anthracite coal, to the pure carbon
remains. The presence of water limits the diffusion mineral graphite. This process is referred to as
rate of gasses that leads to low oxygen availability. coalification.
The large heat capacity of water also induces a low Some perspectives on the necessary time scales
temperature, which inhibits decomposing and behind the formation of peat and coal resources
decomposition-facilitating organisms and leads to are indispensable for a better understanding of the
reduced rates of decay. whole picture. The surface part of peat below the
Peat formation is typically initiated by two living ground layer, being less than 300 years old,
mechanisms. One is called terrestrialization, amounts on average to 10.2% of the total peat
which is when bodies of water are replaced by carbon volume [17]. Only the deeper and basal
peat-forming mires. The other mechanism is parts of the peat are thousands of years old. In
paludification, which is replacement of dry land comparison, coal requires much longer time
by mires due to rising groundwater table or some periods to form. Analysis of thick bituminous
other hydrological change in the local environ- coal bed layers indicates that optimum peat-
ment. Peat-producing wetlands are divided into forming conditions must have been maintained
ombrogenous peatlands, which are rain-fed sys- for 5,000–10,000 years for every meter of clean
tems, and topogenous peatlands, which owe their bituminous coal [18]. Other studies have shown
origin to a place and its surface/groundwater that the global consumption of fossil fuel in 1997
regime [14]. Moore defined several important sub- was created from ancient plant matter
categories such as bogs, bog forests, marsh, fens, corresponding to over 400 times the net primary
swamps, floating swamps, and swamp forests [15]. production of the world’s current biota [19]. Coal
The plant debris undergoes several chemical cannot be formed at the same rate society uses it
changes due to bacterial decay, compaction, heat, and is therefore a nonrenewable fuel. Peat requires
and time, which eventually leads to formation of shorter time scales but is still slowly formed in
peat. Pristine plant parts such as roots, barks, or terms of human time scales.
spores are possible components of peat and can be Studies have shown similarities between the
used to determine which type of plants originally extensive coastal plain peat deposits in Southeast
made up the peatland. The peat harvested in the Asia and the North American/European Carbonif-
Northern Hemisphere was chiefly formed during erous coal-bearing sequences in terms of sediment
the Holocene epoch just after the retreat of the accumulation, mineralogy, and geochemistry
colossal glaciers that covered most of Europe. [20–23]. Several models provide a good under-
Consequently, most of the constituent peat- standing of the cyclic nature of coal sequences,
producing plant species are still thriving and con- the lateral continuity of coal beds, as well as phys-
tinue to produce peat as of today. ical and chemical properties. However, no single
The peat must be buried by additional sedi- model can provide a comprehensive explanation of
ments to be transformed to coal. Burial leads to all aspects regarding occurrence, development, and
compaction of the peat and squeezes out water types of coal.
Coal and Peat: Global Resources and Future Supply 313

Coal and Peat: Global

Resources and Future
Supply, Fig. 2 Principle
mechanism of the formation
of coal and peat. From
Steve Greb and Kentucky
Geological Survey
(Reproduced with
permission)

There are many books and articles on peat and certain coal components. Geological age distribu-
coal geology that can provide more detailed tions of the world’s principal black coal and lig-
descriptions of sedimentation, structural changes, nite deposits have been compiled by Walker [26].
and other factors and how they influence the for- During the Carboniferous Period, most of what
mation process. The formation of coal and peat is today the Northern Hemisphere was located
can be described in a simplified manner (Fig. 2). near the equator and consisted of tropical peat
mires, which later formed the coal basins of West-
Occurrences of Coal and Peat ern and Eastern Europe, the Eastern USA, and
The oldest known land plant remains have been former Soviet Union. Coals were also formed in
dated to the Ordovician/Silurian Era [24] but did the areas that today make up South America,
not produce any noteworthy accumulations. Cer- southern Africa, India, Australia, and Antarctica,
tain deposits of Devonian age are the oldest but this formation occurred under cooler, more
known coal accumulations, but it was not until temperate conditions. Coals formed at this time
the Carboniferous/Permian Period that the world’s now make up the bulk of the world’s black coal
plant cover was sufficiently rich to produce and reserves. They are generally high in rank but may
preserve significant amounts of coal. Although have undergone significant structural change.
coal and peat are continuously being formed, The climate changed during the Permian
there are a few major episodes of coal accumula- Period and the lush vegetation of the northern
tion throughout history. In total, three major coal areas eventually ceased to deposit coal-forming
formation periods have occurred, and these coals sediments. New but less abundant coal formation
form the bulk of the world’s coal reserves. They occurred during the Triassic, Jurassic, and Creta-
occurred during the Carboniferous/Permian ceous Periods. Such coals can chiefly be found in
Period, the Jurassic/Cretaceous Period, and the Canada, China, Russia, and the USA. Changes in
Tertiary Period. floral types, such as the onset of angiosperm
Plate tectonics and climate have been floras, as well the variety of deposition environ-
described as important factors for answering ments generally lead to the formation of coals that
when and where coal is deposited [25]. Places are more complex in their chemical makeup.
with favorable conditions have changed over Analysis of macarels and other constituents in
time as continental plates have drifted or when coal provides insight in the specific conditions
climatic properties have been altered over the long that were present during their accumulation.
time scales involved in the formation process. The third major period of coal formation took
Changes in vegetation types can also be traced to place during the Tertiary Period, and those coals
314 Coal and Peat: Global Resources and Future Supply

form the bulk of the world’s brown coal reserves Anthracite: 30 MJ/kg
as well as important parts of the black coals cur- Bituminous coal: 18.8–29.3 MJ/kg
rently being mined. Tertiary coals are mostly lig- Subbituminous coal: 8.3–25 MJ/kg
nite, but in some areas, temperature change has Lignite: 5.5–14.3 MJ/kg
resulted in higher-ranking coals.
The age and stratigraphy of all major coal Some of the world’s present peat deposits were
deposits have been studied extensively, in partic- formed as early as a few hundred thousand years ago,
ular those formations that are of economic inter- although the main part of the global peat resources
est. Meticulous surveys on individual fields are were formed during the last 10,000 years [30]. Occur-
important for development and actual production, rences of peat are linked to climate, vegetation, local
but such details for a single field can easily fill an environment, and precipitation that allow peat-
entire bookshelf! More general overviews of the forming conditions to arise. There are also less peat
world’s most important coal regions have been occurrences in the Southern Hemisphere, as there is
made by others [26, 27]. Reviews on various less land there. Present-day peat occurrences can be
pollutants in coal, such as arsenic or metals, can predicted well by suitable models [31].
also be found in more specialized literature. Peat is present all over the world, and estimates
Sometimes it is convenient to express coal conclude that 2–3% of the global surface is covered
deposits in energy terms, especially if they are by peat-forming areas, equivalent to about 4 million
going to be used as energy sources. Depending km2. This figure includes some 200,000 km2 of
on specific properties of the coal in a deposit, the tropical peats in Southeast Asia. A major proportion
actual energy content may vary significantly. Coal of the world’s peatlands are located in North Amer-
is divided into different ranks depending on the ica and the northern parts of Asia, but significant
degree of coalification as it matures from peat to deposits are also found in Northern and Central
anthracite. Coal is broadly divided into these four Europe. In total, over 90% of the world peatlands
ranks, namely, anthracite, bituminous, subbitumi- are in the temperate and cold belt in the Northern
nous, and lignite [28, 29], although much more Hemisphere. The remainder is located in tropical
detailed systems can be found. and subtropical latitudes and mostly located under
There are many classification systems used all forests. Such accumulations have been identified in
over the world, each with their own set of defini- the tropical regions of Africa, Latin America, as well
tions. The terms brown coals, black coals, hard as Indonesia and southern/eastern Asia.
coals, stone coals, thermal coals, and steam coals Estimates of the total area of the world’s
are commonly used, but their meaning may vary peatlands vary considerably from study to study.
from country to country. Lignite and subbituminous The basic problem is to separate areas where the
coals are commonly seen as low-ranking coals, peat layer is 30 cm thick at minimum, i.e.,
while bituminous coal and anthracite are considered peatlands, from areas only covered with wetland
high-ranking coals. Caution should always be vegetation, such as floodplains, coastal lagoons,
exercised when analyzing global coal data to avoid mangroves, and swamps. Hence, only a few coun-
mixing up different classification schemes. tries have performed accurate peatland mappings.
It should be noted that the ranks tend to overlap More comprehensive discussions on peat and
in terms of energy and that the calorific value can closer description of specific formations are avail-
vary within a rank depending on local geological able in reference literature, scholarly journals, and
properties. Relatively high moisture levels and a reports from various geological surveys.
low carbon content and calorific value characterize
low-rank coals, such as lignite and subbituminous Coal and Peat Resource and Reserve
coal. Higher-rank coals, such as anthracite and Classifications
bituminous coal, have a higher carbon and energy It is first necessary to look closer at classifications
content and a lower level of moisture. The follow- and what the terms actually mean in practice before
ing energy contents of coal ranks are common: defining and discussing global resources of coal
Coal and Peat: Global Resources and Future Supply 315

and peat. Far from all coal and peat being available It is necessary to investigate and explore any
for utilization by society, some formations are tech- coal or peat deposits to ascertain whether some-
nically unfeasible or just too economically chal- thing can be mined economically and that a prod-
lenging to recover or may even prove to be uct can be obtained that is marketable. Two
undesirable to consumers, even though they are common viewpoints are typically used to assess
geologically available. As a result, it is important resources and reserves:
to be precise when dealing with terms related to
future supply of coal and peat as possible energy 1. Degree of geologic assurance
sources for society. This is especially true for out- 2. Degree of economic feasibility
looks related to sustainable use and responsible
stewardship of these resources. The method of estimation along with the defi-
Concepts such as endowment or potential are nition and expression of geological assurance is a
multifaceted, and the meaning of such terms is key factor for the reliance of coal tonnage esti-
disputed by different schools of thought. One mates. Details surrounding the topography, tec-
school argues that resource endowment is crustal tonic variability, structural alteration, and
abundance, while the other school maintains that postdepositional erosion provide uncertainties and
the concept is relatively meaningless for supply should be taken into account in resource mappings.
assessments and should be avoided. As always, For example, heavily disturbed coals with great
the truth is located somewhere in between the variation in thickness will need many more data
two extremes. Crustal abundance is solely deter- points for an accurate and reliable assessment than
mined by geological processes, which are rea- relatively undisturbed coals of constant thickness.
sonably well understood by geologists. Data point spacing criteria and how far trends are
Furthermore, the physical in situ amounts will extrapolated from measurement points are essential
also be virtually fixed as a result of the slow for the degree of geologic assurance (Fig. 3). Eco-
underlying formation processes, even though in nomic feasibility is dependent on the content of
situ estimates might change in reports as a result sulfur and other unwanted pollutants, thickness,
of new information. depth, environmental constraints, coal price versus
Available coal accumulations can be divided mining costs, and other parameters that affect
into reserves and resources. Reserves are defined recovery and consumer desirability.
as being proved and recoverable under present Sadly there is no international and uniform
technology and economic conditions. Resources method for assessment, categorization, and desig-
include additional discovered but uneconomic or nation of coal resources and reserves. Several
unavailable amounts together with undiscovered agencies, countries, and organizations have devel-
inferred, assumed, and speculative quantities. oped their own sets of definitions and methods to
This distinction is justified by the idea that over calculate resources and reserves. However, it is
time, and perhaps with new technology and higher possible to roughly convert assessments made
prices, production and exploration activities will within one system to another and form aggregated
allow a reclassification of some resources into pictures of the entire world situation. This section
reserves [32]. The terms resources and reserves will focus on the system used by the United States
are often used synonymously by laymen, but it Geological Survey (USGS), while closer descrip-
should be absolutely crystal clear that this is a tions on other systems can be found in more
completely erroneous and misleading practice! specialized literature. For comparison, the
Being specific and using the right term are essen- Australian and Russian systems are conceptually
tial to paint a realistic picture in any supply assess- similar but differ in details. In contrast, the UK
ment or when planning for a sustainable future. system is excellent for mining operation but not
Studies have highlighted this problem and the very functional for large areas that may contain
need to include non-geological factors in future coal but are not mined. USGS use the following
supply estimates [13, 33]. definition for coal resources [34]:
316 Coal and Peat: Global Resources and Future Supply

Coal and Peat: Global

Resources and Future
Supply, Fig. 3 Reliability
categories based solely on
distance from drilling
measurement points.
Measured and indicated
coal together form the
demonstrated coal category
(Based on USGS resource
classification system [34])

Coal resources are naturally occurring concentra- adjacent to the mine or property boundaries, or
tions or deposits of coal in the Earth’s crust, in such (3) simply left because for a variety of other rea-
forms and amounts that economic extraction is
currently or potentially feasible. sons [34]. Mining and production practices also
place constraints based on thickness, depth, rank,
This category is divided into a number of sub- and other properties on the recoverable amount of
divisions, such as inferred, indicated, or measured the in situ resource. USGS has chosen the follow-
resources, depending on geological certainty, ing definition for the reserve base:
depth, and thickness. Technology and proper Reserve base: those parts of the identified resources
investment can, at least in theory, recover all that meet specified minimum physical and chemical
available in situ coal if the required economic criteria related to current mining and production
practices, including those for quality, depth, thick-
conditions exist. However, resource numbers are ness, rank, and distance from points of measure-
inadequate as they do not take into account eco- ment. The reserve base is the in-place demonstrated
nomics, the fundamental utilization properties, or (measured plus indicated) resource from which
even legal issues surrounding extraction in a geo- reserves are estimated. The reserve base may
encompass those parts of a resource that have a
graphical region. Recoverability is the crucial reasonable potential for becoming economically
property for actual production, and this leads to a available within planning horizons beyond those
variety of attempts to express the recoverable that assume proven technology and current eco-
volumes of coal that can be extracted from the nomics. The reserve base includes those resources
that are currently economic, marginally economic,
available resources. some of those that are currently sub-economic, and
Far from all coal being recovered from a coal some of the resources that have been or will be lost-
bed, some amounts will be unavoidably lost in the in-mining but whose attributes indicate possible
recovery process for technical reasons. This future recovery.
includes coal that is (1) left to support under- Recoverable volumes are also affected by many
ground mine roofs, (2) used as barrier pillars other factors that make them dynamic. Higher
Coal and Peat: Global Resources and Future Supply 317

prices, new exploration, or improved technology resources and reserve terminology, as used by
yield increased amounts of recoverable coal. In a EIA, USGS, and US Bureau of Mines, can be
similar way, increased freight rates, extraction obtained from the EIA [39]. The estimated recov-
costs, or introduction of various restrictions can erable reserves category corresponds to proved
decrease recoverable volumes. Likewise, political reserves discussed in the BP Statistical Review
influences can greatly influence demand patterns or of World Energy [40], proved recoverable
even cause direct physical restrictions for mining. reserves used by World Energy Council (WEC)
USGS [34] finally uses the following terminology [41], or reserves as used by the German Federal
to denote reserves: Institute of Geosciences (BGR) [42].
Reserves: Virgin and/or accessed parts of a coal
reserve base which could be economically extracted
Coal and Peat Resources in the World
or produced at the time of determination consider-
ing environmental, legal, and technologic To make accurate coal and peat resource esti-
constraints. mates, exploration geologists make detailed stud-
ies of relevant accumulations. They take
Important restrictions are legal prohibitions on
measurements at appropriate intervals and extrap-
coal mining in certain areas and/or the use of a
olate using the appropriate classification catego-
particular mining methodology, such as strip min-
ries, knowledge of the local geology, and good
ing. This can prevent the extraction of coal,
judgment. In theory, there is not a fundamental
despite the fact that the coal might be technically
difference between underground and surface for-
and economically recoverable. For example, land-
mation maps. However, shallow drilling is
use regulations and federal laws prevent mining
cheaper and makes it possible to drill more
near homes, public buildings, or federally funded
holes, which makes surface mappings generally
highways [30]. Accessibility, availability, and
more reliable. Because deep boreholes are costly
restriction factors are linked to the society and its
(which means less are drilled thus making the
legal and political climate. This can have a major
assessment less reliable), the exploration geolo-
influence on the volumes practically available for
gist uses various geophysical methods to supple-
recovery. Detailed discussions of how availability
ment the drilling data.
of recoverable coal is estimated by the USGS can
Resources are commonly expressed as in situ
be found in other works [35–37]. The Energy
tonnage, which is the total amount occurring in
Information Administration (EIA), the energy
place. It can be calculated from the thickness of
authority of the USA, utilizes the following defi-
the deposit, the area of extent, and the relative
nition to include restrictions and accessibility in
density. Thickness is commonly determined at
their reports and assessments of American coal
each point of measurement, while the area is mea-
commonly used by non-geologists, planners, and
sured on a map or plan. The relative density is
policy-makers:
normally taken from the section that will be
Estimated recoverable reserves: cover the coal in mined, or alternatively it might be estimated
the demonstrated reserve base considered recover- using the known average ash content of a coal or
able after excluding coal estimated to be peat seam. The formula for in situ tonnage is:
unavailable due to land use restrictions or coal
currently economically unattractive for mining

(and after applying assumed mining recovery thicknessðmÞ  area m2
rates).

 relative density tons=m3
EIA creates this category by applying eco- ¼ in situ tonnage
nomic feasibility criteria factoring downward
from the reserve base. The coal reserve term At regular intervals, scientists and governmen-
used by EIA may better be described as potential tal agencies conduct and publish world coal
coal reserves, as they are generally not proved by resource assessments. In particular, since its estab-
any detailed drilling [38]. A full discussion of coal lishment in 1923, the World Energy Council
318 Coal and Peat: Global Resources and Future Supply

(WEC) has reviewed all forms of energy resources resource estimate of around 16,000 billion tons
including coal and peat. Data from the WEC come (reported as recoverable reserves plus estimated
from reports by the WEC member committees and additional in situ amounts). In subsequent assess-
constitute a sample that reflects the information ments, their world coal resources have varied
available in particular countries. German Federal from 10,000 to 14,000 billion tons, with most
Institute for Geosciences and Natural Resources estimates closer to the lower part of the range. In
(BGR) has also released a number of assessments comparison, BGR estimates of world coal
on global coal resources. These agencies will resources have varied from 6,000 to 18,400 bil-
serve as the main data source for the resource lion tons. Using either estimate, it is clear that
figures presented here. global coal resources are large and the world
Peatland resource estimates are commonly production of coal (>7 billion tons) is seemingly
quoted on an area basis because initial quantifica- insignificant in comparison. However, there are
tion normally arises through soil survey programs reasons to doubt these agencies’ resource assess-
or via remotely sensed data that focuses on geo- ments. First, there are large fluctuations in the
graphical extent. Estimates of the world’s peatland reported resource amounts as definitions and
area vary considerably (Table 1). This is partly due reporting practices change, coal-bearing areas
to different wetland classification systems and are added or subtracted, and new geological infor-
partly due to insufficient survey data. Peat mation becomes available.
resources have been estimated to be 5,000–6,000 For instance, total coal resources in the USA
billion cubic meters but are hard to quantify in any decreased by more than 60% between 1913 and
detail [30]. WEC gives an estimate of 3,500–4,000 1962. In contrast, between 1968 and 1974, coal
billion cubic meters [41]. Uncertainties in water estimates were revised upward, in part due to
content, ash content, and many other factors influ- improved accuracy in assessing the amount of
ence the result significantly when converting to lower-rank coal [33] but also by including
tonnage. This can be seen as a reason for undiscovered resources in the estimates. Further
circumventing tonnage figures and preferring vol- exploration may show that a substantial part of
umetric units in resource assessments. However, it the undiscovered resources do not exist [33]. Sim-
is clear that the world possesses a huge tonnage of ilar modifications, as well as some brief reasoning
peat overall. The important question is how much behind the revisions, can be found in recent world
of those resources can realistically be recovered in coal resource estimates from the BGR (Table 3).
the future. In 2005, BGR reported that they changed coal
World coal resource estimates can be traced resource estimates because there was insufficient
back to 1936, when the WEC gave a world substantiating information for many countries and
many countries had skipped reporting years. In
2006, they incorporated new resources in China
Coal and Peat: Global Resources and Future Supply,
and former Soviet Union that lead to a 115%
Table 1 The distribution of peatlands and other terrestrial
wetland ecosystems in the world (Data source [30]) increase in world hard coal resources and a
200% increase in lignite resources. In 2007,
Peatlands Wetlands Total
Region [km2] [km2] [km2] BGR made yet another upward revision, increas-
Africa 58,000 282,000 340,000 ing world hard coal resources by 68% and lignite
Asia 1,119,000 1,149,000 2,268,000 resources by 36%. In both cases, the majority of
Australia and 14,000 10,000 24,000 the increase came from new resources identified in
Oceania the USA by a recent USGS study [43]. Additional
Central and 102,000 330,000 432,000 upward revisions came from recent reevaluations
South America of Chinese, Australian, and South African coal
Europe 957,000 n.a. 957,000
data in subsequent BGR publications.
North America 1,735,000 657,000 2,392,000
Another important observation and fact of this
World 3,985,000 2,428,000 6,413,000
world is that the peatlands are unevenly
Coal and Peat: Global Resources and Future Supply 319

distributed with the vast majority concentrated in of this world and a function of favorable geology
a few countries (Table 2). This trait can also be in certain parts of the world, bestowing significant
seen in the assessment literature of world coal deposits of peat and coal on relatively few coun-
resources (Table 3). This inequality is a sad fact tries. However, uncertainties are present, and

Coal and Peat: Global Resources and Future Supply, Table 2 Distribution of peat resources for selected countries
and regions (Data source [41])
Region Country Peatlands [km2]
2007 2013
Africa Guinea 5,250 1,950
Nigeria 7,000 1,600
South Africa 9,500 n.a
Uganda 14,200 n.a
Zambia 11,060 12,200
Total Africa 58,410 56,170
Asia China 10,440 53,120
Indonesia 270,000 206,950
South Korea 6,300 n.a
Malaysia 25,360 25,900
Burma 9,650 1,230
Total Asia 331,880 1,490,360
Australia and Oceania Australia 150 1,350
New Zealand 2,600 3,610
Papua New Guinea 6,850 10,990
Total Australia and Oceania 9,640 8,050
Central and South America Brazil 15,000 23,860
Chile 10,470 10,470
Falkland Islands 11,510 11,510
Guyana 8,140 8,140
Venezuela 10,000 10,000
Total South America 62,400 130,800
Europe Belarus 23,970 23,500
Estonia 9,020 9,020
Finland 89,000 89,000
Germany 14,200 13,000
Iceland 10,000 8000
Ireland 11,800 11,800
Latvia 6,400 6,600
Norway 23,700 28,010
Poland 12,000 12,500
Russia 568,000 1,390,000
Sweden 64,000 66,000
Ukraine 10,080 8,000
UK 19,260 27,500
Total Europe 876,510 525,670
North America Canada 1,113,280 1,113,280
Cuba 6,580 5,290
Mexico 10,000 1,000
USA 214,000 625,000
Total North America 1,354,220 1,762,270
320 Coal and Peat: Global Resources and Future Supply

Coal and Peat: Global Resources and Future Supply, Table 3 Estimated total coal resources (recoverable reserves
plus other in situ resources), as published by the BGR. All numbers in million metric tons (Mt) (Data source [42])
Country 2004 2005 2006 2007 2015
USA 1,090,404 1,057,921 1,359,404 6,719,682 6,680,329
China 1,089,787 1,089,789 4,367,000 4,367,000 5,462,672
India 259,217 291,863 252,359 229,867 260,544
FSU 2,295,275 2,326,121 2,915,146 2,931,293 2,957,655
Australia 298,400 304,801 193,700 148,200 1,598,761
South Africa 163,851 163,851 48,751 28,559 213,560
Germany 87,827 91,497 84,573 83,065 82,982
Poland 70,262 101,281 179,459 179,459 178,913
UK 5,220 7,220 190,005 187,132 186,770
Indonesia 10,581 64,829 26,839 23,948 109,825
World n.a. 6,038,991 9,553,840 15,510,618 18,412,036

there is a problem with a lack of available data for it is possible to assume a universal recovery factor
some countries and regions. Also, as seen in based on historic recovery factors for similar
Tables 2 and 3, significant differences may appear deposits. The US Bureau of Mines performs this
from one assessment to another depending on type of calculation regularly [33]. However, the
scope, used definitions, and areas included. most reliable method is a deposit-by-deposit anal-
To summarize, resource estimates are impor- ysis because it can incorporate local economic
tant from a geological point of view and are vital conditions as well as the technical recoverability
for understanding where recoverable amounts of the minable coal.
may be located. However, it is essential to under- Coal resources are primarily of academic inter-
stand that society does not have economic access est and unfortunately provide little information
to all the reported resources. In situ resources about what quantity of the resource can be practi-
include all amounts, even those located deep or cally recovered. For example, let us look at the
of poor quality, making resource estimates largely Gillette coalfield in Wyoming and USGS’ com-
an academic figure [33]. Resources must be prehensive investigation of its coal supply.
converted into reserves by a detailed evaluation Luppens et al. [44] assessed the Gillette coalfield
of the degree of certainty of the size of the partic- and its 11 coal beds and estimated the original coal
ular coal resource and the economic, legal, and in place to be 182 Gt, with no restrictions applied.
environmental constraints placed on its exploita- Available coal resources, which are the part of the
tion. Vast resources do not necessarily mean sig- original resource that is accessible for potential
nificant future production, as production is mine development after subtracting all restric-
dependent on many more factors than just geolog- tions, are about 148 Gt. Recoverable coal, which
ical availability. This naturally leads to the next is the portion of available coal remaining after
section in which reserves are discussed. subtracting mining and processing losses, was
determined for a stripping ratio of 10:1 or less
Coal and Peat Reserves in the World and resulted in a total of 70 Gt. Applying current
Reserves are usually calculated from resources by economic constraints and using a discounted cash
applying suitable and relevant constraints. Legal, flow at 8% rate of return, the coal reserve estimate
environmental, and geological factors prohibit the for the Gillette coalfield is only 9.1 Gt. While
recovery of the entire resource so reserve esti- economic conditions definitely have the largest
mates will always be less than resource estimates. impact, reductions caused by technical and avail-
Rather than perform a detailed constraint analysis, ability restrictions are also significant.
Coal and Peat: Global Resources and Future Supply 321

Extraction of peat and coal is primarily an the coalfields they studied. More importantly, they
economic activity that recovers valuable products demonstrated that the generous EIA reserve base
from certain deposits with a profit. Prevailing and would be far less than the reported figure using
anticipated price, location of the deposit relative their methodology. The EIA estimate did not
to the market, transportation costs, specific use, include the effects of mined-out prime reserves
and the production cost are all important factors and environmental, industrial, economic, and
for determining the feasibility of exploitation. social considerations. Nor did it include the tech-
Forecasting costs, estimating revenues, and han- nical aspects of the minability and washability of
dling inflation over long time periods are an the individual seams. In contrast, second-
uncertain business. Additional uncertainty comes generation assessments of available coal amounts
from the regulatory environment, particularly in the Pittsburgh coal bed were performed and
environmental laws, tax policies, and similar. In described by Watson et al. [48]. Their assessment
essence, profitability is a dynamic factor and can was that half of the original coal resource in place
quickly change. To obtain an accurate estimate of had already been extracted leaving only 15 Gt and
the recoverable volume of a deposit requires a those resources were increasingly being restricted
meticulous economic study in combination with by new environmental laws, encroaching popula-
comprehensive physical surveys. tion, and affected by technological factors.
An often overlooked but fundamental factor is Several organizations regularly compile peat
the net energy analysis. Hubbert [45] wrote about and coal reserve estimates. Many national energy
net energy in relation to oil but the same principle agencies and geological surveys keep reserve
applies to all energy resources: “there is a different records for their own country. The World Energy
and more fundamental cost that is independent of Council and the BGR collect country level data
the monetary price. That is the energy cost of and publish global coal reserve estimates. The BP
exploration and production. So long as oil is Statistical Review of World Energy [40] is another
used as a source of energy, when the energy cost commonly referenced source of reserve estimates,
of recovering a barrel of oil becomes greater than but it only reproduces the numbers compiled by
the energy content of the oil, production will cease World Energy Council without backdating the
no matter what the monetary price may be.” In the estimates. Table 4 shows the latest reserve figures
case of coal and peat, a poor net energy return will from these three sources. Several studies have
prevent deposits that are too deep, too compli- criticized the reserve estimates found in published
cated, or too thin from being usable as energy literature for not adequately accounting for all
sources. Mining companies leave those resources resource constraints [13, 33, 49–51].
behind because they require more energy to Changes in policy can also cause rapid and
exploit than they can yield in return. Several significant changes to reserve estimates. For
researchers have used net energy analysis to deter- example, when Germany removed their mining
mine mining boundaries in an objective subsidy, Germany’s coal reserves were greatly
manner [46]. reduced [52, 53]. In Russia’s case, Malyshev
Rohrbacher et al. developed a methodology for [54] shows that according to the best estimates
describing recoverability using geological, tech- by experienced geologists, mining engineers, and
nical, and economic factors [47]. They concluded economists, around 40% of the Russian coal
that past coal estimates did not address the amount reserves cannot be regarded as of any practical
that might realistically be available for production interest because they are highly unprofitable.
after environmental and technological restraints Highly reliable estimates of coal reserves are
are considered. They showed how air quality reg- rare in part because the factors that influence
ulations, mine reclamation legislation and regula- recoverability change frequently. Changes in clas-
tions, new safety and health standards, and sification systems, economics, regulations, and
advancements in mining technology had dramat- other factors can cause significant changes in
ically changed the economic competitiveness of reserves over time. In the case of the USA,
322 Coal and Peat: Global Resources and Future Supply

Coal and Peat: Global Resources and Future Supply, BP dates them differently. All numbers in million metric
Table 4 Estimated coal reserves for selected countries. tons (Mt) (Data sources [40–42])
Note how WEC and BP numbers are identical, except that
Country or region WEC BGR BP
Proved at end 2007 Proved at end 2007 Proved at end 2008
Australia 76,200 76,900 76,200
Brazil 7,059 6,596 7,059
Canada 6,578 6,578 6,578
China 114,500 192,000 114,500
Germany 6,708 40,936 6,708
India 58,600 73,510 58,600
Indonesia 4,328 6,981 4,328
Japan 355 355 355
Kazakhstan 31,300 18,947 31,300
Poland 7,502 16,329 7,502
Russia 157,010 161,553 157,010
South Africa 30,408 28,559 30,408
Ukraine 33,873 34,375 33,873
UK 155 432 155
USA 238,308 262,971 238,308
World 826,001 989,913 826,001
Proved at end 2011 Proved at end 2014 Proved at end 2015
Australia 76,400 62,095 76,400
Brazil 6,630 6,596 6,630
Canada 6,582 6,578 6,582
China 114,500 131,614 114,500
Germany 40,548 36,321 40,548
India 60,600 90,276 60,600
Indonesia 28,017 25,668 28,017
Japan 347 350 347
Kazakhstan 33,600 25,605 33,600
Poland 5,465 21,632 5,465
Russia 157,010 160,364 157,010
South Africa 30,156 9,893 30,156
Ukraine 33,873 34,375 33,873
UK 228 70 228
USA 237,295 253,124 237,295
World 891,530 984,624 891,531

significant downward revisions of the reserves conflicts, or indirectly, such as taxation. In the
occurred when mining restrictions and other USA, the Surface Mining Control and Reclama-
non-geological factors decreased recoverability tion Act of 1977 directly prohibits surface mining
[13]. Despite the problems with reserve estimates, in national parks, forests, wildlife refuges, trails,
they are still the best available estimates of how wild and scenic rivers, wilderness, and recreation
much of a resource can be economically and tech- areas. Many national parks have been established
nically recovered. over the years, further limiting land available for
Regulations can directly affect coal recover- coal production. Petsch [55] studied Europe and
ability, such as prohibitions due to land use the Ruhr area and concluded that people have
Coal and Peat: Global Resources and Future Supply 323

become sensitive to intrusions made on their envi- Coal and Peat: Global Resources and Future Supply,
ronment, for instance, from mining or colliery Table 5 Peat reserves for selected countries. All numbers
in million metric tons (Mt) (Data source [61])
spoil. Physical limitations from land use regula-
tions have generally increased over time in many Peat reserves [Mt]
countries. Historically, air quality regulations Country 2008 2015
have increased the demand for low-sulfur coal Belarus 400 2,600
Canada 720 720
[13, 56–58]. Expanded environmental awareness
Estonia 60 60
in the 1960s and 1970s lead to heated debates and
Finland 6,000 6,000
stricter regulations.
Ireland n.a. n.a
In 2009, President Obama stated that “at a time
Latvia 76 190
of such great challenge for America, no single Lithuania 190 190
issue is as fundamental to our future as energy.” Moldova n.a. n.a
Part of that challenge is reducing the dependence Russia 1,000 1,000
on imported oil and political pressure to limit Sweden n.a n.a
greenhouse gas emissions. Governments may Ukraine n.a n.a
very well tighten coal regulations and pursue USA 150 150
future efforts for addressing anthropogenic cli- Other countries 1,400 1,400
mate change and CO2 emissions. Generally, the World total 10,000 12,000
air quality regulations in the USA have been an
environmental success story although it has done
little to mitigate CO2 emissions [59]. Interestingly, USGS Mineral Commodity Survey on peat [61]
tight regulations are not the sole factor in future contains some reserve estimates (Table 5). Only a
emissions. Regulatory uncertainty can be signifi- small portion of peatlands are available for extrac-
cant as businesses delay investments until they tion due to environmental constraints and other
can make solid return on investment decisions factors.
[60]. Exactly how future CO2 policies will affect Only a few countries have a peat industry of
coal reserves remains to be seen, but it is likely significant importance, so the global peat reserves
that such policies will reduce the recoverable pro- should be regarded as highly uncertain. Econom-
portion of coal resources. ics, demand, environmental regulations, and other
Sustainable development, environmental stew- factors are still uncertain for many regions. The
ardship, and the importance of natural resources for world possesses large amounts of peat, but that
the continued well-being of mankind are driving should not be seen as guarantee for future produc-
forces for future policies, at least in any foreseeable tion. If demand for peat increases, reserve esti-
outlook. Historical trends point toward increased mates will in all likelihood become more accurate.
restrictions, and this implies that future energy As with coal resources, global peat reserves are
policies will result in significantly decreased coal very unevenly distributed. A small number of
availability. However, politics are constantly nations control the vast majority of the world’s
shifting, and sudden change in any direction cannot coal reserves. The USA, Russia, China, Australia,
be ruled out. To summarize, recoverability is a India, and South Africa control over 80% of the
complex issue, and future supply of coal will be world’s coal reserves (Table 4) – a remarkable
dependent on many factors ranging from geologi- concentration of resources. Peat shows a similar
cal availability and quality to socioeconomic distribution with just a few countries as the
acceptability and energy policies. world’s major reserve holders (Table 5).
Because peat is a globally minute energy
source, reserve estimates are difficult to find, and Global Coal and Peat Production
there is no known comprehensive global study. Production is the activity that actually transforms
Volumetric data can be converted to tonnage by coal and peat into something useful for consumers
using the average bulk density of produced peat. and society. Some of the produced volumes are
324 Coal and Peat: Global Resources and Future Supply

inherently lost in preparation, washing, or trans- Producing coal requires more complex opera-
formation to specialized products such as bri- tions and often underground mining methods.
quettes to meet customer needs. Generally coal Coal mining can be done from the surface by
and peat are burned to create heat, to produce using draglines and other machines to remove
steam, and then to transform steam into electricity. the overburden and expose the coal for recovery.
They are also used in industrial chemical pro- Surface mining methods can generally recover a
cesses, steel making, and horticulture. high percentage of the coal and are relatively cost-
The world produced 7.8 Gt of coal in 2015 [40] effective. However, there are limits to the amount
and a comparatively tiny amount of peat – only of overburden that can be practically removed to
around 28 million tons [61]. Peat’s role as an obtain access to the coal. Above certain stripping
energy source is negligible in the global energy ratios (the amount of overburden that must be
context, while coal is the second most important removed divided by the amount of recoverable
fuel, second only to oil. As a result, this section coal), the surface method becomes unworkable
will briefly cover peat and coal mining and high- and is replaced by deep mining.
light some important aspects of production. In comparison, underground mining is more
Currently, just a few countries are the principal complicated and expensive and recovers less
producers of peat, and all those nations are located coal since certain parts must be left to support
in Europe. Finland, Ireland, Belarus, Russia, Swe- the roof. Up to 40% of the total coal in a coal
den, Ukraine, and the Baltic states are the largest seam may be left as supporting pillars, although
producers of peat. Finland is perhaps the most this sometimes can be recovered at a later stage.
significant peat consumer and derives 6% of its Worldwide about 40% of the production origi-
total energy from peat as a way to replace nates from surface mines, while 60% stems from
imported fuels [62]. In Finland, over 90% of the underground operations. Surface mining domi-
total peat production goes to energy production. nates in some important producing countries
In contrast, the USA uses most of its peat for [63]. The chosen production method will have a
horticultural uses [61]. Peat is often used as a significant impact on the economic as well as
practical alternative for smaller heat and/or technical restrictions that affect the recoverability
power plants. There is also significant trade of the coal.
flows of peat among certain countries, but most Since the 1950s, the six largest coal reserve
of the peat is still being consumed by the same holders have increased their share of world pro-
country that produced it. duction (Fig. 4). China in particular has rapidly
Wetlands that contain peat have to be drained increased its coal production and now accounts
before they can be exploited. Peat in situ con- for nearly 50% of world coal production. The
tains roughly 90% water, and drainage can USA is the second largest coal producer, followed
reduce this significantly. Additional drying by by India, Australia, Russia, and South Africa.
the wind and sun will further reduce moisture Even though coal is mined in over 50 different
content down to 40–50%. The traditional pro- countries [63], the bulk of production comes from
duction method was to hand-cut peat into sods these six nations.
or strips, but this has been replaced by mechan- It is impossible for coal and peat production to
ical means. Another production technique avoid affecting and disturbing the local environ-
yields fine peat granules by using a mechanical ment. Measures that restrict or even prevent mining
miller to disturb and granulate the top layer of in environmentally sensitive and other protected
the peatland. Some of the produced peat goes to areas are affecting legal and economic feasibility
the production of peat briquettes, which is a of new mines. Mine tailings, acid mine drainage,
more practical fuel for households. WEC [41] leakage of associated substances like sodium and
states that the bulk of peat production is heavy metals, land subsidence, impact on local
obtained from milling, which is more easily wildlife, and landscape changes are problems that
used for electricity and heat plants. are intrinsically associated with production.
Coal and Peat: Global Resources and Future Supply 325

Coal and Peat: Global

Resources and Future
Supply, Fig. 4 Total
world production
1950–2015. Prior to 1985,
the coal production from the
Donetsk basin is not
included in Russia. In 2015,
just six countries accounted
for slightly above 80% of
the total world coal
production. These countries
will determine the shape of
future global coal
production (Data source
[40, 64])

Avoiding addressing such issues could stir up polit- New inventions will likely continue to improve
ical or legal controversies [13, 55]. production even further as research and develop-
In addition, safety regulations are another com- ment continues.
plicating factor. Mining companies must agree to However, there is a danger in expecting too
follow specific safety laws to receive a mining much from new technology. Increasing depletion
permit. Some major hazards are explosions of the best and most easily accessible coal seams
(caused by methane gas and dust), roof collapse, will force future production to be derived from
and machine failure. Surface mining is generally less favorable deposits. This will increase mining
safer since the risk of explosions is lower, and the costs which in turn will make energy more expen-
risk of roof collapses is almost completely absent. sive across the whole economy. Some studies
Since many coal deposits are not reachable by even show that technological progress will not
surface mining, the company must resort to under- meaningfully be able to offset depletion
ground mining. Closer discussions on safety- [69–71]. The study authors reason that new tech-
related issues have been made by others [65, nologies and mining methods that boost efficiency
66]. Good safety brings additional costs, while and decrease labor requirements will be unable to
poor safety will hurt the public acceptance of the offset the higher production costs of more com-
mining operations as well as potentially risking plicated and labor-intensive deposits.
the demand for the “unsafe” coal. Poor safety may Coal has been produced in many countries for
also result in governmental intervention and the many decades, which enables us to perform lon-
closure of mining operations if required safety gitudinal studies of production data. Several coun-
regulations are not followed. tries are currently producing only a small share of
New technology and recovery methods can what they have produced historically (Fig. 5).
change or potentially even revolutionize produc- Much of the reduced production can be explained
tion. Improved mechanization, use of long-wall by the move to other energy sources like petro-
techniques that minimize the coal left behind to leum or uranium, depletion of commercially fea-
support the roof, and other measures have sible deposits, energy policy changes, and similar.
improved productivity and economic competi- Countries like the UK still have a significant
tiveness of mining operations [67, 68]. These amount of coal resources left (Tables 3 and 4) but
new mining methods have made previously not enough to suggest that coal will be the fuel of
unreachable deposits feasible for development, choice again. The remaining coal is generally
especially when they reduce production costs. located deeper than the previously mined coal
326 Coal and Peat: Global Resources and Future Supply

Coal and Peat: Global

Resources and Future
Supply, Fig. 5 Historical
coal production of the UK,
France, and Japan. All
countries experienced
production disturbances
around the two world wars
and the great depression
(Data source [40, 64])

and will be more expensive to mine in terms of others use physical or statistical models derived
labor and capital investment. Given the draw- from natural sciences. The life of the reserves and
backs of coal, especially in terms of climate whether production can match demand are impor-
change because of its high carbon content, it is tant questions for planning and sustainability.
not at all certain that demand for coal will ever The economist Adelman [72] has said that the
grow again. This naturally leads to the question of amount of mineral in the earth is an irrelevant,
future production of coal and peat. nonbinding constraint to production. He further
says that the amount of resource in the ground is
Future Production probably unknowable but surely is unimportant
Future production of peat and coal is an open and and of no economic interest. This is correct in the
complex question. The resources might be vast, very limited sense that we will never recover the
but only a small amount is recoverable once legal, very last amounts of coal or any other resource
environmental, and economic constraints are from the earth. The ultimate recovery is undefined
accounted for. In particular, it would be a mistake and governed by non-geological circumstances.
to think that the huge worldwide resource esti- Although the recovery factor might be undefined
mates mean that coal and peat will provide a that does not make it unlimited, it is bound
steady energy supply in to the future. In addition between 0% and 100% of the resource in the
to the previous factors, production also depends earth’s crust, which is clearly defined but not
on social acceptability (i.e., who would want to known exactly. Thus, the fact that the amount of
see mountain top removal), technical restrictions, recoverable coal is undefined and unknown can-
the development and competitiveness of alterna- not be an argument against the existence of future
tive energy sources, and expected demand. production limitations. The slow creation of coal
Although the following overview is unable to and peat in comparison to their rapid extraction
discuss all the factors that will affect future pro- from mining imposes an inherent finiteness to the
duction, it will illuminate some of the complexity amount of coal and peat available.
and leading indicators of future production Hubbert [73] proposed a model that extrapo-
projections. lates the production curves of finite resources such
There are many methodologies for creating as petroleum or coal. His model assumes that
production and supply projections for coal or production begins at zero, just before production
peat. Some methods are quantitative and rely on has started, and ends at zero, when the resource
field data, while others are qualitative. Some ana- has been fully exhausted. In between, the produc-
lysts base their projections on economic theories; tion curve passes through one or more maxima.
Coal and Peat: Global Resources and Future Supply 327

The actual shape of the production curves may production history since current production
vary, but they are ultimately limited by the amount reached 8 Gt per year in 2013 and production
of resource that was deemed to be recoverable. increase appears to have halted.
Hubbert initially proposed a bell-shaped curve for Hubbert’s method is sensitive to the assumed
an idealized production behavior, representing ultimate reserve that constrains the curve so it is
various development stages without giving any essential to have the best possible resource esti-
mathematical description for it. Later, he derived mate. It has been described as a reasonable model
a mathematical framework chiefly based on the provided that it is acknowledged that reserves are
logistic function and curve fitting to production dynamic and that other factors also influence pro-
and discovery data [74]. duction [75]. Growth curve methods can provide
Figure 5 shows a simple example of how actual insight into the general long-term flow of finite
production can be described by mathematical energy resources, but perturbations caused by
curves. When doing long-term projections, a suit- sudden and unforeseen near-term economic or
able curve is fitted against historical production political changes cannot be predicted. Therefore,
and constrained by a reasonable estimate on the long-term life cycle projections should not be
recoverable volumes that may be ultimately pro- used as a substitute for meticulous economic stud-
duced. The total resource in the earth’s crust will ies that forecast perturbations in coal production
ultimately place a limit on future production, but over the upcoming years or decades [76]. Studies
the real limit will be lower since it is unrealistic to have also used the development of regulations,
assume that all geologically existing volumes can restrictions, and recoverable volume estimates to
be recovered. It still remains to be seen if global create production projections [13].
coal production levels recovers from its peak in Other quantitative assessments that have relied
2013 (Figs. 4 and 6). on reported reserve and resource assessments
Hubbert himself [73] used 2,500 Gt as the found that coal output can increase until 2050 or
ultimate world coal reserves and placed the pro- even 2100 [32, 76]. Thielemann et al. [32]
duction peak around 2,150 at an annual output of assumed that resources will simply be converted
6 Gt. This implies that there are no near-term into reserves and then produced due the self-
problems for coal production. However, his sim- regulating price-driven supply cycle. This model
ple approach does not consider all aspects of gives no foreseeable production limitations within
recoverability. It also agrees poorly with actual the next century. Mohr and Evans [77] point out

Coal and Peat: Global

Resources and Future
Supply, Fig. 6 Historical
production of anthracite in
Pennsylvania fitted with
both logistic and Hubbert
curves (Data source [13])
328 Coal and Peat: Global Resources and Future Supply

that shrinking coal resource and reserve estimates Increasing global competition for the remaining
over the last 25 years seems to indicate that factors petroleum might cause countries to increase their
other than the self-regulating supply mechanism reliance on domestic energy sources. In summary,
are playing a role. In any case, future production there are many factors that can affect future pro-
will still be limited to the countries and regions duction, and without a holistic approach, it is
where the resources are located. difficult to make reliable projections.
Watson et al. [48] concluded that future coal The current world dependence on fossil fuels
production from the Pittsburgh coal area is depen- also brings about another problem. Since fossil
dent on a complex set of factors, including future fuels dominate the existing energy system, they
transportation costs and environmental policies. will have to power any possible shift to renewable
Without a holistic view of all factors affecting or other energy sources [78]. The effect this has on
recoverability and geological availability, no rea- fossil fuel reserves will depend largely on the size
sonable supply projection can be made. Although and speed of the shift, since fossil fuel energy used
predictions of future production will always be will be additional to that already in use.
uncertain, it is important to remember that produc- In conclusion, there are a lot of coal and peat
tion is dependent on much more than just geolog- available in terms of resources. However, far from
ical existence. Claims that abundant resources will all the resources are suitable for production and
automatically provide future energy supply need to future production will largely depend on the
be treated with suspicion – they oversimplify the choices we make. The model and assumptions
production situation. about recoverability one uses have a significant
For example, there are significant unexploited impact on future production projections. The key
coal reserves located in eastern Siberia and far point is that huge worldwide resource estimates
eastern Russia. However, these deposits are far for coal and peat do not automatically translate
from major population centers and will need into an assured supply. Vast resources are present
major investments in infrastructure before they if society demands coal and peat and is prepared to
can be exploited. How fast this can occur is an pay the price for recovering them.
open question and depends primarily on eco-
nomic and political factors. Also, Russia might
never develop those deposits and may use other Future Directions
energy sources instead. A similar point can be
made for the USA and the vast coal resources Coal will remain as one of the most important
locked up in Alaska [43]. Underlying assumptions global energy sources for in the foreseeable
can greatly affect future production projections, future. Future development of coal production
and one must always remember that production is will be linked to the development of global energy
a complex issue with many factors involved. supply, which will affect virtually all aspects of
Recent concern for the CO2 emissions that society in some way. However, production is a
come from burning fossil fuels may turn coal complex issue that requires a holistic treatment.
and peat into undesired fuels for the future in This chapter has highlighted several of the indi-
some countries. On the other hand, countries that cators and parameters that affect future production
care less about the environment might turn to coal of coal and peat.
and peat as cheap, but polluting, alternatives to It is also clear that more work needs to be done
imported oil. Clean coal technologies and carbon in exploration and geology to better understand
capture and storage might alleviate much of the the occurrence of coal and peat. Better survey data
emission problems from combustion, which can and increased geologic certainty would also help
improve coals attractiveness. Cheap and reliable to improve the accuracy of resource
wind and solar energy or improved nuclear power estimates – especially in those countries and
might also deliver cheaper energy to consumers regions where few assessments have been made
than coal and peat if carbon emissions are taxed. or where only old data is available. There is no
Coal and Peat: Global Resources and Future Supply 329

substitute for a good geological understanding of 13. Höök M, Aleklett K (2009) Historical trends in Amer-
depositional systems and detailed surveys when ican coal production and a possible future outlook.
Int J Coal Geol 78(3):201–216. doi:10.1016/j.coal.
making resource estimates. 2009.03.002
Estimation of reserves and recoverable vol- 14. Diessel CFK (1992) Coal-bearing depositional sys-
umes involve consideration of numerous eco- tems. Springer-Verlag, Berlin
nomic, technical, legal, and social parameters 15. Moore PD (1987) Ecological and hydrological aspects
of peat formation. In: Scott AC (ed) Coal and coal-
relative to specific markets and expected demand bearing strata: recent advances, vol 32. Spec. Publ.
during a present and period with economic and Geo. Soc, London, pp 7–15
regulatory uncertainty. Emission and climate 16. Nadon GC (1998) Magnitude and timing of peat-to-
goals, such as the Paris Agreement, will also be coal compaction. Geology 26(8):727–730. https://doi.
org/10.1130/0091-7613
instrumental in determining future trajectories of 17. Mäkilä M (2006) Lot of peat deposits under 300 years
world coal use. Improved understanding and han- old in Finland. Geological Survey of Finland, Peat
dling of this complex set of factors are required for Research Report 59/2006.
accurate and reliable assessments in the future. 18. Ryer TA, Langer AW (1980) Thickness change
involved in the peat-to-coal transformation of bitumi-
Recoverability is not a topic that should be taken nous coal of Cretaceous age in central Utah.
lightly or oversimplified. Instead it needs holistic J Sediment Petrol 50:987–992
treatment to provide reasonable outlooks for wise 19. Dukes JS (2003) Burning buried sunshine: human
decision-making with regard to securing future consumption of ancient solar energy. Clim Chang
61(1–2):31–44. https://doi.org/10.1023/A:102639131
world supply of coal and peat. 7686
20. Neuzil SG, Supardi CCB, Kane JS, Soedjono K (1993)
Inorganic geochemistry of domed peat in Indonesia
Bibliography and its implication for the origin of mineral matter in
coal. Geol Soc Am Spec Pap 286:23–44
21. Cecil CB, Dulong FT, Cobb JC, Supardi K (1993)
Primary Literature Allogenic and autogenic controls of sedimentation in
1. Ehrlich PR, Ehrlich AH, Holdren JP (1970) Ecoscience: the central Sumatra basin as an analogue for Pennsyl-
population, resources, environment. W.H Freeman and vanian coal-bearing strata in the Appalachian Basin.
Company, San Francisco Geol Soc Am Spec Pap 286:2–22
2. Cook E (1977) Energy: The ultimate resource? Resource 22. Gastaldo RA, Allen GP, Huc AY (1993) Detrial peat
papers for college geography, Issue 77–4, 42 p foundation in the tropical Mahakam River Delta, Kali-
3. Simon J (1966) The ultimate resource 2. Princeton Uni- mantan, Eastern Borneo: sedimentation, plant compo-
versity, New Jersey sition and geochemistry. Geol Soc Am Spec Pap
4. Bromley DA (2002) Science, technology, and politics. 286:107–118
Technol Soc 24:9–26 23. Ruppert LF, Neuzil SG, Cecil CB, Kane JS
5. Einstein A (1905) Ist die Trägheit eines Körpers von (1993) Inorganic constituents from samples of a
seinem Energieinhalt abhängig? Ann Phys domed and lacustrine peat, Sumatra, Indonesia. Geol
323:639–641 Soc Am Spec Pap 286:83–96
6. IEA, 2015. Key world energy statistics 2015, see also: 24. Wellman CH, Osterloff PL, Mohuiddin U (2003) Frag-
http://www.iea.org ments of the earliest land plants. Nature 425:282–285.
7. Pimentel D, Hurd LE, Bellotti AC, Forster MJ, Oka IN, https://doi.org/10.1038/nature01884
Sholes OD, Whitman RJ (1973) Food production and 25. Butler J, Marsh H, Goodarzi F (1988) World coals:
the energy crisis. Science 182(4111):443–449 genesis of the world’s major coal fields in relation to
8. Green MB (1978) Eating oil: energy use in food pro- plate tectonics. Fuel 67(2):269–274
duction. Westview Press, Boulder 26. Walker S (2000) Major coalfields of the world. IEA
9. Pfieffer DA (2006) Eating fossil fuels: oil, food and the Coal Research, London
coming crisis in agriculture. New Society Publishers, 27. Saus T, Schiffer HW (1999) Lignite international.
Gabriola Island Rheinbraun AG, Cologne
10. Pimentel D (2007) Food, energy, and society. CRC, 28. American Society for Testing and Materials (2005).
Boca Raton Standard Classification of Coals by Rank. ASTM
11. Akinlo AE (2002) Energy consumption and economic D388–05, ASTM International
growth: evidence from 11 sub-Saharan African coun- 29. Carpenter AM (1988) Coal classification. IEA Coal
tries. Energy Econ 30:2391–2400 Research, London
12. Hondroyiannis G, Lolos S, Papapetrou E (2002) 30. Lappalainen E (1996) General review on world
Energy consumption and economic growth: peatland and peat resources. In: Lappalainen E (ed)
assessing the evidence from Greece. Energy Econ Global peat resources. International Peat Society,
24:319–336 Jyskä
330 Coal and Peat: Global Resources and Future Supply

31. Lottes AL, Ziegler AM (1994) World peat 47. Rohrbacher TJ, Teeters DD, Sullivan GL, Osmonson
occurrence and the seasonality of climate and vege- LM (1993) Coal resource recoverability – a method-
tation. Palaeogeogr Palaeoclimatol Palaeoecol ology. U.S. Bureau of Mines Circular 9368. http://
106(1–4):23–37 pubs.usgs.gov/usbmic/ic-9368/
32. Thielemann T, Schmidt S, Gerling PJ (2007) Lignite 48. Watson WD, Ruppert LF, Tewalt SJ, Bragg LJ
and hard coal: energy suppliers for world needs until (2001) The Upper Pennsylvanian Pittsburgh Coal
the year 2100 – an outlook. Int J Coal Geol 72:1–14 Bed: resources and mine models. Nat Resour Res
33. van Rensburg WCJ (1982) The relationship between 10:21–34. doi:10.1023/A:1011529430807
resources and reserves estimates for US coal. Res 49. Blackmore G, Ehrenreich SB (1987) Reserve data
Policy 8(1):53–58 base report of the National Coal Council: advisory
34. Wood GH, Kehn TM, Carter MD, Culbertson WC report to the U.S. Department of Energy. National
(1983) Coal Resource Classification System of the Coal Council, Arlington
U.S. Geological Survey. US Geological Survey Circu- 50. National Petroleum Council (2007) Facing hard truths
lar 891. http://pubs.usgs.gov/circ/c891/ about Energy. http://www.npchardtruthsreport.org/
35. Eggleston JR, Carter MD, Cobb JC (1990) Coal 51. U.S. National Academies (2007) Coal: research and
resources available for development – a methodology development to support national energy policy.
and pilot study. US Geol Surv Circ 1055. http://pubs. National Academies Press, Washington, D.C
usgs.gov/circ/c1055/ 52. Storchmann K (2005) The rise and fall of German hard
36. Carter MD, Gardner NK (1989) An assessment of coal coal subsidies. Energ Policy 33(11):1469–1492
resources available for development, central Appala- 53. Frondel M, Kambeck R, Schmidt CM (2007) Hard
chian region. US Geol Surv Open-File Rep 89–362. coal subsidies: a never-ending story? Energ Policy
http://pubs.usgs.gov/of/1989/of89-362/ 35(7):3807–3814
37. Luppens JA, Rohrbacher TJ, Haacke, JE, Scott DC, 54. Malyshev YN (2000) Strategy for the development of
Osmonson LM (2006) Status report: USGS coal the Russian coal industry. J Min Sci 36(1):57–65
assessment of the Powder River, Wyoming. U.S. 55. Petsch G (1982) Environmental problems of coal pro-
Geological Survey Open-File Report 2006–1072. duction in the federal republic of Germany with par-
http://pubs.usgs.gov/of/2006/1072/ ticular reference to the Ruhr. Environ Geochem Health
38. American Association of Petroleum Geologists 4:75–80
(2007) Unconventional energy resources and 56. Tobin RJ (1984) Air quality and coal – the US expe-
geospatial information: 2006 review. Nat Resour Res rience. Energ Policy 12:342–352
16:243–261 57. Yeager KE, Baruch SB (1987) Environmental issues
39. Energy Information Administration (1996) U.S. Coal affecting coal technology: a perspective on US trends.
Reserves, Appendix A, Specialized resource and Annu Rev Energy 12:471–502
reserve terminology. http://tonto.eia.doe.gov/ftproot/ 58. O’Brien B (1997) The effects of Title IV of the Clean
coal/052995.pdf Air Act Amendments of 1990 on Electric Utilities: an
40. BP (2016) BP Statistical Review of World Energy. Update. EIA report DOE/EIA-058297 distribution cat-
http://www.bp.com egory UC-950. ftp://ftp.eia.doe.gov/pub/electricity/
41. World Energy Council (1924) Survey of energy ef_caau1.pdf
resources 2013 and previous reports and statistical 59. Ackerman F, Biewald B, White D, Woolf T, Moomaw
yearbooks from previous world power conferences. W (1999) Grandfathering and coal plant emissions: the
World Energy Council, London. http://www. cost of cleaning up the Clean Air Act. Energ Policy
worldenergy.org/ 27:929–940
42. German Federal Institute of Geology and Natural 60. Patiño-Echeverri D, Fischbeck P, Kriegler E (2009)
Resources (1980–2015) Reserves, resources and Economic and environmental costs of regulatory
availability of energy resources. Various editions uncertainty for coal-fired power plants. Environ Sci
43. Flores RM, Stricker GD, Kinney SA (2004) Alaska Technol 43:578–584
coal geology, resources, and coalbed methane poten- 61. U.S. Geological Survey (2016) Mineral Commodity
tial. USGS report. http://pubs.usgs.gov/dds/dds-077/ data – Peat Statistics and Information. http://minerals.
44. Luppens JA, Scott DC, Haacke JE, Osmonson LM, usgs.gov/minerals/pubs/commodity/peat/
Rohrbacher TJ, Ellis MS (2008) Assessment of coal 62. Geological Survey of Finland (2009) Peat resources of
geology, resources, and reserves in the Gillette coal- Finland. http://en.gtk.fi/Resources/peat_resources.html
field, Powder River Basin, Wyoming. U.S. Geological 63. World Coal Institute (2005) The coal resource – a
Survey Open-File Report 2008–1202. http://pubs. comprehensive overview of coal. http://www.
usgs.gov/of/2008/1202/ worldcoal.org/
45. Hubbert MK (1982) Response to David Nissens 64. Mitchell B (2003) International historical statistics
remarks. http://www.hubbertpeak.com/Hubbert/to_ 1750–2000. Palgrave MacMillan, London
nissen.htm 65. Kecojevic V, Nor ZD (2008) Hazard identification for
46. Kurleyna MV, Tanaino AS (1997) Open-pit and under- equipment-related fatal incidents in the U.S.
ground mines – energy analysis of open-pit mining. underground coal mining. J Coal Sci Eng (China)
J Min Sci 33(5):453–462 15(1):1–6
Coal and Peat: Global Resources and Future Supply 331

66. Grayson RL (2008) Improving mine safety technology 78. Mohr SH, Evans GM (2009) Forecasting coal produc-
and training in the U.S. recommendations of the mine tion until 2100. Fuel 88:2059–2067
safety technology and training commission. J Coal Sci 79. Moriatry P, Honnery D (2009) What energy levels can
Eng (China) 14(3):425–431 the Earth sustain? Energ Policy 37:2469–2474
67. Szwilski AB (1988) Significance and measurement of
coal mine productivity. Min Sci Technol 6(3):221–231
68. Kulshreshtha M, Parikh JK (2002) Study of efficiency Books and Reviews
and productivity growth in opencast and underground 80. Thomas L (2002) Coal geology. Wiley, New York
coal mining in India: a DEA analysis. Energy Econ 81. Cobb CJ (1994) Modern and ancient coal-forming
24(5):439–453 environments. Geological Society of America Special
69. Tilton JE (2003) Assessing the threat of mineral deple- Paper
tion. Miner Energy 18:33–42 82. Lappalainen E (1996) Global peat resources. Interna-
70. Rodríguez XA, Arias C (2008) The effects of resource tional Peat Society; Geological Survey of Finland,
depletion on coal mining productivity. Energy Econ cop. 359 s
30:397–408. doi:10.1016/j.eneco.2007.10.007 83. Yudovich YE, Ketris MP (2005) Arsenic in coal: a
71. Topp V, Soames L, Parham D, Bloch H (2008) Pro- review. Int J Coal Geol 61(3–4):141–196
ductivity in the mining industry: measurement and 84. Seredin VV, Finkelman RB (2008) Metalliferous
interpretation. Productivity Commission, Melbourne coals: a review of the main genetic and geochemical
72. Adelman MA (1990) Mineral depletion, with special types. Int J Coal Geol 76(4):253–289
reference to petroleum. Rev Econ Stat 72:1–10 85. Schissler AP (2004) Coal mining, design and methods
73. Hubbert MK (1956) Nuclear energy and the fossil of. In: Cleveland CJ (ed) Encyclopedia of energy.
fuels. Shell Development Company, Houston Elsevier Academic, San Diego
74. Hubbert MK (1959) Techniques of prediction with 86. Schilstra AJ, Gerding MAW (2004) Peat resources. In:
application to the petroleum industry. Shell Develop- Cleveland CJ (ed) Encyclopedia of energy. Elsevier
ment Company, Dallas Academic, San Diego
75. van Rensburg WCJ (1975) ‘Reserves’ as a leading 87. George H, Meech J, Workman L (1986) Towards
indicator to future mineral production. Res Policy reducing the physical environmental impact of North
1:343–356 American surface coal mines; a review of potential
76. Milici RC, Campbell EVM (1997) A predictive pro- selective overburden handling techniques. Min Sci
duction rate life-cycle model for southwestern Virginia Technol 3(2):81–94
coalfields. Geol Sur Circ 1147 http://pubs.usgs.gov/ 88. Suárez-Ruiz I, Crelling JC (2008) Applied coal petrol-
circular/c1147/ ogy: the role of petrology in coal utilization. Aca-
77. Ion DC (1979) World energy supplies. Proc Geol demic, New York
Assoc 90:193–202
 Acid (rock or mine) drainage Many metal ore
Coal and Other Mining bodies and coal deposits contain significant
Operations: Role of quantities of sulfide minerals – often including
Sustainability the ore minerals themselves. When such min-
erals are brought to the surface, they react
Sandip Chattopadhyay1,2 and chemically with air and water producing sulfu-
Devamita Chattopadhyay3 ric acid, which may dissolve other minerals
1
Tetra Tech Inc., Cincinnati, OH, USA containing potentially toxic elements. This
2
US Environmental Protection Agency, Office of acid drainage from coal and metal mining
Pollution Prevention and Toxics, Washington, around the world can pollute water and the
DC, USA surrounding land, affecting plant and animal
3
ERG (Eastern Research Group, Inc.), Chantilly, life. Acid drainage is known as acid mine
VA, USA drainage when it is closely associated with
mining activities, and acid rock drainage
when this phenomenon occurs naturally, with-
Article Outline out human intervention. Both phrases are in
common use, although particular stakeholder
Glossary groups have particular preferences related to
Definition of the Subject and Its Importance the controversial nature of this issue.
Introduction Acidophile An organism that thrives in a rela-
Mining Life Cycle tively acidic environment.
Coal Mining Ammonification The biochemical process
Restoration Process whereby ammoniacal nitrogen is released
Sustainability and Mining from nitrogen-containing organic compounds.
Irrigation with Lime-Treated Acid Mine Drainage Amorphous Irregular, having no discernible
(AMD) order or shape. Rocks or minerals that possess
Application of Phytotechnology no definite crystal structure or form, such as
Biomining and Bioprocessing amorphous carbon.
Bioremediation Bioleaching Extraction of metal from solid min-
Metal Recovery: Biotic and Abiotic erals into a solution is facilitated by the metab-
Water Quality and Acid Mine Drainage: Pre-mine olism of certain microorganisms.
Predictions and Post-Mine Comparisons Biomining Extraction of specific metals from
Recent Sustainable Technologies their ores through biological means, usually
Future Directions bacteria. It is an actual economical alternative
Bibliography for treating specific mineral ores, involving
percolation and agitation techniques.
Glossary Community The people living around the mine
who are directly affected (both positively and
Abandoned mines Mines for which the owner negatively) by the mine’s activities.
cannot be found, or for which the owner is finan- Contaminated land/water Land/water containing
cially unable or unwilling to carry out cleanup. concentrations of potentially toxic materials
They may pose environmental, health, safety, and (organic or inorganic) elevated above the natural
economic problems to communities, the mining background concentrations in a particular area. In
industry, and governments in many countries. relation to mining, land or water contamination

© Springer Science+Business Media, LLC, part of Springer Nature 2020 333

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_864
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2019, https://doi.org/10.1007/978-1-4939-2493-6_864-4
334 Coal and Other Mining Operations: Role of Sustainability

may occur through fuel spills, runoff from waste Archaea are unique because unlike most life on
rock dumps, leaks from tailings impoundments, Earth that rely on oxygen and complex organic
windblown dust from tailings and waste rock, compounds for energy, Archaea rely on simple
smelter emissions, and drainage from mine work- organic compounds (e.g., acetate) and hydrogen
ings. Contaminated groundwater is caused by the for energy.
seepage of contaminated waters into aquifers. Mining life cycle The processes of exploration,
Crystalline A substance in which the constituent mining development, extraction, processing,
atoms, molecules, or ions are packed in a regularly refining, smelting, and mine closure.
ordered, repeating three-dimensional pattern. Ore Mineral-bearing rock that can be mined and
Denitrification A microbially facilitated process treated profitably under the existing economic
of nitrate reduction that may ultimately pro- conditions, or those conditions which are
duce molecular nitrogen through a series of deemed to be reasonable.
intermediate gaseous nitrogen oxide products. Reclamation Process of converting derelict land
Dissimilatory reduction Sulfate-reducing bac- (land that requires intervention before benefi-
teria reduce sulfate in large amounts to obtain cial use) to usable land and may include engi-
energy and expel the resulting sulfides as neering as well as ecological solutions.
waste; this is known as dissimilatory sulfate Remediation Environmental cleanup of land
reduction. They are anaerobes, which use sul- and water contaminated by organic, inorganic,
fate as the terminal electron acceptor. or biological substances.
Electrowinning The recovery of metal by elec- Restoration Seeks to artificially accelerate the
trolysis. An electric current is passed through a processes of natural succession by putting back
solution containing dissolved metals, and this the original ecosystem’s function and form.
causes the metals to deposit on a cathode. Tailings Mineral wastes produced from the pro-
Extractant An immiscible liquid used to extract cessing operations after the valuable minerals
a substance from another liquid. have been extracted.
Gypsum A sedimentary rock consisting of Waste rock The mineral wastes produced during
hydrated calcium sulfate. mine development – including overburden and
Heap leaching To dissolve minerals or metals out barren rock – and those parts of an ore deposit
of an ore heap using chemicals. For example, a below the economic cutoff grade. Often, and
cyanide solution percolates through crushed ore particularly in some metal deposits, the waste
heaped on an impervious pad or base pads during rock may contain sufficient sulfide mineral con-
heap leaching of gold. centrations to generate long-term acid drainage
Macrophyte An aquatic plant that grows in or problems.
near water and is either emergent, submergent,
or floating. They provide cover for fish and
substrate for aquatic invertebrates, produce oxy- Definition of the Subject and Its
gen, and act as food for some fish and wildlife. Importance
Mesophile An organism that grows best in mod-
erate temperature (typically between 15
C and Mining, minerals, and metals are important to the
40
C). The term is mainly applied to microor- economic and social development as they are
ganisms. It is also used to describe mesophilic essentials for modern living. However, supplies
anaerobic digestion, which takes place optimally of minerals, such as coal, are limited, and sustain-
around 37–41
C or at ambient temperatures able management of natural resources requires the
between 20
C and 45
C, where mesophiles maintenance, rational and enhanced use, as well
are the primary microorganisms present. as a balanced consideration of ecology, economy,
Methanogenesis Production of methane by bio- and social justice. Mining industry’s recognition
logical processes that are carried out by and acceptance of its sustainable development is
methanogens. A methanogen is a single-celled growing. In the mining and metals sector, this
microorganism and is a member of the Archaea. means that investments should be:
Coal and Other Mining Operations: Role of Sustainability 335

Technically Environmentally Financially Socially

Appropriate Sound Profitable Responsible

Leaders in the global mining business commu- efficiency and the need for more equitable access
nity have begun to assign a new strategic signifi- to resources worldwide. The environmental and
cance to the term sustainability. No longer health impacts of different mineral products in use
confined to the economic realm, sustainability need to be carefully managed. Where the risks
embraces a broad spectrum of organization char- associated with use are deemed unacceptable or
acteristics related to social and environmental are not known, the costs associated with using
responsibility. Profitability alone is inadequate as certain minerals may outweigh the benefits.
a measure of success, and that many of the non- The importance of sustainable development in
financial concerns associated with sustainability mining industry includes actions at all levels to:
are fundamental drivers of long-term shareholder
value. Failure to recognize these strategic issues • Support efforts to address the environmental,
threatens the very survival of a business economic, health, and social impacts of mining
enterprise. throughout the life cycle, including workers’
Large amounts of material are involved in health and safety. A range of partnerships, fur-
large-scale mining and minerals extraction. The thering existing activities at the national and
problems arising from the change in the chemis- international levels, among interested govern-
try of million tons of natural ore during the ments, intergovernmental organizations, min-
processing steps (such as, grinding, calcining, ing companies and workers, and other
and roasting operations) and their resultant bio- stakeholders to promote transparency and
availabilities are not well understood. Mining accountability for sustainable mining and min-
produces large volumes of waste, and decisions erals development.
regarding waste handling and other aspects of • Enhance the participation of stakeholders,
operations are often difficult and expensive to including local and indigenous communities,
reverse; they need to be made correctly initially to play an active role in minerals, metals, and
through mine closure planning. Another chal- mining development throughout the life cycles
lenge is the environmental legacy left by mining. of mining operations. This includes efforts
The environmental issues of current mining after mine closure for rehabilitation purposes,
operations are daunting enough. But in many in accordance with national regulations and
ways far more troubling are some of the continu- taking into account significant transboundary
ing effects of past mining and smelting. The loss impacts.
of biodiversity is the other great challenge of • Foster sustainable mining practices through the
mining sustainability. The loss of biodiversity provision of financial, technical, and capacity
is an irreversible loss. Conservation practices building support to developing countries and
that guarantee a minimum impact on biodiver- countries with economies in transition for the
sity must be adopted and implemented. mining and processing of minerals. Improving
As important as the methods of mining and value-added processing, upgrading scientific
beneficiation is how the minerals are used for and technological information, and reclaiming
sustainable development. An integrated approach and rehabilitating degraded sites might be con-
for the use of minerals must be developed. Current sidered for large and small-scale mining
patterns of minerals used raise concerns about operations.
336 Coal and Other Mining Operations: Role of Sustainability

Introduction focused on effectively integrating environment,

social, and economic impacts.
Population growth, continued urbanization and The quantity of mining waste produced fluctu-
industrialization, and the challenge of meeting ates yearly, and as the individual mines and
global demand for commodities are generally quarries manage their wastes according to local
agreed to be relatively predictable in their evolu- conditions there are no definitive statistics. After
tion while at the same time are expected to play being removed, waste rock, which often contains
key role in shaping the future context for the acid-generating sulfides, heavy metals, and other
mining sector. Today’s mining of mineral contaminants, is usually stored above ground in
resources faces the challenges related to the large free-draining piles. This waste rock and the
more complex extraction conditions, e.g., great exposed bedrock walls from which it is excavated
depths, natural hazards, environmental restric- are the primary sources of pollution caused due
tions, which all increase the production costs. to mining. The US mining industry produces
The main constraints to sustainability in the min- approximately 8,000,000 tons per year (ton/year)
ing sector derive from the ever-increasing demand of process residue that may contain hazardous
for mined resources, the consumption of resources species as well as valuable by-products. These
(mostly energy and water) needed to extract and process residues are generated by smelter off-gas
process metals, and the increasing pollution gen- cleaning at approximately 5,500,000 tons/year,
erated by the extraction process. The intensity of and bag house dust and wastewater treatment at
the environmental impact depends on what is approximately 2,100,000 tons/year [28]. Compa-
being mined, where, and how. Many environmen- rable statistics was obtained for other countries. It
tal issues are an inseparable part of any kind of was estimated that 96.4 million tons of mining and
mining activity and influence overall economic quarrying waste was produced in 2004 in the UK
system of mining industry. Sustainable manage- [4]. The Canadian mineral industry typically gen-
ment of environment in any way affected by min- erates one million tons of waste rock and 950,000
ing activity has crucial influence on a matter of tons of tailings per day, totaling 650 million tons
national environmental significance in countries of waste per year [24]. The right technology may
where mining of fossil fuels and valuable minerals be able to recover marketable by-products from
is under development. Over the years, mining and process residues to generate revenue and reduce
exploration companies have been highly criti- disposal costs for the mining industry. The pro-
cized for their destruction of the environment cess residue that cannot be reused can impact the
and unsustainable practices. However, recently water resources, and the effect may be manifest
companies are winning awards for operating throughout the life cycle of the mine and even
within stricter regulations that mean a lower car- long after mine closure.
bon footprint – and are taking pride in successful
reclamation projects that help support their social
license to operate.
The history of mining has been one of boom Mining Life Cycle
and bust periods that were a balance between
available natural resources, production costs, and Minerals are nonrenewable resources, and the
the ability to sustain profitability. In recent years, mines have finite lives. Mining represents a tem-
the mining industry leadership has emphasized porary use of the land, and during this temporary
that the modern day mining business has become use of the land, the mining life cycle can be
more complex with the advent of stricter regula- divided into the following stages: exploration,
tions, better-informed stakeholders, and closer development, extraction and processing, and
in-depth scrutiny of current and/or proposed oper- mine closure.
ations. Increased emphasis is being placed on the Exploration is the work involved in determin-
importance of sustainable operations that are ing the location, size, shape, position, and value of
Coal and Other Mining Operations: Role of Sustainability 337

an ore body using prospecting methods, geologic any mining, or higher or better uses.” Abandoned
mapping and field investigations, remote sensing mines can cause a variety of health-related hazards
(aerial and satellite-borne sensor systems that and threats to the environment, such as the accumu-
detect ore-bearing rocks), drilling, and other lation of hazardous and explosive gases when air no
methods. Building access roads to a drilling site longer circulates in deserted mines, use of these
is one example of an exploration activity that can mines for illegal residential or industrial dumping,
cause environmental damage. and others. Many closed or abandoned mines have
The development of a mine consists of several been identified by federal and state governments
principal activities: conducting a feasibility and are being reclaimed by both industry and
study – including a financial analysis to decide government.
whether to abandon or develop the property –
designing the mine, acquiring mining rights, filing
an Environmental Impact Statement (EIS), and Coal Mining
preparing the site for production. An example of
site preparation impacting the environment is Mining can affect both air and water. A fossil fuel
removal of the overburden (surface material that has been mined intensively is coal. As an
above the ore deposit that is devoid of ore min- example of environmental impact, effects of coal
erals) by excavation prior to mining. mining are presented below.
Extraction is the removal of ore from the Coal is and will continue to be a crucial ele-
ground on a large scale by one or more of three ment in a modern, balanced energy portfolio, pro-
principal methods: surface mining, underground viding a bridge to the future as an important low
mining, and in situ mining (extraction of ore from cost and secure energy solution to sustainability
a deposit using chemical solutions). After the ore challenges. A review of the Annual Energy Out-
is removed from the ground, it is crushed so that look 2010 report from the Energy Information
the valuable mineral in the ore can be separated Administration [5] indicates that world coal con-
from the waste material and concentrated by flo- sumption will increase by 56%, from 132 quadril-
tation (a process that separates finely ground min- lion BTU in 2007 to 206 quadrillion BTU in 2035.
erals from one another by causing some to float in Figure 1 presents the data for coal consumption
a froth and others to sink), gravity, magnetism, or [5] versus total energy consumption [17] between
other methods, usually at the mine site, to prepare 1980 and 2030. Coal’s share of electricity gener-
it for further stages of processing. The production ation stood at 48% in 2008. It fell to 30.4% in
of large amounts of waste material (often very 2016, causing coal production to take its biggest
acidic) and particulate emission have led to plunge on records going back to 1949.
major environmental and health concerns with The growth rate for coal consumption is
ore extraction and concentration. Additional pro- uneven, averaging 1.1% per year between 2007
cessing separates the desired metal from the min- and 2020 and 2.0% per year between 2020 and
eral concentrate. 2035. The slower growth rate for the earlier period
The closure of a mine refers to the cessation of is largely from a decline in coal consumption in
mining at a site. Planning for closure is often 2009 during the global economic recession. With
required to be ongoing throughout the life cycle of economic recovery, world coal consumption
the mine and not left to be addressed at the end of rebounded, and it is expected to return to its
operations. It involves completing a reclamation 2008 level by 2013. The US energy demand is
plan and ensuring the safety of areas affected by expected to grow at an average annual rate of
the operation, for instance, sealing the entrance to an 1.1% for the next 25 years. Among the largest
abandoned mine. The Surface Mining and Control consumers of energy are China and India and
Act of 1977 states that reclamation must “restore the together they accounted for about 10% of the
land affected to a condition capable of supporting world’s total energy consumption in 1990 and
the uses which it was capable of supporting prior to 20% in 2007. In a separate report by Waddell
338 Coal and Other Mining Operations: Role of Sustainability

700

600
Energy Consumption (Quadrillion Btu)

500 World Energy

Coal

400

300

200

100

0
80

85

90

95

00

06

10

15

20

25

30
19

19

19

19

20

20

20

20

20

20

20
Year

Coal and Other Mining Operations: Role of Sustainability, Fig. 1 Global Energy Consumption (1980–2030) and Coal
Consumption in US (2008–2018)

and Pruitt [32], it is projected that coal utilization resources, as well as a balanced consideration of
will increase worldwide by 44% by 2025. ecology, economy, and social justice.
The coal extraction in coal basins may last for The process of coal excavation consists of sev-
decades and not centuries, as the amounts of coal eral cycles: the cycle of excavation, disposal, and
reserves are limited. Sustainable management of recultivation. The first step is reconnaissance of
natural resources, such as coal, requires the main- the terrain that has been identified for mining.
tenance, rational and enhanced use of natural Preparation of the terrain may necessitate
Coal and Other Mining Operations: Role of Sustainability 339

demolition of constructions, felling of trees, relo- the area population not only due to the increased
cation of watercourses, roads, etc. This is usually number of mine workers but also due to the devel-
followed by excavation of the overburden and opment of related industries. Such development
then transporting and disposing the material will often result in disappearance of the arable
using specialized equipment (mining machinery). upper seam due to the building up in infrastructure
In present days, bucket wheel excavators, con- facilities (roads, railroad tracks, water-ways, indus-
veyor belts, and stackers are mainly used. When trial areas, etc.) and of the change in the purpose of
the coal seam appears, such a seam is excavated the soil in the vicinity of the mine.
and transported to the crushing plant. Mining can cause physical disturbances to the
Some of the important characteristics of coal landscape, creating eyesores such as waste-rock
excavation are listed below: piles and open pits. Such disturbances may con-
tribute to the decline of wildlife and plant species
• Notwithstanding the occupancy of large areas in an area. In addition, it is possible that many of
and the change in the purpose of the land, end the premining surface features cannot be replaced
of mining will result in bringing back large after mining ceases. Mine subsidence (ground
portions of the used areas to the previous movements of the Earth’s surface due to the col-
state. It means that though certain areas are lapse of overlying strata into voids created by
temporarily occupied by mining activities, underground mining) can cause damage to build-
for example for a period of ten or more ings and roads.
years, they can be reused as before following Open-cast coal mining leads to the disturbance
this period. of geological layers where different sedimentary
• Area occupied by the external dump, although products are being mixed, which means that a
it is generally higher than the terrain before completely new, anthropogenic land, i.e., substra-
mining activities, can be brought back to the tum, is created, without any resemblance to the
state so that it can be reused. original land, which is now called a deposol. This
• At the end of mining activities, the mine is an example of visual pollution where the char-
becomes the owner of large areas of agricul- acteristics of the landscapes’ appearance have
tural land and forests. The owner of an agricul- changed due to mining. The open-cast coal
tural land and forest is obliged to work the soil exploitation results in morphological modifica-
and control forest husbandry. tions of the terrain, such as the creation of large-
scale depressions and the formation of outside
overburden dumps.
Pollution from Coal Mining
The process of coal excavation causes deterioration
of the original morphological and pedological struc- Pollution of Air
ture of the terrain and soil, and results in the release Coal mines emit constituents that cause or con-
of harmful substances and/or mineral dust into the tribute significantly to air pollution and that may
air. Such a release is primarily related to the deteri- reasonably be anticipated to endanger public
oration of the upper seam structure during mining health and welfare. In each stage of mining,
operations. The negative impact may occur as a from exploration to ore recovery to downstream
result of the excavation of the upper seam and its processing, there is the potential for air quality
inadequate disposal, and of mixing the upper seam impacts due to emissions of particulates (dust,
with the lower one, as well as other barren materials. diesel, and silica). The monochromatic appear-
The impact of lignite excavation also represents a ance of the mine areas is due to generation of
potential contamination of the upper seam due to the large quantities of fugitive dusts during mining
precipitation of dust from the air. operations. Coal-mining areas are black, baux-
In addition to the aforementioned impacts, min- ite and iron-ore rich regions are red, while lime-
ing in a particular area will result in an increase in stone gives a chalky white hue. Dust results
340 Coal and Other Mining Operations: Role of Sustainability

from blasting, handling, processing, or trans- by-product of coalification, or the process by

porting of soil and rock or can arise from bare which organic materials convert into coal. It is
or poorly vegetated areas in combination with stored throughout the surrounding rock strata
air movements. It is one of the most visible, in varying sized pockets and, due to the greater
invasive and potentially irritating impacts of overburden pressures, often increases in con-
mining, and its visibility often raises concerns. centration the deeper the coal seam. Because
Many dusts contain metals which are potentially methane can create hazardous working condi-
hazardous and are known to cause certain dis- tions for miners, it must be removed from
eases. It has the potential to severely affect flora underground mines. While methane escapes
and fauna near the mine and to impact the health during the processing, transport, and storage
of mine workers and local residents. Dust also of coal, 90% of the emissions come from the
affects the agricultural productivity of the area. actual coal mining process from all three cate-
The level of dust generated, its behavior (travel gories of coal mines: surface mines, under-
distance), and types of health and environmental ground mines, and abandoned mines.
risks depend on many factors including mine In 2004, methane accounted for 14.3% of
type, local climate, topography, working the total anthropogenic greenhouse gas emis-
methods, types of equipment used, the mineral- sion load. Since preindustrial times, methane
ogy and metallurgical characteristics of some has contributed to 22.9% of the greenhouse
ores, and the land use of the area around the gas load in the Earth’s atmosphere. Methane
mine. is more abundant in the Earth’s atmosphere
now than it has been at any time during the
• EIA reported that estimated recoverable past 400,000 years, and the average atmo-
reserves of coal of the USA stand at 275 billion spheric concentration of methane has
tons, an amount that is greater than any other increased 150% since 1750 due to human
nation in the world. All of the energy growth activities [29]. The US EPA [31] has con-
forecasts have major carbon dioxide (CO2) cluded that recovering methane from coal
emission consequences. The World Energy mines would significantly reduce the amount
Outlook (WEO) 2007 Reference Scenario pro- of greenhouse gases emitted into the atmo-
jects CO2 emissions to increase from 26.6 sphere, because every ton of methane recov-
gigatonnes (Gt) in 2005 to 34.1 Gt in 2015 ered is equal to approximately more than 20 t
and 41.9 Gt in 2030 [10]. In the Alternative of carbon dioxide emissions. While methane
Policy Scenario [33], CO2 emissions rise to is a danger in coal mines, it can, if captured,
31.9 Gt in 2015 and to 33.9 Gt in 2030. EIA be a valuable commodity: natural gas. To
of the United States Department of Energy date, underground mines have the most
(DOE) in its International Energy Outlook potential to generate a profitable amount of
(IEO) 2007 forecasted that CO2 emissions in methane. Mines that have adopted a “methane
the Reference Case will increase from 26.9 Gt to markets” program as promoted by the US
in 2004 to 33.9 Gt in 2015 and to 42.9 Gt in EPA have been successful at methane capture
2030. In the past, global CO2 emissions rose and sale, because research has been develop-
from 23.5 Gt in 2000 to 27.1 Gt in 2005 [11], ing for over three decades to create the best
which is an increase of over 15%. Based on technologies.
this, one can say that the forecasts made by • Particulate matters (PM), including total
IEA and IEO are optimistic. suspended particulates, PM10 (particle size
• Methane is the second most emitted green- less than 10 mm), and PM2.5 (particle size less
house gas after CO2, and is more than than 2.5 mm), are released during coal mining
20 times more potent than CO2 in terms of its operations. Coal-mining activities that can lead
heat-trapping capabilities. Methane is a to the release of particulate matter are blasting,
Coal and Other Mining Operations: Role of Sustainability 341

truck loading, bulldozing, dragline operation, (CWP) is characterized by development of

vehicle traffic, grading, and storage piles. coal macules, a focal collection of coal dust
• Volatile organic compounds (VOCs), includ- particles with a little reticulin and collagen
ing nonmethane organic compounds, are often accumulation. These lesions may be visible as
vented along with methane, from coal mining small opacities (less than 1 cm in diameter) on
operations. VOCs are precursors to ground- X-rays. Complicated CWP is characterized by
level ozone, a criteria air pollutant, and are lesions consisting of a mass of rubbery well-
regulated under the Clean Air Act. defined black tissue that is often adherent to the
• Nitrogen oxides (NOx) are a group of gases chest wall. This is associated with decrements
that are known to be ground-level ozone and in ventilatory capacity, low diffusing capacity,
PM2.5 precursors and that include nitrogen abnormalities of gas exchange, low arterial
dioxide (NO2), a criteria pollutant. Sources of oxygen tension, pulmonary hypertension, and
NOx at coal mines include fugitive emissions premature death. The disease may progress
from overburden and coal blasting events, tail- after the cessation of exposure.
pipe emissions from mining equipment, point Effects on Plant: Plants exposed to toxic dust
source emissions from stationary engines, exhibit lesser growth, become more prone to
coal-fired hot water generators, and natural- disease attack and rodent attack, their stomata
gas fired heaters. and other holes get chocked, and they also gasp
• For some pollutants, such as sulfur dioxide, the for oxygen. Fruits, vegetables, and cereals
minerals processing industry is the largest from these plants contain toxins.
source of emissions. Effects on Infrastructure: Onset of biochemical
• Certain kinds of ore, for example uranium, are reaction when constituents present in the coal
radioactive, and pose health problems to dust starts reacting with the constituents of the
workers and adjacent communities. Radioac- host surface materials resulting in increase in
tivity follows with whatever radioactive mate- the rate of weathering/weakening.
rials escape into the atmosphere or water and
accompanies this material to its final destina- Pollution of Water
tion. Besides targeted minerals that are radio- The potential impacts of mining on the water
active, the host rock may be radioactive and Younger and Wolkersdorfer [34] are:
pose problems for workers. Radioactivity is
also present in a gas, radon, which can cause • Disruption of hydrological pathways
problems for workers in underground mines, if • Seepage of contaminated leachate into aquifers
it is present. • Depression of the water table around the
dewatered zone
Impacts of Dust • Disposal of saline mine water into rivers
The impacts of dust, especially from coal mine,
area are included below: The impacts of mining arising from the disrup-
tion of hydrological pathways, seepage of con-
Effects on Animals: Coal dust is a tumorigenic taminated leachate into aquifers, and depression
agent in experimental animals. Coal dusts are of the water table tend to be relatively localized
associated with lymphomas and, at the higher and limited compared to disposal of mine water.
dose, adrenal cortex tumors in rats exposed to Disposal of mine water is a worldwide problem,
either 6.6 or 14.9 mg/m3 for 6 h/day intermit- occurring wherever operating mines, both under-
tently for 86 weeks. ground and opencast, are found Pulles [21]. The
Effects on Humans: Coal dust causes pneumoco- quality of the mine water depends largely on the
niosis, bronchitis, and emphysema in exposed chemical properties of the geological materials
community. Coal workers’ pneumoconiosis that come in contact with the water.
342 Coal and Other Mining Operations: Role of Sustainability

The mining industry is a major producer of 2Feþ3 þ 6H2 O $ 2FeðOHÞ3 ðsÞ þ 6Hþ
acidic sulfur-rich water that typically pose a risk
to the environment. Mining increases the exposed 14Feþ3 þ FeS2 ðsÞ þ 8H2 O !
surface area of sulfur-bearing rocks allowing for
2SO4 2 þ 15Feþ2 þ 16Hþ
excess acid generation beyond natural buffering
capabilities found in host rock and water resources.
Pyrite (FeS2) is the major sulfur mineral in coal. In undisturbed natural systems, this oxidation
When coal is mined, fresh sulfur-bearing minerals process occurs at slow rates over geologic time
in the coal and rocks are exposed to air and water. scale. Iron and sulfur-oxidizing bacteria are also
Problematic mine drainage is given many names known to catalyze these reactions at low pH, thereby
including acid rock drainage (ARD), acid mine increasing the rate of reaction by several orders of
drainage (AMD), and mining-influenced water magnitude [20]. Bacteria such as Thiobacillus ferro-
(MIW), among other names. AMD forms during oxidans and Ferroplasma acidarmanus are known
metal or coal mining when sulfur-bearing minerals to specifically accelerate such reactions.
are exposed to water and air, forming sulfuric acid. A few management options for saline mine
Heavy metals leached from rocks can combine water, as summarized by Annandale et al. [1], are:
with the acid and dissolve, creating highly toxic
runoff. The general distinction between ARD and (i) Pollution prevention at source
AMD depends on whether drainage quality has (ii) Reuse and recycling of water to minimize the
been degraded by mining or is of poor quality volume of polluted water
due, in part, to natural causes. MIW is the term (iii) Treatment of effluents, if the problem cannot
used to refer all mining-related water because be solved through prevention, reuse, and
acidic, neutral, and alkaline water can all transport recycling
metals and other contaminants. (iv) Discharge of treated effluent
The origin of acidic metal-rich mine drainage (v) Utilization of gypsiferous mine water for
water is the accelerated oxidation of iron pyrite irrigation
(FeS2) and other sulfidic minerals. The reaction of
pyrite with oxygen and water produces a solution The term “gypsiferous” refers to rocks and
of ferrous sulfate and sulfuric acid. Ferrous iron soils containing more than 2% gypsum, i.e., cal-
can further be oxidized producing additional acid- cium sulfate (CaSO4 • 2H2O). There are concerns
ity. The following reaction shows the reaction of regarding potential use of mine water for agricul-
pyrite with oxygen and water to produce hydro- tural crops [6], due to the amount of salt that
gen ions, sulfate ions, and soluble metal ions. would leach and potentially contaminate ground-
water. The next few sections highlight successes
and limitations with respect to environmental
2FeS2 ðsÞ þ 7O2 þ 2H2 O ! 2Feþ2 þ 4SO4 2 þ 4Hþ
impact and sustainable use of such waters for
irrigation of agricultural crops.
Further oxidation of Fe+2 (ferrous) to Fe+3
(ferric) occurs when sufficient oxygen is
dissolved in the water or when the water is
exposed to sufficient atmospheric oxygen. Restoration Process

2Feþ2 þ ½ O2 þ 2Hþ ! 2Feþ3 þ H2 O It is the obligation of the mining company to

restore the working cavity in the best possible
Fe+3 can either precipitate as Fe(OH)3, a red- way, through overburden disposal activities and
orange precipitate seen in waters affected by technical reclamation, to the existing natural envi-
ARD, or it can react directly with pyrite to pro- ronment in terms of functionality and aesthetics.
duce more Fe+2 and acidity as per the following The following measures can be applied for the
reactions. recultivation of deteriorated surfaces:
Coal and Other Mining Operations: Role of Sustainability 343

• Technical measures – to contribute to the environmental performance, and the duty of care
improvement of resistant and deformable char- required of mine operators by government and the
acteristics of the dump and directly influence community. Particulate emission management
the enhancement of the erosion stability of system can result in cost savings, increased
slopes. profits, and improved government and commu-
• Biotechnical measures – to contribute to faster nity relations, as well as easier access to resources
achievement and maintenance of the dump and financial support in the future.
stability along with the technical measures. There are a number of systems to address
• Biological measures – that imply the imple- emissions of harmful air pollutants from coal
mentation of agricultural and forest improve- mines. It is technologically feasible to capture or
ments that contribute to the stability and flare methane from coal mines, instead of releas-
maintenance of reclaimed areas, but they are ing methane directly into the atmosphere. Many
much more significant from the aspect of area mines have already taken steps toward capturing
revitalization and establishing natural bioceno- methane emissions for economic reasons.
ses (a group of interacting organisms that live Twenty-three mines in Alabama, Colorado, Penn-
in a particular habitat and form a self- sylvania, Virginia, and West Virginia have meth-
regulating ecological community). ane drainage facility and they recover methane
within the range of 3–88% efficiency [7]. Of
The process of reverting an industrial land these 23 mines, 12 sell recovered methane for
back to an agricultural terrain is a very slow pro- natural gas energy use and two mines use the
cess and good planning steps are required. methane to heat mine ventilation air and to gener-
Depending on the size of the coal deposit, the ate onsite power. Flaring is an option if capture is
process of turning an agricultural, forest, or not technologically feasible. Although flaring
urban environment into an industrial (mining) occurs at a small number of mines in the USA,
one, and then, by the recultivation process, back there is a long and safe history of flaring at work-
again into the agricultural or forest one may ing coal mines in the United Kingdom and
require long time, even many decades. Australia. US EPA estimates that nearly 50% of
There are ongoing arguments in favor of pre- all of US coal mine methane emissions, or more
venting further expansion of the utilization of fossil than 1.25 million tons of methane, can be reduced
fuels, such as coal. This is one reason, among at a zero net cost, while nearly 90% can be
others, why it is necessary to perform the reduced at a cost of less than $15/ton. However,
recultivation of a dump even though it has a rela- the benefit of reducing methane could be as much
tively low impact on the environment. As we are as $240/ton of methane reduced. Any efforts to
talking about fertile alluvial lands, it is necessary to address methane will most likely also address
preserve the fertile solum (a set of related soil VOC emissions due to the fact that VOCs and
horizons that share the same cycle of pedogenic methane are often released together from coal
processes) through selective excavation in order to mines.
bring the soil back to agricultural production or Examples of particulate matter emission con-
other use. The forestation of soil and terrains dete- trol measures currently in use include, but are not
riorated by open-cast lignite mining prevents fur- limited to: (a) storing coal in enclosed coal silos or
ther deterioration processes, contributes to the barns; (b) paving coal mine access roads;
maintenance of the ecological balance in nature, (c) watering or treating with dust suppressant
and enhances the absorption of CO2 from the air any unpaved roads; (d) enclosing conveyor trans-
and increases the content of the oxygen therein. fer points; (e) using dust collection baghouses or
The control of particulate emission is a funda- other controls to reduce emissions from transfer
mental part of a mine environmental management points and crushers within processing plants;
plan because of the increasing public awareness of (f) fitting out-of-pit conveyors with hoods or oth-
human health issues and expectations of erwise containing emissions; (g) fitting out-of-pit
344 Coal and Other Mining Operations: Role of Sustainability

dump hoppers with water sprays, a baghouse, or The rising global demand for resources acceler-
other controls; (h) treating haul roads with dust ates this trend. After extraction, the subsequent
control chemicals or water; (i) watering short-term stages in the raw materials’ life cycle entail further
haul roads; and (j) regularly maintaining haul environmental pollution. Comparative analyses of
roads to reduce dust reentrainment. industrial sectors show that the highly resource-
With regards to NOx, mines have reduced intensive industries are associated with above-
emissions by as much as 75% through the use of average levels of emissions of greenhouse gases
borehole liners and changing their blasting agent and other pollutants.
blends. Other mitigation measures that may be Climate change and the overexploitation of
effective at reducing NOx include reduced blast natural resources are two sides of the same coin.
sizes, changed composition of explosive agents, Climate change will impact water supplies, exac-
and changed placement of blasting agents. Reduc- erbating existing pressures on water resources
tion of impacts due to radioactivity consists of caused by population and economic growth.
ventilation for underground mines, dust control, Given the combination of these stressors, the sus-
standard containment measures for solid waste, tainability of water resources in future decades is a
and standard precautions for discharges from concern in many parts of the world. The sustain-
processing. ability of water resources is defined as the main-
tenance of natural water resources in adequate
quantity and with suitable quality for human use
Sustainability and Mining and for aquatic ecosystems. Human needs for
resources, like, water, land, continue to grow
The Earth’s resources have played a vital role for with increasing population, primarily for direct
human communities. Nowadays, the management consumption, but also secondarily for energy pro-
of our natural resources has become an urgent duction, and agricultural and mining activities.
issue at both national and international level. We can enjoy the privilege of mining in the best
The extraction and use of natural resources are possible way, but we have the responsibility of
also causing environmental problems, which looking out for the future generations, preserving
require urgent solutions. For mining operations, some of the natural resources by utilizing reserves
resource extraction not only has a massive impact in compliance with the principles of the sustainable
on ecosystems it also releases pollutants development. We also need to reclaim and to
contained in the rock, and consumes large revive deteriorated surfaces resulting from mining.
amounts of water and energy. The transportation The nine sustainable development challenges
of resources from remote areas requires an eco- that the mining industries face are:
logically intensive transport infrastructure.
Due to the progressive exploitation of mineral • Ensuring long-term viability of the mining
deposits and the availability of new technologies, industry
deposits with lower ore content are now being • Controlling, using, and managing the land
mined. This means that an increasing amount of • Using minerals to assist with economic
ore and other nonusable material has to be extra- development
cted to produce the same amount of metal. For • Making a positive impact on local
example, towards the beginning of the twentieth communities
century, copper ore mined by US mineral industry • Managing the environmental impacts of mines
consisted of about 2.5% of usable metal by • Maximizing the use of minerals so as to reduce
weight; today the proportion has dropped to waste and inefficiency
0.51% [8]. This activity impacts even more • Giving stakeholders access to information to
severely on ecosystems and water resources and build trust and cooperation
increases the volume of mining waste, resulting in • Managing the relationship between large cor-
even more radical changes to entire landscapes. porates and small-scale mining companies
Coal and Other Mining Operations: Role of Sustainability 345

• Sector governance: clearly defining the roles, (b) Information gathering (to emphasize the
responsibilities, and instruments for change importance of collecting and organizing rele-
expected of all stakeholders vant information into a suitable framework)
(c) Identifying risks (systematically reviewing
Mining, Minerals and Sustainable Development impact areas under consideration, and identi-
(MMSD) North America has developed an fying opportunities or threats)
approach to assess how a mining/mineral project (d) Analyzing and evaluating the risks
or operation contributes to sustainability. The (qualitative scales of likelihood and conse-
assessment considered tracking the record of min- quence can be assigned to identified opportu-
ing and minerals in the past and its current contri- nity or threat to create overall risk rating and
bution to and detraction from economic prosperity, prioritized list)
human well-being, ecosystem health, and account- (e) Treating risks (control measures to address
able decision-making. A set of seven questions opportunities or threats considered high
were developed as a means of assessing whether priority)
the net contribution to sustainability over the long (f) Reporting and reviewing to represent a broad,
term of a mining/mineral project or operation will scanning exercise that can be picked up by
be positive or negative, and a way of discovering existing business planning and monitoring
how current activities can be improved and aligned processes. It also provides an opportunity for
with the emerging concept of sustainability subsequent evaluation of relevant metrics for
[12]. The seven-part numbering used by MMSD the issues identified by the process, allowing
(see Table 1) has been intended as an aid to com- operations to measure both impacts and pro-
municate and not to imply a particular sequencing gress towards agreed objectives.
of steps or prioritization of topics. These questions
follows a hierarchy of objectives, indicators, and A responsible mining sector must maximize
specific metrics. In this way a single, initial moti- development benefits while actually improving envi-
vating question cascades into progressively more ronmental and social sustainability (see Fig. 2). This
detailed elements which can be tailored to the can be achieved through the following criteria:
project or operation being assessed for application
throughout its full life cycle. • Incorporating considerations of potential
The focus on applying the seven questions adverse environmental, health and, social
to sustainability approach is not so much on impacts with economic benefits over the min-
how mining can be sustainable (mining as a ing life cycle, including postdecommissioning
discrete activity cannot continue indefinitely) • Enhancing the participation and oversight of
but on how mining can contribute to stakeholders in mining developments, includ-
sustainability [9]. ing local and indigenous communities
Sustainability Opportunity and Threat Analysis • Ensuring economic gains are distributed in
is another simple operational tool addressing the ways that contribute to sustainability
social, economic, and environmental dimensions
of the issues under consideration, and it can be
applied to evaluate the viability of a mining oper- If a mine is to make a positive contribution to
ation and its ability to contribute to sustainable local and regional sustainability, a sufficient por-
development objectives. An impact-based analysis tion of its revenues must be used to ensure that
involves constructing inventory of sources of what is ultimately left behind compensates for
impacts and following key steps including: what is lost and, moreover, improves the sustain-
ability of livelihoods and ecosystems. In addition
(a) Scoping (addressing the reasons for the min- to focusing on actual mining activities and lega-
ing process and the environment, and agree- cies, sustainable mining would also ensure mini-
ing to the scope of the exercise) mal environmental impacts and maximum social
346 Coal and Other Mining Operations: Role of Sustainability

Coal and Other Mining Operations: Role of Sustainability, Table 1 Seven questions to sustainability – how to assess the
contribution of mining and minerals activities
Seven fundamental questions Framework for guiding a sustainability assessment
1. Engagement Processes of engagement are committed to, designed, and
Are engagement processes in place and working implemented so that they:
effectively? Ensure all affected communities of interest have the
opportunity to participate in the decisions that influence
their own future
Are understood, agreed upon by implicated
communities of interest, and consistent with the legal,
institutional, and cultural characteristics of the community
and country where the project or operation is located
2. People Project/operation lead directly or indirectly to
Will people’s well-being be maintained or improved? maintenance of people’s well-being:
During the life of the project/operation
Postclosure
3. Environment Project or operation lead directly or indirectly to the
Is the integrity of the environment assured over the long maintenance or strengthening of the integrity of
term? biophysical systems so that they can continue in
postclosure to provide the needed support for the well-
being of people and other life forms
4. Economy Assurance of the financial health of the project/company
Is the economic viability of the project or operation and contribution of the project or operation to the long-
assured, and will the economy of the community and term viability of the local, regional, and global economy
beyond be better off as a result? in ways that will help ensure sufficiency for all and
provide specific opportunities for the less advantaged
5. Traditional and Nonmarket Activities Project or operation contribute to the long-term viability
Are traditional and nonmarket activities in the community of traditional and nonmarket activities in the implicated
and surrounding area accounted for in a way that is community and region
acceptable to the local people?
6. Institutional arrangements and governance Institutional arrangements and systems of governance in
Are rules, incentives, programs, and capacities in place to place that can provide certainty and confidence that:
address project or operational consequences? Capacity of government, companies, communities, and
residents to address project or operation consequences is
in place or will be built
Capacity will continue to evolve and exist through the
full life-cycle including postclosure
7. Synthesis and continuous learning Placement of an overall evaluation and periodic
Does a full synthesis show that the net result will be reevaluation based on:
positive or negative in the long term, and will there be Consideration of all reasonable alternative
periodic reassessments? configurations at the project level
Consideration of all reasonable alternatives at the
overarching strategic level for supplying the commodity
and the services it provides for meeting society’s needs
Synthesis of all the factors raised in this list of
questions, leading to an overall judgment that the
contribution to people and ecosystems will be net positive
over the long term

benefits from the extraction, manufacturing, use, well-being, especially the well-being of people
and disposal of mineral products. To make a pos- now suffering material deprivation. The mining
itive contribution to sustainability at a broader developments must go beyond economic consid-
scale, mining products and revenue must be erations, as important as they are, to become more
used, directly or indirectly, to enhance lasting inclusive and holistic. Every stakeholder –
Coal and Other Mining Operations: Role of Sustainability 347

Health and Effective

Economic Engagement
Safety
Benefit

Continuous Factors Respect for

Learning and Affecting Indigenous
Adaptation Mining People
Sustainability

Resource
Use Environment
Efficiency Operation
Life Cycle

Coal and Other Mining Operations: Role of Sustainability, Fig. 2 Cycle of factors necessary for developing a
sustainable mining

including miners, local, state and federal govern- useful and reliable model. Since optimization of
ments, and community groups – must understand, mine design, and in particular coal mine design, to
accommodate, and respect other stakeholders. For address sustainability along with other parameters
the mining industry, it’s an opportunity not only to has not been widely practiced, identifying the
gain social license but to tackle some of the appropriate parameters for measurement and the
toughest challenges facing the industry and mathematical or logical relationships between
move toward being a more sustainable sector. these parameters is not a trivial task.
Most mine designs are based on traditional Another important constraint on mining oper-
mining engineering factors, such as the quality ations is the statutory and regulatory frameworks
of the commodity being mined, geology, topogra- within which they are required to operate. Many
phy, hydrology, land ownership, geography, infra- of these legal and policy structures tend to create
structure, etc. Environmental compliance and an adversarial approach by instituting a system
sustainability are to be considered in mine design where one or more parties must respond to
and operation as a modifying factor to those actions, proposals, decisions, etc., of another
designs. While practices may become more party. The participation is often late in the design
responsive to sustainability, mine design con- process, or after design has been completed, and
tinues to be governed by established mining engi- thus any change to mine or reclamation design
neering approaches. necessary as a result of public or regulatory
A review of the available literature on engi- agency input creates a retrofitted design, which
neering optimization does not reveal any focus on cannot possibly be optimal. To optimize the
mine design, environmental protection associated design of mining and reclamation operations, tra-
with mines, and sustainability. Mathematical ditional mining engineering considerations and
multicriteria optimization approaches, however, environmental and sustainability goals must be
have been used in resource management. Mine accounted for simultaneously.
design optimization would need to consider all Several technologies are discussed below
constraints, system parameters and characteris- that are considered to promote sustainable
tics, and desired outcomes in order to build a development.
348 Coal and Other Mining Operations: Role of Sustainability

Irrigation with Lime-Treated Acid Mine much of an opportunity for gypsum precipitation,
Drainage (AMD) which is able to drastically reduce the salt load of
the receiving water in the case of calcium and
Higher crop yields were obtained under sprinkler sulfate-rich mine water. The application of cal-
irrigation with treated mine water compared to cium nitrate as a nitrogen source to the crop adds
dryland production, without any foliar injury to calcium to the soil and removes some sulfate from
the crop. Possible nutritional problems, for exam- the water system by enhancing gypsum precipita-
ple deficiencies of potassium, magnesium, and tion. Measurement of the hydraulic conductivity
nitrate, occurring due to calcium and sulfate dom- of the soil is recommended to monitor the effect of
inating the system, can be solved through fertili- the water on the infiltration rate of the soil, as high
zation. Such soils need to be managed and sodium levels are likely to cause deflocculation or
fertilized differently to those on which crops are dispersion of clay particles.
produced under normal farming conditions.
Sugar-beans, wheat, maize, and potatoes were
very successfully produced under irrigation with Application of Phytotechnology
calcium sulfate- and magnesium sulfate-rich mine
water. In an experimental setting, soil salinity Phytotechnology can be applied to address issues
increased with time, but the values of soil satu- related to stabilization of tailings and hydraulic
rated electrical conductivity stabilized at rela- control for drainage so that human and ecological
tively low levels, due to gypsum precipitation exposures to contaminants associated with mining
[1]. Measurements taken between 1997 and solid wastes and mine impacted waters are low.
2007 showed that soil salinity increased from a Implementation of phytotechnology is a common
low base and oscillated around 250 mS m1. component of mining reclamation and restoration
Land preparation and fertilization management projects by the establishment of a plant cover as a
are, however, critical for successful crop produc- final remedy. In certain cases, application of
tion, especially on rehabilitated soil. During short phytotechnology can be used for removal of
to medium term (up to 8 years) irrigation with metals from contaminated media. Establishing
gypsiferous mine water, negligible impact was phytotechnology requires careful plant species
noticed by Annandale et al. [1] on groundwater selection and soil amendments that equates to an
quality. They operated the system with flexibility initial investment; however, these systems, once
and managed with the multiple objectives, like, established, can be maintained with minimal
maximum crop production, water use, job crea- effort.
tion, economic return or maximum gypsum pre- There are six basic phytoremediation mecha-
cipitation, and minimum salt leaching. nisms that can be used to clean up contaminated
Gypsum precipitation was also shown to be sites: phytosequestration, rhizodegradation,
taking place in the soil. The presence of gypsum phytohydraulics, phytoextraction, phytodegra-
did not create any physical and/or chemical prop- dation, and phytovolatilization (Table 2).
erty changes that could adversely affect crop pro- The particular phytotechnology mechanism
duction and soil management. Crop production used to address contaminants depends not only
under irrigation with coal-mine water rich in cal- on the type of contaminant and the media affected
cium (Ca), magnesium (Mg), and sulfate (SO4) is, but also on the cleanup goals. Typical goals include
therefore, feasible, and sustainable if properly containment through stabilization or sequestration,
managed. remediation through assimilation, reduction,
Pasture production with sodium sulfate-rich detoxification, degradation, metabolization or min-
mine effluent is also feasible, but requires a well- eralization, or both. Applying phytotechnology to
drained profile and a large leaching fraction to impacted sites entails selecting, designing,
prevent unsustainable build up of salt in the soil. installing, operating, maintaining, and monitoring
Unfortunately, this type of water does not present planted systems that use the various mechanisms
Coal and Other Mining Operations: Role of Sustainability 349

Coal and Other Mining Operations: Role of Sustainability, Table 2 Phytotechnology mechanisms to sequester constit-
uents of interest [13]
Mechanism Description Cleanup goal
Phytosequestration Ability of plants to sequester selected contaminants in the rhizosphere Containment
zone through exudation of phytochemicals and on the root through
transport proteins and cellular processes
Rhizodegradation Exuded phytochemicals can enhance microbial biodegradation of Remediation by
contaminants in the rhizosphere destruction
Phytohydraulics Ability of plants to capture and evaporate water off the plant and take Containment by
up and transpire water through the plant controlling hydrology
Phytoextraction Ability of plants to take up contaminants into the plant with the Remediation by
transpiration stream removal of plants
Phytodegradation Ability of plants to take up and break down contaminants in the Remediation by
transpiration stream through internal enzymatic activity and destruction
photosynthetic oxidation/reduction
Phytovolatilization Ability of plants to take up, translocate, and subsequently transpire Remediation by
volatile contaminants in the transpiration stream removal through
plants

mentioned above. The goal of the system can be (h) Favorable public perception
broadly based on the remedial objectives of con- (i) Improves aesthetics, including reduced noise
tainment, remediation, or both. Furthermore, the (j) Applicable to remote locations, potentially
target media can be soil/sediment, surface water, without utility access
or groundwater, and these can be either clean or (k) Provides restoration and land reclamation
impacted. In some cases, groundwater trans- during cleanup and upon completion
itioning to surface water can be addressed as a (l) Can be cost-competitive
riparian situation where target media are combined.
One of the main advantages of phytotechnology The benefits of using the phytotechnology-
is that this technology can be applied to a variety of based techniques are the relative lower costs,
metals in mining sites and to impacted soil/sedi- labor requirements, and safer operations com-
ment, surface water, and groundwater. In addition, pared to the more intensive and invasive conven-
it can be applied to various combinations of chem- tional techniques. Phytotechnology generally
ical types and impacted media simultaneously. provides long-term remedial solutions. Planta-
Additional advantages are: tions may require irrigation, fertilization, weed
control (mowing, mulching, or spraying), and
(a) Considered a green and sustainable pest control. Establishment of phytotechnology
technology systems include various expenditures, such as
(b) Does not require supplemental energy, earthwork, labor, planting stock, planting
although monitoring equipment may use method, field equipment, heavy machinery
solar power (typically farming or forestry equipment), soil
(c) Can improve the air quality and sequester amendments, permits, water control infrastruc-
greenhouse gases ture, utility infrastructure, fencing, security,
(d) Minimal air emissions, water discharge, and etc. About 10–15% of the initial capital costs
secondary waste generation can be added as a contingency for
(e) Lower maintenance, resilient, and self- replanting [13].
repairing Phytotechnology are appropriate only under spe-
(f) Inherently controls erosion, runoff, infiltra- cific conditions. The major limitations are depth,
tion, and dust emissions area, and time. The physical constraints of depth
(g) Passive and in situ and area depend on the plant species suitable to the
350 Coal and Other Mining Operations: Role of Sustainability

site (i.e., root penetration) as well as the site layout insoluble (such as pyrite and arsenopyrite). Two
and soil characteristics. Phytotechnology typically routes (the “thiosulfate” and “polythionate”
require larger tracts of land than many alternatives. mechanisms) have been proposed for the biolog-
Time can be a constraint since phytotechnology ical oxidation of these sulfide minerals [23].
generally take longer than many other alternatives Acid-soluble sulfides are readily degraded by
and are susceptible to seasonal and diurnal changes. sulfur-oxidizing acidophiles. The mineral is first
Additional limitations include plant’s tolerance to subjected to proton-mediated dissolution, forming
specific constituents of concern or site conditions, the free metal and hydrogen sulfide:
availability of water as irrigation source, climate
(challenging for plant establishment in areas with MS þ 2 Hþ ! M2þ þ H2 S
short growing season or in arid environments), and
pests, infestations, or other nuisances. The hydrogen sulfide so formed is microbially
oxidized to sulfuric acid, allowing the process to
continue:
Biomining and Bioprocessing H2 S þ 2 O2 ! H2 SO4

Mining industries are increasingly aware of the Acid-soluble sulfides may also be attacked by
potential of microbiological approaches for recov- ferric iron, producing ferrous iron and polysulfide:
ering base and precious metals from low-grade
ores, and for remediating acidic, metal-rich waste- MS þ Fe3þ þ Hþ ! M2þ þ 0:5 H2 Sn þ Fe2þ :
waters that drain from both operating and aban-
doned mine sites. Biological systems offer a Polysulfide may be further oxidized by ferric
number of environmental and economical advan- iron to produce elemental sulfur, which, in turn, is
tages over conventional approaches, such as pyro- oxidized to sulfuric acid, and will further acceler-
metallurgy, though the microbial application is ate mineral dissolution via proton attack:
not appropriate in every situation. Biomining
(metal extraction) and bioprocessing are currently 0:5 H2 Sn þ Fe3þ ! 0:125 S8 þ Fe2þ þ Hþ
utilized in full-scale operations to process low-

grade deposits and reprocessing earlier metal- 0:125 S8 þ 1:5 O2 þ H2 O ! 2Hþ þ SO4 2
containing wastes. This usually results in the pro-
duction of less chemically active tailings, lower Sulfides that are resistant to proton attack are
energy inputs, and other environmental benefits oxidized by ferric iron, producing thiosulfate as an
(zero production of noxious gases). Recently, initial by-product:
there have been major advances in the field of

microbiology, which will allow greater control FeS2 þ 6Fe3þ þ 3 H2 O ! 7Fe2þ þ S2 O3 2 þ 6 Hþ
of bioleaching operations, resulting in greater effi-
ciencies and faster rates of metal extraction. Min- Accelerated oxidation can result in low pH,
eral processing using microorganisms have been high concentrations of dissolved metals, and,
exploited for extracting gold, copper, uranium, in some cases, elevated temperatures. These con-
and cobalt. Engineering systems ranging from ditions limit the diversity of life forms that occurs
crude heap leaching systems to temperature- in commercial bioleaching operations. Single-
controlled bioreactors have been used, depending celled organisms live only in extremely acidic
on the nature of the ore and the value of the liquors (pH < 1–4) and are obligate acidophiles,
metal product. A typical example of mineral some are thermophilic (to varying degrees), and
bioprocessing mechanism is discussed below. some can fix carbon dioxide. The primary micro-
Sulfide minerals may be divided into acid sol- organisms involved in mineral oxidation are
uble (such as zinc sulfide or sphalerite) and acid- those that catalyze the oxidation of ferrous iron
Coal and Other Mining Operations: Role of Sustainability 351

and/or reduced sulfur while others contribute 6HCO3  ðaqÞ þ 6H2 O ! C6 H12 O6 þ 6O2 þ 6OH
to the process indirectly by, for example, re-
moving materials that accumulate during ore The reduction of soluble iron (ferric iron) does
dissolution [14]. not decrease solution acidity, however, the reduc-
Engineering approaches used in biomining are: tion of solid phase (crystalline and amorphous)
(i) irrigation-based principles (dump- and heap- ferric iron compounds does.
leaching, and in situ leaching) and (ii) stirred tank
processes [22]. The dump leaching involves gath- FeðOHÞ3 þ 3Hþ þ e ! Fe2þ þ 3H2 O,
ering low-grade copper-containing ore of large
rock/boulder size into vast mounds or dumps where e represents an electron donor, which is
and irrigating these with dilute sulfuric acid to generally an organic substrate.
encourage the growth and activities of mineral- Bacteria that catalyze the dissimilatory reduc-
oxidizing acidophiles, primarily iron-oxidizing tion of sulfate to sulfide generate alkalinity by
mesophiles. Copper can be precipitated from the transforming a strong acid (sulfuric) into a rela-
metal-rich streams draining from the dumps by tively weak acid (hydrogen sulfide).
displacement with iron. Other developments on
the engineering and hydrometallurgical aspects of 
SO4 2 þ 2CH2 O þ 2Hþ ! H2 S þ 2H2 CO3
biomining have involved the use of thin layer
heaps of refractory sulfidic ores (mostly copper, Besides the ameliorative effect on AMD
but also gold-bearing material) stacked onto brought about by the resulting increase in pH,
water-proof membranes, and recovery of solubi- the reduction of sulfate is an important mecha-
lized copper using solvent extraction coupled with nism for removing toxic metals from AMD, since
electrowinning. In situ bioleaching has been these metals (e.g., zinc, copper, and cadmium)
developed to scavenge for uranium and copper form highly insoluble sulfides.
in otherwise worked out mines.
Zn2þ þ H2 S ! ZnS þ 2Hþ
Bioremediation The bioremediation technologies, like
constructed wetlands and compost bioreactors for
Water-draining active and abandoned mines and
AMD, are passive systems. In case of the constructed
mine wastes are acidic and often contain elevated
wetlands, the solid-phase products of water treat-
concentrations of metals (iron, aluminum and
ment are contained within the wetland sediments.
manganese, and others) and metalloids (arsenic,
The key major advantages of these passive bioreme-
chromium, selenium, and uranium). The basis of
diation systems are their relatively low maintenance
bioremediation of AMD derives from the abilities
costs. The limitations of these technologies are:
of some microorganisms to generate alkalinity
and immobilize metals, thereby essentially revers- (a) They are often relatively expensive to install.
ing the reactions responsible for AMD. (b) May require more land area than is available
Microbiological processes that generate or suitable.
net alkalinity are mostly reductive processes, and (c) Their performance is less predictable than
include denitrification, methanogenesis, sulfate chemical treatment systems.
reduction, and iron and manganese reduction. (d) The long-term fate and stability (in the case of
Ammonification (the production of ammonium compost bioreactors) of the deposits that
ion from nitrogen-containing organic compounds) accumulate within them is uncertain [15].
is also an alkali-generating process. Photosyn-
thetic microorganisms, by consuming a weak Aerobic wetlands are generally more effective
base (bicarbonate) and producing a strong base to treat mine waters that are net alkaline. The
(hydroxyl ions), also generate net alkalinity: reaction involves oxidation of ferrous iron, and
352 Coal and Other Mining Operations: Role of Sustainability

subsequent hydrolysis of the ferric iron produced, (SRB) from AMD via sulfate reduction and pro-
which is a net acid-generating reaction. duction of biogenic hydrogen sulfide from the
four-stage and six-stage metal sequential separa-
4Fe2þ þ O2 þ 4Hþ ! 4Fe3þ þ 2H2 O tion and biorecovery units. The four-stage selec-
4Fe3þ þ 4H2 O ! 4FeðOHÞ3 þ 12Hþ tive sequential batch type metal precipitation
process was able to separate metal sulfides and
In the event the pH of the mine water decreases hydroxides at reasonably high recoveries and pre-
significantly as a result of these reactions, addi- cipitate purities. Copper sulfide and zinc sulfide
tional amendments (such as an anoxic limestone were precipitated in Stage 1, aluminum hydroxide
drain) may be required. To maintain oxidizing was precipitated in Stage 2, ferrous sulfide was
conditions, aerobic wetlands are relatively shal- precipitated in Stage 3, and other metals were
low systems that operate by surface flow. Though precipitated in Stage 4. The effluent water
macrophytes are planted for aesthetic reasons, contained only calcium and magnesium in trace
they also regulate water flow, accelerate the rate concentrations and both sulfate and metal sulfide
of ferrous iron oxidation, and stabilize the accu- concentrations were below detectible limits.
mulating iron precipitates. Von Fahnestock [35] recently developed a
In contrast to aerobic wetlands, the compost breakthrough technology to efficiently remove
bioreactors (installations that are enclosed entirely sulfate ions and metal cations from AMD by
below ground level and do not support any mac- using the Acid Mine Drainage Value Extraction
rophytes) are anaerobic. The microbially cata- Process (AMD VEP), which is a novel adaptation
lyzed reactions in compost bioreactors generate of liquid-liquid extraction process. The process
net alkalinity and biogenic sulfide, and thus can converts AMD to purified water as well as sale-
treat mine waters that are acidic and metal-rich. able products, such as potassium sulfate and iron
The reductive reactions that occur within compost sulfate. This technology results in simultaneous,
wetlands are driven by electron donors that derive cost-effective isolation and concentration of use-
from the organic matrix of the compost itself. The ful metals and sulfate ions from mine pool water.
indigenous iron- and sulfate-reducing bacteria are A 30-gpm (113.5-L/m) demonstration plant was
generally considered to have the major roles in built and was operational within 15 months in
AMD remediation in compost bioreactors. St. Michael, PA. In this technology AMD water,
laden with sulfate and iron, feeds into the water
purification stages where the AMD water is
sequentially contacted with an extractant solution
Metal Recovery: Biotic and Abiotic in a counter-current flow path. The extractant
solution is formulated to efficiently pull the sulfate
Innovative and alternative techniques that facili- and iron from the aqueous phase.
tate the economic control and recovery of metal The Water Purification Stages are composed as
values are beneficial not only for protection of a set of four mixer-settler units, which are two-
human health and the environment, but also for compartment tanks. The different stages are
the recovery of these commodities and resource described below:
conservation. AMD, rich in metals, should be
considered not only as a serious environmental 1. In the first compartment, the extractant and the
problem, but as an important resource of metals AMD are mixed in a chamber. The residence
of considerable value to many industrial concerns. time is between 60 and 90 sec.
Recently developed metal recovery processes 2. In the second stage, the combined effluent flows
from AMD and other acidic streams enable us to from the mixing chamber into the settling cham-
recover and recycle these metals cost-effectively. ber where the organic extractant phase disen-
A two-stage process was developed by Tabak gages from the water phase. The extractant,
et al. [27] to separate the sulfate-reducing bacteria containing iron and sulfate, overflows an exit
Coal and Other Mining Operations: Role of Sustainability 353

weir into the settling chamber and is separated highways for 6 months a year, by changing oper-
cleanly from the water phase, which underflows ating conditions and reagent feeds.
the same weir and exits as a separate stream
with proportionately less iron and sulfate.
3. In the recovery stages of the AMD VEP, the Water Quality and Acid Mine Drainage:
extractant, loaded with iron and sulfate, flows Pre-mine Predictions and Post-Mine
sequentially from the water purification stages Comparisons
to the metals recovery and sulfate recovery
stage of the process. Iron and sulfate are sequen- Necessity of accurate prediction of acidic drain-
tially recovered as usable products during these age from proposed mines is recognized by both
stages. The metals-recovery stages are a set of industry and government as a critical requirement
two to three metal cation recovery tanks. The of mine permitting and long-term operation. Sub-
extractant flows counter-clockwise with an stantial emphasis has been placed on prediction of
aqueous sulfuric acid solution to form an iron acid drainage associated with coal development in
sulfate concentrate that is harvested for reuse. the Eastern USA [2], and metal mining in the
The sulfuric acid and the extractant are mixed Western USA and in Canada [16]. The prediction
and separated in a similar manner to that in the of acid-generating potential from any geologic
mixer-settler tanks of the extraction section. formation is dependent on the ability to character-
4. The sulfate-recovery stages remove sulfate ize the presence and quantity of both acid-forming
from the extractant, which again flows through minerals and neutralizing minerals in the mate-
a series of mixer-settler tanks in a counter- rials that are expected to be unearthed during
current fashion. The extractant is contacted mining operations. Typically, samples are col-
with potassium carbonate (a basic aqueous lected by drilling during exploration, analyzed,
solution) to produce a potassium sulfate con- and interpreted with respect to their risk of acid
centrate. The extractant exits the last mixer- formation. In these analyses, the amount of sulfur
settler of the sulfate-recovery section regen- present in geologic materials is measured and
erated and is ready to contact a new stream of attributed to being either an acid-forming mineral
AMD feedwater in the water purification such as pyrite (FeS2) or nonacid-forming mineral
stages. The potassium sulfate (K2SO4) concen- such as gypsum (CaSO4 • 2H2O). The relative
trate leaves the sulfate-recovery stages and is amount of acid-forming minerals is then
collected and stored for sale as fertilizer or contrasted to the amount of neutralizing minerals
reuse in a product tank. such as calcite (CaCO3) to develop a prediction of
the probability of acid generation.
An engineering analysis indicated that the The acid base accounting (ABA) of a material
maximum cost to treat and recover metals by is the balance between total acid-generating
AMD VEP process is $8.00 dollars/1000 gallons potential (AP), which is the total amount of acidity
[3] dependent on product values and reagent pric- that would be produced if all sulfide in a material
ing. AMD VEP, based on abiotic processes, is one is completely oxidized, and total acid-neutralizing
of the few recent technologies to purify AMD potential (NP), which is the amount of acid that
water sufficiently to enable its discharge to surface could be consumed by neutralizing minerals. AP
waters. In addition, the product water can be made and NP are converted to CaCO3 equivalents and
useful for industrial, municipal, agricultural, resi- reported as grams of CaCO3 per kilogram rock.
dential, and other uses. These process by products ABA is typically calculated from analysis of sul-
enable the system to focus on meeting seasonal fide S and carbonate C, assuming a 1:1 molar ratio
and regional demands. For example, potassium of sulfide S (AP) and carbonate C (NP).
sulfate fertilizer can be produced for 6 months a Converting chemical analysis for sulfide
year for the turf grass industry, and the same S (SFeS2) and carbonate C (CCaCO3) can be
system can produce sodium sulfate deicer for expressed as [30]:
354 Coal and Other Mining Operations: Role of Sustainability

AP ¼ SFeS2 ð10Þ  ð3:12Þ • Mining from tailings: Sometimes amounts of

the metals being mined end up in the wastes.
NP ¼ CCaCO3  ð10Þ  ð8:33Þ Reprocessing these wastes can result in more
usable material.
where • Dust suppression techniques: During the mining
process, large amounts of dust are released into
SFeS2 = Concentration sulfide sulfur in sample the atmosphere. This can be minimized by pre-
(weight % S) wetting the areas to be blasted with high volume
3.12 = molecular weight of CaCO33 / molecular sprinklers. Once the dust is in the air, the only
weight of sulfur way to remove it is through mist cannons. Cur-
CCaCO3 = Concentration carbonate carbon in rent mines using open pit mining techniques can
sample (weight % C) reduce dust through by purchasing high volume
8.33 = molecular weight of CaCO3 / molecular sprinklers and wet earth fog cannons.
weight of carbon • Liquid membrane emulsion technology: This
is a separation technique that can be used to
The ratio of neutralization potential (NP) to extract usable metal from highly toxic or acidic
acid potential (AP) is commonly presented in waste water produced by a mine.
graphical interpretations with the inference that • Sulfuric acid leaching extraction process: Cur-
geologic materials with an abundance of NP are rently, bastnaesite (carbonate-fluoride mineral)
unlikely to generate acidic drainage. Skousen is generally processed in a hydrochloric acid
et al. [26] reported that NP:AP ratios <1 com- leaching process. This process leaves fluorine
monly produce acidic drainage, NP:AP ratios and thorium contaminated wastes and it creates
between 1 and 2 may produce either acidic or cerium of low purity. However, a sulfuric acid
neutral drainage, and NP:AP ratios >2 should leaching extraction process can produce high-
produce alkaline water. purity cerium while also allowing thorium to
However, this index does not always accurately be separated for use and better containment.
predict the resultant acid generation from a mine. • Impermeable tailings storage: Tailings are
Out of 56 mines evaluated by Skousen et al. [26], often stored in some form of tailings pond.
11% did not conform to the expected results based Radioactive elements in this pond can contam-
on NP:AP ratios, including four sites with ratios >2 inate the water and if leaks occur, the surround-
that eventually produced acidic drainage. Mineral- ing area. Tailings will be required to be stored
ogical variation between each geologic domain in impermeable lined ponds. Alternatively, the
causes dissimilar reactivity to weathering conditions tailings can be dehydrated into a paste as done
and leads to laboratory variability in assessment. by Molycorp. Mountain Pass, a mine in the
Forecasting future water quality impacts from Clark Mountain Range in California, United
AMD based on laboratory and field data should States, was shut down in 2002. At the time, it
not be considered routine and robust, rather they was owned by Chevron. At its peak, it was
should be considered an area of uncertainty and producing most of the world’s supply of rare
ongoing research. earth metals. When the Mountain Pass mine
was operating at full capacity in the 1990s, it
produced as much as 850 gallons of salty
Recent Sustainable Technologies wastewater per minute, which also contained
radioactive thorium and uranium. The mine
Research and development of sustainable mining was closed due to a large environmental mis-
technology in the areas of processing, clean water, hap when waste leaked into the desert.
and energy efficiency will be continued. Exam- Molycorp acquired the land in 2008, and
ples of the recent technologies include: adopted sustainable mining policy. One of the
Coal and Other Mining Operations: Role of Sustainability 355

processing technologies include change in tail- 5. EIA (2010) Annual energy outlook 2010 with projec-
ings storage. Paste-tailings are presently dried tions to 2035. U.S. Energy Information Administra-
tion, Office of Integrated Analysis and Forecasting,
out and laid flat in a facility, which makes them U.S. Department of Energy, Washington, DC.
less volatile and more compact. This improve- DOE/EIA-0383(2010)
ment had the following results: (a) reduction in 6. Du Plessis HM (1983) Using lime treated acid mine
the supply of fresh water need as the water water for irrigation. Water Sci Technol 15:145–154
7. Earth Justice (2010) Petition for rulemaking under
from tailings can be recycled, (b) reduction in the clean air act to list coal mines as a source cate-
chemical usage in tailings that can be recycled gory and to regulate methane and other harmful air
along with the water, and (c) elimination of emissions from coal mining facilities under section
120 evaporation ponds. 111. Denver
8. Gardner G, Sampat P (1998) Mind over matter: recast-
• Using waste heat from mining to generate ing the role of materials in our lives. Worldwatch
steam and power can decrease the carbon Institute, Washington, DC, p 18
footprint. 9. Hodge RA (2004) Mining’s seven questions to sus-
• Choice of ventilation and diesel engines: 10% tainability: from mitigating impacts to encouraging
contribution. Episodes 27(3):177–184
of mining costs come from electricity con- 10. IEA (2007a) World energy outlook 2007 – China and
sumption. Better ventilation systems and diesel India insights. OECD/IEA, Paris
engines could be purchased by mines to max- 11. IEA (2007b) CO2 emissions from fuel combustion,
imize efficiency. 1971–2005, 2007 edn. OECD/IEA, Paris
12. IISD (2002) Seven questions to sustainability: how to
assess the contribution of mining and minerals activities.
Task 2 work group, mining, minerals and sustainable
development North America (MMSD) North America.
Future Directions International Institute for Sustainable Development,
Winnipeg
The starting point is typically the acceptance of 13. ITRC (2010) Technology overview of phytotechnologies
sustainable developmental principles at board- for mining solid waste and mine impacted waters. The
Interstate Technology & Regulatory Council
room level as corporate goals, and then informing 14. Johnson DB (2001) Importance of microbial ecology
the workforce, investors, and others of that com- in the development of new mineral technologies.
mitment. Relevant employees need to be engaged Hydrometallurgy 59:147–158
as a first step in the practical application of sus- 15. Johnson DB, Hallberg KB (2002) Pitfalls of pas-
sive mine drainage. Rev Environ Biotechnol
tainable development principles, followed by the 1:335–343
gradual extension of training in sustainable 16. MEND (2001) List of potential information require-
methods of working to the workforce as a whole. ments in metal leaching/acid rock drainage assessment
and mitigation work. Mining Environment Neutral
Drainage Program. Price WA. CANMET, Canada
Centre for Mineral and Energy Technology.
Bibliography 17. Mishra PC, Jha S (2010) Dust dispersion modeling
in opencast coal mines and control of dispersion
1. Annandale JG, Beletse YG, Stirzaker RJ, Bristow KL, in Mahanadi coalfields of Orissa. Bioscan 2:479–500
Aken ME (2009) Is irrigation with coal-mine water 18. Norris PR, Burton NP, Foulis AM (2000) Acidophiles in
sustainable? In: Proceedings of international mine bioreactor mineral processing. Extremophiles 4:71–76
water conference, Pretoria 19. Okibe N, Gericke M, Hallberg KB, Johnson DB
2. Brady KBC, Smith MW, Schueck J (1998) Coal mine (2003) Enumeration and characterization of acido-
drainage prediction and pollution prevention in Penn- philic microorganisms isolated from a pilot plant
sylvania. Pennsylvania Department of Environmental stirred tank bioleaching operation. Appl Environ
Protection, Harrisburg Microbiol 69:1936–1943
3. Conkle HN (2008) Engineering analysis report, value 20. Nordstrom DK, Southam G (1997) Geomicrobiology-
recovery from mine or rock drainage water, with water interactions between microbes and minerals. Miner
purity enhancement, using innovative, low cost tech- Soc Am 35:261–390
nology. Battelle, Columbus 21. Pulles W (2006) Management options for mine water
4. DEFRA (2007) Waste strategy for England 2007. Pre- drainage in South Africa. In: WISA, mine water
sented to Parliament by command of her majesty. Depart- division (ed) Mine Water Drainage-South African
ment for Environment, Food and Rural Affairs. PB12596
356 Coal and Other Mining Operations: Role of Sustainability

Perspective. Water Institute southern Africa, of Solid Waste. Washington, DC. RCRA Docket
Johannesburg No. F-96-PH4A-S0001
22. Rawlings DE (2002) Heavy metal mining using 29. U.S. EPA (2006a) Global mitigation of non-CO2
microbes. Annu Rev Microbiol 56:65–91 greenhouse gases. Office of Atmospheric Programs,
23. Schippers A, Sand W (1999) Bacterial leaching of Washington, DC. EPA 430-R-06-005
metal sulfides proceeds by two indirect mechanisms 30. U.S. EPA (2006b) Management and treatment of water
via thiosulfate or via polysulfides and sulfur. Appl from hard rock mines. Engineering issue. Office of
Environ Microbiol 65:319–321 Research and Development National Risk Management,
24. SDWF (2011) Mining and water pollution. Available Research Laboratory, Cincinnati. EPA/625/R-06/014
from Safe Drinking Water Foundation. http://www. 31. U.S. EPA (2009) Identifying opportunities for meth-
safewater.org/PDFS/resourcesknowthefacts/Mining+ ane recovery at U.S. Coal mines: profiles of selected
and+Water+Pollution.pdf gassy underground coal mines 2002–2006. Coalbed
25. Sherlock EJ, Lawrence RW, Poulin R (1995) On the Methane Outreach Program. EPA 430-K-04-003
neutralization of acid rock drainage by carbonate and 32. Waddell S, Pruitt B (2005) The role of coal in climate
silicate minerals. Environ Geol 25(1):43–54 change: a dialogic change process analysis. The gen-
26. Skousen J, Simmons J, McDonald LM, Ziemkiewicz P erative dialogue project launch meeting, New York,
(2002) Acid-base accounting to predict post-mining 6–8 Oct
drainage quality on surface mines. J Environ Qual 33. WEO (2007) World energy outlook 2007: China and
31:2034–2044 India insights. International Energy Agency, Paris.
27. Tabak HH, Scharp R, Burckle J, Kawahara FK, 61 2007 01 1 P1
Govind R (2003) Advances in biotreatment of acid 34. Younger PL, Wolkersdorfer C (2004) Mining impacts
mine drainage and biorecovery of metals: 1. Metal on the fresh water environments: technical and mana-
precipitation for recovery and recycle. Biodegradation gerial guidelines for catchment scale management.
14:423–436 Mine Water Environ 23:S2–S80
28. U.S. EPA (1995) Identification and description of 35. Von Fahnestock, M. (2010) AMD value extraction
mineral processing sectors and waste streams. Office process. Water Cond. Percificat 52(1):38–40
 clean coal yield by the theoretical maximum
Coal Preparation yield attainable at the same ash content
according to washability analysis.
Gerald H. Luttrell1 and Rick Q. Honaker2 Washability A laboratory procedure that uses
1
Department of Mining and Minerals dense liquids (usually organic) to partition par-
Engineering, Holden Hall Virginia Polytechnic ticles into various density fractions that repre-
Institute and State University, Blacksburg, VA, sent the ideal separation potential of coal (also
USA called float–sink analysis).
2
Department of Mining Engineering, Mining and Yield An indicator of separation performance
Mineral Resources Engineering University of (usually reported as a percentage) that is calcu-
Kentucky, Lexington, KY, USA lated by dividing the clean coal tonnage by the
feed coal tonnage.

Article Outline
Definition of the Subject
Glossary
Definition of the Subject
Coal preparation, which may also be called
Introduction
washing, cleaning or processing, is the method-
Assessment of Cleaning Characteristics
ology by which coal feedstocks are upgraded in
Coal Processing Operations
order to reduce freight costs, improve utilization
Beneficial Impacts of Coal Preparation
properties and minimize environmental impacts.
Challenges for Coal Preparation
The upgrading, which occurs after mining and
Future Directions
before transport of the cleaned product to market,
Bibliography
is achieved using a variety of low-cost
solid–solid and solid–liquid separation pro-
Glossary
cesses. Examples of processing technologies
used by this industry include screening, classifi-
Ash A measure of coal purity that consists of the
cation, dense medium separation, gravity con-
noncombustible residue that remains after coal
centration, froth flotation, centrifugation,
is completely burned.
filtration and thickening. Several of these pro-
Cut size The particle size that has an equal prob-
cesses also play an important role in environmen-
ability of reporting to either the oversize or
tal control for the preparation facility.
undersize product of a sizing device such as a
screen or hydraulic classifier.
Cut point The density corresponding to a parti-
cle that has an equal probability of reporting to Introduction
either the clean coal or reject product of a
density-based separator. According to the annual census of coal prepara-
Dense medium An aqueous suspension of high- tion plants conducted by Coal Age [1], the USA
density micron-sized particles (usually magne- operates 286 coal preparation plants in 12 states.
tite) used to separate coal from rock based on This number is relatively small by comparison to
differences in density. the worldwide fleet which is estimated to be 2283
Organic efficiency An indicator of coal cleaning plants [2]. The capacity of the plants can range
performance calculated by dividing actual from less than 200 t/h for small operations to

© Springer Science+Business Media, LLC 2012 357

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_431
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media LLC, 2012
https://doi.org/10.1007/978-1-4419-0851-3_431
358 Coal Preparation

6000 t/h or more for large industrial plants. The Assessment of Cleaning Characteristics
need for such facilities can be attributed to the
fact that freshly mined coals contain a heteroge- The capability of coal preparation to improve
neous mixture of both organic (carbonaceous) coal quality varies widely due to inherent differ-
and inorganic (mineral) matter. The inorganic ences in the liberation characteristics of mined
matter includes noncombustible materials such coals. The degree of liberation is determined by
as shale, slate and clay. These impurities reduce the relative proportion of composite particles
the coal heating value, leave behind an undesir- (i.e., particles of intermixed coal and rock that
able ash residue, and increase the cost of trans- are locked together) present in a particular coal.
porting coal to market. The presence of unwanted The composite particles make it impossible
surface moisture also reduces the heating value to physically separate all of the carbonaceous
and can lead to handling and freezing issues for matter from the host inorganic matter. Conse-
consumers. Therefore, essentially all coal supply quently, plant operators are often forced to sacri-
agreements with electrical power stations impose fice coal recovery by discarding some composite
strict limitations on the specific energy (heat), particles as waste in order to improve clean coal
ash and moisture contents of purchased coal. quality to a level that can meet customer
Moreover, as the first step in the power cycle, specifications.
coal preparation plants improve the environmen- The theoretical trade-off between coal recov-
tal acceptability of coal by removing impurities ery and quality can be quantified in the labora-
that may be transformed into harmful gaseous or tory using washability (float–sink) analysis. An
particulate pollutants when burned. These pollut- example of experimental data collected from a
ants typically include particulates (fly ash) and float–sink analysis is shown in Fig. 1. The anal-
sulfur dioxide (SO2) as well as toxic trace ele- ysis is performed by sequentially passing a coal
ments such as mercury. The presence of mineral sample through containers containing liquids
impurities can also influence the suitability of (usually organic) of increasingly higher densi-
coal for high-end uses such as the manufacture ties [3]. Pure coal has a relatively low specific
of metallurgical coke or generation of petro- gravity (SG < 1.3) and is collected as a float
chemicals and synthetic fuels. As such, coal product from the first container, whereas pure
preparation is typically needed to achieve the rock is much denser (SG > 2.2) and is collected
high levels of coal purity demanded by these as a sink product from the last container. Com-
secondary markets. posite particles report as float products in the
This entry provides a brief overview of some of intermediate containers filled with liquids hav-
the technological systems used in coal prepara- ing densities between that of the first and last
tion. In addition, this entry highlights the benefits containers. After density partitioning, the prod-
resulting from coal preparation activities as well ucts from this procedure are dried, weighed and
as the issues associated with the sustained opera- analyzed for quality (ash, sulfur, mercury, etc.).
tion of coal processing facilities. This review sug- Washability data are very useful for predicting
gests that coal preparation will continue to have a and analyzing the performance of coal prepara-
significant impact on the cost, recovery and qual- tion plants since most cleaning processes sepa-
ity of coal supplies. However, future improve- rate coal and rock based on differences in
ments in separation technology and practices are density.
needed to provide further reductions in waste gen- Coal washability has a tremendous impact on
eration and downstream environmental impacts. how effectively a preparation plant can upgrade
In many cases, these improvements may also gen- a particular coal feedstock. However, the types
erate incremental revenue due to the recovery of of processes employed and practices used for
usable coal from waste streams, which could pro- operation and maintenance (O&M) can also
vide a financial incentive for pursuing these greatly influence the performance of the prepa-
activities. ration facility. This effectiveness is typically
Coal Preparation 359

100.0 47.8
1.30
(28.4) (4.2)
SG Legend
15.6 Mass
1.40
(14.5) (Ash)
SG
6.6
1.50
(22.8)
SG
2.2
1.60
(31.2)
SG
2.1
Example: A perfect separation at 1.6 SG would 1.70
(39.6)
provide 72.2% yield of clean coal at 9.0% ash SG
5.6
and 27.8% yield of reject at 78.9% ash. Thus, a 1.90
(62.5)
plant operating with a 90% organic efficiency SG
would be expected to provide a 65.0% clean coal 20.1
yield at 9.0% ash (i.e., 72.2% x 0.9 = 65.0%).] (87.5)

Cumulative float C Sink

Specific gravity Mass Ash Mass Ash Mass Ash
Sink Float (%) (%) (%) (%) (%) (%)
1.30 47.8 4.2 47.8 4.2 100.0 28.4
1.30 1.40 15.6 14.5 63.4 6.7 52.2 50.5
1.40 1.50 6.6 22.8 70.0 8.3 36.6 65.9
1.50 1.60 2.2 31.2 72.2 9.0 30.0 75.4
1.60 1.70 2.1 39.6 74.3 9.8 27.8 78.9
1.70 1.90 5.6 62.5 79.9 13.5 25.7 82.1
1.90 2 0. 1 87.5 100.0 28.4 20.1 87.5
Totals 100.0 28.4

Coal Preparation, Fig. 1 Float–sink (washability) analysis for a 28.4% ash run-of-mine coal

reported as an organic efficiency, which is Coal Processing Operations

defined as the yield of clean coal actually pro-
duced by the separation divided by the theoreti- During the early part of the last century, coal
cal maximum yield of clean coal that could be mines made use of labor intensive sorting
achieved at the same ash content according to methods such as manual hand-picking to remove
washability analysis. Organic efficiencies are unwanted rock from mined coal. This inefficient
often in the high 90 percentiles for well designed approach was soon replaced by simple mechani-
and well run operations, although values well cal sorting processes that reduced misplacement
below this desired level are not uncommon for and provided higher levels of productivity. These
problematic plants. Inefficient processes or prac- early plants typically cleaned only the coarser
tices, which misplace significant amounts of particles (usually larger than >5–10 mm) and
clean coal into waste and/or rock into the clean either recombined the untreated fine particles
product, have a large negative impact on profit- back into the washed product or discarded the
ability. For example, a decrease in organic effi- fines as waste. These inefficient systems dumped
ciency of just one percentage point can large tonnages of coal into waste piles, some of
potentially represent a reduction in plant profit- which are being remined and reprocessed today.
ability of 20% or more. These historic periods were followed by many
360 Coal Preparation

Sizing Cleaning Dewatering

D&R2 Sizer
screens
Plant
Coarse

feed D&R2
Raw coal
screens
screens Dense medium Stoker
vessel centrifuge
(Optional)

D&R2
screens
D&R2
Small

screens
Deslime Centrifugal
screens1 Dense medium dryer
cyclones

Sieve Bend3
Spirals (Sieve
Fine

Dewatering screens)
screen
Classifying
cyclones
Ultrafine

Froth Screen-Bowl
Deslime flotation Thickener
cyclones Refuse/ Clean
(Optional) waste coal

Notes: 1 -Raw coal and deslimescreen decks may be on the same machine, 2 –D&R screens primarily used to recover
dense medium, 3 -Sieve bends may also be combined with clean coal classifying cyclones.

Coal Preparation, Fig. 2 Simplified flowsheet for a modern coal preparation plant showing unit operations for sizing,
cleaning and dewatering of four different size fractions required for efficiency coal upgrading

decades of technology development that ulti- 0.15 mm), and ultrafine (minus 0.15 mm) feed
mately led to the design and operation of rela- material. Although many commonalities exist,
tively efficient plants that are capable of the final selection of what number of circuits to
complete or partial upgrading of the entire size use, which types of unit operations to employ, and
range of mined coals. Many modern coal prepa- how they should be configured, is highly subjec-
ration plants operating today are as complex as tive and dependent on the characteristic properties
industrial facilities once employed only by the of the feed coal in terms of size and composition.
chemical processing industry.
Figure 2 shows a simplified process flowsheet Unit Operations for Particle Sizing
for a modern plant that includes operations for Coal produced by mechanized mining operations
particle sizing, cleaning and dewatering. This contains particles as small as fine powder and as
sequence of operations, which is called a circuit, large as several hundred millimeters. Particles too
may be repeated for different size fractions since large to pass into the plant are crushed to an appro-
the processes used in coal preparation have a priate upper size, or rejected where insufficient
limited range of applicability in terms of particle recoverable coal is present in the coarse size frac-
size (Fig. 3). Modern plants may include as many tions. The crushed material is then segregated into
as four separate processing circuits for treating the groups having well defined maximum and mini-
coarse (plus 10 mm), small (10  1 mm), fine (1  mum sizes. Screens are typically employed for
Coal Preparation 361

(a) Sizing Grizzly

Vibrating screens
Sieve bend
Frequency sieves
Classifying cyclones

(b) Cleaning Dense medium vessel

Coarse jig
Dense medium cyclones
Water-only cyclone
Spirals
Teeter bed
Froth flotation

(c) Dewatering Screens

High frequency screen
Vibratory centrifuge
Screen-scroll centrifuge
-
Screen-bowl centrifuge
Disc vacuum filter
0.01 0.1 1 10 100
Particle diameter (mm)

Coal Preparation, Fig. 3 Effective range of particle sizes treated by various coal preparation processes

sizing coarser particles, while combinations of sta-

tionary sieves and classifying cyclones are used for
sizing finer particles. Figure 3a shows the typical
sizes of particles that can be effectively handled by
common types of industrial sizing equipment.
Screens and Sieves Screens are mechanical
sizing devices that use a mesh or perforated plate
to sort particles into fine (particles that pass
through the screen openings) and coarse
(particles that are retained on the screen surface).
Screens are used for a wide range of purposes
including the extraction of large rock and trash
prior to treatment, the production of optimum size
fractions for separation devices and the rinsing/
dewatering of products from cleaning operations.
The types of screens commonly used in coal prep-
aration include grizzly, vibrating and high- Coarse
frequency screens. A grizzly is a series of evenly
spaced parallel bars that are positioned in the
direction of material flow. This simple device is
used only for gross sizing of very coarse particles
(>50 mm) and is generally applied only for the Fines
protection of equipment from large oversize mate-
rial. Vibrating screens are the most common Coal Preparation, Fig. 4 Vibrating screen used for feed
equipment for sizing particles and are used in coal sizing
362 Coal Preparation

Flow
Fines

Coal Preparation, Fig. 5 High-frequency screen used Coal Preparation, Fig. 6 Sieve bend used to remove fine
for dewatering of fine rock “slimes” from coal

almost every aspect of coal preparation. Vibrating the sieve is smaller than the spacing between bars.
screens (Fig. 4) utilize an out-of-balance rotating Sieve bends are commonly used for fine particle
mechanism to create a vibrating motion to sort dewatering and may also be used ahead of vibrat-
particles and to move material along the screen ing screens to enhance sizing performance.
surface. High-frequency screens (Fig. 5) are nor- Classifying Cyclones Classification is used
mally used for dewatering fine coal or rock by in coal preparation for fine sizing where conven-
retaining slurry between two opposing inclines. tional screening and sieving becomes impractical
As the name implies, this type of screen operates or too expensive. Classification exploits differ-
at a high vibrating frequency, which is beneficial ences in the settling rates of particles of different
for dewatering fine solids. size (i.e., smaller particles settle slower than larger
The sizing of fine coal is particularly difficult particles). Classification is usually applied to
due to the increased likelihood of plugging and dilute suspensions of <1 mm diameter particles.
the large number of particles that must pass This technique is generally less efficient than
through the screen surface. As such, fine sizing screening or sieving since the separation can be
is performed in many cases using sieve bends. influenced by particle shape and density as well as
A sieve bend (Fig. 6) consists of a curved panel particle size. The most common type of classifi-
constructed from inverted trapezoidal bars (profile cation device used in the coal industry is the
or wedge wire) placed at right angles to fluid flow. classifying cyclone (Fig. 7). In this device, dilute
The capacity of the sieve is relatively high since slurry is pumped under pressure to multiple par-
the leading edge of the trapezoidal bars “slice” allel units (banks) of cyclones. The rotating flow
material from the flowing stream. However, due to field within the cyclone creates a centrifugal force
the unique geometry, the particle size retained by that increases the settling rate of the particles.
Coal Preparation 363

High SG
rock

Smaller Low
particles SG
Magnetite coal
suspension

AIR CORE Feed

Larger
particles

Coal Preparation, Fig. 7 Classifying cyclone used to Coal Preparation, Fig. 8 Chain-and-flight dense
hydraulically size fine coal medium vessel used to separate coarse coal and rock

properties that vary according to mineralogical

Larger particles are forced to the wall of the
composition of individual particles. Some of the
cyclone where they are carried to the discharge
common properties that are used to separate coal
port (apex) at the bottom of the cyclone, while
and rock include size, density and wettability. As
finer particles exit out the top of the cyclone
shown in Fig. 3b, the effectiveness of different
(vortex finder). Classifying cyclones are com-
types of separators is limited to a relatively narrow
monly applied to obtain cut sizes of
range of particles sizes due to limitations associ-
0.10–0.15 mm and represent the only practical
ated with equipment efficiency and geometry.
option for sizing ultrafine particles (cut size of
Dense Medium Separators The most com-
50 mm or less). Smaller cyclones are typically
mon process for upgrading coarse coal
used to size finer particles since these units pro-
(>12.5 mm) is the dense medium vessel. While a
vide a higher centrifugal force. Classifying
wide variety of different designs exist in industry,
cyclones are very popular in the coal industry
one of the most popular is the chain-and-flight
since they offer a high unit capacity, operational
vessel (Fig. 8). This density-based separator con-
flexibility and low maintenance.
sists of a large open tank through which an aqueous
suspension of finely pulverized magnetite (SG =
Unit Operations for Coal Cleaning 5.2) is circulated. The magnetite creates a suspen-
The primary purpose of coal preparation is to sion with an artificial density that is between that of
separate valuable carbonaceous material from coal and rock. Consequently, lighter coal particles
contaminants in the surrounding host rock. The introduced into the medium float to the surface of
separation is typically accomplished using low- the vessel where they are transported by the over-
cost processes that exploit differences in physical flowing medium to a collection screen. Waste rock,
364 Coal Preparation

or more cyclones. Dense particles of rock sink in

the rotating medium and are pushed toward the
wall of the cyclone where they are eventually
rejected out the bottom. Lighter particles of coal
rise to the surface of the rotating medium and are
carried with the bulk of the medium flow out the
top of the cyclone. Recent advances in cyclone
technology have allowed the upper particle size
limit for DMCs to increase in many current appli-
cations to 60–75 mm. This development has made
it possible to simplify plant circuitry in some
facilities by eliminating the need for a dense
medium vessel.
The primary disadvantage of dense medium
processes is that they require considerable ancil-
Magnetite lary equipment to support the operation. A typical
suspension Low SG circuit requires that the clean coal and rock prod-
coal ucts pass over drain-and-rinse (D&R) screens to
High SG
rock wash the medium from the surfaces of the prod-
ucts and to dewater the particles. The medium
Feed recovered from the drain section is circulated
back to the separator by a pump, while medium
from the rinse section is diluted by spray water
Coal Preparation, Fig. 9 Dense medium cyclone used to and must be concentrated before being fed back.
separate small coal and rock The concentration is achieved using magnetic
separations that effectively recover magnetite
which is much denser, sinks to the bottom of the from nonmagnetic particles that would otherwise
vessel where it is collected by a series of mechan- contaminate the circulating medium.
ical scrapers called flights. The flights travel across Water-Based Separators A wide variety of
the bottom of the vessel and eventually drag the water-based density separators, which avoid the
waste rock over a discharge lip at one end of the need to utilize a artificial dense medium, are also
separator. The unit is highly flexible since the qual- available for separating coal and rock in the par-
ity of the clean coal product can be readily adjusted ticle size range between 0.15 and 1 mm. The most
by varying the density of the medium from 1.3 to common methods include water-only cyclones
1.7 SG by controlling the amount of magnetite and (WOCs) and spirals. A WOC is similar to a clas-
water introduced into the suspension. sifying cyclone, but typically has a stubby wide-
Particles smaller than about 0.5 mm float or angled conical bottom (Fig. 10). Separations
sink too slowly in a static vessel to be effectively occur in a WOC due to differences in the settling
separated. Therefore, dense medium cyclones rates of coal and rock in the centrifugal field
(DMCs) are typically used to clean coal and rock within the cyclone. The separation is also
in the size range between 0.5 and 12.5 mm. These enhanced by the formation of autogenous medium
high-capacity devices (Fig. 9) make use of the created by the natural fines already in the feed
same basic principle as dense medium vessels, slurry. These units are often employed in two
i.e., an artificial magnetite-water medium is used stages or in combination with other water-based
to separate low-density coal from high-density separators to improve performance.
rock. In this case, however, the rate of separation Another water-based separator for treating fine
is greatly increased by the gravitational effect coal that has grown tremendously in popularity is
created by passing medium and coal through one the spiral concentrator (Fig. 11). A spiral consists
Coal Preparation 365

Low SG
coal
High SG
rock

Refuse Midds Clean

Coal Preparation, Fig. 10 Bank of twin water-only Coal Preparation, Fig. 11 Bank of spirals used to sep-
cyclones used to separate fine coal and rock arate fine coal and rock

of a corkscrew shaped conduit with a modified combination with WOCs to improve the efficiency
semicircular cross-section. During operation, feed of separating fine coal.
slurry is introduced to the top of the spiral and is Froth Flotation Separators The smallest
permitted to flow by gravity along the helical path size fraction of coal (<0.15 mm) cannot be effec-
to the bottom of the unit. Particles in the flowing tively separated using density-based separators.
film are stratified such that lighter coal particles are For these particles, a process known as froth flota-
forced to the outer wall of the spiral, whereas tion must be used. Flotation is a physicochemical
heavier particles are forced inward to the center process that separates particles based on differ-
of the spiral. The segregated bands of heavy to ences in surface wettability. Flotation takes place
light materials are collected at the bottom of the by passing finely dispersed air bubbles through an
spiral. Adjustable diverters (called splitters) are aqueous suspension of particles. A chemical
used to control the proportions of particles that reagent, called a frother, is added to promote the
report to the various products. A three-product formation of small bubbles. Typical addition rates
split is usually produced, giving rise to three pri- are in the order of 0.1–0.5 lb. of reagent per ton of
mary products containing clean coal product, coal feed. Coal particles, which are naturally
refuse and misplaced “middlings.” Because of the hydrophobic, selectively attach to air bubbles and
low unit capacity (2–4 t/h), spirals are usually are buoyed to the surface of the pulp where they are
arranged in groups fed by an overhead radial dis- collected from a coal-laden froth bed that forms
tributor. To save space, several spirals (two or atop the cell. Most of the impurities that associate
three) may be intertwined along a single central with coal are hydrophilic and remain suspended
axis. Spirals have been successfully utilized in until they are discharged as a dilute slurry waste.
366 Coal Preparation

Coal

Feed
Reject
Air

Coal Preparation, Fig. 12 Conventional flotation bank Coal Preparation, Fig. 13 Column-type flotation cell
used to clean ultrafine coal used to clean ultrafine coal

and impeller design. A disadvantage of conven-

Another chemical additive, called a collector, may tional flotation cells is that they allow a small
be added to improve the adhesion between air amount of clay slimes to be recovered with the
bubbles and coal particles. Collectors are com- water that reports to the froth product. Therefore,
monly hydrocarbon liquids such as diesel fuel or in cases where high clay concentrations are
fuel oil. In some plants, clay slimes (<0.03 mm) expected, advanced flotation processes such as
are removed ahead of flotation using small diame- column cells (Fig. 13) are often used. Column
ter classifying cyclones to minimize the carryover cells virtually eliminate the recovery of clay
of the high ash slimes into the froth product with slimes by washing the froth with a countercurrent
the process water. flow of water. This feature allows columns to
Most industrial installations make use of produce a higher purity coal product without
mechanical flotation machines (Fig. 12). These sacrificing coal recovery.
machines consist of a series of agitated tanks
(4–6 cells) through which fine coal slurry is Unit Operations for Coal Dewatering
passed. The agitators are used to ensure that larger Solid–liquid separators are used downstream of
particles are kept in suspension and to disperse air coal cleaning processes to remove unwanted sur-
that enters down through the rotating shaft assem- face moisture from the sorted products. Excess
bly. The air is either injected into the cell using a moisture lowers the heating value of the coal,
blower or drawn into the cell by the negative may lead to handling/freezing problems, and
pressure created by the rotating impeller. Most increases the overall costs of transporting the
commercial units are very similar, although coal to customer sites. As shown in Fig. 3c, sev-
some variations exist in terms of cell geometry eral different types of mechanical dewatering
Coal Preparation 367

Feed

Water Water

Dewatered Dewatered
coal coal

Coal Preparation, Fig. 14 Vibratory centrifugal dryer

used to dewater small/fine coal Coal Preparation, Fig. 15 Screen-bowl centrifuge used
to dewater fine coal

methods are required depending on the size of another popular design, called a screen-bowl
particles to be treated. The removal of water centrifuge, may be used. The screen-bowl centri-
from the surfaces of coarser (>5 mm) coal is fuge is a horizontal unit that consists of both a
predominantly carried out using screens, where solid bowl section and a screen section (Fig. 15).
the prime force involved is gravity. Finer particles Feed slurry is introduced into the bowl section
have a higher surface area and tend to have corre- where most of the solids settle out under the
spondingly higher moisture content. There finer influence of the centrifugal field. A rotating scroll
particles are typically dewatered using centrifugal transports the settled solids up a beach and across
methods or filtration systems. the screen section where additional dewatering
Centrifugal Dewatering Centrifugal dryers, occurs before the solids are discharged. Solids
which consist of a rotating screen or basket, use that pass through the screen section are typically
centrifugal force to pull water away from the added back to the machine with the fresh feed.
surfaces of coal particles. These devices operate This unit is capable of providing low product
in much the same fashion as the spin cycle in a moistures, although some ultrafine solids are
washing machine. For coarser fractions, two lost with the main effluent that is discarded
popular centrifugal dryers are the vibratory cen- from the solid bowl section.
trifuge and the screen-scroll centrifuge. The Filtration Dewatering Filtration processes
vibratory centrifuge (Fig. 14) uses a vibrating may be used to dewater fine coal whenever a
mechanism to move solids through the unit, high coal recovery is desirable. Filtration involves
while the screen-scroll design uses a screw con- the entrapment of fine solids as a cake against a
veyor to move the solids. For finer particles porous filtering media. Traditionally, flotation
(<1 mm), both the screen-scroll centrifuge and concentrates have been dewatered using some
368 Coal Preparation

Coal Preparation, Fig. 16 Disc vacuum filter used to Coal Preparation, Fig. 17 Thermal dryer used to dry
dewater fine coal coal to low moisture contents

form of vacuum filtration. These units are capable

of high coal recoveries (>97%) while generating most popular design is the fluidized bed dryer
product moisture in the 20–30% range. The most (Fig. 17), which uses coal, oil or gas as the fuel
popular type of vacuum filter is the disc filter source to heat the incoming air stream. The
(Fig. 16). This device consists of a series of cir- amount of fuel required depends on the amount
cular filtration discs that are constructed from of water that must be evaporated which, in turn,
independent wedge shaped segments. At the depends on the amount of coal fed to the dryer and
beginning of a cycle, a segment rotates down the percentage of water in the dewatered product
into the filter tub filled with fine coal slurry. Vac- [4]. When operating correctly, thermal dryers can
uum is applied to the segment so that water is reduce the clean coal moisture to less than 6% by
drawn through the filter media and a particle weight at a unit cost proportional to the tonnage of
cake is formed. The resultant cake continues to water evaporated [5]. Unfortunately, thermal
dry as the segment rotates out of the slurry and dryers require high capital costs, often in excess
over the top of the machine. The drying cycle ends of several million or more. Dryers can also suffer
when the vacuum is removed and the cake is from emission problems associated fugitive dust
discharged using scrapers and a reverse flow of and poor opacity. Emissions of nitrous oxides,
compressed air. sulfur dioxide, volatile organic compounds
Thermal Dewatering Thermal dryers have (VOCs) may also present issues for some sites
been successfully employed in the coal industry to seeking operating permits. Moreover, thermal
reduce clean coal moisture to single-digit values drying of combustible particles of coal can present
whenever mechanical dewatering processes are safety hazards resulting from accidental fires or
incapable of meeting contract specifications. The dust/gas explosions.
Coal Preparation 369

increase settling rates. The pH levels must also

be carefully monitored and controlled. Conven-
tional thickeners, which are typically 50–200 ft.
diameter, require constant monitoring to ensure
that overflow clarity and underflow density is
maintained and the rake mechanism is not over-
loaded. A torque sensor on the central drive is
typically used to monitor rake loading, while a
nuclear density gauge is often used to monitor
underflow density.

Waste Disposal and Storage

The separation processes used in coal preparation
often generate large volumes of waste rock and
slurry. As a result, the final step in coal preparation
involves the disposal and permanent storage of
these waste products into various types of surface
and/or underground repositories. Although many
variations exist, a common approach is to use
separate refuse piles and slurry impoundments.
Refuse piles are designed to receive coarse parti-
cles of waste rock that can be easily dewatered by
Coal Preparation, Fig. 18 Conventional thickener used screens or sieves. This material is relatively easy
to clarify process water and thicken solids to handle and can be transported by truck or belt
haulage systems to the disposal area. On the other
hand, fine coal wastes are difficult to dewater and
Thickening and Clarification Another are typically discarded in slurry form. The waste
essential solid–liquid separation process used in slurry contains water, coal fines, silt, clay and
coal preparation is thickening. Thickening other fine mineral particles from the processing
(or clarification) processes are used to treat the plant. In most cases, the slurry is discarded into an
process water so that it can be recycled and reused impoundment (Fig. 19). An impoundment is an
within the plant. A thickener consists of a large engineered structure consisting of a large volume
tank in which particles are permitted to settle, earthen settling basin formed behind a man-made
thereby producing a clarified overflow and thick- dam or embankment. The dam or embankment is
ened underflow (Fig. 18). Thickeners are usually usually constructed from compacted coarse refuse
fed dilute slurry (<5–10% solids) that is rejected material. The waste slurry is transferred to the
from the fine coal cleaning circuits. The clarified impoundment by pumping thickened underflow
water that overflows along the circumference of from the plant thickener through a pipeline. The
the thickener is recovered and reused as process volume of the impoundment must be sufficiently
water. The thickened underflow, which often con- large to ensure that fine particles settle by gravity
tains 20–35% solids, is transported to the center of before the clarified water at the surface is recycled
the thickener by a series of rotating rakes. The back to the plant for reuse.
thickened sludge is then pumped to an appropriate
disposal area or is further dewatered prior to dis-
posal. Chemical additives (coagulants and floccu- Beneficial Impacts of Coal Preparation
lants) are usually introduced ahead of the
thickener to promote the aggregation of ultrafine Coal preparation provides several attractive ben-
particles so as to enhance water clarity and efits from both an economic and environmental
370 Coal Preparation

60%
N
77%

Pennsylvania

52%

WEST
VIRGINIA

46%
39% 34%
45%
59% 39%
45% 44%
Kentucky 40% 41%
54%
55%
46% 44%
44%
47% 39% Virginia
24%
49% 43%

38%

Coal Preparation, Fig. 20 Examples of clean coal yield

values for a selected sampling of eastern US plants
Coal Preparation, Fig. 19 Active slurry impoundment
used for fine waste disposal (before reclamation)

of the quality demands imposed by the coal con-

perspective. The most recognized of these include sumer. The quality of the inplace can be affected
increased coal reserves, lower transportation by small changes to mining practices that directly
costs, improved utilization properties and abate- influence out-of-seam dilution. Variations in
ment of pollution. washability, which reflect the selective liberation
of composite particles of intermixed coal and
Coal Reserves rock, can also make estimations of coal yield
Perhaps no other technology has had a greater unreliable when coal is subjected to size
impact on expanding the reserve base of econom- reduction.
ically recoverable coal than coal preparation. The average yield produced by coal prepara-
Preparation plants employ low-cost physical sep- tion plants has steadily declined over the years due
aration processes to convert run-of-mine coal to a depletion of lower ash feeds and increased
resources into saleable coal reserves. The percent- mining mechanization. For example, Fig. 20
age of clean coal tonnage generated from each ton shows the yield of clean coal currently obtained
of run-of-mine coal is commonly referred to as from a random survey of several major plants
plant yield. This parameter is one of the most operating in the eastern USA. It is not uncommon
difficult to estimate when evaluating the reserve for eastern operations to experience yields under
base for a previously undefined coal property. The 30–35%, thereby producing only 1 t of clean coal
yield can be influenced by several factors includ- from three or more tons of mined product. An
ing the quality of the in-place coal, the washability estimate complied from production records sup-
(separating characteristics) of the run-of-mine plied by coal producers suggests that the average
coal, the efficiency of the separation processes yield is now less than 50% (i.e., 49.8%  3.5%)
used by the preparation facility and the strictness for the total USA. This situation is expected to
Coal Preparation 371

worsen as eastern reserves become thinner and improvements, less maintenance, workforce
more challenging to mine. A study reported by reductions, etc.) usually need to be closely exam-
Weisenfluh et al. [6] indicated that nearly 52% of ined to ensure that clean coal yield is not adversely
the remaining eastern Kentucky coalfield impacted by these cost-cutting measures.
resources are located in coal seams that are
0.35–0.70 m in thickness, while 31% are in Utility Performance
0.70–1.1 m thick seams. Likewise, a study of the The thermal efficiency of a power station is very
Virginia coalfields found that 30% of the total important. A higher efficiency is obviously bene-
reserve base (more than 200 million tons) exists ficial from an economic perspective since this
in seams with a thickness less than 0.70 m provides a proportional improvement in generated
[7]. Consequently, ever increasing amounts of revenue for the power station. Moreover, a higher
rock from out-of-seam dilution are being mined, efficiency also reduces the production of green-
loaded and hauled to preparation plants for house gases and other pollutants of environmental
removal and disposal. concern since less coal fuel needs to be burned per
unit of electrical power generated.
One method for improving thermal efficiency
Coal Transportation is to use washed coals of higher quality can sig-
Most of the coal consumed in the USA is used for nificantly improve the thermal efficiency of a
the production of electricity. The cost of trans- boiler [9–12]. Higher quality coals are more reac-
porting the coal to the power station is usually tive and require less excess air for effective com-
borne by the utility and paid based on the deliv- bustion, thereby improving efficiency via a
ered tonnage. The mine operator is also paid by reduction in heat lost with the flue gas. Higher
the utility based on tonnage, although the unit quality coals also improve efficiency by avoiding
price is typically adjusted up or down to account fouling/slagging problems in the boiler, which
for the actual heat content of the supplied coal tend to raise the flue gas temperature and increase
fuel. In most cases, this simple pricing structure heat losses. The extent to which the proper appli-
provides the base economic justification for the cation of coal preparation technology improves
operation of coal preparation facilities. The high- thermal efficiency is highly case specific and dif-
ash rock rejected by coal preparation plants ficult to predict from purely theoretical consider-
contains insufficient heating value to justify the ations. The most reliable data for quantifying
shipment of this material to the utility. The savings efficiency improvements are typically reported
in transportation costs are directly proportional to based on actual plant studies. One such classic
the increase in heating value. study [13] monitored improvements to boiler per-
Cost–benefit studies suggest that the econom- formance resulting from switching from an
ics of coal production are more sensitive to trans- untreated coal (15% ash and 3.5% sulfur) to
portation cost than to any other factor. For washed coal (9% ash and 2.8% sulfur) from the
example, a study by Norton [8] was one of the same mine. Despite the modest improvement in
first to demonstrate how unfavorable changes in coal quality, the boiler efficiency increased from
the cost of mining, processing and transportation by about 1.5 absolute percentage points as a result
affect profitability. The study concluded that the burning better quality coal. The capacity factor
cost of coal transportation had the greatest overall also rose by almost 10% due to less fouling/slag-
impact on total revenue. This study also indicated ging problems.
that the clean coal yield was second only to trans- The use of coal preparation technologies to
portation cost in determining revenues. In con- improve boiler efficiency has also been a major
trast, increases in the capital and operating costs benefit reported in other nations. The International
of the coal preparation plant had only a minor Energy Agency (IEA) estimates that coal-fired
impact on revenue. Therefore, any steps taken to power plants in India can increase thermal effi-
reduce preparation costs (fewer capital ciency by up to 10 percentage points by switching
372 Coal Preparation

from unwashed to washed coal [14]. China also such as the 1990 Clean Air Act Amendment
expects to make greater use of coal preparation (CAAA) and 2005 Clean Air Interstate Rule
technology to improve thermal efficiencies and (CAIR). Moreover, coal preparation has a benefi-
environmental performance [15]. The average cial impact on reducing greenhouse gas emissions
thermal efficiency in China has been reported to by increasing the thermal efficiencies.
be less than 29% [16], compared to around 38% in Noncombustible impurities present in the feed-
OECD countries. As such, coal preparation is stocks supplied to coal power stations generate
expected to continue to have a large impact on waste steams as either bottom ash/slag or fly ash.
the international community. Of these, the finest particles of fly ash emitted to
Another benefit of coal preparation is that it the atmosphere are considered to be of greatest
removes impurities that have a significant influence environmental concern due to their potential
on the O&M costs for a coal-fired boiler. Studies adverse impact on human respiratory health
have demonstrated that the removal of abrasive [24]. Power stations make use of several types of
mineral impurities such as pyrite and quartz can effective control technologies to minimize fine
substantially reduce wear rates and increase the particulate emissions. These postcombustion
throughput capacity of utility pulverizers [17, technologies include efficient processes such as
18]. The impacts associated with the abrasive wear electrostatic precipitators (ESPs), fabric filters
and slagging/fouling of boiler tubes can also be (FFs), cyclones and wet scrubbers. Modern con-
mitigated to a large extent by utilizing washed trol systems typically achieve better than 99.5%
coals that have been properly cleaned to remove removal of all particulates and exceed 99.99% in
unwanted mineral matter [19–21]. Vaninetti and some cases. However, standards for particulate
Busch [22] have provided a detailed description of emissions continue to become increasingly strin-
these problems and have developed empirical for- gent as reflected by expanded regulations by the
mula that can be used to assist in the evaluation of US EPA to include particles finer than 2.5 mm in
changes to coal quality in specific types of boilers. new ambient air quality standards [25]. As such,
The handling characteristics of solid coal are there is continued interest in removing greater
also an important issue. A poor handling coal may amounts of particulates upstream of other emis-
hang in railcars, plug chutes and bins and stick to sion controls using coal preparation technologies.
conveyor belts. These problems may result in The separation processes used in coal prepara-
unscheduled shutdowns, thereby reducing power tion plants remove noncombustible minerals that
station availability. Washed coals from prepara- ultimately affect the amount and type of particulate
tion plants typically have superior handling char- matter (PM) that passes downstream to emission
acteristics to run-of-mine coals, especially if all or control systems. For these systems, proper levels
a portion of the ultrafines have been removed. of coal washing can be identified that effectively
These coals also typically present fewer problems reduce ash loading and improve removal efficien-
in terms of unwanted dust generation, solids run- cies. A recent presentation by American Electric
off during precipitation events and freezing prob- Power to the Asia Pacific Partnership [26] con-
lems during colder months [23]. cluded that “Consistent and proper quality coal is
best tool to improve plant operating performance
Pollution Abatement and reduce PM and SO2 emissions. Removal of
Coal preparation plays an important role in reduc- some of the coal ash (includes rocks) at the mine is
ing the emissions of pollutants that associate with more economic than in the pulverizer, boiler, pre-
the mineral matter contained in coal. These emis- cipitator and scrubber.” Washing also minimizes
sions normally include solid particulate emissions the total amount of high-surface-area fly ash that is
such as fly ash as well as gaseous emissions of more hazardous to dispose due to its high reactivity.
precursors associated with acid-rain and air Even in cases where particulate controls are cur-
toxics. These emissions are strictly regulated for rently deemed adequate, greater use of coal prepa-
coal-fired utilities through various legislative acts ration may be required in the future to compensate
Coal Preparation 373

including improved market flexibility, lower

Northern
Appalachia scrubber loading and concurrent removal of
Organic other impurities (ash, trace elements, moisture,
Southern Pyritic etc.). Although coal preparation does not directly
Appalachia
impact the nitrogen content of coal, washing has
been shown to help reduce NOx emissions by
Alabama
providing a consistent high quality fuel that
makes it easier to control the combustion
Eastern
Midwest
environment [31].
Coal preparation also has a tremendous impact
Western on the amount of air toxics released to the atmo-
Midwest sphere. The 1990 CAAA contained provisions
that established new emission standards for a
Combined variety of air toxics, which are commonly referred
to as hazardous air pollutants (HAPs). Many of
0 1 2 3 4 5 6
the HAPs identified in the CAAA are present as
Sulfur content (%)
trace elements in coal [32, 33]. Some of the most
Coal Preparation, Fig. 21 Distribution of sulfur types noteworthy of these are antimony, arsenic, beryl-
typically in US coals. (After Cavallaro et al. [29]) lium, cadmium, chlorine, chromium, cobalt, fluo-
rine, lead, manganese, mercury, nickel, selenium
for deterioration in feedstock quality as higher qual- and radionuclides. The concentrations of these
ity coal reserves become depleted. elements are known to vary considerably from
Coal preparation has also played an important seam to seam, and in some cases within the same
role in environmental compliance by reducing the seam [34, 35]. During combustion at electrical
sulfur contained in coal feedstocks prior to com- utilities, these elements may be released to the
bustion [27]. Sulfur occurs in coal as three distinct atmosphere as solid compounds with the fly-ash
forms (e.g., sulfate, organic and pyritic). Sulfate and in the vapor phase with the flue-gas. Existing
sulfur is present in very small quantities and is not post-combustion control technologies, such as
considered a serious problem. Organic sulfur, electrostatic precipitators, can be reasonably
which is part of the basic structure of coal, is not effective in reducing the concentration of trace
removable by conventional cleaning. The pre- elements associated with fly-ash. These com-
combustion removal of organic sulfur is possible monly include elements such as antimony, beryl-
using chemical cleaning methods; however, these lium, cadmium, cobalt, lead and manganese.
elaborate processes are not economically compet- Capture efficiencies of >97% have been reported
itive with flue gas scrubbing. Pyritic sulfur is for electrostatic separators [36]. On the other
present as discrete inclusions of iron sulfides dis- hand, trace elements such as arsenic, chlorine,
tributed within the coal matrix. As such, this form mercury and selenium have the potential to vola-
of sulfur has the potential to be removed by phys- tilize and are less effectively controlled by post-
ical cleaning processes such as coal preparation combustion methods.
[28]. Fortunately, many of the high sulfur coals in Studies have shown that many of the HAP
the USA also contain a high proportion of pyritic precursors identified in the CAAA are associated
sulfur (Fig. 21). with the mineral matter commonly rejected by
Coal preparation plants have been reported to coal preparation plants. This approach to HAPs
remove up to 90% by weight of pyritic sulfur, control is attractive since the waste rock rejected
although rejections are typically in the 30–70% by coal preparation plants is coarser and has a
range due to liberation constraints [28–30]. When lower reactivity than the high-surface-area ash
compared to the postcombustion control of sulfur, generated by power stations [37]. Inplant sam-
coal preparation offers several distinct advantages pling campaigns conducted by various
374 Coal Preparation

Ash
in the food chain as higher species consume lower
life forms exposed to mercury contamination
Sulfur
[42]. Data reported by the US Environmental Pro-
As
tection Agency (EPA) indicate that coal-fired util-
ities are currently the largest human-generated
Be
source of mercury releases in the USA. It is esti-
Cd mated that these plants release approximately 48 t
Co annually [43]. In order to curb these emissions,
Cr
EPA issued the world’s first-ever rule to cap and
reduce mercury emissions for coal-fired power
F
plants. Compliance options available to utilities
Hg include postcombustion capture of mercury by
Mn existing or new flue gas scrubbing technologies
Northern App.
Ni
as well as precombustion control of mercury by
Southern App.
coal preparation and coal switching [44].
Pb Illinois Basin
A recent study by Quick et al. [45] showed that
Sb the mercury content of coal delivered to utilities
Se (based on ICR data) was lower than that of the
0 20 40 60 80 100 in-ground coal resources in the USA (based on
Reduction (%) COALQUAL data). This comparison is shown in
Fig. 23 for the primary eastern coal producing
Coal Preparation, Fig. 22 Reduction in trace element
states. Based on this study, Quick et al. [45] con-
content after coal preparation. (After Fonseca et al. [36])
cluded that “. . .selective mining and more exten-
sive coal washing may accelerate the current trend
researchers [36, 38] suggest a good correlation toward lower mercury content in coal burned at
between the rejection of mineral matter and the US electric utilities. . .” and “. . .since recent
removal of trace elements during physical reductions of sulfur emissions from coal-burning
cleaning. These findings are also supported by electric utilities are largely due to a declining
laboratory float–sink tests performed using a vari- sulfur content of delivered coal, rather than from
ety of eastern US coals [39–41]. These data sug- scrubbing combustion gases, these simple, low-
gest that trace elements are typically rejected at cost approaches to reduce Hg emissions should
levels of 40–70% by weight using conventional not be overlooked.”
preparation technologies. These values appear to According to Alderman [46], cleaning can
be in good agreement with earlier values reported reduce mercury by more than 50% in many east-
by Fonseca et al. [36], which showed an average ern and western coals and lignites, excluding
trace element removal by conventional coal prep- southern Powder River basin coals. Greater rejec-
aration of approximately 64% for six different tions of mercury by coal preparation appear to be
coals (Fig. 22). On the other hand, the large degree limited by inadequate liberation and/or the pres-
of variability observed in the data from these and ence of organically associated mercury. Several
other studies suggest that the rejections of trace studies have suggested that mercury has some
elements by coal preparation are very site specific degree of association with the iron sulfides present
and need to be quantified on a case by case basis. in many run-of-mine coals [47–49]. In fact, the
Of the air toxics reduced by coal preparation, data summarized in Fig. 24 suggest that the con-
mercury is perhaps the trace element in coal of centration of many metallic elements found in
greatest environmental concern [35]. Mercury can coal often correlate well with the presence of
be released during coal combustion and subse- sulfide minerals [50]. As a result, the rejection
quently deposited in the environment. Ecological level for mercury is often in the same range as
studies have shown that mercury bio-accumulates the pyritic sulfur rejection for coals subjected to
Coal Preparation 375

Coal Preparation,
Fig. 23 Comparison of All
mercury contents for
delivered (ICR) and
in-ground (COALQUAL) AL
coals. (After Quick et al.
COALQUAL
[45])
ICR
KY

State of origin
OH

PA

th
25th 50 75th
VA

th
WV 10th Mean 90

0 10 20 30 40
Mercury content (lb Hg / 1012 BTU)

1000.0 10.000 100.0

1.000
Cadmium (ppm)
Arsenic (ppm)

100.0 10.0
Cobalt (ppm)

0.100

10.0 1.0
0.010

1.0 0.001 0.1

0.1 1.0 10.0 100.0 0.1 1.0 10.0 100.0 0.1 1.0 10.0 100.0
Sulfides (%) Sulfides (%) Sulfides (%)

1000.0 1000.0 10.0

100.0 1.0
Mercury (ppm)

100.0
Nickel (ppm)
Lead (ppm)

10.0 0.1

10.0
1.0 0.0

0.1 1.0 0.0

0.1 1.0 10.0 100.0 0.1 1.0 10.0 100.0 0.1 1.0 10.0 100.0
Sulfides (%) Sulfides (%) Sulfides (%)

Coal Preparation, Fig. 24 Effect of sulfide mineral content on the concentration of trace metals in a sample of
Pittsburgh seam coal. (After Luttrell et al. [50])

coal preparation. While liberation can be associated with the fine grinding of coal
improved by reducing the topsize of the feed [28]. Moreover, fine particles are also difficult
coal [51], this approach is difficult to justify in and costly to upgrade, dewater and handle in
today’s marketplace due to the high costs existing coal preparation facilities.
376 Coal Preparation

Another important impact of coal preparation is 10

in the reduction of greenhouse gas emissions 14 Mesh
9
resulting from improvements in thermal efficiency,
i.e., less CO2 is produced per unit of electricity 8
generated. Calculations indicate that a one percent-

Coal availability (Billion Tons)

7
age point improvement in thermal efficiency pro- 3/8 In.
vides a two to three percentage point reduction in 6
CO2 emissions for a typical coal-fired utility. An
investigation conducted by Couch [52] indicated 5

that there are more than 4000 coal-fired boilers 4

(>50 MW capacity) worldwide that could improve
thermal efficiencies and reduce CO2 emissions by 3 1-1/2 In.
using coal preparation to improve coal quality. 2
Moreover, von Hippel and Hayes [53] have noted
that coal preparation reduces the amount of energy 1
consumed (and CO2 generated) during coal trans-
0
portation by increasing the specific heating value 1 10 100
of the delivered coal. Smith [13] also found that Particle Top Size (mm)
freshly mined coals subjected to coal preparation
Coal Preparation, Fig. 25 Effect of decreasing top size
processes tended to display slower releases of on low-sulfur coal availability. (After Cavallero et al. [30])
methane, another greenhouse gas.

processes are considerably less efficient and sub-

Challenges for Coal Preparation stantially more costly. This is unfortunate since the
amount of high quality recoverable coal within a
Coal preparation offers may attractive benefits for given reserve can theoretically be increased by
coal-fired power generation. These typically crushing. The size reduction improves the liberation
include lower transportation costs, improved by reducing the population of intermixed composite
properties for coal utilization and reduced emis- particles of coal and rock. A study conducted by
sions of particulate and gaseous pollutants. How- DOE [30] indicates that the reserves of compliance
ever, the industry also faces several challenges coal in central Appalachia could be nearly doubled
that need to be resolved to ensure that preparation by efficiently cleaning at a particle size of 1 mm
plants can continue to operate at a profit without (Fig. 25). Although a systematic assessment has not
damage to the environment. These concerns can been performed to date for trace elements, the size
be generally classified as technical factors that reduction would also be expected to substantially
relate to shortcomings in current processing sys- improve the removal of coal-related pollutants other
tems, environmental factors that involve improv- than just ash and sulfur.
ing waste handling and disposal, or health and Unfortunately, the solid–solid separation pro-
safety factors that may affect workers or citizens cesses used industrially to treat fine coal represent
in the surrounding area. the single greatest loss of potentially recoverable
coal in a preparation facility. Field studies indicate
Technical Issues that the froth flotation process, which is normally
Cleaning of Fine Particles The freshly mined used to recover <0.2 mm coal, typically recovers
coals that are fed to coal preparation plants are only 60–80% of the organic matter in this size
typically crushed to liberate rock before washing range. This surface-based separation is inherently
and to limit the largest size of particles that enter less effective in removing pyritic sulfur than
the plant. Operators prefer to keep the particles as density-based processes used to treat the coarser
large as possible (usually >50 mm) since fine coal sizes of coal [54]. As such, the desulfurization of
Coal Preparation 377

fine particles is also often poor in many operating in an attempt to lower the product moisture. In
preparation facilities. Therefore, the development addition, these processes are typically the most
of effective, low-cost processes for treating fine energy intensive processes used in coal prepara-
coal is a major need for the preparation industry. tion, often consuming 6–12 kW/t/h of dry solids
Effective solutions need to be found for improving processed [58]. Thermal dryer systems can
the recovery, selectivity and capacity of froth flota- effectively reduce moisture; however, these mas-
tion processes. This goal may be achieved through sive units require very large capital expenditures
fundamental and applied studies that seek to under- that are difficult to justify in the coal industry.
stand and improve flotation chemistry, equipment Also, indirect thermal drying systems (e.g.,
design and process control. In addition, new types Holoflite and Torus Disc) typically require
of density separators need to be developed for 200–400 kW/t/h of product for drying fine coal
treating fine coal. Centrifugal separators such as solids to single-digit moistures [59]. Moreover,
enhanced gravity concentrators, which have been stringent air quality standards make it impossible
successfully applied in the gold industry, may in many cases to obtain new operating permits for
prove useful for this purpose. These devices have thermal dryers. Therefore, the coal preparation
the potential to reject significantly greater amounts industry needs to develop new mechanical
of pyritic sulfur (and mercury) that is not efficiently solid–liquid separation processes that are substan-
removed using surface-based separators such as tially more efficient in terms of removing moisture
froth flotation [55, 56]. However, these machines and less expensive to purchase, operate and main-
have several limitations including low throughputs tain. Innovative systems are critically needed,
and control difficulties. which may require fundamental studies to identify
Dewatering of Fine Particles The controlling mechanisms that can lead to the devel-
solid–solid separation processes employed by opment “breakthrough” technologies.
modern coal preparation plants require large Sizing of Fine Particles There are many
amounts of process water. After cleaning, the technical challenges for the coal preparation
unwanted water must be removed from the sur- industry that directly relate to size–size separa-
faces of the particles using mechanical dewatering tions. Size–size separations are required ahead of
equipment. Inefficient removal of moisture lowers solid–solid separations since these units are only
the heating value, increases transport costs and effective within a narrow particle size range.
creates handling/freezing problems for the Vibrating screens are generally efficient and cost
cleaned coal. Coarse particles can be readily effective for sizing and dewatering coarser parti-
dewatered using simple screening systems, while cles. On the other hand, screening systems for
finer particles require more complicated unit oper- finer particles, particularly those finer than
ations such as centrifuges and filters. Unfortu- 0.5 mm, tend to blind easily, wear quickly, and
nately, the mechanical systems used to dewater suffer from low throughput and low efficiency.
fine coal are inefficient and costly [57]. Fines The misplacement of incorrectly sized particles
often represent as little as 10% of the total run- into equipment not designed to handle such sizes
of-mine feed; however, this size fraction may can have a large adverse impact on both the sep-
contain one third or more of the total moisture in arating performance and maintenance require-
the delivered product. ments for a preparation plant. Another important
The availability of low-cost mechanical issue with screening is the desliming of coal prod-
dewatering equipment that can efficiently remove ucts to remove ultrafine mineral sediments that are
moisture from fine coal is widely considered to be detrimental to quality and moisture. Many in the
an important need for the coal preparation indus- industry believe that the ability to screen ultrafine
try. Existing technologies for fine coal dewatering particles at sizes of 0.15 mm and smaller is par-
tend to produce unacceptably high moistures ticularly important.
exceeding 25–35% by weight or intentionally In addition to new screening systems, break-
sacrifice half or more of the ultrafines as a waste through technologies in ultrafine classification are
378 Coal Preparation

also needed in the coal preparation industry whole. Today, mining companies accept these envi-
[60]. Classification is the separation of particles ronmental standards as good business practice and,
due to differences in settling velocities, which in some cases, may go beyond simple compliance
depends not only on particle size, but also on to promote goodwill and to set an example for
particle density and shape. Firth and O’Brian others. Nevertheless, potential environmental
[61] noted that while existing classification sys- issues associated with coal processing still exist.
tems were adequate for the coal preparation indus- These issues have been generically classified by the
try of the past, “. . .it is apparent that further World Bank to include air emissions, wastewater,
improvements in yield/ash/moisture relationship hazardous materials, solid wastes and noise [64].
achieved by coal preparation plants will require Coarse Waste Disposal Of the various chal-
increased efficiency in this size separation step.” lenges facing the coal preparation industry, per-
The ability to classify and better utilize ultrafine haps none are as significant as those which relate
particles will increase industry productivity and to waste handling and disposal. This importance
reduce the generation of wastes. Many in the can be attributed to the fact that coal cleaning
industry believe that currently there is a lack of operations produce large volumes of waste that
efficient ultrafine sizing and desliming technolo- must be discarded into refuse piles or impound-
gies for separating at cut sizes of 100 mm and finer. ments. Refuse piles are designed to receive coarse
On-Line Analysis and Control Tremendous particles of waste rock that can be easily
strides have been made in the automation and dewatered. This material is relatively easy to han-
control of coal preparation plants during the past dle and can be safely transported by truck or belt
several decades [62]. The application of on-line haulage systems to the disposal area with little or
sensors together with programmable logic control- no potential for environmental damage. On the
lers (PLCs) has allowed modern plants to operate other hand, the waste contains solid and liquid
more efficiently and to improve safety by reducing components that may present long-term disposal
manpower requirements. On the other hand, the problems depending on the sizes, types and quan-
industry continues to struggle with the determina- tities of minerals present and the conditions under
tion in real-time of the quality of their coal products which the wastes are stored (dry vs wet, loose vs
[63]. Analyzers are commercially available for the compacted, etc.). The factors play a key role in
on-line analysis of many quality parameters for establishing the structural integrity (slope stabil-
coal including ash, sulfur and moisture, although ity, surface water run-off, sediment containment,
the measurement accuracy is often poor due to seepage, etc.) and chemical nature (acid genera-
sampling and calibration issues. On-line analyzers tion, metal dissolution, etc.) of the wastes. Solid
do not currently exist that can be used to determine sediments and dissolved may be transported by
important data such as particle size distributions rainwater where it would pollute streams or
and washability in real-time. Therefore, improving groundwater. Many of these issues can be effec-
automation, control and sensor technologies is a tively managed via proper disposal practices and
key challenge for the industry to overcome. monitoring programs. On the other hand, uncer-
Advances in on-line analysis would enhance the tainties related to the intricate biochemistry and
industry’s ability to maximize the recovered energy complex hydrology of the waste warrant contin-
in a marketable product and minimize the genera- ued investigation to fully assess the potential for
tion of unwanted wastes. negative impacts associated with long-term dis-
posal of coarse waste. Overall, improved waste
Environmental Issues characterization, including better methods to
Environmental compliance in the coal mining define the nature of wastes from coal preparation
industry is strictly controlled by government regu- operations, is considered by many to be a high-
lations. History suggests that these regulations priority need for the coal preparation industry.
were necessary in many cases to ensure uniform Slurry Disposal and Storage The handling
environmental stewardship across the industry as a and disposal of fine slurry waste is widely
Coal Preparation 379

considered to be one of the most difficult chal- into streams and rivers in late 2000. A number of
lenges facing the coal preparation industry. Fine small accidental releases of black water have also
wastes have historically been discarded into been reported at various plant sites.
earthen impoundments for permanent disposal. Several alternatives to impoundments have
An impoundment is an engineered structure been employed by the coal industry in an attempt
consisting of a large volume settling basin formed to avoid any future potential for environmental
behind a man-made dam or embankment. The damage. These alternatives have been examined
waste, which is difficult to dewater, is normally extensively by Gardner et al. [67]. For example,
pumped from the preparation plant thickener to some mines use new modes of slurry disposal
the impoundment as slurry. The slurry contains such as slurry cells and underground injection
water, coal fines, silt, clay and other fine mineral wells. Slurry cells have been used successfully
particulates from the processing plant. In most in some cases (albeit at higher cost), but limita-
cases, the slurry is retained behind a man-made tions associated with maintaining less than
embankment (earthen dam) constructed from 20 acre-ft of settling area make this alternative
compacted refuse material. The impoundment is difficult to apply in all cases. Likewise, the use
designed to have a volume that is sufficiently large of injection wells has raised public concerns about
to ensure that fine particles settle by gravity before groundwater contamination and well water qual-
the clear water at the surface is recycled back to ity [68, 69]. To overcome these problems, various
the plant for reuse. In some cases, chemical addi- types of mechanical solid–liquid separators have
tives may be used to promote settling and to recently been investigated as a means of more
control pH. According to the National Research fully dewatering the fine solids prior to disposal.
Council [65], the coal industry discards 70–90 Notable examples include deep-cone (paste)
million tons of fine wastes each year into existing thickeners and different types of filters (pressure,
impoundments. There are 713 active impound- vacuum, belt press, plate-and-frame, etc.). Deep-
ments and ponds, most of which are located in cone thickeners have been shown to be capable of
central Appalachia. In addition, it has been esti- producing a paste of 45–55% solids as underflow
mated that more than 2.5 billion tons of fine coal in waste coal applications [70]. Ideally, the paste
waste has been discarded in the past into existing can be discarded as a stacked pile, thereby
and abandoned impoundments. avoiding the need for impoundments to handle
Impoundments, like any body of water waste slurry. This technology has already been
contained behind a dam, can pose a safety and demonstrated projects at two mining sites in the
environmental risk if not properly constructed, eastern USA [71, 72]. Despite these advance-
monitored and maintained. Potential problems ments, the Committee on Coal Waste Impound-
include structural failures, seepage/piping, over- ments [65] that recently examined the issue of
topping, acid drainage and black water dis- slurry disposal concluded that “. . .although there
charges. Since the well-known Buffalo Creek are alternatives to disposing of coal waste in
dam failure in 1972, strict engineering standards impoundments, no specific alternative can be
have been mandated by government agencies to recommended in all cases.” Therefore, in the
regulate the design and operation of impound- absence of a preferred disposal method, continued
ments. A detailed report on the design and evalu- development of new and improved processes and
ation of tailings dams for the mining industry has practices for slurry disposal is critically needed in
been published by the EPA [66]. No failures of the coal preparation industry.
impoundment dams or overspills have occurred Process Water Quality The overwhelming
since this legislation was enacted. However, sev- majority of cleaning processes used in coal prep-
eral breakthroughs of slurry into old mine work- aration require large amounts of process water.
ings beneath impoundments have occurred. The Nearly all of the process water is supplied by
most notable was the Martin County incident, thickening units, which settle out ultrafine
which released about 309 million gallons of slurry suspended solids and recycle clarified water back
380 Coal Preparation

into the plant. A small amount of fresh make-up 0.01–0.05% solutions with water to improve perfor-
water from an external source is usually required mance. These chemical are also widely employed in
to satisfy the balance between moisture contents the purification of drinking water and food as well as
of solids entering and exiting with the plant. The in the treatment of sewage, paper-making, oil recov-
clarification and recycling of process water pro- ery, storm water run-off and many other types of
vides an effective means of reducing fresh water industrial wastewater [76]. Finally, coal operations
demands and lowing environmental impacts. On may occasionally use small amounts of dust sup-
the other hand, plant operators are faced with the pressants or freeze inhibitors to address seasonal
difficulty of avoiding the buildup of suspended problems that may arise at the plant site.
solids and dissolved substances in the process Several studies have been carried out in recent
water. The effects of these components on sepa- years to assess the ecotoxicological impacts of
rating performance and plant maintenance (e.g., chemicals used by processing plants. For example,
rusting and scaling) are not well understood. Dete- several studies have raised concerns that seepage
rioration of process water quality is known to from impoundments containing mine influenced
reduce sizing efficiency, lower flotation recovery waters (e.g., residual chemicals, particulates, acid
and increase magnetite losses [73]. Evidence also drainage) may be impacting the population of fresh
exists suggesting that dissolved ions also water muscles/mollusks in tributaries around coal
adversely impact the performance of dewatering mining and processing facilities [77]. Many of
processes [74]. Detailed studies are needed to these issues were addressed in a recent symposium
better understand these problems so that effective hosted by the Nature Conservancy that addressed
solutions can be identified and implemented the coexistence of coal mining and healthy aquatic
before they impede preparation plant operations. ecosystems [78]. Another recent study commis-
There are also growing concerns by the public sioned by the New Zealand Auckland Regional
that chemicals used in the process of coal may be Council (ARC) examined the effects of residual
harmful to the environment. The vast majority of coagulants and flocculants on natural waters
coal is upgraded without being contacted with any [79]. The study concluded that the negative
chemical additives using density-based separation impacts of these chemicals were “. . . low level
processes. On the other hand, finer coal particles and not likely to be significant in relation to other
(typically <0.2 mm) are processed using froth flo- factors which govern the health of aquatic commu-
tation circuits, which require small dosages of nities. The benefit of reduced sediment levels in
reagents known as collectors and frothers. Collec- discharges is considered to outweigh the risk of any
tors consist of oily hydrocarbons, such as diesel fuel, low level impacts attributable to residual floccu-
kerosene and fuel oil, which are insoluble in water lants.” However, the study noted that improper
and coat fine coal particles. If required, dosages are application of these chemicals associated with mis-
typically less than 0.3–2 lbs. of collector per ton of usage or overdosing may create an environmental
fine coal processed. Likewise, frothers are added to risk. Therefore, to avoid these concerns, the devel-
all flotation systems to promote the formation of opment of new technologies that eliminate and/or
small air bubbles and to create a stable froth. significantly reduce the additions of process
Frothers are typically various types of alcohol reagents is desirable as a long-term solution to
and/or polyglycol surfactants [75]. Addition rates this potential problem. In the near-term, it is likely
for frothers are typically less than 10–15 ppm in that chemical manufacturers will continue to
the process water. In addition, plants add water develop low-risk “green” reagents for use in coal
treatment chemicals into their thickening circuits to preparation facilities. Many plants that dispose of
enhance settling rates and improve water clarity. waste slurry via underground injection have suc-
These reagents commonly include various types of cessfully switched to natural flotation collectors,
natural and synthetic coagulants and flocculants. such as blends of canola oil, vegetable oil, soybean
Polymer flocculants, which are added in very oil, etc., as environmentally friendly replacements
small amounts, are diluted in strength to for petroleum-based hydrocarbon collectors [80].
Coal Preparation 381

Air Quality and Dust The US EPA promul- in this region was of sufficient quality to not
gated standards of performance for air quality for require any coal preparation except for simple
all new and modified coal preparation plants crushing and sizing. More recently, however,
under the 1976 Clean Air Act. These New Source increased levels of rock dilution have been noted
Performance Standards (NSPS) address all types for coals mined in this region, largely due to the
of particulate emissions, including fugitive dust, mining of more challenging reserves and the use
which may result from producing, handling, trans- of larger mining equipment that are less selective
porting and storing coal. Operations impacted by [83]. This trend is pushing some coal producers to
such regulation include crushers and breakers, consider coal washing for the first time. In addi-
sizing equipment, cleaning systems, thermal tion, new federal and state clean air quality
dryers, conveying systems, coal storage areas requirements are pressuring utilities and coal
and coal transfer/loading systems that are directly companies to use precombustion cleaning as a
part of the coal preparation facility. For some means of reducing SOx and trace element emis-
processes, such as thermal dryers, opacity less sions [84]. One example is the proposed
than 10% must be maintained [81]. Dusting prob- Springerville power plant in New Mexico, which
lems are not uncommon around material handling has been tentatively approved contingent upon the
transfer points where ultrafine particles have the use of precombustion cleaning to improve the
opportunity to become airborne. These locations quality of their coal feedstock.
may include truck and railcar load-outs as well as There are many challenges in using pre-
conveyor and chute transfer points in and around combustion cleaning to upgrade western coals.
the preparation plant. Many operations utilize The processes traditionally used to wash eastern
water trucks and water sprays in combination coals cannot be readily adopted in the west since
with additions of various chemical and/or crusting water resources are lacking and low rank coals
agents to lower dusting problems. In addition, often disintegrate in water. As a result, current
some operations make use of enclosures, such as industry efforts are focusing on the development
silos and bins, to avoid exposure of coal to the of dry cleaning processes to upgrade these lower
environment. The use of inflatable structures over rank reserves. Dry coal separators, such as pneu-
coal storage piles have also been attempted with matic jigs and air tables, are already finding appli-
limited success [82]. Despite these efforts, fugi- cations at selected mine and utility sites
tive dust still seems to be an issue for many plant [85–87]. Other developing technologies that may
sites, often leading to disputes with the public and be applicable for this purpose include various
creating concerns for worker health. Therefore, types of electrostatic [88, 89] and magnetic sepa-
work is needed to develop better practices and rators [90, 91]. The continued development of
control technologies for reducing dust emissions. these low-cost units is believed to be important
The reconstitution of fines via agglomeration and for upgrading many of the western coal reserves
briquetting technologies would be expected to where water is scarce. The development of auto-
substantially reduce dust emissions, but these pro- mated sorters, which use optical, electromagnetic
cesses are currently too expensive to be practical or X-ray detection, to identify and extract rock
in the coal industry. from coal, also show considerable promise for
dry coal concentration [92]. Optical sorters are
used for separating particles that are liberated at
Future Directions relatively large sizes (>10 mm). These devices
are already being used for separating diamond,
Dry Processing gold, uranium and sulfide ores as well as for sepa-
The low-sulfur coal reserves in the western states rating plastic bottles in the recycling industry. Such
have become the most important supply of domes- devices are particularly useful for preconcentration,
tic fossil fuel in the USA during the past few which can increase throughput and result in energy
decades. Historically, the majority of coal mined savings for other types of solid–solid separators.
382 Coal Preparation

Continued R&D is necessary to improve separation emissions of nitrous oxides by 10–20%, carbon
efficiency (particularly for finer sizes), increase dioxide by 8–12% and mercury by as much as
reliability and lower costs for this new generation 70% [97]. A commercial-size demonstration plant
of coal preparation technology. owned by Evergreen Energy is currently being
operated in Gillette, Wyoming. Specific concerns
Conversion Processes associated with this technology include the high
Coal preparation activities have traditionally been cost of thermal processing as well as problems
limited to include only those processes that related to disposal of process wastewater and
involve physical separations. These processes spontaneous combustion of the treated products.
include unit operations for particle sizing, concen- Another widely discussed coal conversion
tration of organic matter and dewatering/disposal technology is the Encoal Process developed by
of plant products. These processes are generally SMC Mining Company and SGI International.
considered to be inherently benign since they do The technology uses a two-step thermal treatment
not alter the chemical structure of the individual process to produce an enhanced solid coal fuel
particles contained in the coal. Conversion pro- (char) as well as some derived liquid fuel. In the
cesses, which include carbonization, gasification first processing step, the low rank feed coal is
and liquefaction operations, are by this definition heated until a completely dry solid is produced.
not considered to be part of the coal preparation The temperature is then increased in a second
industry. On the other hand, a new generation of processing step to promote decomposition and
coal preparation technology is being developed drive off gases via mild gasification. According
and commercialized that bridges the gap between to published reports [98], the Encoal process gen-
traditional coal cleaning and coal conversion pro- erates about one-half ton of solid fuel and one-half
cesses. These gateway technologies have the barrel of condensed liquids from each ton of feed
potential to reduce transportation costs and coal supplied to the thermal reactor. The products,
improve utilization properties for many of the as alternatives to existing fuel sources, are capable
low rank coals located in the Powder River of lowering sulfur emissions in coal-fired boilers
basin. Under this expanded definition, conversion nationwide [99]. The gaseous products that are
processes that are geared to the production of not condensed into useful liquid are burned to
enhanced solid fuels may also be included under supply thermal energy for the process. A 1000 t/
the coal preparation umbrella. Perhaps the best day demonstration plant was successfully oper-
known of these processes include the K-Fuel, ated between 1992 and 1998 under DOE sponsor-
Encoal and SynCoal processes [93–95]. Some of ship. However, concerns associated with this
these solid fuel production facilities may also process include high treatment costs, excess
produce gaseous or liquid byproducts that are of fines production, dusting problems, wastewater
value in the synthetic fuel market. The upgrading generation and need for coal-char stabilization
of low rank coals, which are abundant in the USA, (to prevent spontaneous combustion).
is believed to be an attractive means of producing Another intriguing conversion processes that
low-emission coals capable of meeting the 2010 couples thermal upgrading with physical cleaning
CAIR standards. This approach is particularly to improve the quality of low rank coals is the
attractive if the process simultaneously produces SynCoal technology. In this process, high mois-
syncrude oil as a byproduct [96]. ture coal is processed through vibrating fluidized
One of the most highly developed and well bed reactors in three sequential stages, two
demonstrated technologies for upgrading low heating stages followed by an inert cooling
rank coal is the K-Fuel process. This technology stage. These reactors remove chemically bound
is a decarboxylation process that uses heat and water, carboxyl groups and volatile sulfur com-
pressure to modify the structure of subbituminous pounds. After thermal upgrading, the coal is put
coals. By driving off moisture and oxygen, the through a deep bed stratifier cleaning process to
process has been shown to be capable of reducing separate the pyrite rich ash from the coal. When
Coal Preparation 383

fed a typical low rank western coal, the SynCoal application of the coal quality impact model:
process can provide a product with a heating value CQIM™. EPRI Report GS-6393, Electric Power
Research Institute, Palo Alto
of up to 12,000 BTU/lb. with moisture and ash 11. Kehoe DB, Parkinson JW, Evans RJ, Levasseur AA
contents as low as 1% and 0.3%, respectively (1990) Environmental and economic benefits of
[100]. A demonstration plant (45 t/h) was success- using clean coals in utility boilers. In: Joint ASME/
fully operated near Western Energy Company’s IEEE power generation conference, Boston, 21–25
Oct 1990
Rosebud coal mine near Colstrip, Montana. 12. Harrison CD, Kehoe DB, O’Connor DC, Stallard GS
Although the plant closed in 2001, the facility (1995) CQE: integrating fuel decisions. In: Fourth
generated more than 2 million tons of products annual clean coal technology conference, Denver,
during its lifespan [101]. Similar to the other mild 5–8 Sept 1995
13. Smith SR (1988) Tennessee Valley Authority’s expe-
conversion technologies, concerns associated rience with switching to improved quality coal. In:
with this process include high treatment costs, Proceedings, conference on reducing electricity gen-
dusting problems and product instability. eration costs by improving coal quality, 5–6 Nov,
Electric Power Research Institute, Palo Alto
14. Bhaskar U (2007) NTPC looking for sites to set up
coal washeries. Wall Street Journal December 3, 2007
15. Glomsrod S, Taoyuan W (2005) Coal cleaning: a
Bibliography
viable strategy for reduced carbon emissions and
improved environment in China? Energy Policy
Primary Literature 33:525–542
1. Fiscor S (2010) U.S. preparation plant census 2010. 16. Blackman A, Wu X (1999) Foreign, direct invest-
Coal Age 9:38–46 ment in China’s power sector: trends, benefits and
2. Kempnich RJ (2003) Coal preparation – a world barriers. Energy Policy 27:695–711
view. In: Proceedings, 20th international coal prepa- 17. Corder WC (1983) For lower power costs, use high
ration exhibition and conference, Lexington, quality coal. In: Proceedings, 3rd European coal uti-
pp 17–39 lization conference, COALTECH, Amsterdam,
3. ASTM (1994) Standard test method for determining 1:221–240
the washability characteristics of coal, Annual Book 18. Scott DH (1995) Coal pulverisers—performance and
of ASTM Standards, 5.05, No. D4371. American safety. IEACR/79, IEA Coal Research, London,
Society for Testing and Materials, Philadelphia pp 1–83
4. Miller I (1998) Use of coal mine methane in coal 19. Raask E (1983) Ash related problems in coal-fired
dryers. Coalbed Methane Outreach Program Techni- boilers. In: Proceedings, engineering foundation con-
cal Options Series, U.S. EPA, Air and Radiation ference on fouling of heat exchange surfaces. White
Draft 6202J, pp 1–4 Haven, Engineering Foundation, New York,
5. Meenen G (2005) Implications of new dewatering 31 Oct–5 Nov 1982, pp 433–460
technologies for the coal industry. In: CAST work- 20. Couch GR (1994) Understanding slagging and foul-
shop, 26–28 July. Virginia Tech, Blacksburg ing during PF combustion. In: International energy
6. Weisenfluh GA, Andrews WA, Hiett JK (1998) Avail- agency, clean coal center, Gemini House, London,
ability of coal resources for the development of coal. IEACR/72, pp 1–118
Kentucky Geological Survey, Report, U.S. Department 21. Hatt R (1995) Correlating the slagging of a utility
of Interior Grant 14–08–0001–A0896, pp 1–32 boiler with coal characteristics. In: Engineering foun-
7. Sites RS, Hostettler KK (1991) Available coal dation conference, Waterville Valley, 16–22 July
resource study of Appalachia 7.5-minute quadrangle 22. Vaninetti GE, Busch CF (1982) Mineral analysis of
Virginia-Kentucky. Virginia Division of Mineral ash data: a utility perspective. J Coal Qual 1(2):22–31
Resources, Publication 118, pp 1–51 23. Jones C (1998) Fuel management: solve common
8. Norton G (1979) The economic role of coal prepara- coalhandling problems. Power 142(6):31–32
tion in the production and utilization of coal for the 24. Smith IM, Sloss LL (1998) PM10/PM2.5 – Emis-
market. Norton-Hambelton Consultants, Ann Arbor, sions and effects. CCC/08. International Energy
pp 1–24 Agency, Coal Research, London, pp 1–75
9. Harrison CD, Hervol JD (1988) Reducing electricity 25. EPA (1997) The benefits and costs of the Clean Air
generation costs by improving coal quality. EPRI Act, 1970 to 1990. U.S. Environmental Protection
Report CS-5713, Electric Power Research Institute, Agency (EPA), pp 1–28
Palo Alto 26. Doherty M (2006) Coal quality impact on unit avail-
10. Davidson PG, Galluzzo NG, Stallard GS, Jennison ability and emissions. American Electric Power Site
KD, Brown RE, Jonas TS, Pavlish JH, Mitas DW, Visit, Asia Pacific Partnership on Clean Development
Lowe DA, Arroyo JA (1990) Development and and Climate, Columbus, 30 Oct–4 Nov, pp 1–10
384 Coal Preparation

27. Couch GR (1995) Power from coal: where to remove technical conference on coal utilization & fuel sys-
impurities? International Energy Agency, Coal tems, Coal Technology Association, Gaithersburg,
Research, London, IEACR/82, pp 1–87 pp 1365–1376
28. Kawatra SK, Eisele TC (2001) Coal desulfurization: 42. Trasande L, Schechter CB, Haynes KA, Landrigan
high efficiency preparation methods. Taylor & PJ (2006) Mental retardation and prenatal methyl-
Francis, New York, pp 1–376 mercury toxicity. Am J Ind Med 49(3):153–158
29. Cavallaro JA, Deurbrouck AW (1977) An overview 43. EPA (2001) Results of Part III of information collec-
of coal preparation. In: Proceedings of the coal desul- tion request. http://www.epa.gov/ttnatw01/combust/
furization: chemical and physical methods sympo- utiltox/utoxpg.html
sium, New Orleans, 23 Mar 1977, pp 35–57 44. Pavlish JH, Sondreal EA, Mann MD, Olson ES,
30. Cavallaro JA, Deurbrouck AW, Killmeyer RP, Galbreath KC, Laudal DL, Benson SA (2003) Status
Fuchs W, Jacobsen PS (1991) Sulfur and ash reduc- review of mercury control options for coal-fired
tion potential and selected chemical and physical power plants. Fuel Process Technol 82:89–165
properties of United States coals. U.S. Department 45. Quick JC, Brill TC, Tabet DE (2002) Mercury in
of Energy, DOE/PETC/TR-91/6, pp 1–309 U.S. coal: observations using the COALQUAL and
31. Couch GR (2003) Coal preparation. International ICR data. Environ Geol 43:247–259
Energy Agency, Clean Coal Center, pp 1–181 46. Alderman K (2007) Market opportunities for coal
32. Obermiller EL, Conrad VB, Lengyel J (1993). Trace preparation. Presentation to the Report Committee
element contents of commercial coals. In: Chow W, of the National Commission on Energy Policy –
Connors KK (eds) Managing Hazardous air pollut- Coal Study, Denver, 17 Sept
ants: state-of-the-art. Electrical Power Research Insti- 47. Tkach BI (1975) On the role of organic matter in
tute/CRC Press, Washington, DC/Boca Raton, concentration of mercury. Geochem Int 11:973–975
pp 73–92 48. Finkelman RB, Stanton RW, Cecil CB, Minkin JA
33. Meij R, te Winkel H (2007) The emissions of heavy (1979) Modes of occurrence of selected trace ele-
metals and persistent organic pollutants from modern ments in several Appalachian coals. Am Chem Soc
coal-fired power stations. Atmos Environ Div Fuel Chem 24(1):236–241
41(40):9262–9272 49. Minken JA, Finkelman RB, Thompson CL, Chao ET,
34. Zubovic P (1966) Physicochemical properties of cer- Ruppert LF, Blank H, Cecil CB (1984) Micro-
tain minor elements as controlling factors in their characterization of arsenic- and selenium-bearing
distribution in coal. In: Coal science. Advances in pyrite in Upper Freeport coal. Scan Electron Microsc
Chemistry, vol 55, American Chemical Society, 4:1515–1524
Washington, DC, pp 221–231 50. Luttrell GH, Venkatraman P, Yoon R-H
35. Swaine DJ (1990) Trace elements in coal. (1998) Removal of hazardous air pollutant precursors
Butterworth, London, pp 1–278 by advanced coal preparation. Coal Prep 19:243–255
36. Fonseca AG, Tumati RP, DeVito MS, Lancet MS, 51. Hucko RE (1984) An overview of U.S. Department
Meenan GF (1993) Trace element partitioning in coal of Energy coal preparation research. In: SME-AIME
utilization systems. In: SME-AIME annual meeting, annual meeting, Los Angeles, 26 Feb–1 Mar, Pre-
Reno, 15–18 Feb, Preprint No. 93–261, pp 1–25 print No. 84–105, pp 1–6
37. Jacobsen PS, Blinn MB, Wan EI, Nowak MA 52. Couch GR (2000) Opportunities for coal preparation
(1992) Role of coal preparation in the precombustion to lower emissions. International Energy Agency
control of hazardous air pollutants. In: Proceedings, Coal Research, CCC/30, London, pp 1–46
8th annual coal preparation, utilization, and environ- 53. von Hippel D, Hayes P (1995) Prospects for energy
mental control contractors conference, PETC, Pitts- efficiency improvements in the Democratic People’s
burg, pp 412–419 Republic of Korea: evaluating and exploring the
38. Ford C, Price A (1982) Evaluation of the effect of options. Nautilus Institute for Security and Sustain-
coal cleaning on fugitive elements. Final report, able Development, Berkeley, pp 1–31
Phase III, BCR Report L-1304 54. Adel GT, Wang D (2005) The assessment of fine coal
39. Akers DJ (1996) Coal cleaning controls HAP emis- cleanability. Int J Coal Prep Util 25(3):129–140
sions. Power Eng 100(6):33–36 55. Honaker RQ (1998) High capacity fine coal cleaning
40. Akers DJ, Raleigh CE, Lebowitz HE, Ekechukwu K, using an enhanced gravity concentrator. Miner Eng
Aluko ME, Arnold BJ, Palmer CA, Kolker A, 11(12):1191–1199
Finkelman RB (1997) HAPs-RxTM: precombustion 56. Honaker RQ, Wang D, Ho K (1996) Application of
removal of hazardous air pollutant precursors. Final the falcon concentrator for fine coal cleaning. Miner
Report, DOE Contract DE-AC22-95PC95153, U.S. Eng 9(11):1143–1156
Department of Energy, pp 1–115 57. Le Roux M, Campbell QP, Watermeyer MS, de Oliveira
41. Palmer CA, Luppens JA, Finkelman RB, Luttrell S (2005) The optimization of an improved method of
GH, Bullock JH (2004) The use of washability stud- fine coal dewatering. Miner Eng 18(9):931–934
ies to predict trace-element reductions during coal 58. Yoon R-H, Eraydin M, Keles S, Luttrell GH
cleaning. In: Proceedings, 29th international (2007) Advanced dewatering methods for energy
Coal Preparation 385

savings in the mineral processing industry. In: SME 74. Yoon RH, Luttrell GH (2008) Advanced dewatering
annual meeting and exhibit, Denver, 25–28 Feb, Pre- systems development. U.S. Department of Energy,
print 07–124, pp 1–5 Final Report for Contract No. AC26-98FT40153,
59. van den Broek JJM (1982) From metallurgical coal 31 July, pp 1–225
tailings to thermal feed. AIME Trans 272:49–52 75. Laskowski JS (2001) Coal flotation and fine coal
60. Mohanty MK, Palit A, Dube B (2002) A comparative utilization. Developments in mineral processing,
evaluation of new fine particle size separation tech- vol 14. Elsevier Science, Amsterdam, pp 1–350
nologies. Miner Eng 15(10):727–736 76. Gregory J, Bolto B (2007) Organic polyelectrolytes
61. Firth B, O’Brian M (2003) Hydrocyclones circuits. in water treatment. Water Res 41(11):2301–2324
Coal Prep 23(4):167–183 77. Beaty B (2007) The Nature Conservancy, Clinch
62. Couch G (1996) Coal preparation: automation and Valley Program, Personal Communication
control, IEAPER/22. International Energy Agency, 78. Nature Conservancy (2007) Conservation manage-
London, pp 1–64 ment of the Clinch and Cumberland river systems: a
63. Belbot M, Vaurvopoulos G, Paschal J (2001) collaborative discussion on coal mining and the
A commercial elemental on-line coal analyzer using aquatic environment. 5–7 September, Abingdon.
pulsed neutrons. In: CAARI 2000: 16th international http://www.cpe.vt.edu/cmrs/agenda.html
conference on the application of accelerators in 79. ARC (2004) Overview of the effects of residual floc-
research and industry, AIP Conference, Denton, culants on aquatic receiving environments. Publica-
pp 1065–1068. https://doi.org/10.1063/1.1395489 tion TP226. Auckland Regional Council (ARC),
64. Anon (2007) Environmental, health, and safety Auckland, pp 1–31
guidelines for coal processing. International Finance 80. Skiles KD (2003) Search for the next generation of
Corporation (IFC), World Bank Group, Good Inter- coal flotation collectors. In: 20th annual international
national Industry Practice (GIIP) Technical Refer- coal preparation exhibition and conference, Lexing-
ence Document, Apr 30, pp 1–22 ton, 29 Apr–1 May, pp 187–204
65. NRC (2002) Coal waste impoundments: risks, 81. EPA (2009) Coal preparation and processing plants
responses and alternatives. National Academies new source performance standards (NSPS), US
Press, Washington, DC, pp 1–244 Environmental Protection Agency (EPA), Fact
66. EPA (1994) Design and evaluation of tailings dams. Sheet. http://www.epa.gov/airtoxics/nsps/coal/cpp_
Technical Document, EPA 530-R-94-038, 59pp. http:// nsps_fr_fs_092509.pdf
nepis.epa.gov/EPA/html/Pubs/pubtitleOSWER.htm 82. Bowling B (2003) Elk Run may use plant – agree-
67. Gardner JS, Houston KE, Campoli A (2003) Alter- ment similar to one made after first dome rupture.
natives analysis for coal slurry impoundments. In: Charleston Daily Mail, Charleston, 29 April 2003
SME annual meeting, 24–26 Feb, Cincinnati, Pre- 83. Bethell PJ (2007) Coal preparation: current status and
print 03–032, pp 1–5 the way ahead. Presentation to the Report Committee
68. Breen T (2007) Lawmakers want study of coal slurry of the National Commission on Energy Policy – Coal
in drinking water. Charleston Daily Mail, Charleston, Study, Denver, 17 Sept 2007
9 Jan 2007 84. Honaker RQ, Luttrell GH, Bratton RC, Saracoglu M,
69. Wilcox L (2007) Coal industry official disputes Thompson E, Richardson V (2007) Dry coal cleaning
research about health of people living near mines. using the FGX separator. In: 24th International coal
The Exponent Telegram, Clarksburg Publishing, preparation conference, Lexington, pp 61–76
Clarksburg, 30 Oct 2007 85. Kelley M, Snoby R (2002) Performance and cost of
70. Parekh BK, Patil D, Honaker RQ, Baczek F (2006) air jigging in the 21st century. In: 19th annual inter-
Improving the densification of fine coal refuse slurries national coal preparation exhibition and conference,
to eliminate slurry ponds. In: Proceedings, XV inter- Lexington, 30 Apr–2 May, pp 175–186
national coal preparation congress, China University 86. Lu M, Yang Y, Li G (2003) The application of com-
of Mining and Technology Press, Beijing, pp 549–554 pound dry separation technology in China. In: 20th
71. Bethell PJ, Gupta BK, Gross R (2008) Replacement annual international coal preparation exhibition and
of belt presses with a deep cone thickener at Lone conference, Lexington, pp 79–95
Mountain Processing. In: 25th annual international 87. Weinstein R, Snoby R (2007) Advances in dry
coal preparation exhibition and conference, Lexing- jigging improves coal quality. Min Eng 59(1):29–34
ton, pp 41–50 88. Yoon R-H, Luttrell GH, Adel GT (1995) POC-scale
72. Henry CD (2008) Fine coal recovery project: Pinna- testing of a dry triboelectrostatic separator for fine
cle mine Smith Branch impoundment. Technical Pre- coal cleaning. In: 11th annual coal preparation, utili-
sentation, Marshall University CBER Summit, zation, and environmental control contractors confer-
pp 1–39. http://www.marshall.edu/cber/summit/ ence, 12–14 July, U.S. Department of Energy,
S2%20-%201%20Henry%20Beard%20Tech.pdf Pittsburgh, pp 65–72
73. Osborne DG (1988) Flotation, agglomeration, and 89. Stencel JM, Schaefer JL, Ban H, Li JB, Neathery JK
selective flocculation. In: Coal preparation technology, (2002) Triboelectrostatic coal cleaning: mineral mat-
vol 1. Graham and Trotman, London, pp 415–477 ter rejection in-line between pulverizers and burners
386 Coal Preparation

at a utility. In: Impact of mineral impurities in solid Department of Energy (DOE), National Energy
fuel combustion. Springer, New York, pp 645–651 Technology Laboratory (NETL) Clean Coal Tech-
90. Oder RR (2005) The co-benefits of pre-combustion nology Compendium, Dec 2003, 2:1–149
separation of mercury at coal-fired power plants. In: 99. DOE (2002) The ENCOAL mild coal gasification
98th annual conference and exhibition. Air & Waste project: a DOE assessment. DOE National Energy
Management Association, Minneapolis, Paper 990 Technology Laboratory, Technical Report, Clean
91. Oder RR (2005) The use of magnetic forces to Energy Document No. 91, Mar 2002, pp 1–40
upgrade low rank coals. In: Upgrading low rank 100. DOE (1997) Clean coal technology: upgrading of
coals symposium, coal & aggregate processing exhi- low-rank coals. U.S. Department of Energy, Topical
bition and conference, Lexington, 2–6 May Report No. 10, Aug 1997, pp 1–28
92. Jong TPR, Mesina MB, Kuilman W (2003) Electro- 101. DOE (2006a) Advanced coal conversion process
magnetic deshaling of coal. Phys Sep Sci Eng demonstration: project performance summary –
12(4):223–236 clean coal demonstration program. Report DOE/FE-
93. Collins KR (2007) EEE: delivering cleaner coal 0496, Mar 2006, pp 1–16
today. In: 2nd annual capital one southcoast energy
conference, New Orleans, 3–5 Dec, pp 1–21
94. Federick JP, Knottnerus BA (1997) Role of the liq- Books and Reviews
uids from coal process in the world energy picture. Arnold BJ, Klima MS, Bethell PJ (2007) Designing the
In: 5th annual clean coal technology conference, coal preparation plant of the future. Society for Mining
Tampa Metallurgy and Exploration (SME), Littleton, pp 1–205
95. Sheldon RW (1997) Rosebud SynCoal partnership: Couch GR (2003) Coal preparation. IEA Clean Coal Cen-
syncoal demonstration technology update. In: 5th ter, London, pp 1–182
annual clean coal technology conference, Tampa Kawatra KS, Eisele TC (2001) Coal desulfurization:
96. Skov ER, England DC, Henneforth JC, Franklin GR highefficiency preparation methods. Taylor & Francis,
(2007) Syncrude and syncoal production by mild New York, pp 1–360
temperature pyrolysis processing of low-rank coals. Kona BB (1997) Coal preparation. Golden jubilee mono-
American Institute of Chemical Engineers, Spring graphs. Allied Publishers, Buffalo, pp 1–221
National Meeting, Houston, Session TB004, 22–26 Leonard JW, Hardinge BC (1991) Coal preparation. Soci-
Apr, pp 1–16 ety for Mining, Metallurgy and Exploration (SME),
97. Wingfield G (2007) Cleaner coal? Forbes, 17 Apr Littleton, pp 1–1131
2007. http://www.forbes.com/2007/04/16/clean-coal- Miller BC (2005) Coal energy systems. Elsevier Aca-
congress-biz-wash-cx_bw_0417coal.html demic, Burlington, pp 1–526
98. DOE (2003) Clean coal technology programs: com- Sanders J (2007) Principles of coal preparation, 4th edn.
pleted projects 2003. Clean Coal Technology, U.S. Australian Coal Preparation Society, Dangar, pp 1–560
 Indirect coal liquefaction (ICL) The conver-
Coal to Liquids Technologies sion of coal to liquids via the intermediacy of
synthesis gas
Marianna Asaro1, Ronald M. Smith2 and Light distillate A distillation cut of low molec-
Burtron H. Davis3 ular weight and low boiling range, obtained
1
SRI International, Menlo Park, CA, USA from refining of hydrocarbonaceous feed-
2
SRI Consulting, Menlo Park, CA, USA stocks, used to produce liquefied petroleum
3
Center for Applied Energy Research, University gas (LPG), gasoline, and naphtha
of Kentucky, Lexington, KY, USA Middle distillate A distillation cut of mid-range
boiling point, obtained from refining hydrocar-
bonaceous feedstocks, containing hydrocar-
Article Outline bons ranging from C5 through about C20 or
C22. When further distilled, the portion of mid-
Glossary dle distillates containing C5 through about C15
Definition of the Subject is often referred to as naphtha, and the portion
Introduction containing C16 through up to C22 is referred to
Survey of Specific Coal Liquefaction as diesel. The naphtha is often distilled further
Technologies to produce gasoline and kerosene/jet fuel, or
Indirect Liquefaction can be used as feed for a naphtha cracker unit to
Future Directions make light olefins. (Less commonly, the gaso-
Bibliography line cut is initially collected along with the light
distillates.)
Glossary Pyrolysis (of coal) A mild gasification process
wherein coal is heated in the absence of oxygen
Coal-to-liquids (CTL) The conversion of coal at sufficiently mildly elevated temperature as
to liquid fuels and/or chemicals to produce a mixture of hydrocarbonaceous
Coprocessing (of coal) The simultaneous con- gases, liquids, and solids suitable for use as
version of coal and waste carbonaceous feed- chemical precursors or fuel
stocks such as petroleum-based residual oil or Synthesis gas (syngas) A mixture of primarily
tar, plastics, or rubbers via once-through direct hydrogen and carbon monoxide produced by
liquefaction into liquid, solid, and gaseous gasification or reforming of hydrocar-
hydrocarbonaceous materials intended primar- bonaceous materials, used to synthesize fuels
ily for use as fuel or chemicals
Direct coal liquefaction (DCL) The conversion Water gas shift reaction The chemically revers-
of coal to liquids via dissolution and/or ible, catalyzed conversion of water and carbon
hydroprocessing, without first gasifying the coal monoxide to hydrogen and carbon dioxide
DOE United States Department of Energy
Fischer-Tropsch reaction The catalytic conver-
sion of synthesis gas to primarily hydrocar- Definition of the Subject
bons, the discovery being credited to Franz
Fischer and Hans Tropsch Like oil and natural gas, coal is a fossil resource
Gasification (of coal) The conversion of coal to and as such is not what first comes to mind when
primarily synthesis gas plus ash at elevated considering sustainable fuel sources. Coal con-
temperature and pressure of oxygen and steam tributes to sustainability as a complement to

© Springer Science+Business Media, LLC, part of Springer Nature 2020 387

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_73
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_73-3
388 Coal to Liquids Technologies

maintaining continuous availability of sufficient homes and industry, fuel gas for town lighting,
fuel to meet growing global demand [1–3]. process gas for industry, coke for the iron industry,
As for other fossil sources, conversion of coal to and fuel for steam locomotives and shipping dur-
transportation fuel is currently more affordable ing the first quarter of the twentieth century.
than conversion of renewable resources such as Although electric power has supplanted coal gas
wind and solar, which are technically far away for public lighting (town gas), the electric power
from availability at even a fraction of the scale industry remains significantly coal based today.
required for sustainability over the coming However, the current availability of natural gas
decades. Given that proved coal reserves in North combined with environmental constraints leads
America alone are currently estimated as capable of the United States to decrease coal use of the gen-
producing more than 840,000,000,000 equivalent eration of electricity.
barrels of oil (the energy equivalent of 32 cubic Coal resources have been used to produce liq-
miles of oil) [4], the abundance of coal is a boon to uid transportation fuels by several process routes
a sustained global liquid fuel supply if used wisely. collectively referred to as coal liquefaction or,
Just as renewable fuel sources will be more generally stated, as coal to liquids (CTL).
benchmarked against fossil sources in terms of Early records of coal conversion to liquid fuel date
economics, so also coal will be benchmarked to 1913 and an extraction process developed by
against renewable fuel sources in terms of envi- Friedrich Bergius in Germany [7]. Bergius treated
ronmental sustainability. Today, modern coal-to- a mixture of pulverized bituminous coal and coal-
liquids (CTL) technology can produce synthetic derived heavy oil with hydrogen, at high temper-
fuels that burn more cleanly than conventionally ature and pressure, to give a liquid product
produced fuels. CTL technology can produce a consisting of heavy oils, middle distillates, gaso-
virtually sulfur-free diesel fuel, much cleaner than line, and gases. Various metal compounds have
conventional diesel – reducing smog and acid been used to catalyze the process. This type of
rain – and can also produce gasoline, jet fuel, or CTL technology is called direct coal liquefaction
chemicals. (DCL).
Most coal use today is for production of power DCL technology involves making a partially
or for steelmaking. Coke plants use coal to pro- refined synthetic crude oil directly from coal,
duce high purity carbons for use in steel manufac- which is then further refined into synthetic gaso-
ture. Combustion of coal provides heat and power line and diesel as well as liquefied petroleum gas
for industrial process operations. Production of (LPG). Raw coal oil is highly aromatic and
liquid fuels and chemicals from coal are lower requires further upgrading. Heavy residues gener-
volume uses; these applications will tend to ated in the process, referred to as asphaltenes, also
increase in importance as the ease of oil recovery include ash from the coal and are either disposed
decreases and its price increases. of or gasified to produce part of the hydrogen
required for the upgrading. The hydrocarbon fuel
products of DCL are similar to hydrocarbon fuels
Introduction derived from crude oil. A slate of partially refined
gasoline-like and diesel-like products, as well as
Coal has been mined and used as a source of propane and butane, are recovered from the syn-
energy for centuries. The first recorded use of thetic crude oil mainly by distillation. Liquid
coal was in China, with indications that the Chi- yields in excess of 70% by weight of the dry,
nese began using coal for heating and smelting mineral matter-free coal feed have been demon-
about 2500 years ago, and most historians believe strated under favorable circumstances. Overall
that coal was used in various parts of the world thermal efficiencies (% calorific value of the
3000–4000 years ago during the Bronze Age input fuel converted to finished products) for
[5]. A system for mining coal was present in modern processes are in the range of 60–70% if
England by 1180 [6]. Coal provided heat for allowance is made for generating losses and other
Coal to Liquids Technologies 389

non-coal energy imports. These processes gener- The main focus of DCL research and develop-
ally have been developed to process development ment continues to be making the process more
unit (PDU) or pilot plant scale. However, no dem- economically competitive with crude oil refining
onstration or commercial-scale plant had yet been by decreasing the cost of production, particularly
built until very recently, primarily because petro- the capital cost of the plant, and the related chal-
leum fuels had been available at lower lenge of finding more efficient ways to deal with
overall cost. tar and solid by-products of the liquefaction.
Development of the DCL approach in the In addition to the price of oil, increasingly strin-
twentieth century proceeded with ambition and gent environmental limitations on the content of
then reserve. DCL was first commercialized in aromatic compounds and sulfur in motor fuels
the 1920s (the Pott-Broche/IG Farben Process). have presented challenge. Fuels generated by DCL
In the Second World War, Germany produced are rich in high-octane aromatics, which is an asset
more than four million tons of fuels by DCL. to engine performance but also an environmental
Plants were also built during this time in the issue. Coal-based processes also require desulfuri-
United Kingdom, France, and Japan. A Japanese- zation. The challenge of altering the product slate to
influenced pilot unit at Fushun Coal Mine was meet environmental regulations is compounded by
operated by the South Manchurian Railway Com- the resultant loss of benefits imparted by aromatics:
pany from 1938 [8] until 1943 and processed in facilitating the liquefaction chemistry itself and
20,000 ton/year of coal. During this same period, then in promoting engine performance when using
the Korean Artificial Petroleum Company oper- the final fuel product. Desulfurization presents its
ated a 100 ton/day plant at its Agochi factory [9]. own performance challenges, because sulfur func-
DCL underwent a new phase of development tions as an engine lubricant.
in the 1970s in response to oil shocks. Major In the 1920s, Franz Fischer and Hans Tropsch
research efforts in the United States during the of the Kaiser Wilhelm Institute developed a cata-
1970s and 1980s in particular involved work by lyst to convert hydrogen and carbon monoxide
Hydrocarbon Research Inc. (HRI, now Hydrocar- obtained from gasification of coal to hydrocarbon
bon Technologies Inc., HTI), cosponsored by the liquids. Because the coal is first gasified before
US Department of Energy (DOE). conversion to liquid fuels, the Fischer-Tropsch
Research on DCL in Japan began in 1974 with (F-T) process for CTL is known as a type of
the Sunshine project and sponsorship by the New indirect coal liquefaction (ICL).
Energy and Industrial Technology Development ICL is a completely different approach from
Organization (NEDO) [10]. Early players DCL for providing liquid fuels from coal. ICL
included Sumitomo Metals and Hasaki, and Nip- processes first convert coal to a gaseous interme-
pon Steel and Kimitsu. The program experienced diate, which is then converted to liquid fuel. Coal
delays in the next decade, presumably resulting gasification technologies rely on incomplete com-
from the drop in oil price. A 150 ton/day pilot bustion, wherein coal is heated with a quantity of
plant was built at Sumitomo Metal Industries air that is insufficient to allow complete combus-
Kashima Steelworks and operated during 1997 tion (formation of carbon dioxide and water). The
and 1998, with a longest run of 1921 h products are condensables and tars, flammable
[11]. Research on DCL in Britain included the gases, and a solid fuel (or waste). Most coal-
two-stage liquid solvent extraction (LSE) and the derived chemicals are made from the tars,
supercritical gas extraction (SGE) processes from although synthesis gas from coal gasification
British Coal. Research on DCL in Germany has also can be used for the production of chemicals
included the single-stage Kohleöl process, as such as olefins, hydrogen, sulfur, methanol, and
improved with the Integrated Gross Oil Refining ammonia. Coal-derived fuels are obtained from
(IGOR+) development by Ruhrkohle AG and the condensables of gasification, which are dis-
Veba Öl AG, as well as the single stage Imhausen tilled and in some cases hydrotreated to give gas-
high-pressure process [12]. oline, diesel, and fuel oil cuts.
390 Coal to Liquids Technologies

Of particular utility is the gasification of coal to Shocks in the price of oil during the 1970s
provide the process intermediate synthesis gas, or prompted renewed efforts; particularly in
syngas, a mixture of primarily CO (carbon mon- South Africa, the United States, Japan, and Brit-
oxide) and H2 (hydrogen). The ratio of H2/CO ain; to promote an alternative to crude oil: a syn-
varies with the source within the range of 0.7–1 thetic crude (syncrude) that would be cost-
for coal. Syngas can be used to generate liquid competitive with oil for refining to transportation
fuels via the aforementioned Fischer-Tropsch fuels. Trade embargoes lasting until the mid-
(F-T) technology, which provides hydrocarbon 1980s effected large-scale application of F-T
fuels that resemble crude oil-derived products. ICL in South Africa, at that point up to 60% of
The F-T processes produce synthetic middle dis- the country’s transportation fuel requirements
tillates that can be used either directly as diesel or being met by coal. With the benchmark price of
in blends with petroleum-derived diesel. oil at US$30 per barrel, Sasol brought two F-T
The CTL Fischer-Tropsch process was first taken ICL plants on line, in 1980 and 1983, with the
forward commercially in the 1930s by the German total capacity of the company’s three plants at
company Ruhrchemie, in conjunction with other 6000 thousand tons per year [14].
partners including Lurgi, to produce gasoline and As a result of a drop in the price of crude oil,
oil. By 1938 there were nine German plants in worldwide developments in indirect coal liquefac-
operation having a total capacity of 660 thousand tion were largely put on hold after the 1980s, with
tons per year [13]. Individual reactors yielded about the exception of those in South Africa where Sasol
5000 L of gasoline/day, with 100 such units used in has been using its F-T Advanced Synthol (SAS)
the first German F-T plants. However, the Bergius reactors since 1989. The ICL approach underwent
process DCL plants were chosen in 1939 when an R&D resurgence in the 1990s, then in response
Germany expanded production during the Second to an increase in the price of oil [15]. All of these
World War, because these were further developed early plants utilized high temperature Fischer-
and could be more readily scaled to larger size, Tropsch (HTFT) technology with an iron-based
eventually processing up to 350 tons of coal and catalyst. In 1993, Sasol successfully proved its
yielding 250,000 L of gasoline per day. low temperature Fischer-Tropsch (LTFT) technol-
By the mid-1940s, natural gas and oil production ogy with an iron-based catalyst to convert coal-
had become more developed and cost-competitive derived syngas to liquid products. The plant is a
with coal, and technology for production of syn- relatively small-scale, 2500 bpd commercial facil-
thetic transportation fuels was not considered eco- ity that was recently converted to use natural gas
nomic after the Second World War. Since then, as feed. Sasol’s newest generation of LTFT uti-
research on CTL processes has waxed and waned lizes high performance cobalt based catalysts in
largely in response to surges and ebbs in the price the Sasol Slurry Phase Distillate (SSPD) process
of oil. However, a perception by some at the time [14], now used in Qatar to convert natural gas to
that reserves of crude oil were limited led to con- middle distillates. Shell has developed a plant in
tinued low-level development of CTL. Sasol pur- Qatar that produces 140,000 bpd liquid products
chased rights from the German companies and in addition to the 120,000 bpd equivalent of light
developed its own CTL process. Construction hydrocarbons.
began in 1950 and an F-T plant came on line in Currently the best option for using coal to
Sasolburg, South Africa, during the mid-1950s produce a high cetane number diesel fuel virtually
using Sasol-Lurgi coal gasifiers and Sasol Synthol free of aromatics and of S and N compounds is the
reactors to produce gasoline and diesel fuels as F-T process coupled with the applicable down-
well as chemical feedstocks from coal. Even so, in stream workup of the crude F-T product [16]. Gas-
the late 1950s and 1960s the huge oil fields of the ification of coal provides syngas, which is
Middle East dominated fuel supply chains to the converted to a range of hydrocarbons. Alterna-
point that further development of CTL technology tively the syngas can be converted to liquid spe-
worldwide essentially disappeared. cialty [17, 18] or commodity chemicals including
Coal to Liquids Technologies 391

methanol or dimethylether (DME), either or both fuels and chemicals, including demonstration at
of which also can be used as automotive fuels. the DOE’s Alternative Fuels Development Unit
Methanol can be used alone, or preferably as a (AFDU) at LaPorte, Texas [24]. This work led to
bend with conventional gasoline, in a spark igni- the development of a slurry phase process by East-
tion internal combustion engine (but not in a die- man Chemical Co. to produce methanol and then
sel engine, because methanol’s cetane number is acetic acid or acetic anhydride. The Eastman facil-
too low). A separate fuel distribution network for ity started up in 1983 and has undergone two
methanol also would be required as well as con- expansions. Today the plant produces 2.5 million
sideration of toxicity (and preventing people from tons/year of acetic acid and acetic anhydride.
drinking it as a replacement for ethanol). In con- In Japan, work on DME was intensified in the
trast to methanol, DME can be used as a fuel or 1990s by NKK Corporation [25]. NKK first
fuel additive for automotive transportation in a researched the one-step slurry DME synthesis at
diesel engine. bench scale and then teamed with the Center for
DME has an inherent combination of both oxy- Coal Utilization, Taiheiyo Coal Mining, and
genate functionality, for antiknock characteristics, Sumitomo Metal Industries to pilot a 5 ton/day
and saturated hydrocarbon functionality, for fuel DME plant with funding from METI (the Ministry
energy content. DME also imparts improved cold- of Economy, Trade, and Industry). The long-range
start characteristics as an additive to methanol fuel plan at that time was described as producing DME at
and has almost twice the energy content of metha- the site of coal mining and then transporting it to
nol on a volume basis. DME can be used as a utility user sites. In terms of fuel value per unit weight, this
fuel or can serve as a blending component or would be much more efficient than transporting coal
replacement fuel for diesel engines. DME is also itself.
an intermediate in production of other fuel, as in Substantial research and development efforts
ExxonMobil’s methane-to-gasoline (MTG) pro- have been devoted to optimizing product yields
cess [19]. DME is a potential feedstock for novel and process efficiencies in conversion of syngas to
oxygenated fuels [20] and is an alternative source fuels and chemicals. Mechanistic and kinetic stud-
of lower olefins such as propylene [21]. Using low ies have been conducted to interpret the funda-
CO content syngas, as derived from coal, a one- mentals of specific conversion processes.
step DME process is more efficient and achieves Methods and hardware to impart temperature con-
higher syngas conversion than using two or three trol and stability in conversion reactors are under
separate unit operations. DME can be produced continuous development because of the large
with methanol in a dedicated plant, or it can be excess heat of reaction. In the case of syngas
used as the fuels production arm of a cogeneration conversions, this research has included discovery
plant to produce power and clean fuels (including of catalytic systems having optimized formula-
even cooking fuel [22]) from coal [23]. tions containing metals with high activity in com-
Primary early research toward one-step conver- bination with cocatalysts or additives that
sion of syngas to methanol in the liquid phase was improve activity and selectivity for the desired
performed by ChemSystems Inc., by the Univer- type of product slate. Reactor design and engi-
sity of Ohio at Akron and the University of neering have proceeded in tandem with the cata-
Missouri-Columbia, and by Air Products and lyst developments in syngas conversions.
Chemicals Inc. (APCI). This development and
demonstration work in the United States was
performed predominantly in the 1980s and 1990s. Survey of Specific Coal Liquefaction
Air Products’ work with the Alternative Fuels Pro- Technologies
gram of the US DOE additionally focused on dual
catalysts for one-step DME production from coal- A primary requirement of all coal liquefaction
derived syngas and led to demonstration of pro- processes is that the fuel value inherent in the
cesses for conversion of coal-derived syngas to coal feedstock must be increased to be more
392 Coal to Liquids Technologies

similar to that of conventional liquid fuels. The high-temperature coking process are fuel gas,
ratio of elemental hydrogen to carbon, H/C, or crude tar and light oils such as benzene, toluene,
even just simply the content of hydrogen itself, and xylene (BTX), naphtha, and ammonia. Most
is important to determining the fit between feed- coal- derived chemicals today are obtained from
stock and process options. Coal has a hydrogen the processing of tars. Coke manufacturers burn
content of about 5% whereas that of refined crude coal tar in coke ovens, refine it to tar-based
petroleum-based transportation fuel is in the products, or sell it to coal tar distillers. Tar-derived
range of 12–15%. Fuel derived from petroleum products include ammonium sulfate, sulfuric acid,
has an average H/C ratio of about 2 whereas that phenols, cresols, BTX, naphthalene, pyridine,
of coal itself ranges from about 0.7–1. phenanthrene, anthracene, creosote, road tars,
Table 1 shows the full range of process options roofing pitches, and pipeline enamels. Today, syn-
that have been studied for converting coal to liq- thesis gas from coal gasification also can be used
uids [26]. These approaches include pyrolysis, for the production of petrochemicals such as
direct liquefaction, coprocessing, and indirect BTX, sulfur, methanol, and ammonia as well as
liquefaction. hydrogen.
The Liquids from Coal (LFC™) process uses
Pyrolysis mild coal gasification to upgrade low-rank coals
In pyrolysis processes, the H/C ratio of coal is to two fuels: a stable, low-sulfur, high-BTU solid
raised by rejecting carbon. The liquid products fuel similar in composition and handling proper-
have little value, but pyrolysis is used on a mas- ties to bituminous coal (referred to as process-
sive scale to turn coal into coke for metallurgy, derived fuel, PDFTM) and a low-sulfur industrial
especially steelmaking. The coking process fuel oil (referred to as coal-derived liquid, or
involves heating coal in closed vessels to very CDL™) [27, 28]. The process was originally
high temperatures (up to 2000
C) in the absence developed by SGI International and then taken
of oxygen. Without oxygen present to cause com- further by the ENCOAL Corporation. Develop-
bustion, at these high temperatures, the coal is ment was about 50% financially supported by $90
broken down into lighter molecular weight sub- million from the US DOE.
stances that leave the vessel in the gaseous state. A PDU was constructed in 1986 and then a
A hard, porous residue containing mostly carbon 1000 ton/day demonstration unit was constructed
and inorganic ash is left behind. The amount of and operated for over 4 years near Gillette, WY, in
volatiles is typically 25–30 weight% of the coal. the mid 1990s. Later configurations included a
Pyrolysis is also used to make chemicals. vibrating fluidized bed (VFB) as part of a deacti-
In coking, the coal beds are deep and the tem- vation loop for solid product, where quenched
peratures are kept very high to promote solids are partially fluidized and exposed to oxi-
resolidification of the decomposing coal. When dative deactivation to reduce the tendency for
making chemicals, the coal beds are thinner and spontaneous ignition. The residence time, oxygen
temperatures are lower, than for a more complete content, and temperature of the gas stream are
gasification, near the melting or the decomposi- selected to deactivate the coal within the VFB
tion temperature of coal to promote production of unit. Hydration further stabilizes the solid PDF
liquids and gases. In both of these processes, fuel product, and after treatment in the VFB system
gas containing carbon and hydrogen is formed in the solids are cooled and a controlled amount of
significant quantities and high quality as mea- water is added to rehydrate the PDF to near its
sured by the gas heating value (BTU/scf). Where equilibrium moisture content. The CDL portion of
possible, this gas has been used for process heat the product is condensed. The technology appears
but any excess gas also could be sold. to be moving forward, with the transformation of
Chemical products can be obtained from ENCOAL into the Saskatchewan-based company
high-temperature coking by-products, synthesis NuCoal, which has memos of understanding with
gas, and coal liquids. The by-products of the several potential partners including Sinopec.
Coal to Liquids Technologies 393

Coal to Liquids Technologies, Table 1 Coal to liquids processes

Mild Two-stage Indirect
gasification/ Single-stage direct direct Indirect liquefaction,
pyrolysis liquefaction liquefaction Coprocessing liquefaction, F-T other
TOSCOAL – H-coal – HRI, US Catalytic two- MITI mark I and Sasol – Methanol-to
Tosco, US Kohleöl/ stage mark II – Japan South Africa gasoline
Char-oil IGOR+  Ruhrkohle liquefaction, Cherry P – Gasel – Axens, (MTG) –
energy & Veba Öl, Germany CSTL – Osaka gas, Japan France ExxonMobil,
development, (today BP, UK) HTI/DOE, US Solvolysis GTL.F1 (Statoil, US
COED – Solvent refined coal H-coal CSTL – –Mitsubishi, PetroSA, Lurgi), Direct DME
FMC, US (SRC-I and SRC-II) – Axens, France Japan Switzerland –JFE, Japan
Liquids from Gulf oil, US Liquid solvent Mobil – US Syntroleum – US Direct DME
coal, LFC – Exxon donor solvent extraction, Pyrosol Rentech – US –NEDO,
Encoal, US EDS – US LSE – British –Saabergwerke, Shell middle Japan
EERC/ NEDOL – NEDO, coal, UK Germany distillates, LPDME – Air
AMAX – Japan Brown coal Chevron – US (SMDS) – Royal products, US
Univ. Imhausen high- liquefaction, Lummus crest – Dutch Shell, Topsøe DME
N. Dakota, pressure – Germany BCL – NEDO, US Netherlands –Haldor
US Conoco zinc Japan Alberta research BP, UK Topsøe,
Lurgi- chloride – US Chevron coal council, ARC ExxonMobil, US Denmark
Ruhrgas – Suplex – Sasol, liquefaction, –Canadian ConocoPhil-lips,
Germany South Africa CCLP – US energy dev., US
Pyrolysis of Lummus Canada
coal – Inst. Of integrated two- CANMET
gas Technol., stage AOSTRA –
US liquefaction, Canada
Twin screw – ITSL –US Rheinbraun –
Coal Technol. Kerr-McGee Germany
Corp., US ITSL –US Technical
Amoco University
CC-TSL – US Clausthal,
Pyrosol – TUC – Germany
Saabergwerke, UOP slurry-
Germany catalyzed – US
Mitsubishi HTI – Us
Solvolysis –
MHI, Japan
Consol
synthetic fuel,
CSF –
Consolidated
coal, US
Supercritical
gas extraction,
SGE – British
coal, UK

Direct Liquefaction with the exact cuts depending on product specifi-

In DCL, the H/C ratio of coal is raised by adding cations. The liquid products are then refined into
hydrogen. The objective of hydrogenation is to diesel, gasoline, and other petroleum products.
maximize the yield of distillate fractions that can The traditional mechanistic view has been that
be subsequently converted into fuels. Distillate upon heating inherently weak bonds in coal are
materials comprise the naphtha fraction, typically broken and that the resulting and then hydrogen is
boiling C5–215
C, plus the middle distillate frac- rapidly transferred from the solvent to cap the
tion typically boiling between 150
C and 370
C, resultant free radicals. Rapid hydrogenation is
394 Coal to Liquids Technologies

indeed critical to liquid yield, preventing retro- Coal to Liquids Technologies, Table 2 Typical operat-
gressive reactions leading to formation of solid ing conditions for direct coal liquefaction
char. While this picture is largely true, in-depth Typical operating
research showed that it is insufficient. Concurrent Operating parameter conditions
with this pathway are solvent-mediated Temperature,
F 750–850
hydrogenolyses of strong C-C bonds – bonds too Pressure, psia 1500–3000
Residence time, hr 0.5
strong undergo simple thermolysis under reaction
Solvent to coal ratio, lb./ 1.5/2
conditions – engendered by addition of a free lb
hydrogen atom or a bimolecular H-transfer from
solvent species and their radicals [29]. Inclusion
of these pathways helps rationalize many other-
wise inexplicable phenomena in coal liquefaction, and hydrogen consumption of 5–7%wt. When an
and also helps identify conditions that maximize external catalyst is used, it is typically some com-
utilization of hydrogen for liquids production bination of cobalt, nickel, and molybdenum on a
rather than produce the less desirable light gases. solid acid support such as silica-alumina [30].
Hydrogen is also critical in the DCL process to DCL process concepts have been based on a
reduce oxygen, sulfur, and nitrogen in the coal variety of operating conditions and catalysts and
feedstock. These elements are removed from the can be performed in a single stage or two stages.
crude liquid product as water (H2O), hydrogen In a typical single stage process, coal is contacted
sulfide (H2S), and ammonia (NH3), respectively. with hydrogen or with hydrogenated solvent at
Oxygenates are reduced to hydrocarbons, because ca. 450
C. The yield of fuel oil is about three
their oxygen content would otherwise lower the barrels of oil/ton of coal with bituminous coals.
fuel value. The nitrogen and sulfur compounds are The light oil fraction is recycled to prepare the
removed because they would otherwise poison the coal slurry feed. The product is difficult to refine
cracking catalysts in refining operations down- because of its high aromaticity and nitrogen con-
stream of the DCL unit. DCL products are only tent. Single stage liquefaction is not as effective
partially refined at the DCL plant. They must be for sub-bituminous coals as for higher rank coal.
further refined into finished liquid fuel products at Most current DCL processes have settled on a
conventional refineries, where additional H2 is multistage approach. Multistage DCL typically
added to bring the H/C ratio up to about 2 for the uses dispersed catalysts in the liquefaction stage
fuel products. followed by use of supported catalysts in subse-
Many factors affect the rate and extent of coal quent upgrading stages, with different pressures
liquefaction, including temperature, hydrogen and temperatures for the different stages. The
partial pressure, residence time, coal type and technologies were demonstrated in several labo-
properties, solvent properties, solvent to coal ratory and pilot plants and the products were more
ratio, ash composition, and catalyst. In general, easily refined than many crude oils. There are
liquefaction is promoted by increasing the tem- many variations of two-stage DCL processes,
perature, hydrogen partial pressure, and residence including use of catalyst in the two stages (both
time. However, if the temperature is too high, gas stages being catalyzed, versus thermal and cata-
yield is increased and coking can occur. Solvent to lyzed stages), distillation between stages, whether
coal ratio is important. If the ratio is too low, there solids are separated between the two stages or
will be insufficient hydrogen transfer activity. If only after the second stage, recycle of ashy bot-
the ratio is high, a larger reactor will be necessary toms, and the operating hydrogen balance.
to provide the required residence time. Typical A block flow diagram (BFD) for a typical
operating conditions are listed in Table 2. direct coal liquefaction plant using coal slurry
For the most highly developed processes, coal hydrogenation is shown in Fig. 1. Coal is ground
conversion can be as high as 90% on a mineral ash and slurried with a process-derived solvent, mixed
free basis, with a C5+ distillate yield up to 75% with a hydrogen rich gas stream, preheated, and
Coal to Liquids Technologies 395

H2S,NH2,CO2

Recycled H2 Gas Recovery Methane & Ethane

Make-up H2
Treatment LPG

Coal + Gasoline
Coal Hydrotreating Refining
Catalyst Conversion Unit Diesel Fuel

H-Donor
Slurry
Slurry Heavy Vacuum
Fractionation Gas Oil

Deashed Oil Solvent Gasifier Ash Reject

Deashing

Uncoverted
Coal

Coal to Liquids Technologies, Fig. 1 Block flow diagram of a typical direct coal liquefaction process Ref. [31]

sent to a two-stage liquefaction reactor system Hydrocarbon Technology Inc. (HTI) that con-
where the organic fraction of the coal dissolves sumed three metric tons of coal per day. Among
in the solvent and the dissolved fragments react the many liquefaction processes developed and
with hydrogen to form liquid and gaseous prod- tested at pilot plant scale have been Kohleöl
ucts. Sulfur in the coal is converted to hydrogen (by Ruhrkohle, Germany); Exxon Donor Solvent
sulfide, nitrogen to ammonia, and oxygen to (by Exxon, now ExxonMobil, US); Solvent
water. The reactor products go to vapor/liquid Refined Coal-I (SRC-I) and SRC-II (by Gulf Oil,
separation. The gas is cleaned, and after removal US); H-Coal (by HRI, now HTI, and DOE, US);
of a purge stream to prevent buildup of inert Catalytic Two-Stage Liquefaction (CTSL, by HRI
materials, the gas is mixed with fresh hydrogen and DOE, US); Liquid Solvent Extraction (LSE,
and recycled. The liquid is sent to vacuum frac- by British Coal Corporation, UK); Brown Coal
tionation for recovery of distillates. Heavy gas oil Liquefaction (BCL, by NEDO, Japan); and
is recycled as process solvent and vacuum bot- NEDOL (by NEDO, Japan). Only the CTSL,
toms are gasified for hydrogen production. LSE, BCL, and NEDOL processes continued in
Ash from the gasifier is sent to disposal. Heavy development at pilot plant scale beyond the late
direct liquefaction products contain polynuclear 1980s. Between 1975 and 2000, the United States
aromatics that are carcinogenic. This problem can alone invested $3.6 billion in direct liquefaction
be solved, however, by recycling to extinction all technologies; some are now licensed to
material boiling above the desired product end- China [31].
point. A recent development, shown in Fig. 1, is to H-Coal was an early single-stage DCL process
put the ash residue and carbon from recycled piloted by HRI and Ashland Synthetic Fuels, Inc.
solvent directly through the gasifier to produce in a 600 ton/day pilot plant adjacent to Ashland’s
hydrogen (and power). refinery in Cattlesburg, Kentucky [32]. The
The longest experience base with DCL in the process consists of slurry preparation followed by
United States was a PDU at the R&D facility of catalytic hydrogenation/hydrocracking at a single
396 Coal to Liquids Technologies

temperature and pressure, 450
C and 2200 psia, first stage of the CTSL process is to promote
respectively, in an ebullating bed reactor. The hydrogenation of the solvent, stabilization of
ebullating bed is a backmixed three-phase isother- the primary liquefaction products, and hydroge-
mal reactor that keeps catalyst in constant motion. nation of the primary and recycle resid. In the
It uses the upward flow of liquid to expand the second stage, the catalyst promotes heteroatom
catalyst bed and to distribute liquid, gas, and cata- removal and thus product quality improvement,
lyst evenly across the reactor. The reactor has the conversion of resid to distillate, and secondary
characteristics of a uniform-temperature similar to conversion to lighter products. The catalyst also
continuous stirred tank reactor. aids in avoiding dehydrogenation.
The H-Coal process was a starting basis for In the first ebullating bed reactor, coal is
subsequent process development sponsored by dissolved with minimal catalysis at about 400
C
the US DOE over about three decades time. for bituminous coal (or 435–460
C for sub-
A principal focus of this long-term project was bituminous coal) and 170 bar. The catalyst for
R&D of catalysts that were tolerant of, and this first reaction is typically nickel-molybdenum
would not be rapidly deactivated by, coal- on alumina and aids in re-hydrogenating the
derived mineral matter in the reactor. Typical recycled process solvent. The resultant heavy
early catalysts showed rapid deactivation coal liquids are fed to a second ebullating bed
because of coking and metals laydown with reactor and hydrogenated at about 440
C and
loss of surface area. A later process version 170 bar with a catalyst such as that used in the
sought to address this deactivation problem by first reactor, or a high-activity iron catalyst. After
using two coupled reactor stages, one operating separation and depressurization, the product of the
at 400
C and the other operating at 420–425
C. second reactor is sent to a distillation column at
The catalyst was also changed from cobalt/ atmospheric pressure, where distillate up to
molybdenum/alumina to nickel/molybdenum 400
C is removed, with yield from coal of up to
on bimodal alumina. This Catalytic Two-Stage 65%. The distillation residue is unconverted coal
Liquefaction (CTSL) process is shown in a block and asphalt, which are typically gasified for cap-
flow diagram in Fig. 2. The role of catalyst in the tive fuel value or feed hydrogen.

Hydrogen Recycle
Gases
Fresh
Hydrogen Light
Distillate

Middle
Slurry First-Stage Second-Stage Distillate
Mixing Ebullating Ebullating
Reactor Reactor Heavy
Distillate
Coal

Preheating Solids
Separation

Recycle Solvent
Residues

Coal to Liquids Technologies, Fig. 2 Process flow schematic of two-stage CTSL direct liquefaction process Ref. [33]
Coal to Liquids Technologies 397

The CTSL process was further developed over stage liquefaction, utilizing ebullating bed cata-
almost 15 years in a pilot plant at the Wilsonville lytic reactors in both stages. The ebullating bed
liquefaction test facility. Beginning in 1973, a allows use of the substantial heat of reaction to
6 ton/day coal liquefaction plant was built by the preheat coal slurry feeds, and this improves pro-
Edison Electric Institute (EEI) and Southern Com- cess thermal efficiency. The converted and refined
pany Services, and the US DOE became the pri- products are separated at high pressure and tem-
mary sponsor in 1976. Amoco Oil Co. (now BP) perature into lighter oils and gases, and heavy oils
joined the project in 1984. and solids. Following pressure letdown and heat
exchange, the lower boiling oils are fractionated
Developments at the Wilsonville Facility in the while the heavier oils are separated from the
United States unconverted coal and ash and recycled as the
Because of the role of Wilsonville in consistently slurry solvent for the coal. Initial work indicated
demonstrating success or failure of a large number that distillate yields as high as 78% could be
of concepts, particular attention has been paid to obtained by operating the first stage at low sever-
the results from this facility [34–36]. The plant ity (399
C) and by using a large pore bimodal
began operation in 1974 in a single-stage mode Ni-Mo catalyst. Significant improvement in dis-
referred to as Solvent Refined Coal (SRC-II, tillate production can be achieved but again at the
which made liquid fuel as opposed to SRC-I cost of increased hydrogen consumption.
which made a solid fuel from coal) but evolved Also explored were Reconfigured Integrated
to a two-stage operation utilizing two ebullating Two-Stage Liquefaction (RITSL), where solvent
bed catalytic reactors. Initial efforts in two-stage de-ashing was practiced after the hydrocracking
liquefaction focused on catalytic upgrading of the system, and Close-Coupled Integrated Two-Stage
thermal products, referred to as Nonintegrated Liquefaction (CC-ITSL), where two reactors
Two-Stage Liquefaction (NTSL). This configura- (thermal/catalytic) were linked directly without any
tion was termed nonintegrated because the coal- intervening processing steps. Incremental improve-
derived resid hydrocracking step did not interact ments in distillate yield and selectivity were realized
with the primary thermal part of the plant. The by changing the process configuration, but at the
NTSL concept proved excessively inefficient expense of increased hydrogen consumption.
because of the high hydrogen consumption asso- The Wilsonville facility was closed in 1992
ciated with the thermal part of the operation. and was decommissioned shortly thereafter, but
Coal throughput, expressed as space velocity per development work funded by the DOE continued
unit reactor volume, was substantially improved in at the Lawrenceville, NJ, facility. The process has
going from NTSL to Integrated Two-Stage Lique- now evolved into a generic composite of much of
faction (ITSL) mode. In ITSL, a short contact time the liquefaction development work funded by the
thermal reactor is closely coupled to an ebullating DOE in the 1980s and 1990s. Most recently, a
bed catalytic reactor where hydrocracking occurs. close-coupled configuration has been adopted in
The thermal resid produced in the ITSL at short which both stages use an active supported cata-
contact times is more reactive toward expanded- lyst, and HTI has included an in-line hydrotreater
bed hydrocracking, which permits operation of the after the second-stage reactor to improve product
hydrocracker at lower temperature and minimizes quality. In 1997, with support from the DOE, HTI
gas make. Minimizing formation of light hydrocar- signed a 2-year agreement with the Chinese Coal
bon gases in turn would be expected to reduce the Research Institute (CCRI) to carry out a feasibility
rate of catalyst coking, improve hydrogen utilization study for a direct coal liquefaction plant using
efficiency, and increase selectivity for higher distil- Shenhua coal [37].
late. A 35% increased yield from C5+ distillation The Kerr-McGee critical solvent de-ashing
was observed. (CSD) technology (also termed the Residual Oil
From 1985 to 1992, work at Wilsonville Solvent Extraction, or ROSE, process) was also
focused on development of the coupled two tested at Wilsonville, resulting in the development
398 Coal to Liquids Technologies

of alternative methods for solids removal from original hot separator pass to a vacuum distillation
primary liquefaction products. Other DCL process column to recover distillable liquids. These are
developments in the United States have included added to the hydrotreating reactor feed and are
the single-stage Exxon Donor Solvent (EDS) pro- subsequently largely recycled as solvent. The vac-
cess, and two-stage processes have included uum column bottoms consist of pitch, mineral mat-
Lummus’s Integrated Two-Stage Liquefaction ter, unreacted coal and catalyst, which in
(ITSL); Kerr-McGee’s ITSL, Chevron Coal Lique- commercial operation would be used as a gasifier
faction (CCLP), Consol Synthetic Fuel (CSF); and feedstock for hydrogen production. Greater than
Amoco’s (now BP’s) Close-Coupled Two-Stage 90% conversion can be obtained with bituminous
Liquefaction (CC-TSL) process. coals, with liquid yields in the range 50–60% on a
In 1996, Sandia National Labs published dry ash-free coal basis. The product distribution is
results of its study on the efficacy of hydrogena- about 19% C1–C4 hydrocarbons, 25% light oil (C5
tion for removing nitrogen and sulfur from the through 200
C bp), 33% medium oil
CTSL crude product [38]. The kerosene fraction (200–325
C bp), and 22% unreacted coal and pitch.
from HTI’s first proof-of-concept run was used, The LSE process was developed more fully
and denitrogenation and desulfurization were than other European DCL processes, by British
achieved under relatively mild conditions. Pro- Coal Corporation between 1973 and 1995 [12]. In
cessing at the highest severity condition (388
C, the LSE process, coal is slurried with a process-
103 psig H2, 1.5 g/h/g-cat) decreased nitrogen derived recycle solvent, preheated and passed to a
from 645 ppm in the feed to 7.9 ppm in the noncatalytic digestion step, which consists of two
hydrotreated product. Sulfur content of 239 ppm or more continuous-stirred tank reactors. These
in the feed was decreased by hydrotreating to less reactors operate in series at a temperature of
than 20 ppm. 410–440
C and 10–20 bar. The solvent acts as a
In the early 1990s, HTI was engaged in DOE- hydrogen donor, transferring up to 2% by weight
sponsored research on coprocessing of coal and of hydrogen to the coal. The digester product is
plastics waste using DCL. partially cooled, filtered to remove unreacted coal
and ash, and dried. The filtered coal extract is
DCL Developments in Europe distilled to recover the light oil wash solvent and
The Kohleöl DCL process [39] is based on technol- is then preheated, mixed with H2, and sent to one
ogy used at commercial scale in Germany, until after or more ebullating bed reactors in series. There is
the Second World War, and then improved with the no interstage separation and the reactors operate at
advent of the Integrated Gross Oil Refining 400–440
C and about 200 bar. The cooled,
(IGOR+) process [40]. Coal is slurried with a depressurized reactor products are distilled with
process-derived recycle solvent and a “red mud” cut-point adjusted to maintain solvent balance,
disposable iron catalyst. The pressurized and pre- resulting in a product that typically boils below
heated slurry is sent to an up-flow tubular reactor 300
C. The column bottoms are routed partially
operated at about 470
C and 300 bar H2. Products to a vacuum distillation column, used to control
from the top of the reactor pass to a hot separator, the the level of pitch in the recycle solvent. The over-
overheads from which are hydrotreated at heads from this column are recombined with the
350–420
C in a fixed-bed reactor maintained at main atmospheric column bottoms stream and
the same pressure as the main reactor. The recycled as the solvent to the slurrying step. The
hydrotreated products are depressurized and cooled product distribution is about 15% C1–C4 hydro-
in two stages. The liquid product from the first stage carbons, 50% light oil (C5 through 300
C bp),
is recycled for use as solvent in the slurry step. The 50% solvent surplus (300–450
C bp), 24% filter
liquid product from the second stage is routed to an cake organics, and less than 1% pitch. A 2.5
atmospheric distillation column to give light oil ton/day pilot plant was built and operated for
(C5–200
C bp) and a medium oil 4 years at Point of Ayr, North Wales, but has
(200–325
C bp) product. The bottoms from the since been decommissioned.
Coal to Liquids Technologies 399

DCL Developments in Japan processing a total of about 60,000 tons of coal.

The Sunshine project pursued in Japan examined Operations ceased in October 1990. In the 1990s,
three approaches to direct liquefaction of bitumi- NEDO worked with Indonesia on developing
nous coal: solvolysis, solvent extraction, and technology for use with that country’s brown
direct hydrogenation of coal. From these studies, coal. Also in Japan, Mitsubishi Heavy Industries
NEDO found that increasing catalyst activity cor- (MHI) researched a two-stage DCL process
related with increased liquid yield; that hydroge- referred to as Mitsubishi Solvolysis.
nation was most effective for liquefaction at The development of DCL is depicted from a
pressures in the 150–250 atm range and tempera- historical process economic perspective in Fig. 4,
tures in the 400–450
C range and that increased which shows the cost that the oil product would
solvent viscosity aided production of light oil as need to sell for to make specific DCL processes
fuel. The approaches were consolidated over a economically competitive. As the process tech-
time span of 18 years, with demonstration of the nology advanced the liquids yield from bitumi-
resultant NEDOL process completed in 2000. The nous coals increased from 2.7 bbl/ton to 4.3 bbl/
process uses 170 atm H2 in the liquefaction reac- ton. The increase in the yield of liquids are partly
tor at 450
C. A process flow diagram of the due to the fact that the later process designs used
NEDOL process is shown in Fig. 3 [11]. steam reforming of natural gas to provide hydro-
Like H-Coal, the NEDOL process is single gen instead of gasifying a portion of the coal. The
stage, meaning that all liquefaction reactors are costs decreased concomitantly, but today direct
operated at a single temperature and pressure. The coal liquefaction still remains highly capital-
NEDOL process has been further developed, intensive. A major consideration in coal liquefac-
however, and is said to have low severity condi- tion is the cost of money. For example, when the
tions and high distillate yield comparable to those interest rates of about 20% in the late 1970s
of the CC-ITSL and IGOR+ processes. decreased to about 8%, the cost of the liquid
NEDO also developed the BCL process, in the product needed to make a profit decreased to
early 1980s, to use brown coal (lower rank, wetter nearly 50% of the 1970s price [42].
coal), to the scale of 50 tons/day in a pilot-plant Overcoming the techno-economic difficulties
constructed at Morwell in Victoria, Australia of DCL could enable a lower cost overall CTL
[41]. It was operated during 1985–1990, process than ICL, given that DCL does not require

Coal to Liquids Technologies, Fig. 3 Flowchart for the NEDOL process for direct liquefaction of bituminous coal
400 Coal to Liquids Technologies

1999 $/Barrel Equiv. Crude

Coal to Liquids 80.00

Required Selling Price

Technologies, O&M
Fig. 4 Relative economics 60.00 Coal
for representative DCL
Capital Related
processes Ref. [36] (O&M 40.00
is operations and
maintenance) 20.00

0.00
H-Coal ITSL CMSL

the gasification step. DCL has been largely dor- Shanghai Electric Group have set up a coal lique-
mant for over a decade in the United States, faction research center in Shanghai. The center will
Europe, and Japan since about 2000. Direct lique- mainly explore and develop direct and indirect coal
faction is undergoing renewed research effort and liquefaction technologies.
has been most actively pursued in the People’s China is rapidly expanding their chemicals
Republic of China (PRC) during the past 15 years. production, and much of this is based on coal
[44] While the growth today (2017) is at a slower
Recent DCL Developments in China rate than 5 years ago, China is a huge market that
Shenhua has recently constructed the world’s first requires several million tons of growth each year
commercial DCL facility, in the Inner Mongolia for chemicals.
Autonomous Region. Shenhua’s direct liquefac- China’s development of the Fischer-Tropsch
tion process combines elements of German, Japa- synthesis plants has also been scaled back during
nese, and, primarily American technologies with the past 10 years when several large plants were
its own innovations. The single-train reactor sys- under consideration. In the 2000–2010 period a
tem and process plant were designed by HTI and number of Fischer-Tropsch demonstration plants
licensed from Headwaters Inc. (HTI is currently a (3000–3500 bbl/day) were constructed and oper-
subsidiary of Headwaters), and operations of its ated in China, mainly by large Chinese coal com-
first production line commenced in 2008. Shenhua panies (e.g., Lu’an Group, Yankkuana Group, and
has planned expansion in two phases, for a total of Yitai Group). In 2016, Synfuels China began
10 production lines with projected oil equivalent operating a 100,000 bbl/day Fischer-Tropsch
products of 10 million ton/year, including approx- plant that utilizes an iron catalyst in a bubble
imately 150,000 barrels/day of middle distillate column reactor. In the same year, the Yitai
fuels (naphtha + diesel). The engineering design Group received permission to begin a US$4.2
of the plant, called SH-I, calls for a daily input of billion dollar (40,000 bbl/day) Fischer-Tropsch
7000 metric tons of Shenhua’s sub-bituminous plant. Thus, it appears that China is increasing
coal. An early train, operating by 2014, the their use of the coal to liquids conversion using
Shenhua DCL plant was producing about the Fischer-Tropsch process.
13,000 bbl/day liquid products. The plant utilizes A number of reports concerned the general
two-stage reactors at about 455
C and 170 atm. reaction pathway for DCL have been based on
ɣ-FeOOH (25  140 mm) is fed together with lumping the conversion into oil + gas, asphaltene
sulfur so that a ratio of Fe:S = 1:2 and a ratio of + preasphaltene, and unconverted [45–47]. Addi-
Fe/coal = 0.5 to 1.0%. The hydrogen donor solvent tional data has been utilized in more recent publi-
is hydrogenated distillate that boils in the range of cations that show that for both eastern and western
220 to 450
C. The effluent coal liquids are hydro- US coals the reaction pathways for the thermal
genated in a suspended bed reactor with forced and catalytic reaction are the same; the added
recirculation at about 360
C and 130 atm. A sum- catalyst only accelerates the conversion but does
mary of direct coal liquefaction and thermal char- not alter the selectivity [48–50] (Fig. 5). The path-
acterization has recently appeared [43]. Also, way shown in Fig. 5 is for eastern US coals, but
Shenhua Group, Shanghai Huayuan Group, and the pathway is similar for the western US coals
Coal to Liquids Technologies 401

Coal to Liquids
Technologies,
Fig. 5 Reaction pathway
for the thermal and catalytic
conversion of coal to
lumped products

except that the asphaltene + preasphalatene values 280 million automotive tires, and about 1.2 billion
are much lower so that the pathway is more direct barrels of waste oil are generated in the United
from unconverted coals to oil + gas products. States each year [52]. Coprocessing with coal
The lessons learned from the direct coal lique- would allow capture of energy content in these
faction work have been summarized as being waste materials instead of simply disposing of
technically feasible but the process for producing them. It has frequently been stated that synergistic
synthetic crudes is not economic [51]. interactions between coal and heavy oil can result in
In summary, advantages of DCL include: distillate yields and coal conversion levels greater
than expected from individually processing the coal
• High-octane gasoline is produced. and heavy oil. Several companies researched
• More energy efficient than ICL. coprocessing, including Chevron’s 6 ton/day
• Products have a higher energy density than coprocessing plant run in 1983 at the company’s
indirect conversion. Richmond, California site. The research included
synergistic incorporation of pretreated petroleum
Disadvantages of DCL include: vacuum resids with coal as feedstocks, and
demetallation of resids was also performed
• Processing difficulties associated with solids [53]. The process was said to have good operability.
separation The poorer solvent properties of heavy oils
• High aromatic content detracted from the conceptual simplicity and led to
• Low cetane number diesel (45) use of forcing conditions and increased hydrogen
• Potential water and air emissions issues consumption that have adversely affected econom-
• High capital intensity ics of coprocessing [54]. Although developments in
• May have higher operation expenses than ICL coprocessing have been essentially on hiatus for
• Lower feedstock flexibility than ICL over a decade, the topics of energy and reuse of
• High requirements for external hydrogen- materials remain prominent. Improvements in tech-
supply (which has to be provided by, e.g., nologies for gasification, solvation, and CO2 capture
additional coal gasification or natural gas) might lead to a resurgence of this CTL approach.

Coprocessing
Coprocessing combines waste materials having Indirect Liquefaction
high hydrogen content (relative to coal) with coal
as feed for conversion to liquid fuel products. Over Indirect coal liquefaction typically refers to pro-
44 billion pounds of plastic waste materials, cesses that first generate synthesis gas by
402 Coal to Liquids Technologies

Coal to Liquids Technologies, Fig. 6 Key process steps for indirect coal liquefaction Ref. [55]

gasification of coal followed by production of solid, Synthesis of fuels and chemicals from syngas
liquid, and gaseous hydrocarbons via catalytic does not occur in the absence of appropriate cat-
reduction of CO in subsequent stages of the process. alysts. The basic concept of a catalytic reaction is
Frequently the term ICL refers to the synthesis of that reactants absorb onto the catalyst surface,
liquid transportation fuels by means of Fischer- rearrange atoms, and combine into products that
Tropsch synthesis catalyst. Other ICL technologies desorb from the surface. One fundamental func-
include methanol and/or dimethylether from syngas, tional difference between different syngas conver-
or methanol-to-gasoline (MTG). MTG was com- sion catalysts is whether or not the adsorbed CO
mercialized in 1986 and uses ZMS-5 zeolite catalyst molecule is cleaved during reaction at the catalyst
to convert methanol to gasoline. surface. In Fischer-Tropsch synthesis, C–O bond
A basic block flow diagram of key process cleavage is a necessary reaction step. In methanol
steps in F-T ICL is shown in Fig. 6. synthesis, the CO bond remains intact. Hydrogen
After mining and coal preparation, plays two roles in catalytic syngas conversions, as
beneficiated coal (coal that has been subjected to a reactant needed for hydrogenation of CO to give
physical grinding and chemical cleaning methods the hydrocarbon moieties, and as a reductant to
during coal preparation) is first gasified with oxy- activate and regenerate the metal surface.
gen and steam to produce syngas. This syngas is Much of the research and process development
cleaned of all impurities and sent to the F-T reac- on coal liquefaction after the 1990s was aimed at
tors where most is converted to liquid hydrocar- matching the fuels synthesis conditions with mod-
bon fuels. The liquid products from F-T ICL ern, efficient coal gasifiers such as those developed
contain zero sulfur and essentially zero aromatic by Uhde, Texaco (now GE), Shell, and CB&I.
hydrocarbons, and therefore ultra clean diesel or A block flow diagram of a typical F-T CTL process
jet fuels can be produced after additional refining. to produce naphtha and diesel middle distillate cuts
Carbon dioxide is produced mainly during water is shown in Fig. 7, with the liquefaction end of the
gas shift (vide infra) and the F-T synthesis reac- process indicated within the dash-lined boundary.
tions and can be captured for subsequent storage Coal feed is prepared in the milling and drying
or use. The unconverted synthesis gas can be section and then sent to the gasifier where it is
either recycled back to the F-T reactors, to maxi- fully converted to raw synthesis gas (or a coal
mize liquid fuels production, or combusted in a slurry gasifier may also be employed instead).
gas turbine combined cycle power plant to gener- In the F-T synthesis step, the syngas is
ate internal and external power. Thus ICL plants converted to raw product. The lighter ends of the
can be configured to produce liquid fuels only, on raw F-T product are partially recycled internally
a freestanding basis, or to produce a combination and the remainder is sent to the heavy ends recov-
of liquid fuels and power for export. ery unit. Oxygen for coal gasification is supplied
The other major ICL approach produces meth- by an air separation unit (ASU, not shown). A part
anol from the syngas, via a different set of chem- of the raw synthesis gas is treated in the water gas
istries from F-T, and uses the methanol as a CO-shift unit to convert steam and CO to hydro-
chemical or as a fuel, typically as a blend with gen and CO2, after which this treated gas is
petroleum-derived gasoline. The methanol serves remixed with the untreated portion. The split
as oxygenate and fuel extender. Dimethylether between the portions is controlled such that the
(DME) also can be produced from syngas, either required H2/CO ratio of the F-T synthesis is
directly or via dehydration of methanol. always met. The shifted syngas is then submitted
Coal to Liquids Technologies 403

Coal to Liquids Technologies, Fig. 7 Fischer-Tropsch coal to liquids process Ref. [15] (Figure courtesy Uhde
Corporation of America)

to an acid gas removal (AGR) step, where COS, hydrocarbons. This water fraction contains impu-
CS2, H2S, and CO2 are removed to produce clean rities such as oxygenates, which have to be
syngas feed for the F-T synthesis unit. The removed in a water treatment step. Sour gas
Rectisol adsorption process is commercially (H2S, COS) removed in the AGR unit is sent to
proven in CTL applications and is typically a sulfur recovery unit, a Claus plant, where it is
deployed for AGR, applying methanol at low converted into elementary sulfur. The tail gas
temperatures as a solvent. However, a more from the Claus plant is hydrogenated and recycled
energy efficient system, such as polyethylene gly- to the acid gas removal unit, or it can be treated in
col (PEG)- or glycol diether- based processes a SCOT (Shell Claus Off-gas Treating) treatment
(e.g., Selexol or Genosorb), may be used instead unit. Using the SCOT process, the sulfur compo-
of Rectisol, although in these processes the COS nents in the Claus tail gas are catalytically
has to be removed in a separate, COS hydrolysis converted to hydrogen sulfide, which is removed
step. A small fraction of the clean syngas is sent to by solvent extraction and recycled to the Claus
a pressure swing adsorption (PSA) unit for pro- plant, and the off gas from the absorber is incin-
duction of pure hydrogen, as H2 is required in the erated. As an alternative to the recovery of ele-
product workup section, and the major portion of mental sulfur, sulfuric acid can be produced from
the syngas is routed to the F-T plant. the sour gas components.
The tail gas from the heavy ends recovery unit The yield of products in a fully self-sufficient
is sent to a combined cycle plant to produce suf- CTL plant based on hard coal gasification is
ficient power for the CTL complex. The F-T approximately 1.88 barrels per ton of coal. The
waxes and FT condensate products are sent to amount of sulfur produced depends on the con-
the product upgrade section, where they are centration of sulfur in the coal.
hydroprocessed with hydrogen (from the PSA
unit) to produce primarily diesel product. Gasification
A significant amount of water is coproduced Gasification breaks down coal into molecules of
with the hydrocarbons during F-T synthesis much lower molecular weight, CO and H2,
404 Coal to Liquids Technologies

usually using high temperature in the presence of the temperature can be kept below the ash fusion
oxygen at high pressure. Gasification technolo- point. The coal bed typically rests on a rotating
gies introduce air and/or oxygen, often with grate through which dry ash residue falls for
steam to moderate the temperature in the combus- removal. To reduce steam usage, British Gas and
tion zone of the gasifier. The product is syngas, a Lurgi developed a slagging bottom gasifier (BGL
mixture of hydrogen and carbon monoxide. The gasifier) in which the ash is allowed to melt and
syngas is subsequently converted to liquid fuels drain off through a slag tap. This gasifier has over
and chemicals in ICL processes. The three main twice the capacity per unit of cross-sectional area
types of gasifiers are the moving bed, fluidized compared to the dry bottom gasifier. Allied Syn-
bed, and entrained flow reactors. Gasification gas, Advantica, and Envirotherm offer the BGL
technology alternatives are grouped under the technology commercially.
three technology types in Table 3. All three types
work well and are in commercial use.
Fluidized Bed Gasifiers
Moving Bed Gasifiers In the fluidized bed gasifier, the feed is fluidized in
The moving bed may use co-current or counter- air, or sometimes oxygen, and steam with upward,
current flows and is sometimes referred to as a turbulent flow for efficient mixing. The tempera-
fixed bed because the coal bed is kept at a fixed, ture is normally below ash melting temperature.
constant height. The moving bed gasifier first The ash is removed as dry particles, which is
provided a convenient means to handle solid advantageous if the ash is highly corrosive and
residuals (ash) from the gasification. Moving bed would damage the reactor if removed as a molten
gasifiers can accommodate coal with moisture slag (a problem when gasifying high-sulfur coal
content as high as 35%, as long as the ash content or biomass). Throughput of the fluidized bed gas-
is less than 10%. Depending on the temperature at ifier is higher than for the moving bed. The low
the base of the bed of coal, the ash is either dry operating temperature of the circulating fluidized
solid or a molten slag. If excess steam is added, bed reactor technologies of Kvaerner and Foster
Wheeler are particularly appropriate for low-rank
Coal to Liquids Technologies, Table 3 Types of coal coal (or other reactive feed such as biomass). The
gasification technology higher pressure of the HTW gasifier technology is
Gasification suitable for coal (and biomass) [56].
type Technology GTI developed an agglomerating ash fluid-
Moving bed Lurgi – Dry gas (Germany) ized bed gasifier for its U-Gas process in which
FBDB – Sasol (South Africa) crushed limestone for sulfur capture is injected
British gas and Lurgi – BGL (UK) with the coal, and adding the limestone aids in
Fluidized High temperature Winkler (HTW) – processing ashes having very high fusion tem-
bed RWE, formerly Rheinbraun – Marketing peratures. Char and ash that exit the gasifier with
by
Uhde (Germany) the product gas are recycled to the hot agglom-
Kvaerner (Finland) erating and jetting zone, where temperatures are
Foster-wheeler (Switzerland) high enough to pyrolyze freshly introduced coal,
GTI/SES – U-gas (US) gasify the char, and soften the ash particles. The
Entrained SCGP – Shell (US) ash particles stick together and fall to the base of
Koppers-Totzek – Uhde (Germany) the gasifier where they are cooled and removed.
PRENFLO™ – Uhde (Germany) The agglomerating fluid bed gasifier can be
SFG – Siemens (Germany) blown by either air or oxygen. The U-Gas tech-
Ge (us) nology is licensed by Synthesis Energy Systems
ConocoPhillips (US) (SES).
ECUST (East China University of Alkaline metal compounds, especially chlo-
Science and Technology) (China)
rides and carbonates, can be used as catalysts for
Coal to Liquids Technologies 405

coal gasification in a fluidized bed including such scrubbing with water to remove chlorine and
compounds as potassium carbonate, zinc chloride, other soluble components, removal of COS and
and sodium carbonate. HCN in a hydrolysis reactor, and removal of H2S
using a methyl-diethanolamine (MDEA) solution.
Entrained Flow Gasifiers The H2S fraction from the desulfurization unit is
Entrained flow technologies particularly target treated in the Claus unit with tail-gas recycle,
issues of solids handling in coal gasification. where liquid elemental sulfur is produced for
When bituminous coals are heated to sale on the local market. Electrical power is gen-
300–350
C, the particles tend to swell and erated by a single combustion turbine and a single
agglomerate in producing a consolidated cake. steam turbine.
The handling of this caking char and the heavy A slurry feed may be used with low rank coals
tars that accompany it has been critical to the as these can exhibit serious particle decrepitation,
development of gasification processes. The dusting, and even spontaneous combustion when
agglomerate that forms can disrupt gas flow pat- exposed to drying conditions upon storage,
terns and lower thermal efficiency. crushing, and handling. A typical coal water
In the entrained flow reactor technology, coal is slurry contains about 70% mass of coal, pulver-
fed as a dry, pulverized solid or as fines slurried ized to minus 200 mesh or finer, and water with
with liquid and is most commonly gasified with about 1% mass of an additive to alter the rheolog-
oxygen (rather than air) in co-current flow. It is ical properties of the liquid mixture and thus allow
operated above the ash-melting point, which a larger coal loading [57]. Slurry fed entrained
means that the molten ash forms a liquid slag inside flow gasifiers include the GE gasifier and the
the gasifier that runs down the walls of the gasifier E-Gas gasifier (now owned by CB&I). The GE
and is removed through a bottom hole. The use of gasifier coal slurry is fed through single-burner at
oxygen at high pressure and temperature is a source the top of the gasifier, which is a refractory lined
of both high throughput and versatility. Typically a pressure vessel. Refractory stability under slag-
dry-feed gasifier such as SCGP or PRENFLO can ging conditions depends largely on the ash con-
process a wide range of coals, from low-grade soft tent and ash properties. Furthermore, the burner
brown coal to top-grade hard anthracite. These dry- lifetime of slurry-fed burners is considerably
fed entrained-flow gasifiers have single-train lower than that of dry-fed burners. Reactor exit
capacities of about 3500–4000 ton/day. Most of gas is cooled either by direct water injection or by
the ash is removed as slag. a radiative cooler located directly below the reac-
Uhde’s PRENFLO entrained gasifier technol- tor. The E-Gas gasifier differs from other systems
ogy has operated successfully for the last 15+ in that it uses a two-stage reactor, also involving
years in the world’s largest dry-fed IGCC plant, refractory lined pressure vessels. The bulk of the
located in the Elcogas plant in Puertollano, Spain. feed slurry and all the oxygen are sent to the first
PRENFLO employs dry-feeding, multiple side- (horizontal) stage, where the coal is gasified. Hot
fired burners and a cooled membrane wall, for gas flows into the second (vertical) stage, where
robustness under the slagging conditions inside the remainder of the coal slurry is injected. Hot
the gasifier. The raw gas leaving the PRENFLO™ fuel gas is cooled in a fire tube boiler gas cooler
gasifier is first cooled down to approximately and then sent to the F-T unit. The process pro-
900
C in a gas quench system with a recycle duces no ash and recycle by-products for uses
gas stream. The sensible heat of the raw gas is such as road construction materials.
used to produce saturated high pressure steam and Today, both solid (dry) and slurry fed gasifiers
saturated medium pressure steam, both of which are likely to be entrained flow gasification systems.
are ultimately used in the steam turbine of the East China University of Science and Technology
power block to produce power. The raw gas is (ECUST) has been working on gasification for
next sent to gas conditioning operations, about 30 years. Based on MW of synthesis gas
consisting of filtration to remove particulates, produced they are now the size of Lurgi and are
406 Coal to Liquids Technologies

the third/fourth largest supplier of gasification tech- (as H2S) is removed from the syngas via a Rectisol
nology [58]. Their entrained flow gasifier utilizes an unit designed by Linde AG and is recovered as
opposed multi-burner technology (www.netl.doe. elemental S by a combination of Claus sulfur
gov/research/coal/energy-systems/gasification/ recovery units (SRUs) and Shell Claus off-gas
gasifipedia/ecust). The down-flow operation uti- treating (SCOT) processes. CO2 is also removed
lizes a water-quench. Both wet and dry feed by the Rectisol process and is currently vented but
options are available. The newer entrained flow could easily be captured for sequestration. A cold
gasifiers are considerably more efficient, albeit at box, also designed by Linde AG, is used for
the price of making product gas having a higher cryogenic separation of CO from a portion of the
content of syngas, total H2 and CO, but a lower syngas, and some of the syngas itself is used in
H2/CO ratio [59]. A critical aspect of long-term downstream methanol plants based on technology
sustainability of any gasification technology is from Lurgi and Air Products. Mercury contami-
that it be versatile enough to accommodate vari- nant originally present in the coal is removed
ous ranks of coal as well as other feeds such as using a packed bed of sulfur-impregnated acti-
biomass. vated carbon. Originally built for production of
The first commercial scale and currently lon- chemicals from coal, the Eastman gasification
gest running coal gasification plant in the United process and associated learning curve with a vari-
States was put into service in 1983 by Eastman at ety of coals are representative of what would be
its Kingsport, Tennessee, facility. A schematic of used for future production of power and fuels
the gasification plant is shown in Fig. 8. The from coal in processes such as integrated com-
process uses two high-pressure (1000 psig) GE bined cycle gasification (IGCC) and industrial
gasifiers, one being a spare, to convert about 1250 gasification.
ton/day of bituminous coal into acetyl chemicals Disadvantages of entrained-flow gasification
via methanol and CO. Over 99.9% of the sulfur include a rather high oxygen requirement and a

Coal to Liquids Technologies, Fig. 8 Flow diagram of Eastman’s coal gasification plant (Figure courtesy of Eastman
Chemical Company)
Coal to Liquids Technologies 407

high waste heat recovery duty. However, the abil- elimination, and these either re-hydrogenate dur-
ity to feed coal that is dry reduces the oxygen ing propagation or exit from the catalytic cycle
requirement and increases efficiency compared during termination.
to a single-stage entrained gasifier using
slurry feed. Performance Variables and Parameters, and
Product Distribution
The reaction temperature and catalytic metal
Fischer-Tropsch strongly influence the chemistry and product dis-
The Fischer-Tropsch (F-T) reaction is a polymer- tribution. The reaction is run in either a high
ization or oligomerization of carbon monoxide temperature (HTFT) or low temperature (LTFT)
performed catalytically under reducing condi- regime. The HTFT reaction is always performed
tions. Chemically F-T is single reaction, with with iron as catalyst, at ~340
C, and the LTFT
approximate stoichiometry for linear hydrocarbon reaction is run at ~220
C when using a Co-based
production as shown in Eq. 1. The greater hydro- catalyst. The Fe-catalyzed LTFT reaction can be
gen content of methyl groups at the chain termini run a little hotter, at 230–270
C. The HTFT
increase the stoichiometric ratio of H2:CO slightly reaction is traditionally used to make light syn-
beyond 2:1, with the actual F-T stoichiometry for crude for production of gasoline and diesel. The
paraffinic hydrocarbon production being about LTFT reaction produces primarily high molecular
2.1:1. weight hydrocarbons, referred to collectively as
F-T wax, that are typically hydrocracked to make
n CO þ ð2n þ 1Þ H2 ⇆ CH3  ðCn2 H2n4 Þ diesel [61].
 CH3 þn H2 O DH 298  38 kcal=mol CO Much novel work on stable cobalt catalysts for
(1) slurry bed gas to liquids (GTL) applications has
been patented in the past 15 years. While the GTL
Mechanistically, the F-T reaction consists of plants in Qatar use a Co catalyst, the CTL plant in
four steps: (1) formation of metal-bound C1 spe- China uses an iron catalyst they developed
cies from feed CO (or possibly CO2 if present); [62]. This is mainly because unlike iron, cobalt
(2) chain initiation wherein metal-bound C2 spe- does not undergo the water gas shift reaction
cies form from intramolecular or intermolecular (WGSR) at the conditions needed to produce the
coupling of the metal-bound C1 species; (3) chain supplemental hydrogen needed to make saturated
propagation wherein the initial metal-bound C2 hydrocarbons from coal. It is also the case that,
species grows as it couples with newly formed while essentially all examples of Co patents use
metal-bound C1 species; and (4) chain termination GTL conditions, many of these Co patents state
wherein metal-bound Cn species undergo hydro- that various feed sources including syngas from
gen transfer reactions that disable further propa- coal may be used instead of natural gas
gation [60]. Mechanistic specifics remain in [63]. Major research domains of emphasis in F-T
debate but it is well-accepted that hydrogen trans- CTL catalysis include understanding and fixing
fer reactions play key mechanistic roles through- problems of iron catalyst instability and attrition at
out the reaction, including metal-catalyzed slurry LTFT conditions [64, 65].
hydrogenation of CO, b-hydride elimination The parameters used to assess and report per-
from the growing chain to give terminal olefins, formance of F-T catalysts are those typically used
and intramolecular reductive elimination of for catalytic reactions: conversion, selectivity,
hydrocarbon (R) and hydrogen (H) fragments product distribution, yield, reactor productivity,
from the catalytic center to give products during and catalyst fate or lifetime. The selectivity of a
chain termination. Hydrogenation is not always catalyst with a given feed composition is the main
complete and oxygenation products such as lower determinant of measured product distribution
alcohols and carbonyl compounds also form. Ole- (although product distribution is also influenced
fins form throughout the process via b-hydride by process design). The yield of a given cut is the
408 Coal to Liquids Technologies

product of overall conversion and selectivity for (carbon basis). Selectivity data also may be sup-
that cut. Yield is usually expressed as a function of plied in absolute quantities, as reactor productiv-
time and catalyst quantity in these continuous ities for specific cuts. The weight fractions of
processes. Catalyst fate relates to the mechanism various product cuts obtained within the F-T
and timeframe of catalyst deactivation and there- kinetic regime are shown in Fig. 9. The product
fore determines its useful lifetime. slate is influenced by many reaction variables,
The conversion for an F-T process is generally particularly the feed composition, the reaction
reported as %CO converted to any product, temperature, the catalytic metal and reaction pro-
although conversion of total feed (CO + H2) is moters, and the support material if present.
sometimes used. Coproduction of the undesired F-T product slates include gaseous and liquid
by-product methane typically rises directly with fractions. The gaseous reaction product includes
conversion. F-T synthesis with Co is typically run hydrocarbons boiling below about 340
C (tail
at about 60% conversion, and conversions in gases through middle distillates). The liquid reac-
LTFT are typically in the range of 50–70%. tion product (the condensate fraction) includes
The reactor productivity is the quantity of hydrocarbons boiling above about 340
C
product produced per quantity of catalyst per (vacuum gas oil through heavy paraffins). Com-
unit time, expressed by weights of product and mercially, the product boiling below 340
C is
catalyst (e.g., kgprod/kgcat/h) or by volumes (Lprod/ typically separated into a tail gas fraction and a
Lcat/h). Productivity may be reported for bulk condensate fraction of about C5–C20 normal par-
product or for specific cuts, or even as the quantity affins plus higher boiling, branched hydrocar-
of CO converted per quantity of catalyst per unit bons. Branching may be advantageous to a
time. Productivity of the LTFT reaction performed number of end-uses, particularly when increased
in a slurry bed reactor is higher than for a fixed octane values and/or decreased pour points are
bed, primarily as a result of improved heat man- desired for lube oil additives. The degree of isom-
agement in the slurry system. The continuous erization is preferably greater than 1, and more
motion of the liquid-catalyst matrix allows suffi- preferably, greater than three moles of isoparaffin
cient heat transfer to achieve commercial slurry per mole of n-paraffin. If used in a diesel fuel
reactor productivities of about 200 volumes of CO composition, the products preferably have a
converted/vol catalyst/hr. cetane number of at least 60.
The product distribution reflects the selectivity The F-T reaction produces a wide range of satu-
of the catalytic system at the reaction conditions, rated and olefinic hydrocarbons. Figure 10 shows
typically reported as a percentage of the total the percentage proportions of olefins, alkanes, oxy-
product or as a percentage of the CO converted genates, and aromatics in upgraded syncrude

a 100 b 1.0
90 C1 0.9 CH1
80 0.8
70 0.7 Wax
Weight Fraction

(C35-C120)
60 C35-C120
0.6
Selectivity

50 0.5 Gasoline
C5-C11 (C5-C11)
40 0.4
Diesel
30 C2 0.3 C2
(C9-C25)
C12-C18
20 0.2 C3
C24-C35
C3
10 0.1
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1.0
Probability of Chain Growth Probability of Chain Growth

Coal to Liquids Technologies, Fig. 9 Product distribution plots for the Fischer-Tropsch reaction Ref. [67] (a) general
distribution, and (b) additional breakdown of the C12+ cuts
Coal to Liquids Technologies 409

Coal to Liquids 100

Technologies,
90
Fig. 10 F-T product slates
dependence on temperature 80
and catalytic metal 70
(Figure courtesy of SRI
Consulting) Ref. [67] 60
50
40
30
20
10
0
Co-LTFT Fe-LTFT Fe-HTFT

Olefins Oxygenates
Alkenes Aromatics

obtained from the F-T reaction at varied conditions W n ¼ ð1  aÞ2 nan1 (2)
of temperature regime and catalytic metal. The
HTFT reaction gives aromatics, while the LTFT The parameter “a” is the Schultz-Flory distri-
reaction does not give aromatics with any catalytic bution factor and represents the ratio of the rate of
metal. Cobalt is a better hydrogenation catalyst than chain propagation to the rate of chain propagation
iron, and therefore products of the Co-catalyzed F-T plus the rate of chain termination. The parameter
reaction tend to have a lower olefin/paraffin ratio “a” is the chain growth probability, and 1-a is the
than those of iron. (Not shown in Fig. 10, the high probability of chain termination. In using this
temperature reaction also gives more branched equation, a and a are considered as independent
hydrocarbons, and with a higher degree of of chain length. Either value can be obtained
branching than the low temperature reaction using graphically using this equation in log form, with
either Fe or Co.) The LTFT and HTFT both give analytical data for Wn as obtained for example by
oxygenates, typically 10–15% for the HTFT reac- gas chromatography. In practice, the parameter a
tion and about half that for the LTFT reaction with is used most often to characterize product
Fe (and much less with Co). Most oxygenates pro- distribution.
duced in the LTFT reaction are alcohols. The HTFT Values of a > 0.9 are representative of pro-
reaction gives alcohols and ketones, plus a lesser cesses producing high amounts of wax, a values
amount of carboxylic acids and aldehydes. The of about 0.8 maximize diesel cuts, and a values of
LTFT reaction gives more methanol than the about 0.7 are characteristic of naphtha production.
HTFT reaction, and the two temperature regimes For the intermediate molecular weight cuts of
give comparable amounts of ethanol. relevance to fuels production, diesel and naphtha,
Product slates for F-T catalysis are controlled the ASF distribution reveals the limitations on
by polymerization kinetics, with fairly constant selectivity shown in Fig. 9. Because selectivity
chain growth probabilities that fix the ranges of for diesel cuts does not exceed about 22% but
possible product distributions [68]. The moles of wax selectivity can reach 90 + %, much industrial
species with carbon number n decrease exponen- research has focused on wax production with
tially with n. The weight fraction of product hav- downstream hydroprocessing of wax to diesel
ing carbon number n, Wn, is given by the (or to lube oil base stocks). The a values for
Anderson-Schultz-Flory (ASF) distribution oxygenated products appear to be about the
(Eq. 2): same as those for hydrocarbons, which suggests
410 Coal to Liquids Technologies

that the products all form by the same or similar HTFT

chain propagation steps and with similar rates of The HTFT reaction has been run in a fixed bed
chain termination, albeit with different chain ter- reactor (FBR) or fluidized bed reactor. Tradition-
mination chemistries for the oxygenates, olefins, ally Sasol used a circulating fluidized bed reactor
and paraffins. (CFBR), in its Synthol process. The Synthol
In practice there are several deviations from the CFBR was replaced in 1989 with the Sasol
ASF product distribution inherent to the F-T pro- Advanced Synthol (SAS) reactor, based on a con-
cess. The amount of methane produced is often ventional fluidized bed and providing superior
higher than would be predicted from this equation, RAMO (Reliability, cost-effective Availability,
because methane is formed by a chain termination Maintainability, and efficient Operability).
mechanism that is separate from those forming the A block flow diagram of the F-T and downstream
higher molecular weight products (methane is portion of Sasol’s Secunda plant using the SAS
formed by addition of 2 H atoms to a metal- technology is shown in Fig. 11.
bound C1). Catalysts that preferentially catalyze The process is employed to produce a range of
hydrogenation over chain initiation, particularly fuels and also olefins. The fluidized, iron-based
nickel, increase the probability of termination to catalyst gives primarily C1–C20 raw product. The
give methane. A second deviation from the ASF C2-rich stream is split into ethylene and ethane.
distribution is that the amount of C2 species formed Ethane is cracked in a high-temperature furnace to
is often lower than expected based on the ASF ethylene, which is then purified. Propylene from
equation, although the reason for this is not clear. the light hydrocarbon gases is purified and used in
It has been suggested that an adsorbed ethylene can the production of polypropylene. Large quantities
react with a surface methylene from either side to of olefins in the C5–C11 range are produced at
propagate the chain; whereas higher alkyl adsorbed these HTFT conditions and the alpha-olefins
species would show a lesser rate of chain propaga- (straight chain olefins with an ene group in the
tion due to steric constraints. terminal position) are recovered for use as
Some approaches to catalyst design have chemicals (C5, C6, and C8). Oxygenates in the
succeeded in controlling the F-T product distribu- aqueous stream are separated and purified in the
tion by altering the value of alpha. One such chemical workup plant to produce alcohols, acetic
approach is the inclusion of alkali metal elements, acid, and ketones including acetone, methyl ethyl
particularly potassium. Oxides of the alkali metals ketone (MEK), and methyl isobutyl ketone
increase a and promote formation of wax in LTFT (MIBK).
catalysis by iron. Use of K2O helps keep the iron
reduced, in the carbide form, which promotes LTFT
growth of longer chains. The early German LTFT plants used fixed bed
With iron catalysts, alpha can exhibit two reactor (FBR) technology with Co catalysts,
regimes of relatively constant value, sometimes through the Second World War. In 1955, Sasol
referred to as the “two alpha” phenomenon. The began using a tubular fixed bed reactor (TFBR),
value of a can increase with chain length rather consisting of a shell containing 2050 tubes of
than remain constant, and typically changes from 12 m length and 5 cm diameter packed with
about 0.60–0.65 to about 0.90–0.93 at roughly extruded precipitated iron-based catalyst.
C20 [69]. A double a is most evident for low A typical catalyst system would be potassium-
temperature catalyst systems with high wax pro- promoted precipitated iron (such as a Sasol-type
ductivity and is not evident for high temperature catalyst consisting of K = 1.9 wt%, Cu = 2.4 wt%,
systems that make diesel or functionalized hydro- Si = 5.9 wt%, Fe = 47.3 wt%, O2 = ~42.5 wt).
carbons. However, deuterium tracer studies show This LTFT is referred to as the Arge process and is
that the two-alpha distribution is due to reactor operated with a shell side temperature of about
design and heavy product hold-up; the reaction actu- 220
C and reactor pressure of 25 bar (early Arge)
ally produces a single alpha distribution [70, 71]. or 45 bar (Arge commissioned in 1987), wherein
Coal to Liquids Technologies 411

Syngas

HTFT
SAS Reactors O2
Autothermal
Reformers Steam
H2

CH4
Water Separation Fuel gas

C4
C5-C10
C2H4 C3H6
Oxygenates Diesel
Recovery
Acids Olig. Separation Hydrofine
Extraction Purification
Pt Reform
LPG Diesel
Water to l-alkenes
Bioworks Gasoline
Gasoline (C5 to C8)

Coal to Liquids Technologies, Fig. 11 Secunda plant block flow diagram Ref. [66]

the heat of reaction is dissipated by generation of Boudouard disproportionation reaction, Eq. 3, is

steam on the shell side. Further scalability could to be avoided, and the heat distribution imparted
be achieved by more than doubling the number of by the slurry bed is advantageous.
tubes, to for example 5000, but the reactor size
eventually becomes unwieldy. Sasol found that a 2CO ⇆ C þ CO2
(3)
slurry phase system is more readily scaled, in part DH 298 ¼ 20:6 kcal=mol CO
due to engineering design and in part because the
Co-based catalysts used are more efficient than Attrition and breakage of catalyst particles is a
the Fe-based catalysts used for the LTFT reaction. more difficult issue to overcome than coking,
Sasol commissioned its first slurry bed reactor particularly for slurry bed catalyst systems
(SBR) (2500 bbl/day) in 1993 and the larger [73]. Attrition refers to abrasive or erosive
(17,000 bbl/day) slurry reactors in Qatar in 2007. removal of surface material that is less than
Table 4 shows typical product slate selectiv- about 50 microns in particle diameter. Breakage
ities for the LTFT, TFBR; the LTFT, SBR; and the refers to a decrease in median particle size caused
HTFT FBR. Both fixed- and slurry-bed LTFT by fracture of catalyst particles. As the concentra-
processes give much more paraffin, less olefin, tion of smaller particles builds up, catalytic activ-
no aromatics, and less oxygenates compared to ity eventually decreases and also the downstream
the HTFT process. LTFT is particularly suited for separation of wax from solid catalyst becomes
a modern coal to diesel (CRD) plant. inefficient. Catalyst particles are subjected to
As indicated in Eq. 1, the F-T reaction is highly hydrodynamically induced mechanical stress at
exothermic. The reaction heat must be removed slurry bubble column conditions that has severely
rapidly, to avoid temperature increases that would limited the lifetime of LTFT iron catalysts.
result in undesired formation of high levels of Precipitated, unsupported iron catalysts are
methane, light hydrocarbons, and, in the extreme, particularly prone to disintegration into smaller
catalyst deactivation due to coking, sintering, and particles. Supporting the iron imparts strength to
disintegration. Deposition of carbon by the the particle to decrease fracture, but at the expense
412 Coal to Liquids Technologies

Coal to Liquids Technologies, Table 4 Carbon selectivity of selected LTFT and HTFT processes (%) Ref. [72]
Product TFBR (LTFT) SBR (LTFT)
CH4 (methane) 4 7
C2–C4 olefins 4 24
C2–C4 paraffins 4 6
Gasoline 18 36
Middle distillate 19 12
Heavy oil, waxes 48 9
Polar oxygenates 3 6
TFBR (LTFT) SBR (LTFT) SASR FBR (HTFT)
C5–C12 C13–C18 C5–C12 C13–C18 C5–C12 C13–C18
n-Paraffins 95 93 96 95 55 60
Paraffins 53 65 29 44 13 15
Olefins 40 28 64 50 70 60
Aromatics 00 00 5 15
Oxygenates 77 76 12 10

of decreased volumetric productivity due to dilu- produced by gasification of coal is lower than
tion of the active catalyst by the support material. that produced by steam reforming of methane in
Making spherical particles by spray drying [74] GTL (gas to liquids), by about a factor of 3. With-
can help with fracture or with attrition, but not out the WGSR, other sources of hydrogen must be
always with both [75, 76]. provided in CTL, if hydrocarbon fuels are the
The pursuit of mechanically robust iron F-T intended product slate as per Eq. 1, and additional
catalysts remains an area of active research. Tra- sources of hydrogen in turn typically require addi-
ditionally, Fe catalysts have been disposed of tional use of power that itself generates CO2.
rather than regenerated because iron is inexpen-
sive. However, a few reports of potential use and CO þ H2 O ⇆ CO2 þ H2
(4)
even regeneration [77] of Fe slurry catalysts have DH 298 ¼ 9:8 kcal=mol CO
recently appeared, notably from Rentech [78],
presumably because management of even low The in situ WGSR occurs concurrently during
hazard waste volumes is of increasing interest. It the iron-catalyzed F-T reaction. Cobalt does not
is also the case that using a fluidized slurry bed catalyze the WGSR under F-T conditions and
catalyst system allows longer runs, since therefore iron catalysts produce more CO2 in the
deactivated catalyst can be removed and fresh FT reactor than do cobalt catalysts for the FT
catalyst added without taking the reactor off-line reaction but produce the same amount of CO2
as is required for fixed bed reactors. On the other for the total synthesis process. In practice, the
hand, an advantage of fixed bed reactors is that WGSR is one of several factors that affect the
since the wax produced simply trickles down the overall efficiency of an F-T CTL process when
bed, there is no problem separating the wax from using Fe versus Co as catalyst, another being the
the catalyst. In the slurry system, special internal greater efficiency of the smaller reactor used for
and/or external devices are required to continu- Co compared to Fe, and each must be taken into
ously remove wax from the finely divided catalyst account by process engineers when designing
in the slurry. CTL plants.
Another key aspect of the WGSR is that it can
Water-Gas Shift Reaction run in reverse. A potential use for some of the
The water-gas shift reaction (WGSR), Eq. 4, is by-product heat typically available in hot outlet
important for CTL because the H2:CO ratio gases from high temperature gasifiers is to
Coal to Liquids Technologies 413

generate steam heat for an external water gas shift catalyst, is the most efficient approach. Lighter
reaction, to be conducted downstream of the gas- and heavier fractions are usually less desirable
ifier, wherein some H2 (from recycle or tail gas) due to limited markets and/or lower prices.
consumes the undesirable by-product CO2. The Co-catalyzed process is typically chosen
for maximum wax production in LTFT gas to
Downstream Product Upgrading liquid (GTL) processes; however, the use of
Upgrading technologies for F-T syncrude are iron-based catalysts is generally preferred for
modifications of those developed for crude oil CTL because coal-derived syngas is deficient in
refining. Refinery process options for producing H2 and cobalt does not catalyze the WGSR that
fuels from F-T middle distillates include wax supplements H2 production during the iron-
hydrocracking, distillate hydrotreating, catalytic catalyzed F-T reaction. When operating in the
reforming, naphtha hydrotreating, alkylation, LTFT mode, iron catalyst can produce up to
and isomerization. The many technology licen- 50–75% linear wax products. LTFT-derived
sors for hydroconversion processes include Chev- feeds can be hydrocracked under much milder
ron, UOP, IFP (Axens), Clariant Sasol Catalysts, conditions than typical crude oil derived feeds,
and Haldor-Topsøe. Catalysts can be obtained such as vacuum gas oils, which require pressures
from these companies as well as from AkzoNobel, as high as 150 bar to prevent coking of the catalyst
Clariant, and Axens among others. by aromatic compounds. This is not an issue with
Reforming and alkylation of gaseous products paraffinic feeds, and pressures between 35 and
to produce liquid fuel-range products are appro- 70 bar are used to hydrocrack LTFT products
priate when upgrading synthetic crude from the using commercial hydrocracking catalysts. The
Fe-catalyzed HTFT reaction. Skeletal modifica- relatively low levels of oxygenates present in
tion technologies most important to upgrading F-T waxes, mainly alcohols and lesser amounts
synthetic crude from HTFT include isomerization, of acids or carbonyls, are easily and completely
mild hydrocracking, and even fluidized catalytic hydrodeoxygenated.
cracking. Catalytic oligomerization of olefins is After hydrotreatment, the straight run diesel
especially applicable to the product from an iron- fuel cut has a cetane number ranging from 70 to
catalyzed, HTFT process because the bulk of the 75. If both hydrotreating and hydrocracking are
crude product consists of these olefins, particu- used, they may be operated at the same pressure
larly C3-C5 olefins (gases). The Sasol Synfuels using a common hydrogen loop. Alternatively, the
CTL refinery in Secunda uses an alkene oligomer- hydrotreater can be operated in a once- through
ization process based on solid phosphoric acid mode, at lower pressure, and the hydrogen tail gas
(SPA) catalysis to convert most of the light alkene then subsequently pressurized and fed to the
HTFT product into liquid fuels [79]. hydrocracker loop. For a 30,000 bpd equivalent
Currently, the production of a high quality product upgrading system involving both hydro-
diesel fuel is preferred to the production of gaso- cracking and hydrotreating, the size of the unit is
line in F-T fuels production. This is because the comparable to those used in oil refineries. There
same factors that count against F-T gasoline, such are many process configurations possible
as product linearity and low aromatic content, depending on site-specific factors including plant
serve as very positive factors in favoring high capacity, available land, and possible integration
quality, high cetane number in F-T diesel fuel. with other facilities.
Straight run F-T diesel makes up about 20% of When using the slurry phase LTFT reactor,
the total F-T product and has a cetane number of water is removed from overhead gases in a three-
about 75 because it is predominantly linear. For phase separator and then the remaining conden-
maximum production of high quality diesel, the sate is sent to a low temperature distillation col-
slurry bed reaction operating in the high wax umn, in the heavy ends recovery section, to
selectivity mode with cobalt-based catalyst, or recover condensate from tail gas in the reactor
the high diesel selectivity mode with iron-based overheads. The overheads from the low
414 Coal to Liquids Technologies

temperature fractionation column are used as fuel • Advanced methanol powered and plug-
gas. Recovered condensates from the column are powered vehicles
sent to the product workup section. The waxy • Direct methanol fuel cells
reactor product is withdrawn from the side of the
F-T reactor and continuously separated from the Methanol can be synthesized from syngas
catalyst. The waxy product is then sent to the first according to reaction 5, where the thermodynam-
low temperature column, which sends a heavy cut ics assume gas phase components. The formation
to the hydrocracker and a light cut to the of methanol from syngas is highly exothermic.
hydrotreater. The wax hydrocracker produces the
diesel product, while olefins in the condensate are CO þ 2 H2 ⇆ CH3 OH
(5)
hydrogenated to paraffins in the hydrotreater. The DH 298 ¼ 21:6 kcal=mol CO
hydrocracking step typically uses a silica-alumina
supported Ni-Mo based catalyst and can be The feasible operating temperature for metha-
adjusted to maximize the production of ultraclean nol synthesis is strongly affected by the tempera-
diesel product. The products from the hydrotreater ture sensitivity of the copper-based catalyst,
and hydrocracker are then combined and fraction- which is prone to deactivation via sintering of
ated into diesel and naphtha products. Up to copper or coking [80]. Methanol synthesis occurs
approximately 80% of the product is diesel and at significant rate in the temperature range of
the remaining portion is naphtha. 210–260
C, optimizing at about 237
C
Product upgrading schemes center around [81]. Modern commercial methanol synthesis cat-
three fuel products (diesel, LPG, and naphtha), alysts make use of the high reactivity of copper
but the upgrading plant is quite amenable to dif- oxides combined with the dispersant function of
ferent configurations if market demand so indi- zinc oxide, supported on alumina. This combina-
cates. The ASF distribution means that the F-T tion was pioneered by ICI [82] (Imperial Chemi-
process and its associated refinery also produce cal Industries PLC, acquired by AkzoNobel in
transportation fuels and products other than diesel 2008) and allows methanol synthesis to be
and gasoline. Other saleable products may include performed at low pressure with recycle
solvents, kerosene, jet fuel, illuminating paraffins (35–55 atm) and relatively low temperature
(sold for use in lighting and appliances in (200–300
C, typically 250
C). The component
South Africa), base oils (for lubricant products), contents are in the range of about 40–80% CuO
and chemicals. Maximizing recovery of the higher and 10–40% ZnO, with 5–10% Al2O3. For exam-
boiling, jet fuel while maintaining diesel produc- ple, the unreduced material RPJ-19, from
tion will tend to increase economic return. Clariant, has been described as 55 wt% CuO,
36% ZnO, 8% Al2O3, with 1% SiO2 [81]. Magne-
Methanol and Dimethylether sia may be added to control sintering and increase
Supplementation of liquid fuel supplies with the available surface area of copper.
methanol made from abundant coal reserves has The methanol synthesis catalysts are also sen-
become increasingly attractive in countries that sitive to catalyst poisons. The contaminants sulfur
are oil and natural gas resource poor. Methanol and halide are typically removed upstream to
has a broad range of applications including: <100 ppm. Because catalyst lifetime is signifi-
cantly diminished by excessive temperature, reac-
• Fuel substitute for gasoline tor temperature control is very important.
• Gasoline fuel blend to extend conventional Currently, all commercial processes produce
fuel supplies methanol from synthesis gas in a vapor phase
• Energy storage medium in coproduction of reactor with a synthesis loop, using packed bed
power and fuel or tubular reactors, and temperature moderation is
• Rapid start combustion turbines to meet peak achieved by recycling large quantities of
electricity demand hydrogen-rich gas. Typically, a gas phase process
Coal to Liquids Technologies 415

is limited to about 16% CO in the reactor inlet, as a [84]. Air Products then undertook a significant
means of avoiding excess heating by constraining research effort together with ChemSystems and
the conversion per pass. Cool, unreacted gas is DOE cofunding, to further develop the process in
injected at different stages in the catalyst bed or the 1980s and 1990s, as part of the DOE’s cogen-
internal cooling surfaces are used to provide fur- eration program. While ChemSystems originally
ther temperature control. used an ebullating bed design with an aromatic
As a means to improve throughput per pass (1,2,4-trimethylbenzene) or mineral oil liquid, Air
with temperature control, ChemSystems and Air Products used the more stable mineral oil in a
Products developed a fundamentally different, slurry bubble reactor to further augment heat-
liquid phase reactor system. In this system, and mass transfer.
referred to as the LPMeOH™ (Liquid Phase The LPMeOH process was demonstrated at
Methanol) process, synthesis gas is bubbled large pilot scale during 1997–2002, in a collabo-
through a slurry, consisting of micrometer-sized rative effort between Air Products (technology
catalyst particles and methanol in a suspension of supplier), Eastman Chemical (syngas supplier,
mineral oil. The nonvolatile mineral oil acts as a plant operator, and end user of methanol), and
temperature moderator, transferring the heat of Arcadis (fuel and power testing). This facility in
reaction from the catalyst surface via the liquid Kingsport, TN, operated for nearly 6 years, with
slurry to boiling water in an internal tubular heat continuous runs of up to 3 months. The nameplate
exchanger. As a result of maintaining a constant, capacity of 80,000 gpd was exceeded by 15%.
highly uniform temperature through the entire Eastman syngas, derived from high-sulfur
length of the reactor, the slurry reactor can achieve (3–5%) bituminous coal, was used. Of note is
a much higher per pass syngas conversion than its the ability to produce sulfur-free methanol
gas-phase counterparts and also CO concentra- containing less than 1% water, which makes this
tions in excess of 50% have been maintained methanol directly suitable for transportation as
without adverse effect on catalyst activity. As well as chemical use. Used as a fuel additive,
with slurry F-T systems, the liquid phase methanol methanol also showed promise for reducing
process benefits from the ability to withdraw spent NOx emissions from gas turbines and diesel
catalyst slurry and add fresh catalyst on line peri- engines. Air Products also applied the understand-
odically, without having to shut the system down. ing gained from developing the LPMeOH process
Typical commercial Cu/Zn/alumina methanol in developing the Liquid Phase Dimethyl Ether
synthesis catalysts are also active for the low Process (LPDME™) process, which was demon-
temperature water gas shift (LTWGS), which strated in the early 1990s.
enables use of syngas from lower quality, low- Sunggyu Lee and colleagues pioneered the
hydrogen feedstocks such as coal to be used in mechanistic understanding of the one-step slurry
making methanol. Reduction of CO or CO2 to methanol process and also worked on the MeOH/
methanol with H2 is exothermic, and heat dissipa- DME slurry process [85]. The LPMeOH process
tion in fixed bed reactors is inadequate to protect has three distinctive features and benefits com-
the catalyst or to allow optimal product recovery. pared to prior art using fixed beds. First, the com-
While less exothermic than the methanol synthe- mercial Cu/Zn/alumina methanol synthesis
sis reaction, the WGSR also adds heat to the catalyst is slurried instead of fixed, which greatly
system. improves heat transfer via the inert, liquid
ChemSystems performed pioneering develop- medium. Second, the slurried catalyst is a fine
ment of the LPMeOH process in the 1970s powder, which avoids mass transfer limitations
[83]. The LPMeOH technology was based on involving pore diffusion. Third, the improved
early design and demonstration work funded by mass transfer in the well-agitated slurry allows
EPRI (Electric Power Research Institute) in the much higher single-pass conversion of the syngas.
mid-1970s, wherein methanol is produced in one The initial method devised to combine steps in
step using the high-CO content syngas from coal liquid-phase DME production involved
416 Coal to Liquids Technologies

combining the methanol dehydration catalyst with 2 CH3 OH ⇆ CH3 OCH3 þ H2 O

the slurry phase methanol synthesis catalyst (6)
DH 298 ¼ 5:9 kcal=mol DME
[86]. For DME, typically the methanol synthesis
catalyst is copper on alumina promoted by zinc
CO2 þ 3 H2 ⇆ CH3 OH þ H2 O
oxide. The ratio of CuO/ZnO/Al2O3 is typically (7)
about 2:1:5 with a H2:CO ratio of 2:1 but will vary DH 298 ¼ 11:7 kcal=mol MeOH
with syngas composition. Alumina is the most
common dehydration catalyst. Both CO conver- Even so, the mechanism actually in effect dur-
sion and selectivity for DME are improved by ing direct synthesis of DME is in some debate and
including 10%wt sulfate in the alumina. Many might not involve the explicit intermediacy of
researchers have reported using zeolites, and methanol as per Eq. 6. Figure 12 shows the results
recent research with zeolites has been quite active of modeling done by JFE Holdings (formed in
in Asia [87]. Likewise, silicoaluminophosphate 2002 by the merger of NKK and Kawasaki
(SAPO) molecular sieves have been reported Steel) based on their pilot plant results, which
recently for use as dehydration catalysts in direct give best conversion at a ratio of 1:1 CO:H2.
synthesis of DME. A third catalyst functionality Specifically, DME may form directly from coal-
may be included to enhance the water gas shift derived syngas as per Eq. 8. The stoichiometry of
reaction, although the methanol synthesis catalyst Eq. 8 being noted for coal-derived syngas, it is
is itself also an effective water gas shift catalyst. also the case that catalysts for direct DME synthe-
Long-term stability of one-step DME synthesis sis have been based on methanol synthesis catal-
catalyst systems was noted as a primary difficulty ysis and most processes give some combination of
by Mobil in the early 1980s and continued as an both DME and methanol.
issue in Air Products’ work. The temperature
needs and catalyst stability are considered for 3 CO þ 3 H2 ⇆ CH3 OCH3 þ CO2
(8)
three simultaneous reactions in direct synthesis DH 298 ¼ 19:6 kcal=mol CO
of DME: water gas shift (Eq. 4), methanol syn-
thesis (Eq. 5), and methanol dehydration (Eq. 6) The WGS reaction used for one-step
[88]. The methanol synthesis via Eq. 5 can be DME/MeOH production is of the low temperature
considered in principle as the combination of type (LTWGS) and operates best at about
water gas shift and the reduction of CO2 by hydro- 180–240
C, which overlaps substantially with
gen (Eq. 7). the methanol synthesis temperature range. The

Coal to Liquids
Technologies,
Fig. 12 Equilibrium
conversion of CO and H2 to
DME (280
C, 5 MPa)
Ref. [89]
Coal to Liquids Technologies 417

overall chemistry is more easily controlled if the accommodate the lower viscosity of DME com-
WGS catalyst has little or no F-T activity for pared to diesel, but without modification of the
hydrocarbon production, which means that the engine itself) [93]. Turbine, boiler, and fuel cell
LTWGS catalyst needs to be copper-based instead tests with DME have also been performed [94].
of iron-based. The typical catalyst used for JFE in Japan has been a major driver for devel-
LTWGS in DME or methanol synthesis has the opment and commercialization of one-step slurry
general formula xCu/yZnO/Al2O3 and has no DME production. NKK, Taiheiyo Coal Mining,
hydrocarbon-producing F-T activity. A typical and their partners constructed a 5 ton/day direct
low temperature shift catalyst contains 30–70% syngas to DME pilot plant in 1999 at Taiheiyo’s
CuO (prior to reduction), 20%–50% ZnO, and up site in Hokkaido and operated it intermittently for
to about 40% alumina [81]. about 5 years [94]. In 2002, JFE began a 5-year
The optimal temperature for the methanol development plan for a 100 ton/day pilot plant
dehydration reaction is set not by catalyst stability demonstration in Kushiro, with METI and many
but by the equilibrium dehydration reaction ther- collaborators that together formed the DME
modynamics. The dehydration proceeds best at Development Corp. and the DME International
about 250
C and yield declines as the tempera- Corp. Interestingly, the feed for syngas production
ture is raised further. The catalyst function for actually used is natural gas; however, JFE opts to
dehydration of alcohol to ether is a solid acid. bring the 2:1 H2:CO ratio (produced by reforming
Various options are possible, including acidic alu- natural gas) down to 1:1 (more like the ratio
minas, silica-aluminas, zeolites, and various solid produced from coal) by adding CO2 during
or supported inorganic acids. In practice, reforming, and the work therefore also relates to
d-alumina is preferred to date. The dehydration advancement of the state of the art in producing
rate is directly related to the acid strength of the DME and methanol by ICL of coal.
zeolite and the overall rate of DME synthesis can JFE reported an impressive purity from their
be controlled by the proportion of ion exchange in overall process: 99.5 + % DME and <100 ppm
the dehydration catalyst [90]. Dehydration is rel- combined water and methanol [95]. In this pro-
atively fast above about 240
C and only a few cess, shown in Fig. 13, coal is fed to the gasifier
percent of dehydration catalyst is needed in the and then raw syngas is brought up to a H2:CO
total catalyst formulation [91]. The LPDME pro- ratio of 1:1 via the WGSR using steam from the
cess was demonstrated on a 10 ton/day scale for gasifier. Sulfur and CO2 are removed prior to
25 days at Air Products’ LaPorte facility using DME synthesis. Unreacted CO and H2 are sepa-
commercially available methanol synthesis and rated from liquid products and sent back to the
d-alumina dehydration catalyst in a ratio of DME synthesis reactor. The DME synthesis
95:5 wt%. The temperature and pressure were makes CO2 as a by-product, plus some methanol
approximately 250
C and 50 bar, respectively. and a slight amount of water. The separation of
Deactivation occurred at a rate of 0.7% per day for CO2, DME, methanol, and water involves a low
both catalyst components [92]. temperature extraction (39
C) [96]. After
The concern that LPG used heavily in Asia as a downstream distillation, methanol is cycled back
fuel and a power source will become increasingly to the DME reactor.
unavailable has sparked significant research to A different type of ICL process involving
develop DME as a replacement. Properties of methanol is the gasification of coal followed by
DME and LPG are similar in terms of fuel charac- conversion to methanol and subsequent conver-
teristics, low toxicity, and especially handling, dis- sion of methanol to a marketable gasoline prod-
tribution, and storage. Some use of DME as an LPG uct. A 100,000 ton/year MTG process plant was
substitute has recently begun in China. To demon- started in China’s Jincheng, Shanxi province, in
strate the potential replacement of automotive fuel, 2009, as part of a CTL demonstration complex.
fleet tests with DME have been performed (the fuel Since the production of methanol is on a large
injection system requiring modification to volume basis, it is very sensitive to cost of the
418 Coal to Liquids Technologies

Purge gas 35m3N/h

CO2 CO2 Non-reached gas

Natural gas
DME
Coal mine methane 5t/d

Coal

Oxygen

Steam

Liquid/gas separator

Reaction condition
750m3N/h
250-280°C
3-7Mpa

CO,CO2,H2 Methanol
Synthesis gas generation furnace Gas purification column DME synthesis reactor CO2 separation column DME purification column 0.8t/d

Coal to Liquids Technologies, Fig. 13 Process flow diagram of DME direct synthesis from coal (or methane)

feedstock. Thus, in the 2000s, the high price of alone, it has been estimated that an F-T CTL
natural gas in the United States caused the closing plant without capture would release about 0.78
of most methanol plants and the transfer of its tons of carbon dioxide per barrel of product, in
synthesis to locations with cheap natural gas, comparison to a comparable oil refinery emitting
such Trinidad and China. Now, just 10 years about 0.1 tons of carbon dioxide per barrel of
later in the mid-2010s the price of natural gas in product. For a typical bituminous coal, approxi-
the United States has declined enough so that it is mately 670 pounds of carbon (or 2450 pounds of
cheap compared to most other locations so that carbon dioxide) would be emitted for every barrel
methanol plants are being restarted and new plants of F-T liquids produced. This compares to about
are under construction. Thus, China is now 250 pounds of carbon (or 900 pounds of carbon
investing in and constructing methanol plants in dioxide) emitted for every barrel of petroleum
the United States which will produce methanol to fuels produced by refineries. When carbon
be exported to China. sequestration is employed for both facilities
(92% captured and stored at the F-T CTL plant
Environmental Factors and 40% at the refinery), the carbon dioxide emis-
The overall conversion of coal to liquids not only sions become approximately equal.
produces fuel- or chemical-grade products but
also uses water and produces carbon dioxide, as ð4 þ yÞ CH þ 2ðy  xÞ H2 O
shown in Eq. 9 [97]. ! ð4 þ xÞ CHy þ ðy  xÞ CO2 (9)

Carbon Dioxide The large quantity of CO2 produced currently

The high carbon content of coal combined with poses a significant challenge to development of
the energy use in ICL processes involving gasifi- the industry. At the time of this writing, no
cation means that carbon dioxide emissions on a requirements exist in the United States to manage
well-to-wheel basis are about 1.8 times more for carbon emissions from fossil fuel sources, but
coal than for oil. In terms of fuel preparation should carbon management be required, carbon
Coal to Liquids Technologies 419

dioxide produced during the conversion process Fischer-Tropsch Reaction Product Water
could be captured for subsequent storage in deep Water is a net reactant in CTL processes (Eq. 9)
saline aquifers or sold for use in enhanced oil but is a product of the F-T portion of CTL pro-
recovery (EOR) operations. A range-finding cesses (Eq. 1). On a mass basis, approximately 1.3
study done in 2004 by Mitretek Systems (now tons of water is produced per ton of hydrocarbon
Noblis) for the US DOE considered production product. After separation of the hydrocarbon
of substitute natural gas (SNG) from coal and product, the F-T reaction water typically contains
assumed that the value of CO2 for EOR was numerous dissolved oxygenated compounds such
$12/ton, which would significantly improve the as alcohols, aldehydes, acids, ketones, esters, and
economics of a CTL plant [98]. At present (2016), traces of other hydrocarbons. Purification of F-T
there are over 4000 miles of dedicated CO2 pipe- reaction water is required in commercial pro-
lines within the United States delivering to com- cesses, because all water is reused. Distillation
mercial EOR projects in areas such as the Permian removes most of the nonacid oxygenates to
Basin of West Texas, southeastern New Mexico, leave a bottom stream known as F-T acid water,
the Rocky Mountain Region of Utah, Wyoming rich in carboxylic acids. The recovered oxygen-
and Colorado, Louisiana, and the Weyburn Fields ated compounds can be purified to chemical prod-
in Saskatchewan. With the exception of CO2 from ucts or used as a fuel source. After extraction of
the Dakota Gasification Plant, all of this CO2 is the oxygenates, only low concentrations of
naturally occurring. organic acids remain and these are biodegraded
Total production related emissions from a two by either aerobic or anaerobic digestion to pro-
million barrels/day F-T CTL industry could be as duce water of very high quality, for use as coolant
much as a half-gigaton/yr. There is already a sig- or even for drinking, from the original F-T acid
nificant knowledge base regarding how to handle water stream. High purity water suitable for use as
CO2 and inject it into deep geological structures, boiler feed water can also be produced but
because injection of CO2 is already used commer- requires a third set of treatments, after distillation
cially in the United States to augment natural gas and digestion, such as adsorption of trace
production. The transport of large volumes of CO2 suspended solids, dissolved salts, and organic
via pipeline is also well-established practice, material on activated carbon.
within segments of the oil and gas industry. In addition to distillation-based purification of
Given adequately scaled infrastructure for carbon F-T reaction water, membrane-based, liquid-
capture and storage in place, emissions from CTL liquid extraction, and biological operations have
plants and emissions from use of CTL fuels would been developed to remove and recover carboxylic
be comparable to those associated with the pro- acids from F-T acid water [100]. The resultant
duction and consumption of petroleum-based water stream from these processes is generally of
fuels. If sequestration of carbon dioxide becomes a good quality and can be reused or disposed of to
mandatory, an additional $4–$8 per barrel for the the environment.
price of low-sulfur, light crude oil would be After physical separation of oil and water
required for profitable operation, depending on streams, plus steam stripping and oil extraction
specific site logistics [99]. to remove volatiles and aromatics, respectively,
The high sulfur content of most coals is sulfur recovery is performed by conversion to
addressed at multiple junctures in the F-T ICL elemental sulfur. Tail gas streams are also cleaned
process. Sulfur is removed when the raw synthesis for use in recycle or as fuel.
gas is scrubbed to remove CO2 and sulfur com- The stream of water from the cooling system
pounds (predominantly H2S) and again during (referred to as the blowdown stream) must be
purification of F-T product streams. The H2S cleaned also. This water contains inorganic salts
containing CO2 streams are treated in a Claus or and is treated with coal ash that chemically binds
similar process, where the H2S is oxidized to the salts, before the remaining water is collected
elemental sulfur. by evaporation and condensation.
420 Coal to Liquids Technologies

In addition to fuels, chemicals, and purified The usage level of water in F-T CTL plants is
water, F-T ICL processes generate mineral waste considered a significant logistic issue in geo-
derived from the coal. This nonleachable slag is graphical areas of low rainfall or otherwise limited
suitable for sale as aggregate. Other, low-volume water resources. For example, competing use of
solid wastes include spent catalysts, organic water for agriculture would be a problem in the
sludge from biological wastewater facilities, inor- Western United States, and competing use of
ganic sludge from flue gas desulfurization or water for thermal-electric plants would be an
evaporation ponds, and by-product sulfur if not issue in the Eastern United States. The availability
recovered. These wastes may require specialized of water is already an issue in Montana and Wyo-
treatment to avoid transmission of soluble com- ming, where most of the inexpensive supplies of
ponents if present; such as traces of the elements coal in the United States are located. Where water
arsenic, chromium, mercury, molybdenum, and is in short supply, this logistic requirement
selenium, to local water supplies. However, F-T becomes an environmental and/or economic
CTL plants incorporating modern, state-of-the-art issue and therefore has precluded construction of
gasification plants should accomplish removal of some planned CTL plants. Use of air instead of
these and other potential pollutants within the water as coolant may substantially reduce water
normal plant process operations. usage requirements of the exothermic F-T
reaction.
Utility Water
The CTL process chemistry uses water as a source
of hydrogen and oxygen, during gasification of Future Directions
coal, and also produces water during the F-T
reaction. The overall chemistry consumes water, Current commercialization of CTL technology
but the chemical transformations are not the only serves the global interest in sustainability of fuel
use of water during coal liquefaction. The other supply. Any additional production of fuels from
major uses of water are for cooling, boiler feed alternative resources reduces overall world
water, and in solids handling, including pro- demand for petroleum and benefits all consumers
cessing of fine coal or dust. Most of the water is by easing pressure on crude oil prices. As the price
recovered after purification, but in a DCL plant for of oil increases and its easy availability decreases,
example, 3–5% of the water being circulated in a countries with large coal reserves and insufficient
cooling tower is lost to evaporation, leaks, and domestic petroleum to meet increasing transpor-
blowdown [101]. Thus, for every 1000 gallons tation fuel demands – particularly the PRC and
circulated, 30–50 gallons of makeup water is India – are potential candidates for using CTL
needed. technologies to supplement conventional petro-
The amount of water needed varies with the leum based fuel products.
process design, choice of gasifier, ambient tem- In the case of DCL, contribution to sustained
perature and humidity, and type of coal. fuel supply requires that the Shenhua project be
According to the DOE Idaho National Lab, successful and lead to commissioning of addi-
approximately 12–14 barrels of water are used tional units, given estimates of proven coal
for every barrel of F-T liquid fuel produced Shenhua coal reserves in the Shanxi Province/
[102], meaning that 40 million gallons of water Inner Mongolia Autonomous Region of about
per day are needed for an 80,000 barrel/day CTL 200 billion tons [103]. In this context, it is impor-
facility. That is enough water to meet the domestic tant to note that parallel efforts are also underway
needs of more than 200,000 people (or 1/5 the in the PRC and elsewhere to erect commercial
population of the state of Montana). However, scale indirect coal liquefaction plants. Presum-
most of the 12–14 barrels of water are used for ably, these plants have lower technical risks
cooling and are released so that while the plant because the designs are derived from ICL tech-
uses water it does not consume the water. nology that has already been well proven at full
Coal to Liquids Technologies 421

commercial scale, by Sasol. On the other hand, the evaluate the technical and economic feasibility
direct liquefaction project, while supported by a of novel processes resulting from computational
6 ton/day pilot plant constructed in Shanghai, has research; advanced integrated gasification and
no precedent on a larger scale. The more that F-T processing concepts and cleanup technology;
1000-fold scale-up from pilot plant to commercial advanced reactor types needed to improve the
unit is an ambitious objective that carries more process efficiency when utilizing specific regional
than the typical amount of technical risk. These coals; novel in situ reactions/processing; and
plants will also incorporate technologies specific modeling of the gas-solid-liquid physics of the
to regional coals and environmental regulations. F-T reactor to help achieve the highest throughput
If DCL technology is successful in the PRC, it and liquid product quality.
will garner increased global interest, perhaps in Life cycle analyses performed on a “mine-to-
combination with indirect liquefaction in a hybrid wheels” basis likely will be used to identify those
systems configuration. In this scenario, the low parts of the overall system impacting health and
value residue from the direct liquefaction process the environment now. Estimates of the impact of
would be gasified to provide a low cost source for CO2 on global warming, and estimates of the
synthesis gas production, rather than incurring a impacts of global warming itself, may also be
cost penalty for disposal. A recent US National attempted but these projections appear at present
Coal Council (NCC) report suggests that this as far too broad in range to aid in resolving issues
might be a good time and justification to revisit of energy sustainability. The greatest current push
the direct liquefaction pathway, particularly if the for the future use of coal appears to be in devel-
scale-up in the PRC is successful. The NCC report oping sustainable concepts for capturing and then
also suggests the speculative possibility of hybrid storing or using the CO2, because doing such
direct and indirect plants. obviates a significant portion of the large uncer-
In the case of ICL, the F-T technology is tainty as to potential effects of that CO2 were it not
already well commercialized but currently has a captured. A related type of research involves com-
very limited worldwide aggregate production vol- bining coal use with chemistries or processes that
ume. Computational science is a rapidly generate less (or consume some amount of) CO2,
expanding area of research that offers the potential to decrease the overall output of CO2 per barrel of
to shortcut development time by providing a the- oil equivalent produced.
oretical basis for subsequent R&D and technology In addition to the freestanding F-T CTL pro-
design activities. It provides the opportunity to cess configuration, the basic plant can be
quickly explore novel processing strategies to reconfigured to capture about 90% of the carbon
guide and accelerate experimental research. The that otherwise would be released into the atmo-
computational work focuses on the complex and sphere within the plant boundary. For example,
critical chemical and physical aspects of Sasol captures more than 90% of the CO2 for their
converting coal to premium fuels: the fundamen- plants in South Africa in their gas cleaning pro-
tals of catalyst activity/selectivity; hydrodynam- cess; however, since there is no use for CO2 in the
ics; separation of small catalyst particles from the vicinity of the plant, the captured CO2 is released.
liquid product; optimum system integration to Also, another option, not involving CO2 recycle,
achieve high process efficiency, minimal pollutant would be to design a polygeneration plant that
emissions, and CO2 capture; and computational produces both fuels and power.
frameworks to enable virtual demonstration of the More production from existing mines and the
entire fuel life cycle. opening of new coalmines will be required to
Laboratory research and additional modeling, accommodate growth in the CTL industries as
followed by larger-scale bench and pilot-scale well as growing global demand for electric
testing, could experimentally verify the computa- power. These facilities will require land for con-
tional findings. However, systems engineering is struction and support operations such as roads,
an important component activity needed to railroads, and storage facilities that create changes
422 Coal to Liquids Technologies

in land use, alter topography, and impact ecolog- and mercury will be removed, perhaps by some
ical systems. Mitigation and reclamation strate- combination of pre- and post-combustion pro-
gies would need to be implemented to offset cesses. Using air coolers where possible will min-
some of these changes to land and ecological imize water use, and solids emissions will consist
resources. Currently, there are very strict reclama- of nonleachable slag from the gasification pro-
tion regulations in place and they are being further cess. Because of the sensitivity of the F-T catalyst
enhanced to encourage a reforestation initiative to to poisons, all contaminants must be removed to
better restore the land, while concurrently improv- near-zero levels (ppb levels) thus ensuring that
ing water quality and providing a source for car- overall plant emissions would be close to zero.
bon sequestration. Even if the environmental risks Certainly coal conversion to hydrocarbon liq-
are addressed, there is a very good possibility of uids and electricity is now economically attractive
public reluctance to accept the need for new facil- from a cash cost viewpoint. However, large capi-
ities, particularly these coal-based plants. Mea- tal investments are required and small to medium
sures would need to be taken to involve the coal conversion plants will not become a reality
general public and other state, local, and non- unless there are effective forms of state assistance
governmental entities to assure them that these with the initial capital investment. This would be
plants could effectively protect human health motivated by the desire for independence from
and safety and the environment. The increasing imported energy and the more efficient and
production of natural gas through fracking will cleaner utilization of the local energy resource.
slow the growth of coal utilization. The Chinese government has accepted this
However, F-T CTL plants would be expected approach, and China is rapidly becoming the pio-
to have very low environmental criteria pollutant neer for developing commercial CTL technology
emissions. They could also be designed to accom- options. However, it is the duty of technology
modate carbon capture and pressurization for sub- providers to ensure that technology is made avail-
sequent sequestration in saline aquifers or oil able to allow coal use to be as efficient as possible.
reservoirs for enhanced oil recovery (EOR). Site-specific early design studies would pro-
Therefore, these plants could remove a good vide the ability to obtain information on environ-
deal of uncertainty associated with possible new mental baselines for CTL plants. These plants
environmental regulations. Early plants that would be ready for CO2 separation and capture
would sell or demonstrate CO2 use for EOR and the information obtained would define
would be encouraged. resource requirements. Site-specific information
Regarding methanol and DME, natural gas is would also address where the resources, such as
currently the preferred feedstock for methanol coal and water, are coming from, how they are
production but companies such as Eastman delivered and how waste products are to be reused
Chemical in particular have rendered the coal- or disposed. Additionally, current R&D activities
based process commercial, and cost-effective cosponsored by the DOE and industry are being
technology for CO2 mitigation could pave the pursued to improve CO2 separation and capture
way for significant increase in commercial pro- and define CO2 storage sinks.
duction of methanol and DME from coal in Asia Making fuels from coal via either DCL or ICL
and elsewhere. processes has the advantage that fuel infrastruc-
Future CTL plants will use advanced clean tures already in place for petroleum crude oil
coal gasification technology to produce transpor- products can be used largely unchanged when a
tation fuels and optionally electric power. Pollut- shift is made to coal-derived fuels. The advantage
ant emissions will be minimal because coal- of fuel infrastructure compatibility offered by syn-
derived sulfur will be removed and converted thetic hydrocarbon fuels is not so great in the PCR
into elemental sulfur. Nitrogen oxides will be at present, where a liquid hydrocarbon fuel infra-
minimized using low-NOx burners and selective structure for transportation fuels is in a highly
catalytic reduction (SCR) in the flue gas stream developmental state. The challenges are primarily
Coal to Liquids Technologies 423

economic, as prospective air pollutant regulatory 9. NEDO (2006a) Clean coal technology in Japan. 4A1.
constraints worldwide give high value to clean Coal Liquefaction Technology Development in
Japan. pp 57–58. www.nedo.go.jp/kankobutsu/pam
synthetic fuels having emission characteristics phlets/sekitan/cct2006e.pdf. Accessed 10 Apr 2010
superior to crude oil derived fuels, whereas coal- 10. Sugawara A, Kurosawa S, Hatori H, Saito K,
based synfuels processes incur significant cleanup Yamada Y, Sugihara M, Wasaka S, Yoshida H,
costs. Much, although not all, of the contaminants Seo T, Susuki T, Inoguchi M, Sohnai M (1998)
Coal conversion technologies on the new Sunshine
needing to be cleaned up result from the biological program in Japan. Prepr Symp Am Chem Soc Div
origins of coal itself. In that regard, clean coal Fuel Chem 43(2):330–334
research might also provide insight on more effi- 11. NEDO (2006b) Clean coal technology in Japan. 4A2.
cient processing and cleanup in emerging Bituminous coal liquefaction technology (NEDOL).
pp 59–60. www.nedo.go.jp/kankobutsu/pamphlets/
biomass-to-liquids processes. sekitan/cct2006e.pdf. Accessed 10 Apr 2010
12. Ghosh TK, Prelas MA (2009) Energy resources and
systems: volume 1: fundamentals and non-renewable
resources. Springer, Berlin
Bibliography 13. Dry ME (2002) The Fischer-Tropsch process:
1950–2000. Catal Today 71:227–241
1. Appert O (2008) CTL could provide long-term, 14. Jager B, Espinoza R (1995) Advances in low tem-
transport-fuel supply. SynGas Refiner IV(9):14–18 perature Fischer-Tropsch. Catal Today 23(1):17–28
2. Gray D, White C, Tomlinson G, Ackiewicz M, 15. Radtke K, Heinritz-Adrian M, Marsico C (2006)
Schmetz E, Winslow J (2007) Increasing security New wave of coal-to-liquids. An opportunity to
and reducing carbon emissions of the decrease dependency on oil and gas imports and an
U.S. transportation sector: a transformational role appropriate approach to a partial revival of domestic
for coal with biomass. National Energy Technology coal industries. VGB PowerTech 86(5):78–84
Laboratory, Pittsburgh, p 61. DOE/NETL-2007/1298 16. Williams RH, Larson ED (2003) A comparison of
3. De Klerk A (2009) Overview of coal-to liquids tech- direct and indirect liquefaction technologies for mak-
nology. In: Klerk D (ed) Beyond Fischer-Tropsch. ing fluid fuels from coal. Energy Sustain Dev VII
Elsevier, Amsterdam, p 1. Arno de Klerk comments, (4):103–129
“Coal-to-liquids technology started as an ironic twist 17. Grant-Huyser M, Maharaj S, Matheson L, Rowe L,
of fate seen against the present day drive to move Sones E (2004) Ethoxylations of detergent-range
towards renewable sources of energy, since coal Oxo alcohols derived from Fischer-Tropsch alpha-
replaced whales as feed material for the production olefins. J Surfactant Deterg 7(4):397–407
of lamp oil (kerosene). Coal is not a renewable 18. Price JG (2004) Chemicals from synthesis gas. US
energy source, except when viewed on a geological 6740683 to Sasol technology, 25 May
time scale, but whales are, yet, the whaling industry 19. Degnan TF Jr, Chen NY, Somorjai GA (2009) Heinz
came close to making whales a non-renewable Heinemann’s legacy at ExxonMobil: an illustrious
resource by converting all whales into oil.” The anec- career in industrial catalysis. Catal Lett
dote illustrates a growing understanding among 133(1–2):227–230
energy experts, that even with conservation global 20. Brown D, Bhatt B, Hsiung T, Lewnard J, Waller
demand for energy is growing and can best be met by F (1991) Novel technology for the synthesis of
careful use of both renewable and nonrenewable dimethyl ether from syngas. Catal Today 8:279–304
resources. On this point, see also Crane et al 21. Sardesai A, Lee S (2005) Alternative source of pro-
4. Crane H, Kinderman E, Malhotra R (2010) A cubic pylene. Energy Sources 27(6):489–500
mile of oil. Oxford University Press, New York 22. Larson ED, Yang H (2004) Dimethyl ether (DME)
5. Thomson E (2002) The Chinese coal industry: an from coal as a household cooking fuel in China.
economic history. Routledge, London Energy Sustain Dev VII(3):115–126
6. Lister G (1944) Chronological records of coal min- 23. Jones Jr, G, Holm-Larsen H, Romani D, Sillis
ing, transport etc. in Northumberland and Durham R (2001) DME for power generation fuel: supplying
from a.D. 1180 to 1839. Ramsden Williams, Consett India’s Southern Region. PETROTECH-2001 08–12
7. Perry H (1974) Coal conversion technology. Chem Jan, New Delhi
Eng 81(15):88–102 24. Bhatt B, Schaub E, Heydorn E. (1993) Recent devel-
8. Murakushi N (1981) The transfer of coal-mining opments in slurry reactor technology at the LaPorte
technology from Japan to Manchuria and manpower alternative fuels development unit. In: International
problems : focusing on the development of the technical conference on coal utilization & fuel sys-
Fushun coal mines. Japanese Experience of the tems, LaPorte, pp 197–208. 26–29 Apr
UNU Human and Social Development Programme 25. Shikada T, Ohno Y, Ogawa T, Ono M, Mizuguchi M,
series, Tokyo, p 47 Tomura K, Fujimoto K (1998) Direct synthesis of
424 Coal to Liquids Technologies

dimethyl ether from synthesis gas. In: Parmaliana A, 40. Strobel BO, Loering R (1992) IGOR – taking the
Sanfilippo D, Frusteri F, Vaccari A, Arena F (eds) short cut in coal hydrogenation. Prepr Pap Am
Studies in surface science and catalysis 119 (Natural Chem Soc Div Fuel Chem 37(1):448–455
Gas Conversion V). Elsevier, Amsterdam, The Neth- 41. NEDO (2006c) Clean coal technology in Japan. 4A3.
erlands, pp 515–520 Brown coal liquefaction technology (BCL).
26. Miller CL, Cicero D, Ackiewicz M, Anderson J, pp 61–62. www.nedo.go.jp/kankobutsu/pamphlets/
Schmetz E, Winslow J (2006) Coal conversion – a sekitan/cct2006e.pdf. Accessed 10 Apr 2010
rising star? In: 23rd international pittsburgh coal con- 42. Lumpkin RE (1988). Recent progress in the direct
ference, Pittsburgh, PA, 25–28 Sept liquefaction of coal. Science, 239, p 873–877.
27. White LC, Frederick JP (1995) ENCOAL mild coal 43. Kandiyoti R, Herod A (2016) Solid fuels and heavy
gasification project. In: Proceedings – annual inter- hydrocarbon liquids: thermal analysis. Elsevier,
national Pittsburgh coal conference, 12th, Pittsburgh, Amsterdam
PA, pp 151–156 44. Tremblay J-F (2017) China’s key feedstock: all of the
28. U.S. Department of Energy, National Energy Tech- above. Chem Eng News 95(12):22–24
nology Laboratory (2002) The ENCOAL ® mild coal 45. Keogh RA, Davis BH (1988) Comparison of analyt-
gasification project. A DOE Assessment. ical results obtained for the determination of oils,
DOE/NETL-2002/1171 asphaltenes, and preasphaltenes. J Coal Qual 7:27
29. McMillen DF, Malhotra M (1989) The role of hydro- 46. Keogh RA, Tsai K, Xu L, Davis BH (1991) Lique-
gen transfer in bond-cleavage and bond-forming pro- faction pathways of U.S. bituminous coals. Energy
cesses during coal conversion, Chapter 4. In Fuel 5:625
Schindler HD (ed) Coal liquefaction. A research 47. Keogh RA, Davis BH (1994) Comparison of lique-
needs assessment DE-AC01-87ER30110 final report, faction pathways of a bituminous and subbituminous
vol 2. U.S. Department of Energy, Washington, DC, coal. Energy Fuel 8:289
pp 30–49 48. Davis BH, Hower JC (in press) Coal technology for
30. Haenel MW (2008) Catalysis in direct coal liquefaction. power, liquid fuels and chemicals. In: Kent JA
In: Gerhard E (ed) Handbook of heterogeneous cataly- (ed) Handbook of industrial chemistry and biotech-
sis, vol 6, 2nd edn. Wiley, Weinheim, pp 3023–3036 nology, vol 1, 12th edn. Springer, New York,
31. Miller L (2008) Coal conversion technology. Con- pp 749–895. 13th Ed
gressional noontime briefing. Rayburn House Office 49. Dadyburjor DB, Liu Z, Davis BH (2012) Coal lique-
Building, Washington, DC faction in Kirk-othmer encyclopedia of chemical
32. Eccles RM, DeVaux GR (1982) H-coal commercial- technology. Wiley, New York
ization: current status. Energy Prog 2(2):111–115 50. Davis BH (2017) Coal to liquid conversion process: a
33. Cleaner Coal Technology Programme (1999) (PDF). review. In: Fuels and lubricants handbook, ASTM
Technology status report 010: coal liquefaction. Publication MNL37, 2nd edn. West Conshohocken,
Department of Trade and Industry. http://www.dti. PA
gov.uk/files/file18326.pdf. Accessed 10 Apr 2010 51. Malhotra R (2005) Direct coal liquefaction: lessons
34. Styles GA (1982) The Wilsonville advanced coal leanred, GCEP Advanced coal workshop, BYU,
liquefaction R&D facility – accomplishments and Provo, 16 Mar
process evolution. Energy Prog 2(3):160–162 52. Ramdoss PK, Tarrer AR (1997) Modeling of two-
35. Valente AM, Cronauer DC (2003) Progress in coal stage coal coprocessing process energy fuels. Energy
liquefaction including a discussion of Wilsonville. Fuel 11(1):194–201
Prepr Symp Am Chem Soc Div Fuel Chem 53. Wender I, Tierney JW (1995) Effect of pretreating of
48(1):147–148 host oil on coprocessing. Final report. DOE/PC/
36. Burke FP, Brandes SD, McCoy DC, Winschel RA, 91054–T15, Contract Number AC22-91PC91054,
Gray D, Tomlinson G (2001) Summary report of the Oct. 1 1995 Oct 01
DOE direct liquefaction process development cam- 54. Miller RL, Giacomelli GF, McHugh KJ, Baldwin
paign of the late twentieth century: topical report. RM (1989) Coprocessing of coal and residuum
DE2002–794281/XAB; DOE/PC-93054-94, Depart- under low-severity reaction conditions: effect of
ment of Energy, Washington, DC basic nitrogen promoters. Energy Fuel 3(2):127–131
37. Nolan P, Shipman A, Rui H (2004) Coal liquefaction, 55. Smith R, Asaro M, Naqvi S (2008) Fuels of the
Shenhua Group, and China’s energy security. Eur future: technology intelligence for coal to liquids
Manag J 22(2):150–164 strategies. SRI Consulting, Menlo Park
38. Stohl FV, Lott SE, Diegert KV, Goodnow DC 56. Koornneef J, Junginger M, Faaij A (2007) Develop-
(1996) Results of hydrotreating the kerosene fraction ment of fluidized bed combustion – an overview of
of HTI’s first proof of concept run. U.S. DOE, trends, performance and cost. Prog Energy Combust
SAND–96-0990C; CONF-960807–2 Sci 33(1):19–55
39. Friedrick F, Strobel BO (1983) The KOHLEOEL 57. Senapati PK, Das D, Nayak A, Mishra PM
experimental plant at Bergbau-Forschung. In: Pro- (2008) Studies on preparation of coal water slurry
ceedings of the European coal utilisation conference, using a natural additive. Energy Sources Part
3rd, Amsterdam, The Netherlands, pp 227–235 A 30(19):1788–1796
Coal to Liquids Technologies 425

58. Hagman C (2016) State of the gasification industry: 75. Bartholomew C, Bukur DB, Datye AK, Nowicki L,
commercial applications & research and develop- Pham HN, Xu J (2003) Attrition resistance of sup-
ment, ECUST OMB Technology Conference, Nan- ports for iron Fischer-Tropsch catalysts. Ind Eng
jing, 7 Nov Chem Res 42(17):4001–4008
59. Consulting SRI (2006) Coal gasification. Process 76. Zhao R, Goodwin JG, Jothimurugesan K,
economics report 154A, chapter 4. SRI Consulting, Gangwal K, Spivey JJ (2001) Spray-dried iron
Menlo Park, pp 4–46 Fischer-Tropsch catalysts.1. Effect of structure on
60. Davis BH (2001) Fischer-Tropsch synthesis: current the attrition resistance of the catalysts in the calcined
mechanism and futuristic needs. Fuel Process state. Ind Eng Chem Res 40(4):1065–1075
Technol 71(1–3):157–166 77. Demirel B, Bohn MS, Benham, CB, Siebarth JE,
61. Leckel D (2005) Hydrocracking of iron-catalyzed Ibsen, MD (2005) Method and apparatus for
Fischer-Tropsch waxes. Energy Fuel 19: regenerating an iron-based Fischer-Tropsch catalyst.
p1795–p1180 US 6838487 to Rentech, 4 Jan
62. Davis BH (2007) Fischer-Tropsch synthesis: com- 78. Benham CB, Bohn MS, Yakobson DL (1996) Process
parison of performances of iron and cobalt catalysts. for the production of hydrocarbons US 5504118 to
Ind Eng Chem Res 46:8938–8945 Rentech, 2 Apr
63. Smith R, Asaro M (2005) Fuels of the future: tech- 79. van der Merwe W (2010) Conversion of spent solid
nology intelligence for gas to liquids strategies. SRI phosphoric acid catalyst to environmentally friendly
Consulting, Menlo Park fertilizer. Environ Sci Technol 44(5):1806–1812
64. Bukur DB (2005) Attrition studies with catalysts and 80. Peng X, Toseland B, Underwood T (1997) A novel
supports for slurry phase Fischer-Tropsch synthesis. mechanism of catalyst deactivation in liquid phase syn-
Catal Today 106:275–281 thesis gas-to-DME reactions. In: Bartholomew C, Fuen-
65. Bukur DB, Carreto-Vazquez VD, Pham HN, Datye tes GH (eds) Catalyst deactivation. Elsevier, Amsterdam
AK (2004) Attrition properties of precipitated iron 81. Cai Y, Davies S, Wagner J. (2003) Water gas shift
Fischer-Tropsch catalysts. Appl Catal A Gen catalyst. US 6627572 to Süd-Chemie, 30 Sept
266(1):41–48 82. Cornthwaite D (1972) British patent 1296212 to
66. Dry M (2004) Chemical concepts used for engineer- Imperial Chemicals Inc. (ICI)
ing purposes. In: Steynberg A, Dry M (eds) Studies in 83. Espino RL, Plezke TS (1975) Methanol production.
surface science and catalysis 152. Fischer-Tropsch US 3888896 to ChemSystems, 10 June
technology. Elsevier, Amsterdam, pp 196–257 84. Sherwin M, Blum D (1979) Liquid-Phase methanol.
67. Smith R (2009) SRI Consulting (2006) coal to gaso- Final report. Electric Power Research Institute, EPRI-
line. Process economics report 271. SRI Consulting, AF-1291
Menlo Park. Data for the figure were obtained from 85. Lee S, Sardesai A (2005) Liquid phase methanol and
reference 57 dimethyl ether synthesis from syngas. Top Catal
68. Donnelly TJ, Yates IC, Satterfield CN (1988) Analy- 32(3–4):197–207
sis and prediction of product distributions of the 86. Air Products Liquid Phase Conversion Company
Fischer-Tropsch synthesis. Energy Fuel 2:734–739 (2004) Commercial-scale demonstration of the
69. Donnelly TJ, Satterfield CN (1989) Product distribu- liquid-phase methanol (LPMEOH) process. DOE/
tions of the Fischer-Tropsch synthesis on precipitated FE-0470. U.S. Department of Energy
iron catalysts. Appl Catal 52(1):93–114 87. Joo O, Jung K, Han S (2002) Modification of
70. Zhan X, Davis BH (2000) Two-alpha Fischer- H-ZSM-5 and gamma-alumina with formaldehyde
Tropsch production distribution. A role for vapor- and its application to the synthesis of dimethyl ether
liquid equilibrium. Pet Sci Technol 18:1037–1053 from Syn-Gas. Bull Kor Chem Soc 23:1103–1105
71. Shi B, Davis BH (2004) Fischer-Tropsch synthesis: 88. Lee S, Parameswaran VR, Wender I, Kulik CJ
accounting for chain-length phenomena. Appl Catal (1989) Roles of carbon dioxide in methanol synthe-
A Gen 277:61–69 sis. Fuel Sci Technol Int 7(8):1021–1058
72. Jager B, Kelfkens RC, Steynberg AP (1994) A slurry 89. NEDO (2006d) Clean coal technology in Japan. 4A4.
bed reactor for low temperature Fischer-Tropsch. In: Dimethyl ether production technology (DME),
Curry-Hyde HE, Howe RF (eds) Natural gas conver- pp 63–64. www.nedo.go.jp/kankobutsu/pamphlets/
sion II. Elsevier, Amsterdam, pp 419–425 sekitan/cct2006e.pdf. Accessed 10 Apr 2010
73. Adeyiga AA, Bukur DB, Carreto-Vazquez V, Ma W, 90. Kim J, Park MJ, Kim SJ, Joo O, Jung K (2004) DME
Nowicki L (2004) Attrition resistance and catalytic synthesis from synthesis gas on the admixed catalysts of
performance of spray-dried iron Fischer-Tropsch cat- cu/ZnO/Al2O3 and ZSM-5. Appl Catal A Gen 264(1):37–41
alysts in a stirred-tank slurry reactor. Ind Eng Chem 91. Gogate M, Lee S (1991) A single- stage, liquid-phase
Res 43(6):1359–1365 dimethyl ether synthesis process from syngas. I. Dual
74. Bai L, Chang J, Hao Q, Li Y, Liu F, Wang H, catalytic activity and process feasibility. Fuel Sci
Xiang H, Xu B, Yi F, Zhang C (2007) Effect of Tech Intl 9(6):653–679
reduction temperature on a spray-dried iron-based 92. Tijm, PJ (2003) Development of alternative fuels and
catalyst for slurry Fischer-Tropsch synthesis. J Mol chemicals from synthesis gas. DOE Contract number
Catal A Chem 261(1):104–111 FC22-95PC93052, Final report
426 Coal to Liquids Technologies

93. Hayashi H, Todoroki A, Yasuto A, Ohno Y (2002) Fraser I, Xu L (eds) Proceedings of the 15th annual
NKK DME diesel vehicle development and fleet test conference of the Association for Chinese Econom-
in Japan. In: International symposium on alcohol ics Studies Australia (ACESA). RMIT (Royal Mel-
fuels, 12–15 Nov bourne Institute of Technology) Business Research
94. Ohno Y, Tanishima S, Aoki S (2005) Coal conversion Development Unit, Melbourne, Australia
into dimethyl ether as an innovative clean fuel. In:
International conference on coal science and technol-
ogy, ICCST, 9–14 Oct Reviews
95. Mogi Y, Ohno Y, Ogawa T, Inoue N, Shikada Dry ME (2001) High quality diesel via the Fischer-Tropsch
T (2000) Development of slurry phase dimethyl process- a review. J Chem Technol Biotechnol
ether synthesis technology. In: Pittsburgh coal con- 77(1):43–50
ference, pp 398–408, 11–14 Sept Egan CJ (1976) Method of power generation via coal
96. Ohno Y, Yagi H, Inoue N, Okuyama K, Aoki gasification and liquid hydrocarbon synthesis US
S (2007) Slurry phase DME direct synthesis 3986349 to Chevron Research Company, 19 Oct
technology – 100 tons/day demonstration plant oper- Furstner AE (1995) In: Furstner A (ed) Active metals:
ation and scale up study. In: Noronha FB, Schmal M, preparation characterization applications. Wiley, Hobo-
Sousa-Aguiar EF (eds) Studies in surface science and ken, NJ
catalysis: natural gas conversion VII, vol 167. Iglesia E, Reyes S, Madon RS (1993a) Selectivity control
Elsevier, Amsterdam, pp 403–408 and catalyst design in the Fischer-Tropsch synthesis:
97. Probstein RF, Hicks RE (2006) Synthetic fuels. sites, pellets, and reactors. Adv Catal 39:221–302
Dover Publications, Inc, Mineola Iglesia E, Reyes S, Soled S (1993b) In: Becker ER, Pereira
98. Gray D, Salerno S, Tomlinson G, Marano JJ CJ (eds) Computer-aided design of catalysts and reac-
(2004) Polygeneration of SNG, Hydrogen, power, tors. Marcel Dekker, New York, p 640
and carbon dioxide from texas lignite. Mitretek tech- O’Brien R, Xu L, Bao S, Raje A, Davis B (2000) Activity,
nical report for the DOE, MTR-04-2004-18, Dec selectivity and attrition characteristics of supported iron
99. Camm F, Bartis JT, Bushman CJ (2008) Federal Fischer-Tropsch catalysts. Appl Catal A Gen
financial incentives to induce early experience pro- 196(2):173–178
ducing unconventional liquid fuels. Report prepared Jin Y, Wang J, Wang T (2007) Slurry reactors for gas-to-
for the United States Air Force and the National liquid processes: a review. Ind Eng Chem Res
Energy Technology Laboratory of the DO. http:// 46(18):5824–5847
www.rand.org/pubs/technical_reports/TR586/. Rahmin II (2008) GTL, CTL finding roles in global energy
RAND Corporation. Accessed 10 Apr 2010 supply. Oil Gas J 24:22–31
100. Kohler LPFD, Du Plessis GH, Du Toit FJ, Koper EL, Sloane D (2008) Clean Coal technologies accelerating
Phillips TD, Van Der Walt J (2006) Method of puri- commercial and policy drivers for deployment. Inter-
fying Fischer-Tropsch derived water. US 7153432 to national Energy Agency/Coal Industry Advisory
Sasol technology, 26 Dec Board, 70 pages
101. U.S. Department of Energy (2006) Emerging issues Tijm PA, Waller F, Brown D (2001) Methanol technology
for fossil energy and water. DOE/NETL-2006/1233 developments for the new millennium. Appl Catal
102. Boardman R (2007) Gasification and water nexus. A Gen 22(1–2):275–282
Presented March 14, 2007 at the GTC Workshop on Williams A, Pourkashanian M, Jones, JM (2001) Combus-
Gasification Technologies, Denver tion of pulverized coal and bimoass. Progress in energy
103. Wei W (2003) Current issues of China’s coal indus- and combustion science, pp 587–610
try: the case of Shanxi. In: Coate B, Brooks R,
 IGCC Integrated gasification combined cycle
CO2 Reduction and MIT Massachusetts Institute of Technology
Coal-Based Electricity Mpa Megapascal
Generation MW Megawatt
MWh Megawatt-hour
János Beér NETL National Energy Technology Laboratory
MIT, Cambridge, MA, USA (USDOE)
NOx Nitrogen oxides
NRC National Research Council (USA)
Article Outline OECD Organization for Economic Cooperation
and Development
Glossary Oxy Oxygen-blown combustion
Introduction Oxy/FGR Oxygen-blown combustion with flue
Pulverized Coal in Ultra-Supercritical Steam gas recirculation
Cycle (PC/USC) PC Pulverized coal
Coal Gasification Combined Gas Turbine-Steam PC/SC Supercritical steam/pulverized coal
Cycle (IGCC) PC/USC Ultra-supercritical steam/pulverized
CO2 Capture and Compression in PC and IGCC coal
Plant PM Particulate matter
PC/Oxy/USC psi Pounds per square inch
CFBC with Oxy/FGR R&D Research and development
Concluding Remarks RD&D Research, development, and demonstration
Bibliography SO2 Sulfur dioxide
TCR Total capital requirement
Glossary

C Deg. Celsius Introduction

F Deg. Fahrenheit
AUSC Advanced ultra-supercritical steam Btu The provision of electric power is one of the pre-
British thermal unit requisites of prosperity; there is strong correlation
CCS Carbon capture and storage (sequestration) between electric power-generating capacity and
CFBC Circulating fluidized bed combustion per capita gross domestic product (GDP) (Fig. 1)
CO2 Carbon dioxide [10]. Across the world, economic indicators sig-
COE Cost of electricity nal that there will be continued growth and
DOE U.S. Department of Energy increased electricity demand; also, at least 3.6
EIA Energy Information Administration billion people lack adequate access to electricity
(USDOE) and 1.6 billion have no electricity at all.
EPRI Electric Power Research Institute During the next 15 years, 1200 GW of new
GW Gigawatt capacity is projected to be added to the world’s
Hg Mercury present electric generation capacity of about 4000
GNP Gross national product GW. The EIA projects that by 2030 global demand
HHV Higher heating value for electricity will grow by more than 75%.
HRSG Heat recovery steam generator The U.S. Energy Information Administration
IEA International Energy Agency (EIA) data for 2008 show that 54% of the elec-

© Springer Science+Business Media, LLC 2012 427

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_74
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media LLC, 2012
https://doi.org/10.1007/978-1-4419-0851-3_74
428 CO2 Reduction and Coal-Based Electricity Generation

Prosperity and Electricity

15000

29 Countries
Electric Generating Capacity Per Capita (W)

10000
900 Million
People
5000

2000 35 Countries
800 Million
13%
1000 People
59 Countries 15%
750 2800 Million
People
500
47% 25%
250 35 Countries
1500 Million
125 People

0
500 1000 2000 3000 4000 5000 7500 15000 30000
GNP per Capita (2000 U.S.$)
1 Copyright @ 2003 Electric Power Research Institute, Inc All rights reserved.

CO2 Reduction and Coal-Based Electricity Generation, Fig. 1 Prosperity and electricity. (Source: EPRI [10])

CO2 Reduction and 3,500 History Projections

Coal-Based Electricity
Generation, Fig. 2 US Electricity demand
3,000 4,804
Electricity generation by
fuel, 1970–2020 (billion
2,500 1,392
kilowatt-hours). (Source: Coal
US EIA [6])
2,000 1970 2020 Natural Gas

1,500

1,000

Nuclear
500 Renewables

0 Petroleum
1970 1980 1990 2000 2010 2020

tricity supply in the USA was generated by coal, and, especially, natural gas for power generation
12% by natural gas (increasing to an estimated projected for the period from 2000 to 2020 is to
32% by 2020), and about 2% by oil, with the rest meet growing demand for electricity and to offset
produced by nuclear power (21%), hydropower the projected retirement of nuclear power plants.
(9%), and by renewable solar or wind (2%). Because of the large coal reserves in major
Electricity generation (billion kilowatt-hours) by developing countries such as China, India, and
fuel type from 1970 to 2020 in the USA is illus- Indonesia, where most of the new power-
trated in Fig. 2 [22]. The increase in the use of coal generating plants are being installed, it can be
CO2 Reduction and Coal-Based Electricity Generation 429

expected that coal will remain the dominant US emissions reductions have been significant
source of power generation worldwide during in the face of increasing output of electricity gen-
the first half of this century. There is also a eration coal (Fig. 3) [15].
strongly growing demand for natural gas, a clean The reduction of CO2 emission remains a
fuel capable of being used in power generation challenge. Carbon capture and geological
with high efficiency. sequestration (CCS) is the key enabling technol-
Coal as an energy source has the attraction of ogy for the long-term significant reduction of
broad availability (there are large reserves in sev- CO2 emissions from coal-based power genera-
eral countries around the world), low cost, and tion. It is expected that CCS will become com-
economic utilization by mature technologies. mercial (deployed without significant
The disadvantages are due mainly to environ- government subsidy) for base load power gener-
mental and health impacts at the mining phase and ation around 2020–2025 following the construc-
in the course of coal utilization. Both the mining and tion and operation of several demonstration
the utilization of coal are subject to increasingly plants during the present and next decade.
tightening standards by governmental regulations. Before the advent of commercial CCS, how-
Because of no acceptable practical alternative ever, higher efficiency technology options are
to coal and natural gas, industrial research and available for coal-based power generation.
development programs since the 1970s have Improved efficiency will reduce the amount of
been concentrated on methods of reducing the fuel used for the generation of a given electric
emissions of so-called criteria pollutants (SO2, energy output and hence reduce the emissions of
NOx, Hg, and fine particulates). New environmen- all pollutants. Also, when the time comes for CCS
tal technologies developed and demonstrated deployment, higher efficiency will reduce the
under the Clean Coal Technology Program of energy penalty and make new plant or retrofit of
the U.S. Department of Energy (DOE) during existing plant less expensive [1, 3, 7].
the past 40 years have become commercial and Main higher efficiency technology options
are being applied worldwide. discussed below are:

Emissions from coal have decreased 187%

200% 84% while electricity from coal has
increased 187%

150%

100%

50%

Emissions
0%
Electricity From Coal

–50%

–100%
–84% 1970 - 2007

CO2 Reduction and Coal-Based Electricity Generation, Fig. 3 US emissions of criteria pollutants and electricity
from coal during 1970–2007. (Source: NCC Report [15])
430 CO2 Reduction and Coal-Based Electricity Generation

• Pulverized Coal Combustion in Ultra- of the combustion air through the burners, but this
supercritical Steam Cycle increase in parasitic energy is usually small com-
• Integrated Coal Gasification Combined Gas pared to the efficiency gain due to the reduced
Turbine-Steam Cycle (IGCC) excess air.
• PC or CFB/Oxy/Combustion in Supercritical The boiler exit gas temperature can be reduced
Steam Cycle by appropriate boiler design limited only by the
dew point of the flue gas. There is a close rela-
tionship between the excess air of combustion and
Pulverized Coal in Ultra-Supercritical the low limit of exit gas temperature from a boiler
Steam Cycle (PC/USC) fired by a sulfur-bearing fuel. Higher excess air
leads to an increase in the oxidation of SO2 to
The efficiency of PC/SC power plant can be SO3, with SO3 promoting sulfuric acid formation
increased in small steps to 43% (HHV) and in the combustion products. Sulfuric acid vapor
beyond, as illustrated in Fig. 4 [19]. increases the dew point of the flue gas and hence
The first two steps in the diagram concern the raises the permissible minimum exit gas tempera-
waste gas heat loss, the largest of a boiler’s heat ture. At an exit gas temperature of 130  C (266  F)
losses, about 6–8%. The air ratio, usually called a reduction of every 10  C (18  F) in boiler exit
excess air factor, represents the mass flow rate of temperature increases the plant efficiency by
the combustion air as a multiple of the theoreti- about 0.3%.
cally required air for complete combustion. The Rankine cycle efficiency is proportional to
Excess air increases the boiler exit gas mass flow the pressure and temperature of heat addition to
and, hence, the waste gas heat loss. Improved the cycle, and is inversely proportional to the
combustion technology, e.g., finer coal grinding condenser pressure, and therefore to the tempera-
and improved burner design, permit lowering the ture of the cooling medium. The usual design
excess air without sacrificing completeness of basis for condenser pressure in the USA is 2.0”
combustion. Some of these remedies require addi- Hg abs. (67 mbar). Power plants in Northern
tional expenditure in energy, e.g., for finer coal Europe with access to lower-temperature cooling
grinding, and for increasing the momentum flux water use condenser pressure of 1.0” Hg abs.

CO2 Reduction and 44 0.003 MPa

Coal-Based Electricity
Generation, Fig. 4 Steps 43
Net efficiency, % (HHV)

to improve the efficiency in

PC/Supercritical Coal Plant. 30 Mpa double reheat
42 600°C 0.0065 MPa
(Schilling [19])
(1112F)
41 single reheat
120°
(248F)
40
1.15 25 Mpa
130° 540°C
39 1.25 (266F) (1004F)
stack gas steam condenser
38 air ratio temperature conditions reheating pressure

G
CO2 Reduction and Coal-Based Electricity Generation 431

(30 mbar) pressure. This difference can produce reintroducing it to the turbine. In the usual desig-
an efficiency gain of more than 2 percentage nation of steam parameters the second and third
points [21]. temperature refers to single and double reheat,
As steam pressure and superheat temperature e.g., 309 bar/594 /594 /594  C, respectively.
are increased above 221 bar (3208 psi) and 374  C The average efficiency of the existing US coal-
(706  F) the steam becomes supercritical (SC); it fueled electricity generating fleet is 33%, based on
does not produce a two-phase mixture of water the higher heating value (HHV) of the coal.
and steam as in subcritical steam, but instead Pulverized coal plants with USC parameters of
undergoes a gradual transition from water to 300 bar and 600/600  C (4,350 psi, 1112/1112  F)
vapor with corresponding changes in physical can be realized today, resulting in efficiencies of
properties. 43% (HHV). These plants are 21% more efficient
Ultra-supercritical (USC) steam generally than the mean of today’s US fleet of coal-fired
refers to supercritical steam at more than plants; i.e., they would use 21% less coal for the
1100  F temperature. EPRI’s terminology for same power generation and emit 21% less CO2.
1300 and 1400  F plants is Advanced Ultra- There are several years of experience with these
supercritical (AUSC). 600  C (1,112  F) plants in service, with excellent
The thermodynamic efficiency of the Rankine availability [4, 13].
steam cycle increases with increasing temperature Further improvement in efficiency achievable
and pressure of the superheated steam entering the by higher ultra-supercritical steam parameters is
turbine. It is possible to increase further the mean dependent on the availability of new, nickel alloys
temperature of heat addition by taking back par- for boilers and steam turbines (Fig. 5) [12, 18].
tially expanded and reduced temperature steam Two major development programs in progress,
from the turbine to the boiler, reheating it, and the Thermie Project of the European Commission

Membrane wall Tubes SH outlet header

2010

HCM 12(?), Nickel alloy Nickel alloy

Nickel alloy

NF 12, SAVE 12,

2005
12 CrCoMo (?)

1998 9 - 12%
Cr-steels
7 CrMo VTiB 10 10 Austenite E 911, P 92, P 122
HCM 2S

1995

13 CrMo 4 4 X 20 CrMoV 12 1Austenite X 20 CrMoV 12 1P 91

260 270 290 300 350 260 270 290 300 350 260 270 290 300 350 bar
550 580 600 630 700 550 580 600 630 700 550 580 600 630 700 °C
570 600 620 650 720 570 600 620 650 720 570 600 620 650 720 °C

CO2 Reduction and Coal-Based Electricity Generation, Fig. 5 Stages in materials development and related
advanced steam parameters. (Henry et al. [12])
432 CO2 Reduction and Coal-Based Electricity Generation

and a US program managed by EPRI for the Coal Gasification Combined Gas
USDOE’s NETL and the Ohio Coal Development Turbine-Steam Cycle (IGCC)
Office (OCDO), aim at steam parameters of
375 bar, 700  C/720  C (5,439 psi, 1292  F/ Gasification-based technologies use partial oxida-
1328  F), and 379 bar, 760  C/760  C tion of coal with oxygen as the oxidant to produce
(5,500 psi, 1346  F/1400  F), respectively. The a synthesis gas (syngas) consisting mainly of CO
plant efficiency increases by about 1 percentage and H2. The gas is cleaned to remove contami-
point for every 20  C rise in superheat and reheat nants before it is used as fuel in a combustion
temperature. An advanced 700  C (1,293  F) USC turbine.
plant will likely be constructed during the next The exhaust gas of the gas turbine raises steam
7–10 years constituting a benchmark for 46% in a heat recovery steam generator (HRSG) for a
efficiency (HHV) coal-fired power plant. steam turbine-electric generator set. The com-
It is estimated that before 2020, the time when bined cycle efficiency improves through the
CCS technologies may begin to become commer- reduced effect of the steam condenser’s heat
cially available, about 13 Gigawatts (GW) new loss. As with combustion technologies, higher
coal-based capacity will be constructed in the efficiency results in lower emissions per MWh.
USA (EIA 2010). If more efficient presently avail- While the IGCC concept is being successfully
able (1112  F) USC technology is utilized instead demonstrated in two plants in the USA and two
of subcritical steam plants, CO2 emissions would plants in Europe, utility power generation
be about 570 MMt less during the 30 years’ life- demands introduce new challenges that will
time of those plants, even without installing a CO2 require further RD&D to overcome. The gasifica-
capture system. tion process operates best under steady-state con-
High efficiency coal-based power generation is ditions. The load change conditions associated
also important to long-term solutions of reducing with utility electricity generation will burden the
CO2 emissions by using CCS, as it mitigates the technology. The many chemical syngas cleanup
significant energy cost of CCS application. processes will have to respond to these changes on
Because of the reduced coal use for a given elec- a real-time basis. In addition, the gasifier and
tricity output, the plant has a smaller footprint with associated gas cleanup systems will be exposed
respect to size of coal handling and emission con- to a much larger range of fuel quality than expe-
trol systems. These savings and the use of modern rience has demonstrated. Again, this variation
analytical techniques that enable optimal use of Ni introduces conditions that require more RD&D
alloys can minimize the cost of USPC technology. to commercialize.
Coal consumption and CO2 emission compar- In the prevalent designs the gasifier operates at
isons for presently available 500 MW PC/SubC, high pressure. Bituminous coal is pulverized and
PC/SC and PC/USC plants without CCS are pre- fed as coal-water slurry [9]. The temperature in
sented below in Table 1 [8]. the gasifier is above 1300  C, so that the coal ash

CO2 Reduction and Coal-Based Electricity Generation, Table 1 Comparative coal consumptions and emissions of
air-blown pulverized coal combustion technologies without CCS. (MIT Coal Study [8])
Performance Subcritical PC/SC PC/USC
Heat rate (Btu/kWh) 9950 8870 7880
Gen. efficiency (HHV) 34.3% 38.5% 43.3%
Coal use (MMt/year) 1.54 1.37 1.22
CO2 emitted (MMt/year) 3.47 3.09 2.74
CO2 emitted (g/kWh) 931 830 738
Assumptions: 500 MW net plant output; Illinois #6 coal; 85% capacity factor
CO2 Reduction and Coal-Based Electricity Generation 433

can be removed as liquid slag. The product syngas A National Research Council Panel reviewing
undergoes rigorous cleanup of particulates, sulfur the DOE’s Advanced Gasification Research Pro-
and nitrogen oxides, and mercury, prior to enter- gram estimated the improvements upon success-
ing the gas turbine (GT) combustor. Pre- ful completion of the research. The research areas
combustion cleanup of the syngas is and anticipated benefits are listed in Tables 2 and
economically favorable compared to post- 3 [16].
combustion flue gas cleanup because of the
lower volume flow rate, undiluted by nitrogen,
and elevated pressure. The syngas can be cooled CO2 Capture and Compression in PC and
prior to cleanup by radiant and convective heat IGCC Plant
exchangers which raise steam, and improve plant
efficiency. However, because of the additional CO2 capture from pulverized coal combustion
cost and operational problems due to fouling, the (PC) involves postcombustion cleanup, the separa-
convective heat exchanger is often omitted in tion and recovery of CO2 that is at low concentra-
recent designs at the detriment of the plant tion and low partial pressure in the exhaust gas.
efficiency. Chemical absorption with amines is presently the
Another element of efficient design is the inte- only commercially available technology. The CO2 is
gration of subsystems with the main generating first captured from the exhaust gas stream in an
plant. For example, air can be fed from the main absorption tower. The absorbed CO2 must then be
compressor of the GT to the Air Separation Unit stripped from the amine solution using a large
that produces oxygen for the gasification process, amount of steam, regenerating the solution for recy-
and the nitrogen can be returned to the gas turbine cle to the absorption tower. The recovered CO2 is
combustor as a diluent to reduce NOx formation. cooled, dried, and compressed to a supercritical
The gain in improved efficiency, however, is fluid. It is then ready to be piped to sequestration.
weighed against operating a more complicated The use of steam for CO2 removal reduces the
plant of somewhat reduced availability. steam available for power generation.
The current IGCC units have, and next gener- To maintain constant net power generation the
ation IGCC units are, expected to have electricity coal input, boiler, steam turbine/generator, and
generating efficiencies that are less than or com- emission control equipment must all be increased
parable to those of supercritical PC generating in size. The thermal energy required to recover
units [11]. CO2 from the amine solution reduces the efficiency
Current units typically gasify high-heating by 5 percentage points. The energy required to
value, high-carbon fuels. Polk IGCC with a compress the CO2 to a supercritical fluid is the
Texaco-GE coal-water-slurry gasifier and radiant next largest factor, reducing the efficiency by 3.5
syngas cooling operates at 35.4% (HHV) gener- percentage points. All other energy requirements
ating efficiency. The Wabash River IGCC with a amount to less than 1 percentage point.
coal-water-slurry fed E-Gas gasifier, radiant and R&D is in progress pursuing the use of alterna-
convective syngas cooling, and no integration tive sorbents, such as chilled ammonia to reduce
operates at about 40% generating efficiency. The the energy intensity of the CO2 capture process.
IGCC in Puertollano Spain with a dry-feed shell- The stakes are high because a successful solu-
type gasifier, radiant and convective heat tion would be applicable to both new plants and to
exchangers, and combustion turbine-air separa- the retrofit of existing plants with CCS [7].
tion unit integration has a generating efficiency IGCC lends itself favorably for efficient CO2
of about 40.5% (HHV). Supercritical PC units capture and sequestration because CO2 can be sep-
operate in the 38–40% efficiency range, and arated from a relatively small volume of fuel gas
ultra-supercritical PC units in Europe and Japan (syngas) at high pressure. Without CCS, IGCC is
are achieving 42–46% (HHV) generating more expensive, and has lower efficiency and
efficiency. availability than PC/SC and PC/USC technology
434 CO2 Reduction and Coal-Based Electricity Generation

CO2 Reduction and Coal-Based Electricity Generation, Table 2 Evolutionary improvements due to DOE advanced
gasification research
Change in cost and performance of IGCC system due to R&D
Increase in Capital cost O&M Reduction in Availability
efficiency reduction ($/ reduction COE ($/ improvement
(%) kW) ($/year) MWh) (%) Emissions
DOE gasification project areas
Warm gas cleanup 1–2 70–100 Minimal 1.8 0 500 ppb
sulfura
Instrumentation 0.5–1 0 Minimal Minimal 1–2 –
(temperature
measurements)b
Materials (refractory)c 0 0 2 million 0.5 4–6 –
Non-DOE or non-gasification technical areas
Heat recovery 3
Industry learning and 1–2
evolution
Source: National Research Council
COE cost of electricity, O&M operation and maintenance, kW kilowatt, MWh megawatt-hour
a
Air Products & Chemicals, Inc. analysis.
b
Internal report by Parsons
c
The DOE turbine program is considered a complementary program to advanced gasification program

CO2 Reduction and Coal-Based Electricity Generation, Table 3 Evolutionary improvements due to DOE Advanced
Gasification Research
Change in cost and performance of IGCC system due to R&D
Capital cost O&M Reduction in Availability
Increase in reduction ($/ reduction COE ($/ improvement
efficiency (%) kW) ($/kWh) MWh) (%) Emissions
DOE gasification project areas
Ion transport 1 75 Minimal 1.4 0 –
membrane air
separationa
Stamet pump 0.5 40–100 Minimal 1.4–1.8 0 –
(from backup
material)b
Non-DOE or non-gasification technical areas
DOE turbine 2–3 (for 60–100
programc combined cycle
power island)
Source: Gary Stiegel [20], U.S. Department of Energy National Energy Technology Laboratory (for data on DOE
gasification project area cost and performance)
COE cost of electricity, O&M operation and maintenance, kW kilowatt, kWh kilowatt-hour, MWh megawatt-hour.
a
Air Products & Chemicals, Inc. analysis
b
Internal report by Parsons
c
The DOE turbine program is considered a complementary program to advanced gasification program

but, if CCS were available today, equipped with This could reduce the oxygen required for the
CCS, IGCC would be cheaper. By recirculating vaporization of the water in the slurry and for the
part of the liquid phase supercritical CO2 the attainment of the high temperature necessary for
water in the slurry feed can be replaced by CO2. trouble-free liquid slag removal from the gasifier.
CO2 Reduction and Coal-Based Electricity Generation 435

CO2 Reduction and Coal-Based Electricity Generation, Table 4 CO2 emissions, efficiency, and cost estimates for
advanced generation technologies without and with CCS
PC/SC PC/USC IGCC SCPC/Oxy
Without With Without With Without With With CCS
CO2 emitted g/kWh 830 109 738 94 824 101 104
Efficiency, % HHV 38.5 29.3 43.4 34.1 38.4 31.2 30.6
TCR $/kW 2800 4524 2865 4408 3016 3996 4016
COE c/kWh 6.00 9.64 5.93 9.25 6.44 8.24 8.82
Source: CO2 and efficiency data from MIT 2007. TCR and COE data modified after Booras [5] to include the effect of the
recent increase in construction costs

For coals of lower heating value such as sub- corrosion in the compressor and pipeline requires
bituminous coals or lignite, the COE gap is, how- further exhaust gas polishing, the fivefold reduced
ever, substantially narrowed or is even reversed. flue gas volume leads to strongly reduced capital
It is noteworthy that there is significant cost and and treatment costs relative to those for an air-
performance loss attached to the capture and com- blown combustion plant [14].
pression of CO2 from both combustion and gasifi- The presently available Cryogenic Air Separa-
cation plants in preparation of its sequestration [6]. tion process consumes a significant fraction of the
Results of studies presented in Table 4 provide generating plant’s output and reduces its effi-
information on CO2 emitted, efficiency, estimates ciency by 6.4 percentage points. The development
of total plant cost and cost of electricity for IGCC, of membranetype oxygen processes with greatly
PC/USC technologies without and with CO2 cap- reduced energy requirements are urgent R&D
ture and compression, and for SC/PC/Oxy. targets.

PC/Oxy/USC CFBC with Oxy/FGR

When oxygen, instead of air is used as oxidant for CFBC with an external heat exchanger lends itself
combustion, the mass flow rate of combustion favorably to oxy/fuel application because the
products is significantly reduced and the flue gas solids circulation provides an effective means,
CO2 concentration is greatly increased. In order to additional to flue gas recycle, for controlling the
avoid unacceptably high temperatures in the boiler, combustion temperature. Solids, consisting of sor-
combustion products, mainly CO2, are recirculated bent, coal ash, and coal char particles, are precip-
from the end of the boiler to the combustion cham- itated from the gas stream exiting the combustor
ber. This restores the furnace gas temperature to air (riser) section of the CFB boiler and are split into
combustion levels resulting in an O2 volume con- two solids streams: one that is recirculated to the
centration of about 30%, compared to 21% for air- riser without cooling, and the other that is cooled
blown combustion. This difference is due to the in an external heat exchanger before recirculation.
higher specific heat of CO2 than that of the replaced The cooling of the combustor by the cold solids
nitrogen, and also, to CO2’s high radiative emis- permits the reduction of the rate of flue gas
sivity. Flue gas recirculation (FGR) increases the recirculation, with the result that the O2 concen-
CO2 concentration in the flue gas to beyond 90%, tration in the feed stream can rise to above 40%
the complement being N2, due to air leakage and without exceeding a limiting combustor tempera-
about 3% O2 required for complete burn out of ture level of about 850  C required by the thermo-
coal. This makes the flue gas ready for sequestra- dynamic stability of CaSO4 and by smooth
tion after the removal of condensables but without fluidization. The corresponding lower gas mass
energy-intensive gas separation. If avoidance of flow leads to reduced size and cost of the boiler
436 CO2 Reduction and Coal-Based Electricity Generation

and of the post combustion plan. The advantages and in Europe. Today, without CCS, IGCC is
also include increased fuel flexibility including more expensive, and has lower availability than
the capability of firing biomass, which may high efficiency PC/SC and PC/USC plants but, if
allow additional CO2 credits. CCS were available today, equipped with CCS
The oxygen-fired CFB concept has been vali- IGCC would be cheaper. Also, IGCC has the
dated in Alstom’s 3 MWth pilot scale test facility capability of less expensively lowering further
[17]. The results show that the technology can be criteria pollutant emissions.
developed from that of conventional CFB boilers, Use of oxygen instead of air in combustion
so that a 100–350 MWe oxy-fired demonstration significantly reduces the mass flow rate of com-
plant could be developed in the short term bustion products with the concomitant reduction
followed by a commercial size plant in the of flue gas capital and treatment costs. Because of
midterm. the high concentration of CO2 in the flue gas it is
expected that following the removal of pollutants
and agents that could be corrosive in the compres-
Concluding Remarks sor and the pipeline, the flue gas can be seques-
tered without the separation of CO2.
Coal will continue to play a large and indispens- The RD&D challenge is in the development of
able role in electricity generation, in a carbon- the membrane oxygen system that is less energy
constrained world, under any scenario. The key intensive than the presently available cryogenic
enabling technology for CO2 emissions mitiga- air separation.
tion in coal combustion and gasification plants is There are important technology developments
CO2 capture and sequestration (CCS). in progress that can change the performance and
CCS has to be demonstrated at scale, integrated economics of advanced technology options by the
with power generation, and the legal framework time CCS will be commercial. A broad portfolio
of sequestration has to be developed before CCS of advanced clean coal technology RD&D should
becomes commercial, probably by the 2020–2025 be aggressively pursued to meet the CCS
period. challenge.
Before the advent of CCS there will be about Demonstration of CO2 sequestration at scale
45 GW new coal-based electricity generating and integrated with power generation will give the
capacity constructed in the USA (about 1000 public more confidence that a practical carbon
GW worldwide) and the question arises of what emission control option exists and maintains
the technology options are for these new plants. opportunities for the lowest cost, widely available
Increased efficiency of power generation is the energy source to be used to meet the world’s
most predictable and cost-effective method for pressing energy needs in an environmentally
CO2 emissions reduction. In coal plant without responsible manner.
CCS it is also the only practical method for miti-
gating CO2 emissions now, and it remains impor-
tant for future plants equipped with CCS to reduce Bibliography
the energy cost of CO2 capture.
Pulverized coal combustion in Rankine cycle 1. Amor AF, Viswanathan R (2004) EPRI, supercritical
steam plant is the prevailing utilization technol- fossil steam plants: operational issues and design
needs for advanced plants. In: Fourth international
ogy. Compared to average efficiency of the
conference on advances in materials technology for
existing coal-based fleet, up to 24% reductions fossil power plants, Hilton Head Island
in CO2 and pollutant emissions can be achieved 2. Beér JM (2004) Electric power generation; fossil fuel.
today in commercial ultra-supercritical new In: Cleveland CJ, Ayres RU et al (eds) Encyclopedia of
energy, vol 2. Elsevier, Amsterdam, pp 217–227
plants.
3. Beér JM (2007) High efficiency electric power gener-
Coal gasification combined cycle (IGCC) ation, the environmental role. Progr Energy Combust
plants are successfully demonstrated in the USA Sci 33:107–134
CO2 Reduction and Coal-Based Electricity Generation 437

4. Blum R, Hald J (2002) ELSAM Skaerbaek Wieghardt K (2001) The thermie project of the
5. Booras G (2008) Economic assessment of advanced EC. Powergen Europe, Brussels
coal based power plants with CO2 capture MIT carbon 14. Lavasseur AA, Chapman PJ, Nsakala NY, Kluger
sequestration forum IX. Cambridge F (2009) Alstom's oxy-firing technology development
6. Booras G, Holt N (2004) Pulverized coal and IGCC and demonstration-near term CO2 solution. In: The
plant cost and performance estimates. In: Gasification 34th international technical conference on clean coal
technologies conference, Washington, DC and fuel systems, Clearwater
7. Dalton St (2007) Efficiency of generation-the role of 15. National Coal Council reports 2008 and 2009.
very efficient/low emission coal. In: International con- National Coal Council, Washington, DC
ference on coal utilization and fuel systems, 16. National Research Council (2005) Report of the panel
Clearwater on DOE's IGCC R&D program. National Academies
8. Deutch J, Moniz J (eds) (2007) MIT the future of coal. Press, Washington, DC
MIT Press, Cambridge 17. Nsakala L et al (2003) Greenhouse gas emissions
9. Dooher JP (2009) Physio-chemical properties of low control in circulating fluidized bed boilers. US
rank coal/liquid CO2 slurries as gasifer feed-stocks. In: DOE/NETL [48] Cooperative Agreement, May 2003
The 34th international technical conference of coal 18. Palkes M (2003) Boiler materials for ultra supercritical
utilization and fuel systems, Clearwater coal power plants. Conceptual Design ALSTOM
10. EPRI (2003) Electricity technology roadmap. Electric Approach NETL-DOE, USC T-1
Power Research Institute, PaloAlto 19. Schilling HD (1993) VGB Kraftwerkstechnik
11. EPRI (2006) CoalFleet RD&D augmentation plan for 73(8):564–576 (English Edition)
advanced combustion based power plants EPRI 20. Stiegel G (2005) US DOE Gasification program over-
Report 1013221 view in National Research Council Report.
12. Henry JF, Fishburn JD, Perrin IJ, Scarlin B, Washington, DC
Stamatelopoulos GN, Vanstone R (2004) In: Twenty- 21. Termuehlen H, Empsperger W (2003) Clean and effi-
ninth international conference on coal utilization & cient coal fired power plants. ASME Press, New York
fuel systems, US DOE. ASME, Coal and Slurry Tech- 22. US Energy Information Administration (2000).
nology Association, Washington, DC, pp 1028–1042 Annual energy OUTLOOK
13. Kjaer S, Klauke F, Vanstone R, Zeijseink A,
Weissinger G, Kristensen P, Meier J, Blum R,
 subgroups dependent upon the content of volatile
Pulverized Coal-Fired Boilers material.
and Pollution Control Calorific value Corresponds to the amount of heat
per unit mass when combusted. Can be
David K. Moyeda expressed as gross calorific value, which is the
GE Energy, Irvine, CA, USA amount of heat liberated during combustion
under standardized conditions at constant vol-
ume so that all of the water in the products
Article Outline remains in liquid form, or as net calorific value,
which is the maximum achievable heat release
Glossary obtainable in a furnace at constant pressure.
Definition of the Subject Carbon dioxide (CO2) A heavy, colorless gas
Introduction that results from the combustion of fossil fuels
Air Pollutant Emissions from Coal Combustion and from natural sources.
Particulate Emissions Carbon monoxide (CO) A colorless, odorless
Sulfur Oxides Formation gas produced by incomplete combustion of
Nitrogen Oxides Formation fossil fuels.
Particulate Matter Control Technologies Coal A solid fossil fuel consisting primarily of
Electrostatic Precipitator carbon, hydrogen, nitrogen, oxygen, sulfur,
Fabric Filter Baghouse and nitrogen. Coal also contains ash, minerals
SO2 Emissions Control Technologies which do not burn, and moisture. Coal is typ-
Wet Flue Gas Desulfurization ically classified or ranked by its volatile matter,
Dry Flue Gas Desulfurization fixed carbon content, and calorific value.
NOx Emissions Control Technologies Dry FGD A process that removes sulfur oxides
Low-Nox Burners from the flue gas and results in the formation of
Overfire Air a dry product or waste.
Reburning Electrostatic precipitator (ESP) A device for
Selective Noncatalytic Reduction removing particulate from a gas stream based
Selective Catalytic Reduction upon using an electric field to charge the particles
Future Directions in the gas and move them to a collecting surface.
Bibliography Fabric filter A device for removing particulate
from a gas stream based upon filtering the gas
Glossary through a filer media.
Flue gas desulfurization (FGD) Technologies
Anthracite Coal which typically contains that are used to remove sulfur oxides from the
86–97% carbon. Anthracite is considered the flue gas.
highest rank of coal as it has the highest energy Lignite Coal which contains 25–35% carbon and
content of all coals. which has a lower calorific value than subbitu-
Ash Inorganic residues remaining after combustion. minous and bituminous coals and typically
Baghouse See fabric filter. higher moisture and volatile content. Lignite
Bituminous coal Coal which typically contains is the lowest range of coal.
45–86% carbon. Bituminous coal lies between Low-NOX burners (LNB) Technology for
subbituminous coal and anthracite in terms of reducing NOX emissions by controlling fuel
rank, and is commonly divided into additional and air mixing in the flame.

© Springer Science+Business Media, LLC 2012 439

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_101
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media LLC, 2012
https://doi.org/10.1007/978-1-4419-0851-3_101
440 Pulverized Coal-Fired Boilers and Pollution Control

Nitric oxide (NO) A colorless gas resulting from role in energy production and approximately 41%
the combustion of fossil fuels. of the world power generation was supplied by
Nitrogen dioxide (NO2) A reddish-brown gas coal-fired power plants in 2008 [1]. While
that can be emitted from the combustion of increased discoveries of natural gas and fuel oil
fossil fuels or is formed by atmospheric reac- resources, and growth in renewable energy
tion of nitric oxide (NO) and oxygen (O2). sources, such as wind, solar, and geothermal
NOX Refers to the total nitric oxide (NO) and energy is projected to reduce the use of coal for
nitrogen dioxide NO2 concentration. power generation, energy from coal will continue
Overfire air (OFA) Technology that reduces to be used to satisfy the world’s energy demands.
NOX emissions based upon air staging. One drawback in the use of coal for power
Reburning Technology that reduces NOX emis- production is that it produces high levels of air
sions based upon staging fuel in a fashion that pollutants, such as particulate, sulfur oxides, and
permits fuel fragments to reduce (or reburn) nitrogen oxides. Coal also produces higher carbon
nitric oxide (NO) in the flue gas. dioxide emissions than other fossils such as natu-
Selective catalytic reduction (SCR) Technol- ral gas and fuel oil. In addition, due to the large
ogy that reduces NOX emissions by mixing quantities of coal that are used for power produc-
ammonia into the flue gas and reacting the tion, emissions of toxic metals, such as mercury,
ammonia with NOX over a catalyst. which are contained in very small quantities in the
Selective noncatalytic reduction (SNCR) coal, can also be a concern.
Technology that reduces NOX emissions by Over the past several decades, a number of
mixing an amine-based reagent into the flue technologies have been developed to reduce the
gas at a temperature which selectively pro- air pollutant emissions formed from coal combus-
motes the reaction of amine (NH2) with nitric tion. Modern power plants can be equipped with
oxide to form molecular nitrogen (N2) advanced technologies that reduce particulate,
Subbituminous coal Coal which typically con- sulfur oxide, and nitrogen oxide emissions to the
tains 35–45% carbon and which typically has a most stringent levels. These technologies are the
lower calorific value than bituminous coal and focus of this entry.
higher moisture and volatile content.
Sulfur dioxide (SO2) A colorless, irritating gas
resulting from the combustion of sulfur Introduction
contained in fossil fuels, particularly coal.
Sulfur oxides Refers to sulfur dioxide (SO2) and The abundance of coal throughout the world led to
sulfur trioxide (SO3). its use in China as early as 1000 B.C. and by the
Volatile matter Non-moisture component of Romans in Britain before 400 A.D. [2]. While the
coal that is liberated at high temperature in use of coal in Briton largely disappeared when the
the absence of air. Romans left in the fifth century, coal use in
Wet FGD A process that removes sulfur oxides England increased in the thirteenth century, and
from the flue gas and results in the formation of by the beginning of the seventeenth century, coal
a product or waste that is a solution or slurry. was the dominate source of energy [3]. During the
industrial revolution of the eighteenth and nine-
teenth century, the invention and development of
Definition of the Subject the steam engine led to an increase in coal pro-
duction and use for industrial processes and trans-
Fossil fuels, such as coal, natural gas, and fuel oil, portation [4–6]. In the late nineteenth century,
are used to generate electric power for industrial, advances in electricity and the invention of a
commercial, and residential use. Due to its rela- reliable incandescent lamp set the stage for the
tively low cost and abundance throughout the use of coal to generate electrical power with the
world, coal has historically played a significant first coal-fired central generating plant in the USA
Pulverized Coal-Fired Boilers and Pollution Control 441

established in 1882 [7, 8]. The growing demand respiratory symptoms, particularly for those sus-
for electrical power led to further developments in ceptible to these problems [22, 23]. In addition,
steam-generating boiler technology, such as pul- sulfur and nitrogen oxide emissions can react with
verizing the coal prior to introducing it to the water in the atmosphere to form acids that deposit
boiler furnace. Pulverized coal firing enabled the in lakes and soils leading to acidification [24–26],
construction of larger boilers and power plants which is referred to as “acid rain.” Nitrogen
and became the predominant firing method for oxides also react with volatile organic compounds
large steam-generating power plants beginning to form photochemical oxidants, such as ozone
in the late 1920s [9, 10]. (O3), which present a hazard to human health and
The air pollution associated with coal combus- plants in high concentrations and which cause
tion was recognized in England as early as the visible “smog” in urban areas [27, 28].
thirteenth century, where the burning of coal in This article will introduce how PM, SO2, and
urban areas created a smoky environment and led NOX emissions are formed during coal combus-
to bans on coal use [11]. While improvements in tion and will discuss the technologies that have
combustion methods and the use of chimneys or been developed to control these emissions from
stacks to disperse the smoke overcame some of pulverized coal-fired power plants. The technol-
the objections to coal burning, fundamentally, the ogy discussion will focus on the primary technol-
growing need for low-cost energy offset public ogies that have been applied and are available on a
concerns over the unhealthy aspects of coal com- commercial scale.
bustion, and such bans were largely ignored
[3]. As the world entered the twentieth century,
widespread industrialization and the increased use Air Pollutant Emissions from Coal
of coal resulted in degraded air quality in London Combustion
and other English cities [12]. By the middle of the
century, severely degraded air quality in Los The primary air pollutants regulated from coal-
Angeles, California [13], and air quality disasters fired power plants worldwide are carbon monox-
in London, England [14, 15], and Donora, Penn- ide (CO), sulfur dioxide (SO2), oxides of nitrogen
sylvania [16], heightened public awareness of the (NO and NO2, which are referred to as NOX), and
dangers of air pollution and government recogni- particulate matter (PM) [29]. In general, CO emis-
tion of the need for research to understand the sions from pulverized coal-fired power boilers are
formation of and problems associated with air low (<50–200 ppmv) as the combustion system
pollutants and for the development of regulations tends to be operated with sufficient excess air to
to limit their emissions [17]. maximize combustion efficiency [30]. Local stan-
For coal combustion, the primary air pollutants dards, rather than national standards, are generally
of concern are particulate matter (PM), sulfur established to limit CO emissions and can vary
dioxide (SO2), and oxides of nitrogen (NO and widely. SO2 and NOX emissions are regulated as
NO2, which are referred to as NOX) [18]. These both can be a health hazard in high concentrations
pollutants originate from the ash, sulfur, and nitro- [22, 23] and are contributors to dry and wet acid
gen species present in the coal. Particulate matter deposition [24–26]. In addition, NOx emissions
reduces visibility and contributes to regional haze. contribute to ground-level ozone formation
In addition, “coarse” particles (from 2.5 to 10 mm through the reaction with volatile organic com-
in diameter, PM10) and “fine” particles (smaller pounds [27, 28]. Particulate emissions can lead to
than 2.5 mm in diameter, PM2 5) can accumulate reduced visibility and degraded air quality. Small
within different areas of the respiratory system particulates emitted from combustion (<10 mm)
and aggravate health problems such as asthma or or formed by reactions of SO2 and NOX in the
lead to increased respiratory symptoms and dis- environment (<2.5 mm) are associated with
ease [19–21]. Exposure to sulfur and nitrogen increased respiratory symptoms and disease
oxides in sufficient concentration can lead to [19–21].
442 Pulverized Coal-Fired Boilers and Pollution Control

In addition to these primary pollutants, since coal-fired power plants vary widely and are deter-
the early 1990s, there has been significant interest mined by the coal type and combustion system
in the emissions of hazardous air pollutants design. Table 1 compares the major properties of
(HAPs), such as toxic organic compounds, acid several coals, low, medium, and high-sulfur bitu-
gases, and metals, from coal combustion. Based minous coals; subbituminous coal; and lignitic
upon comprehensive studies performed to charac- coal. As can be seen in the table, the sulfur, nitro-
terize HAPs from fossil fuel–fired power plants gen, and ash content contained in coals can vary
[31, 32], the US Environmental Protection significantly. Bituminous coals are characterized
Agency (EPA) identified mercury as the primary by a higher heating value in terms of energy release
HAP of concern due to the contribution of coal- per mass of fuel consumed than lower rank subbi-
fired power plants to the total anthropogenic emis- tuminous coal and lignite, which have higher mois-
sions of this compound and to the risk of exposure ture contents in comparison to bituminous coal.
to methylmercury through the consumption of Bituminous coals also tend to have higher sulfur
contaminated fish [33]. Methylmercury is a highly and nitrogen content in comparison to the low-rank
toxic neurotoxin that bioaccumulates up the food coals. In the remainder of this entry, the impacts of
chain [34]. The concern over mercury emissions coal ash, sulfur, and nitrogen content and boiler
from coal combustion is shared in other developed design and operation on emissions from coal-fired
areas of the world [35–37]. In addition to mercury, power plants will be discussed.
other HAPs of concern from coal-fired power
plants due to their levels of emissions and poten-
tial health risks include carcinogenic metals, such Particulate Emissions
as chromium, nickel, and arsenic, and acid gases,
including hydrogen chloride (HCl) and hydrogen Particulate emissions from pulverized coal com-
fluoride (HF), which are soft-tissue irritants and bustion result primarily from the mineral matter
can cause respiratory disease [38]. included in the coal. The mineral matter can be
A by-product of the combustion of fossil fuels is classified as being inherent (chemically bound
the generation of carbon dioxide (CO2). In the into the coal matrix), included (present in the
atmosphere, CO2 and other gases, such as water coal matrix), or extraneous (soil and rock mixed
vapor, absorb and reemit infrared radiation into the coal during mining) [44]. The composi-
reflected from the earth’s surface, resulting in a tion of the mineral matter varies according to the
“greenhouse” effect that raises surface temperature geographic location, type of coal, and how much
of the earth [39]. Since coal has a higher carbon-to- extraneous material is entrained into the coal dur-
hydrogen ratio than other fossil fuels, such as nat- ing mining. The primary minerals in coal are
ural gas and fuel oil, CO2 emissions from coal quartz, aluminosilicates, iron sulfides, and car-
combustion are higher than those from the com- bonates [45]. These minerals can be broken
bustion of other fossil fuels [40, 41]. The high down into the major ash constituents shown in
carbon footprint of coal-fired power plants has led Table 2 [46]. The primary constituents are silica
to an increased demand for the generation of power and alumina oxides (SiO2 and Al2O3). The iron
from alternative sources, such as cleaner fuels (e.g., (Fe) and alkali metal (calcium (Ca), magnesium
natural gas) and renewable sources (e.g., biomass, (Mg), and sodium (Na)) content vary widely but
wind, and solar), to the development of high- are important as these metals have a major influ-
efficiency coal-fired power plants (e.g., integrated ence on the slagging and fouling characteristics of
gasification combined cycle (IGCC)), and to the the resulting ash [47]. Slagging refers to the
research and development of methods for carbon buildup of molten or partially fused ash on the
capture and sequestration from large coal-fired furnace walls or radiant heat transfer surfaces of
power plants [42, 43]. the boiler. Fouling refers to the deposit of ash on
This article focuses on the primary pollutants the convective heat transfer surfaces such as the
from coal. The levels of SO2, NOX, and PM from superheater and reheater.
Pulverized Coal-Fired Boilers and Pollution Control 443

Pulverized Coal-Fired Boilers and Pollution Control, Table 1 Typical coal properties
Bituminous low Bituminous medium Bituminous high
Property Units sulfur sulfur sulfur Subbituminous Lignite
Ultimate analysis
Carbon % 69.36 70.06 62.08 65.54 66.15
dry
Hydrogen % 5.32 5.00 4.44 4.15 4.20
dry
Nitrogen % 1.50 1.66 1.07 0.95 0.96
dry
Sulfur % 1.04 3.08 7.40 0.79 0.37
dry
Oxygen % 12.73 7.54 6.15 14.00 20.72
dry
Ash % 10.05 12.66 18.86 14.57 7.60
dry
Proximate analysis
Volatile % a.r. 42.97 38.15 34.87 31.16 27.02
matter
Fixed % a.r. 43.21 43.10 42.79 34.88 33.38
carbon
Ash % a.r. 9.62 11.77 18.05 11.26 4.97
Moisture % a.r. 4.20 6.98 4.29 22.70 34.63
Dry, ash-free basis
Nitrogen % d. 1.67 1.90 1.32 1.11 1.04
a.f.
Sulfur 1.16 3.53 9.12 0.92 0.40
Volatile 49.86 46.95 44.90 47.18 44.74
matter
Fixed 50.14 53.05 55.10 52.82 55.26
carbon
Gross heating value
kJ/kg 27,244 27,386 25,663 20,002 16,866
Btu/ 11,713 11,774 11,033 8599 7251
lb
a.r. as received, d.a.f. dry, ash-free

During combustion, the mineral matter in coal formation of additional ash particles smaller than the
undergoes several transformations to become partic- parent particle [50]. Second, at high temperatures,
ulate matter that is entrained in the flue gas. The size the more volatile metals will vaporize [51]. These
and composition of the particulate matter depends vapors can then undergo homogeneous nucleation
upon the mineral composition, how it is included in to form very fine particulate that grow larger by the
the coal, the presence of other species, and the time- condensation of additional volatile species onto the
temperature history of the coal particles as they are particle surface or can form larger particles by coag-
being burned [48]. There are two primary mecha- ulation and chain agglomerate formation. Ash par-
nisms for ash formation during combustion. First, ticles formed by the first mechanism represent the
ash may remain with the burning coal particle, melt bulk of the fly ash mass and are greater than 1 mm in
at high temperature and coalesce to form liquid diameter, with a typical size range of 3–50 mm, and
droplets, and then agglomerate with other ash parti- those by the second mechanism are less than 0.5 mm
cles [49]. Fragmentation of the coal char leads to the in diameter, with a peak around 0.1 mm.
444 Pulverized Coal-Fired Boilers and Pollution Control

Pulverized Coal-Fired Boilers and Pollution Control, Table 2 Major constituents in ash for typical US coals
State coal Pennsylvania Utah Wyoming North Dakota
type bituminous bituminous subbituminous lignite
Ash composition
SiO2 wt.%, 45.3 60.0 43.3 22.0
S-free
Al2O3 24.2 22.7 17.2 20.4
Fe2O3 20.3 4.1 6.3 11.8
TiO2 1.2 1.2 1.4 0.5
O2O5 0.6 1.5 2.5 0.1
CaO 4.8 4.6 22.7 30.3
MgO 1.1 1.9 4.0 8.0
Na2O 1.4 1.1 1.7 5.1
K2O 1.3 1.8 0.5 1.4

The total particulate emissions in the flue gas and are believed to consist of thiols, sulfides, and
from a coal-fired boiler depend upon the design of thiophenes [53]. The proportions of inorganic and
the combustion system and the amount of organic sulfur vary depending upon the coal; how-
unburned carbon resulting from incomplete com- ever, organic sulfur typically comprises between
bustion. For pulverized coal-fired boilers, 30% and 50% of the total sulfur [54]. The remain-
70–90% of the ash will typically end up in the der is primarily pyrite, as inorganic sulfates are
flue gas (fly ash), and the remaining 10–30% of typically less than 0.1% of the total sulfur.
the ash will collect on the walls of the boiler and During the combustion process, the inorganic
end up as ash removed from the bottom of the and organic sulfur is released and converted to
boiler (bottom ash) [9]. Cyclone-fired boilers and sulfur dioxide, SO2. For pyrite, this process
other boilers with wet bottom designs collect involves decomposition to FeS and subsequent
more bottom ash and generate approximately oxidation [55]. For the organic sulfur, this process
15–30% fly ash. Boiler load can also impact par- involves decomposition and oxidation of the par-
ticulate emissions with lower boiler loads ent compound to release the sulfur-bearing spe-
resulting in lower emission rates depending upon cies. The gas-phase sulfur chemistry is complex
boiler design and fuel characteristics [52]. Periodic [56], but in fuel-lean flames, the released sulfur
cleaning of heat transfer surfaces, such as the species are readily oxidized to SO2. Under fuel-
superheater, reheater, economizer, and air pre- rich conditions, hydrogen sulfide (H2S) and car-
heater, via soot blowing can lead to short-term bonyl sulfide (COS) are also present in low con-
increases in particulate loading in the flue gas. centrations [57, 58], but these species are readily
oxidized to SO2 in the high-oxygen environment
of practical combustion systems.
Sulfur Oxides Formation For bituminous and subbituminous coals,
nearly all of the sulfur in the coal (90–95%) is
Sulfur is present in coal in inorganic and organic released as sulfur oxides, both SO2 and sulfur
forms. The primary inorganic form is as iron sul- trioxide, SO3, with the remainder ending up as
fide (FeS2), which is typically in pyrite (cubic) sulfates in the ash. For pulverized coal-fired
form. A small amount of sulfur is also present as boilers, emissions of SO2 are independent of the
inorganic sulfates, typically as salts of minerals boiler design and operating conditions [59]. For
such as calcium and iron (CaSO4 • 2H2O and lignitic and subbituminous coals containing ash
FeSO4 • 7 H2O). The organic compounds with a high alkali (calcium and sodium) content, a
containing sulfur are larger chain hydrocarbons higher fraction of the sulfur is retained in the ash.
Pulverized Coal-Fired Boilers and Pollution Control 445

Sulfur oxide emissions from lignites can be cor- The nitrogen content in coal typically varies
related with the sulfur, sodium oxide (Na2O), and between 1% and 2%, with some coals having
silica (SiO2) content of the ash [60]. nitrogen contents higher and lower than this
The fraction of sulfur that is emitted as SO3 is range. Analytical techniques suggest that the
small and is typically 0.5–1.5% of the total sulfur majority of the nitrogen is contained in five-
in the coal [61]. While equilibrium favors the member (pyrrolic) or six-member (pyridinic) aro-
formation of SO3 at low temperatures, the reaction matic ring structures [68, 69], with pyrrolic nitro-
kinetics are too slow to permit appreciable SO3 to gen being the dominate form for all coal ranks.
form prior to the flue gas exiting the stack A smaller fraction of nitrogen is present in other
[49]. The SO3 that does form is a concern as, at species which may include quaternary nitrogen
low temperatures, it reacts with moisture to form (i.e., a nitrogen molecule with four bonds) or
sulfuric acid, which can condense out of the flue aromatic amines [70]. The form and distribution
gas and cause corrosion. The coal sulfur content of nitrogen within the coal matrix is expected to
and fly ash composition and boiler design and impact how the nitrogen is released during the
operating variables, such as excess oxygen and combustion process and the potential for its con-
flue gas residence time-temperature profile, are all version to NO.
factors that can influence the concentration of SO3 As a coal particle devolatilizes during combus-
that is emitted [62]. tion, a fraction of the nitrogen is released with the
volatiles, while a fraction of the nitrogen stays in
the coal char [71]. The volatile nitrogen may be
Nitrogen Oxides Formation contained in tars or in the form of cyanic and
amine species, which are believed to be the prod-
During combustion, NOX emissions can form ucts of rapid secondary pyrolysis of the hydrocar-
through three main mechanisms. The first process bon forms. As the tars undergo further
is fixation of nitrogen in the air (atmospheric breakdown, a fraction of the nitrogen is released
nitrogen) via the overall reaction: primarily in the form of HCN [72], and a fraction
is incorporated into soot compounds [73]. The
N2 þ O2 ! 2NO ð1Þ nitrogen species released during devolatilization
can undergo oxidation to form NO via the overall
The detailed reaction is believed to be initiated by reactions:
a free oxygen atom (O) attacking the very stable
nitrogen molecule (N2) to form NO and a free nitro- HCN þ O2 ! NO þ CO þ H ð2Þ
gen atom (N), which can then attack O2 [63]. As this
chain reaction can only be initiated at high tempera- 1
NH3 þ O2 ! NO þ H2 O þ H2 ð3Þ
tures, NO formed from this process is referred to as 2
“thermal NOX.” The second mechanism is through
the reaction between hydrocarbons in the flame front In the absence of oxygen, the fixed nitrogen
with molecular nitrogen [64]. As this reaction can species can be reduced to N2. After the volatiles
produce NO levels higher than those expected from are released, further nitrogen species, primarily as
thermal NO formation alone within the initial stages HCN, can be released from the coal char by ther-
of combustion, it is referred to as “prompt NOX.” mal dissociation of the remaining organic solids.
The third mechanism is through oxidation of fixed As the coal char burns, NO can be formed by
nitrogen species present in the fuel. NO formed from heterogeneous oxidation of nitrogen remaining
this process is referred to as “fuel NOX.” For pulver- in the char [74].
ized coal combustion, the majority of the NOX emis- Pilot-scale studies to evaluate the impacts of
sions result from fuel nitrogen rather than thermal coal type and combustion conditions on the for-
NOX formation [65, 66]. Prompt NOX is estimated to mation of NOX emissions from pulverized coal
be less than 5% [67]. combustion have shown that as the total fuel
446 Pulverized Coal-Fired Boilers and Pollution Control

nitrogen content increases, NOX emissions release per surface area resulting in high NOX
increase, but that the fraction of fuel nitrogen emissions, while more modern designs use a
evolved with the volatiles also impacts NOX emis- lower ratio to reduce NOX emissions.
sions [75]. Large differences in NOX emissions
can result from coals having similar nitrogen con-
tents and burned under the same conditions when Particulate Matter Control Technologies
fuel nitrogen is evolved at different rates. Fuel and
air contacting also has a strong influence on NOX To remove dust from a gas stream, such as fly ash
emissions with rapid and intimate contact of air contained in boiler flue gas, a force must be
with the fuel during devolatilization leading to applied that causes the particles to divert from
higher NOX emissions than those which result the flow direction of the gas long enough for the
from the slow mixing processes typical of a dif- particles to contact a collecting surface. The
fusion flame [76]. The staging inherent in a diffu- collecting force can be gravitational, centrifugal,
sion flame lowers NOX emissions by allowing inertial, direct interception, diffusional, or electro-
fuel nitrogen to evolve in an oxygen-free environ- static [80]. Some collection devices may use a
ment where it can be reduced to N2 rather than combination of these forces. The effectiveness of
being oxidized to NO [77]. In a practical pulver- the collecting force depends upon the characteris-
ized coal flame, coals which evolve fuel nitrogen tics of the particulate matter to be collected;
early in the combustion process tend to produce hence, the particulate matter size distribution and
lower NOX emissions than those that retain more chemical and physical characteristics must be
nitrogen in the char. known to select the most appropriate collecting
The coal type and the boiler design and com- device [81].
bustion system all impact the NOX emissions Modern coal-fired power plants use two pri-
produced from pulverized coal-fired power plants mary devices to remove and collect fly ash: elec-
[75, 78]. Low-rank coals (i.e., subbituminous and trostatic precipitators and fabric filters. Older
lignite) produce lower NOX emissions than high- boilers and industrial boilers may also use
rank coals (i.e., bituminous and anthracite). This mechanical collectors, such as cyclones, to
is primarily due to the lower nitrogen and higher remove larger fly ash particles prior to the
volatile content of the low-rank fuels. For power remaining particulate being collected in an elec-
generation, the majority of pulverized coal fired trostatic precipitator or fabric filters. Venturi
boilers use either an array of circular burners scrubbers have also been used for large coal-
located on one or two walls of the furnace (i.e., fired boilers; however, these are less common
wall or opposed wall firing) or columns of burners and would not be used for compliance with cur-
arranged on the furnace corners firing in a tangen- rent emissions regulations. The remainder of this
tial pattern (i.e., tangential firing). Tangential fir- section will provide an introduction to electro-
ing results in longer, slower mixing flames and static precipitators and fabric filters.
produces lower NOX emissions than wall-fired
boiler designs, which tend to have shorter, faster
mixing flames. Other firing system designs, such Electrostatic Precipitator
as cyclone and arched-fired boilers, that are
designed to remove the coal ash in a molten state If a dust-laden gas is passed through a strong
tend to produce the highest NOX emissions due to electrical field, the dust will become charged and
intense fuel and air mixing and higher tempera- will begin to flow in the direction of the ion flow,
tures. For a given boiler design, firing configura- enabling them to be removed from the gas. To
tion, and coal, NOX emissions can be apply this phenomenon in practice requires four
approximately correlated with the total heat liber- steps: (1) the particles need to be charged, (2) the
ation per unit of cooled surface area [79]. Older particles need to be collected on a surface, (3) the
boiler designs tend to have a high ratio of heat particles need to be removed from the surface in a
Pulverized Coal-Fired Boilers and Pollution Control 447

fashion that minimizes their reentrainment into frequency power supply for application of electri-
the gas flow, and (4) the particles need to be cal power, rappers to remove particulate from the
removed from the device [82]. The collecting collection electrode, and pyramidal-shaped hop-
device that accomplishes these steps is typically pers to remove particulate from the system
referred to as an electrostatic precipitator (ESP). [83]. The collection electrodes or plates are
As shown in Fig. 1, in an ESP, an electric potential arranged in multiple mechanical fields, where
is set up between a discharge electrode and a each field consists of a series of equally spaced
collection electrode, and the dust laden gas is plates perpendicular to the gas flow. The discharge
passed between the electrodes at a relatively low electrodes or wires are suspended in the gas pas-
velocity. When the potential is high enough, sage between each pair of plates. An ESP will
corona discharge is generated near the discharge have one or more mechanical fields. A high-
electrode that begins to ionize the gas molecules; voltage pulsed direct current or direct current is
these molecules then migrate to the lower poten- applied to the discharge electrode. When the volt-
tial collecting electrode. In the process, the mole- age is high enough, a corona discharge will be
cules attached themselves to the dust particles generated that causes particles to be moved to the
which then also migrate to the electrode. Upon collection electrode.
reaching the electrode, the particles lose their The performance of an ESP is impacted by the
charge and stick to the electrode. As the dust design of the ESP, the strength of the electric field,
layer on the collection electrode builds up, the and the particle characteristics. The theoretical
electrical current is reduced, and it becomes nec- collection efficiency of an electrostatic precipita-
essary to vibrate the plate to remove the particles. tor can be expressed by the Deutsch-Anderson
This is accomplished through a rapper system that equation [84]:
strikes the top of the collection plate periodically
to remove the dust. The dust then falls into the  ¼ 1  eðVÞo
A
ð4Þ
collection hopper where it can be removed.
ESPs are generally classified by the method in where  is the collection efficiency for a given
which the particulate is removed from the collec- particle size, A is the surface area of the collecting
tion electrode, e.g., wet versus dry, and by the electrode, V is the volumetric gas flow rate, and o
geometry of the electrodes, e.g., wire in tube or is the migration velocity of the particle. The term
wire and plates. The most common design for (A/V) is referred to as specific collection area
coal-fired power plants is the dry, wire, and plate (SCA) of the ESP. Eq. 4 shows that the collection
type as illustrated in Fig. 2. The main components efficiency of an ESP can be improved by either
of an ESP are a gas-tight casing or outer shell, increasing the surface area of the collecting elec-
discharge electrodes, collection electrodes, a trodes or decreasing the volume of gas to be
high-voltage transformer rectifier or high- treated. The theoretical migration velocity of the

Pulverized Coal-Fired (−)

Boilers and Pollution Collection Electrode
Control, Fig. 1 Principle
Particle Collection
of electrostatic precipitator Gas & −
Particles Charging Clean
operation Particle Gas
Flow Flow
Electrons

Corona
(−) (−)
Discharge Discharge
Electrode Electrode
448 Pulverized Coal-Fired Boilers and Pollution Control

Discharge Electrodes,
Wires (−)
Collection Electrodes,
Rappers
Plates (+)
High Voltage
Plan View
Transformer
Rectifier

Clean
Flow Collection
Electrodes
Gas
Control
(Plates) Outlet
Baffles

Casing

Hoppers
Dusty Gas
Inlet

Particulate

Pulverized Coal-Fired Boilers and Pollution Control, Fig. 2 Overview of electrostatic precipitator

particle can be calculated by comparing the elec- The resistance to electrical conductivity is called
trostatic forces on the particle to the drag force the “resistivity” [86]. As the dust layer builds up
and, with simplifying assumptions, can be on the collection electrode, the voltage at which
expressed as [85]: sparking occurs is lowered due to the increased
electrical field at the dust layer. If the resistivity of
Ec Ep d the ash is too high, charge will build up on the
o¼ ð5Þ
4pmg collected particles and can become high enough to
cause an electrical breakdown which causes ions
where Ec is the charging field strength, Ep is the of the opposite polarity to be injected back into the
precipitating (collecting) field strength, d is the gap which reduces the charge on the particles in
particle diameter, and mg is the gas viscosity. the gas flow and which can cause sparking. This
Eq. 5 shows that the particle migration velocity breakdown is referred to as “back corona”
is proportional to both the electric potential and [87]. The resistivity of fly ash is a function of the
the particle size. As particle size is reduced, volt- particle composition, gas temperature, and con-
age must be increased to maintain migration centrations of water vapor and SO3 in the flue
velocity and collection efficiency. Eq. 5 suggests gas [88].
that continuing to increase the electric field will Properly designed and operating ESPs can
continue to increase the particle migration veloc- remove between 99% and 99.9% of the total par-
ity and, hence, the ESP collection efficiency. ticulate matter. Older equipment is generally less
While this is true, a practical limit is reached efficient. As noted above, the primary factors that
where the field breaks down and sparking influence the ESP performance are the particle
between the electrodes occurs [80]. Sparking resistivity, the particle size distribution, and the
destroys the electrical field and lowers the collec- flue gas temperature and flow rate [83]. As these
tion efficiency. parameters vary from the conditions assumed in
The electrical conductivity of the particulate designing the ESP, the ESP collection efficiency
matter has an impact on the electrical field can be impacted. Other factors that can impact the
between the discharge and collection electrodes. ESP performance are the electrical conditions in
Pulverized Coal-Fired Boilers and Pollution Control 449

the ESP, the gas flow distribution, and the particle The filtration media can be flat and supported in
reentrainment during rapping [89]. The applied a frame or in bags which are supported on cages.
voltage in each field of the ESP needs to be opti- As bags are most common for large systems, the
mized to maximize the acceptable spark rate with containing device is commonly called a baghouse.
the specific ash characteristics. The gas flow dis- In a baghouse or fabric filter, the dust-laden gas
tribution entering the ESP should be optimized to passes through the suspended filtration media. Par-
ensure a uniform flow distribution between the ticles impact on the filter and are held. As collec-
collecting plates, and the ESP should be designed tion proceeds, a deposit begins to build up that also
to minimize the potential for gas flow to bypass serves as a means of collecting particulate. Even-
the plates (sneakage). The rapping system design tually, the deposit must be removed or the pressure
and frequency should be optimized to minimize drop will be too high. As this can be accomplished
the reentrainment of particles into the gas stream. by several means, baghouses are most commonly
classified by the cleaning method [91]. Typical
cleaning methods include mechanical, reverse air-
flow, and pulse-jet cleaning. Mechanical cleaning
Fabric Filter Baghouse
typically involves flowing the gas on the inside of
the bag, stopping the gas flow, and shaking the bag
If a dust-laden gas is passed through a fiber bed or
to remove particulate. Reverse airflow cleaning
filter, the dust will be removed from the gas and
consists of periodically flowing gas in the opposite
will collect on the filter. The principle mecha-
direction to the normal gas flow to remove the
nisms of filtration include impaction, interception,
particulate buildup from the filter. In a pulse-jet
and diffusion, which are related to the relative
system, the gas flows from the outside to the inside
particle and fiber size and velocity through the
of the bag. Periodically, a pulse of compressed air
filter [90]. These mechanisms are illustrated in
is injected down the center of the bag to flex the
Fig. 3. Particles will impact on a fiber when they
bag and remove particulate.
are too large to follow the gas streamlines around
Reverse airflow and pulse-jet baghouses have
the filter. Interception will occur when a particle
been applied to coal-fired boilers throughout the
follows a streamline close to a fiber and is
world [92, 93]. Simplified overviews of the
attracted to the surface by van der Waals forces.
cleaning methodology and main components for
Particles that are very small are influenced by
reverse airflow and pulse-jet fabric filters are illus-
Brownian diffusion and can come into contact
trated in Figs. 4 and 5. The similar components for
with the fiber and be removed from the gas.
each type of fabric filter include a gas-tight casing,
Other forces, such as electrostatic and gravity,
fabric filters, supporting frames, and pyramidal-
can also play a role in removing particulate from
shaped hoppers to remove particulate from the
the gas.
unit. Large-scale systems will be designed with
multiple compartments to facilitate cleaning and
Impaction maintenance. Reverse airflow baghouses include
Particles
a fan and ductwork and valves for flowing cleaned
gas back through the unit. Pulse-jet baghouses
include a compressed air header and distribution
Filter plenum to periodically introduce compressed air
Gas Flow
Element into the throat of the filter bags. While significant
experience exists with the use of reverse airflow
Diffusion
baghouses on large utility boilers, interest in
pulse-jet baghouses is increasing since this design
Interception
uses higher air-to-cloth ratios than the reverse
Pulverized Coal-Fired Boilers and Pollution Control, airflow design, resulting in smaller equipment
Fig. 3 Principal mechanisms involved in fabric filtration size and lower capital costs [94].
450 Pulverized Coal-Fired Boilers and Pollution Control

Clean Gas
Reverse Gas
Outlet
Inlet

Fabric Filter

Compartment

Hopper

Dusty Reverse
Gas Gas
Inlet Outlet
Flow
Control
Baffles
Particulate Particulate

Pulverized Coal-Fired Boilers and Pollution Control, Fig. 4 Reverse air flow baghouse features

Pulverized Coal-Fired Compressed Clean Gas

Boilers and Pollution Air Header & Outlet
Control, Fig. 5 Pulse jet Distribution
baghouse features Plenum

Fabric Filter

Compartment

Support Frame

Dusty
Gas
Inlet Hopper
Flow
Control
Baffles
Particulate

The filtration media used in fabric filters can be selection. For coal-fired power plants, the bag-
woven or felted. Woven fabrics are typically used house is installed after the air preheater, where
in shaker-type baghouses as they have good normal gas temperatures are in the range of
mechanical strength. Felted fabrics tend to be 120–175  C (250–350  F) and where the presence
used for reverse air and pulse-jet fabric filters of sulfur oxides and moisture increases the poten-
[95]. For proper operation, the type of fabric tial for corrosion. In this application, the utility
must be matched to the specific application. The industry has primarily used coated fiberglass fab-
gas stream temperature and chemical composition rics for reverse airflow baghouse designs as these
are critical factors that influence the fabric fabrics are resistant to chemical attack and can
Pulverized Coal-Fired Boilers and Pollution Control 451

withstand temperatures up to 260  C [96]. The cake. The particle size impacts the buildup (and
primary coating in use today is poly- removal) of the dust cake on the filter and, along
tetrafluoroethylene (PTFE), which improves bag with the particle composition, impacts the adhe-
lifetime and provides resistance to acidic attack sion of particles to the filter surface and the cohe-
and thermal excursions. For pulse-jet baghouses sion of particles to each other. Reactions between
equipped units fired with low-sulfur coal, the bags the particles and constituents of the gas can also
are typically made of polyphenylene sulfide impact the cohesiveness of the dust cake. Finally,
(PPS) [97]. the cleaning cycle interval and intensity will also
The performance of a fabric filter is impacted impact the overall performance of a fabric filter.
by the particle properties, fabric properties, oper-
ating characteristics, cleaning method, and inter-
relationships between these parameters SO2 Emissions Control Technologies
[98]. While the fundamental mechanisms
involved in removing particles from the gas are The primary mechanisms for removal of a gas-
understood, the highly complex nature of the col- eous pollutant from a gas stream involve
lection process and the participation of the dust (1) absorption into a liquid, (2) gas/solid reaction,
layer in the collection process make the design of and (3) adsorption onto a solid. The first two
fabric filters more empirical in nature. The two mechanisms are the primary methods for removal
primary design and operating parameters are the of sulfur dioxide from coal-fired boiler flue gas.
air-to-cloth ratio, which sets the velocity ratio Commercial processes based upon these mecha-
passing through the filter and, hence, the relative nisms can be categorized into whether the process
size of the unit, and the pressure drop, which sets is regenerable, where the sulfur compound is sep-
the energy consumption requirements [99]. These arated from the absorbent, or nonregenerable,
two parameters are related as baghouses with a where the sulfur compounds are thrown away
small A/C ratio will have a lower pressure drop with the absorbent [102]. The product from regen-
than baghouses with a higher A/C ratio, resulting erable processes can be concentrated SO2, hydro-
in a trade-off between capital and operating costs. gen sulfide, elemental sulfur, or sulfuric acid. SO2
The bag cleaning method influences the A/C ratio removal processes can be further characterized
and pressure drop as more energetic cleaning into wet processes, where the product is a solution
methods can reduce the A/C ratio (and unit size) or slurry, and dry processes, where a dry product
for a given pressure drop [100]. is produced [103]. The majority of non-
Fabric filters are highly efficient collection regenerable processes produce a throwaway
devices for both coarse and fine particulate, with waste; however, some wet processes can be mod-
typical efficiencies of 99–99.9%. Due to the col- ified to produce a gypsum by-product that can be
lection mechanisms involved in fabric filtration, sold if site-specific conditions permit.
particle removal performance is not as sensitive to The primary technologies for removing sulfur
particle size for these devices as with ESPs. In dioxide from coal-fired power plants are wet lime-
comparing the costs of fabric filters to ESPs, fab- stone (CaCO3) and dry lime (Ca(OH)2) scrubbing
ric filters become more economic where highly [104], typically referred to as wet and dry flue gas
efficient removal of submicron particles are desulfurization (FGD). The preference for these
required [101]. As noted above, the design A/C technologies over regenerable processes is largely
ratio and operating pressure drop impact fabric driven by the low cost and high availability of
filter performance. Fabric filter performance is limestone and lime, by the relative simplicity of
also impacted by flue gas and particle characteris- these processes in comparison to other tail end
tics. The flue gas temperature impacts the volume processes, and by the fact that power plants are
of the flue gas and, hence, the operating A/C ratio, not chemical companies. For wet scrubbing, lime-
while the flue gas moisture content and composi- stone tends to be preferred over lime due to lower
tion can impact the characteristics of the dust cost, particularly if a ready supply is located close
452 Pulverized Coal-Fired Boilers and Pollution Control

to the power plant. In comparison to lime, lime- oxidation of SO3 to SO4 in solution, and crystal-
stone requires finer grinding, has higher transpor- lization of CaSO3 and CaSO4. By controlling the
tation costs, requires larger equipment, and is less solution pH and concentration of CaCO3 in the
responsive with respect to pH control [105]. Thus, solution, high levels of SO2 removal can be
absorbent selection is done by performing a com- achieved. As SO2 absorbs into the solution as
parison based upon site-specific technical and bisulfate (HSO3-), the primary product from the
economic factors. For wet FGD systems that pro- process is calcium sulfite. However, due to the
duce gypsum, which will be sold for wallboard high oxygen content of flue gas (3–10% by vol-
manufacturing, the use of hydrated lime rather ume), a fraction of the calcium sulfite will oxidize
than limestone can lead to improvements in the to calcium sulfate. Typically, the product from this
product quality in some cases [106]. In dry scrub- process is a waste sludge. It is possible to force
bing, lime is required due to the need for rapid oxidation of calcium sulfite into calcium sulfate
reaction times as lime has a higher reactivity than by bubbling air through the solution which
limestone. improves the ability to remove water from the
A wide variety of wet and dry FGD processes waste and permits the material to be upgraded
exist, but the basic processes are similar for use for gypsum (CaSO4 •2HO).
[107–109]. In addition to or instead of lime and The basic components of the limestone wet
limestone, several processes introduce sodium scrubbing process are illustrated in Fig. 6. Flue
compounds, such as sodium carbonate (e.g., gas from the particulate control device (ESP or
trona), where these materials are available locally baghouse) enters the bottom of an absorber tower
to improve process efficiency [110]. Seawater (typically a spray tower) and flows upward. The
scrubbing, which uses seawater to remove the absorbing solution is sprayed into the gas stream.
SO2 and discharge in into the ocean, has also The SO2 is absorbed into the solution where it
been applied to coal-fired power plants located reacts with the limestone to form solids. The
in a coastal environment [111]. While wet and resulting slurry is then captured in an effluent
dry scrubbing are highly efficient technologies, holding tank where ground limestone slurry and
they have high capital and operating costs. This recycled solution are added. This solution is then
has led to interest in lower capital cost SO2 recycled to the absorber. A portion of the solution
removal technologies such as furnace sorbent is sent to the thickener to concentrate the waste
injection [112, 113]. The remainder of this section solids as a sludge. The sludge from the thickener
will discuss the basic wet limestone and dry lime is then dewatered in a vacuum filter to produce a
scrubbing processes. filter cake, which is mixed with fly ash to stabilize
the waste prior to sending it to a landfill. Enhance-
ments to the basic process include the use of a pre-
Wet Flue Gas Desulfurization scrubber upstream of the main absorber to quench
the flue gas and protect the materials used in the
In wet FGD, flue gas containing SO2 is input into absorber, incorporation of a second scrubbing
a scrubbing tower or absorber where it is brought loop into the pre-scrubber, and blowing air into
into contact with an alkaline solution or slurry the effluent holding tank to force oxidation of
containing partially dissolved limestone calcium sulfite to calcium sulfate [114]. This latter
(CaCO3). The SO2 is absorbed into the solution improvement minimizes scale formation,
where it reacts to form calcium sulfite (CaSO3) improves the characteristics of the sludge, and
and calcium sulfate (CaSO4). The detailed chem- eliminates the need for adding fly ash for stabili-
istry of the process is complex and depends upon zation. Forced oxidation is also used to produce
the composition of the slurry and extent of disso- solids that can be used to produce gypsum for
lution of the limestone [105]. The primary steps wallboard manufacturing. If lime rather than lime-
involved in the process are gas/liquid mass trans- stone is used, forced oxidation is not required for
fer of the SO2, dissolution of the limestone, scale control. The cleaned flue gas exiting the
Pulverized Coal-Fired Boilers and Pollution Control 453

Reheater Clean Gas

to Stack

Mist Eliminator

Sprays

Absorber
Flue Gas Tower
from PM
Make Up
Control
Water

Limestone Slurry

Effluent
Thickener
Holding Tank

Fly Ash

Waste to Overflow
Mixer Vacuum Filter
Landfill Tank

Pulverized Coal-Fired Boilers and Pollution Control, Fig. 6 Basic limestone wet scrubbing process

absorber after passing through a mist eliminator most systems operate with a calcium-to-sulfur
designed to remove entrained droplets from the molar ratio of close to 1:1. Liquid pH is monitored
gas, passes through a reheater, and is then sent to and used to control the limestone addition rate to
the stack. A reheater is used to protect equipment maximum limestone utilization, to minimize the
downstream of the scrubber from condensation potential for scale in the system, and to reach a
and corrosion, to reduce the visible plume, and specific emissions target. The limestone grind fed
to improve rise and dispersion of the stack gas. to the FGD system can also be important, with finer
The performance of a wet FGD system is grinds leading to a smaller-sized reaction tank,
impacted by a number of design and operational higher SO2 removals, and an increase in gypsum
factors, most of which are related [114]. The design purity [116].
of the absorber tower must provide for effective Wet FGD systems are capable of achieving
contacting of the flue gas with the absorber solution over 95% SO2 removal, with some systems
and for separation of the cleaned gases from the being designed for 96–98%, and can be applied
liquid. From an economic standpoint, the size of to both low and high sulfur coals. As discussed
the tower is linked closely to the process chemistry above, a number of parameters can impact system
and the relative flows of flue gas and absorber performance and must be optimized to ensure
solution, which is typically expressed as the liquid- effective SO2 control as well as maintain system
to-gas ratio. Operational factors that impact perfor- reliability and availability. Overall, wet FGD sys-
mance include the SO2 concentration in the flue tems represent a significant capital, operating, and
gas, pH of the slurry, solids concentration in the maintenance expense for a coal-fired power plant.
slurry, concentration of other components in the The main operating and maintenance (O&M)
slurry such as magnesium and chloride ions, resi- costs differ for various designs and should be
dence time of the slurry in the reaction tank, and considered in selecting an FGD process
degree of slurry oxidation [105, 115]. High utiliza- [117]. Experience with existing plants has led to
tion of the limestone (or lime) is important and a number of equipment improvements, such as
454 Pulverized Coal-Fired Boilers and Pollution Control

large, single-tower, absorbers with higher flue gas the flue gas, a fraction of the CaSO3 will be
velocities and improved mist eliminator designs, oxidized into CaSO4.
and process improvements, such as the use of fine An overview of the lime dry scrubbing process
limestone and solution buffers [103]. These is shown in Fig. 7. Hot flue gas from the air
improvements can lead to higher SO2 removal preheater outlet enters the top of the cylindrical
efficiencies, lower capital costs, and reduced spray dryer where an atomizer sprays the alkaline
O&M costs. slurry from the feed preparation system into the
flue gas. SO2 is absorbed into the spray droplets
where it reacts with the lime to form a solid. The
water in the droplet evaporates leaving a fine
Dry Flue Gas Desulfurization
powder that can be collected from the spray
dryer and in a subsequent particulate collection
In dry FGD, flue gas containing SO2 is input into a
device. Twin fluid and rotary or spinning disk
spray dryer absorber where it is brought into con-
atomizers can be used; however, most large utility
tact with an alkaline solution containing hydrated
applications use a rotary atomizer. Typically, fab-
lime (Ca(OH)2). The SO2 is absorbed into the
ric filters are used for removal of the fly ash and
solution where it reacts with the alkali material
spent absorbent. Collection of the solids on the
to form solid calcium sulfite (CaSO3). Absorption
fabric filter permits additional SO2 removal to
and reaction occur while the thermal energy of the
occur resulting in increased sorbent utilization.
flue gas vaporizes the water in the droplets to
The spray dryer absorber needs to be designed
produce a fine powder. The primary processes
to maximize SO2 removal, yet produce a dry
involved in spray drying consist of atomization
product. Since absorption and reaction of SO2 is
of the alkaline slurry into a spray of droplets,
fastest when surface liquid is present, it is neces-
contacting of the spray with the flue gas, absorp-
sary to operate as close to the saturation point of
tion of SO2 into the droplets, reaction with the
the flue gas as possible and to maximize the res-
suspended alkali material, drying of the spray by
idence time available for reaction while still yield-
evaporation of the moisture in the droplets, and,
ing dry solids at the absorber walls and outlet
finally, separation of the solids from the gas
[119]. The quality of water that can be used for
[118]. Due to the high oxygen concentrations of
the slurry is limited by the boiler outlet

Pulverized Coal-Fired Fresh Lime

Boilers and Pollution Slurry
Control, Fig. 7 Basic lime Flue Gas
dry scrubbing process

Spray Clean Gas

Dryer to Stack
Absorber Fabric Filter

Make Up Recycle Solids for Solids for

Water Slurry Recycle Disposal

Waste to
Landfill
Pulverized Coal-Fired Boilers and Pollution Control 455

temperature, which is set by the boiler thermal For coal-fired power plants, the primary com-
efficiency, and the desired approach to saturation. bustion modification techniques are low-NOX
The concentration of solids in the slurry is limited burners, overfire air, and reburning technology
by the viscosity of the material which impacts the [121]. Each of these technologies relies on con-
practical aspects of handling and atomizing the trolling the combustion process in the furnace.
material. These two factors constrain the SO2 The primary postcombustion technologies are
removal that can be achieved for a given inlet selective noncatalytic reduction (SNCR) and
SO2. selective catalytic reduction (SCR) [122]. These
Dry FGD systems are capable of achieving technologies involve the use of a reagent that
over 95% SO2 removal and are typically applica- reduced NOX to N2. Combustion modification
ble to low and medium sulfur coals. There are technologies are generally the most cost-effective
technical constraints on the application of this when evaluated against postcombustion technol-
technology to high sulfur coals as an increase in ogies on a cost per mass of pollutant removed
sulfur content requires an increase in the solids’ [123]. SCR technology is the most expensive
content of the slurry and a point can be reached technology but also provides the highest level of
where the viscosity of the material is too high for NOX emission reduction. SNCR technology pro-
proper pumping and atomization [103]. vides a modest level of NOX reduction, with cost-
effectiveness typically between that of combus-
tion modifications and SCR.
In retrofitting existing units with NOX emis-
NOx Emissions Control Technologies sions control technologies, selection of the most
cost-effective technology or set of technologies
Technologies to reduce NOX emissions from generally requires a boiler-specific analysis that
combustion systems can be placed into two includes assessment of the NOX emission goals or
broad categories: (1) combustion modification limits and the expected performance and costs of
technologies which reduce the formation of NOX the specific technologies. Retrofit costs for NOX
during combustion and (2) postcombustion tech- control technologies are very site specific and
nologies which eliminate NOX from the flue gas depend upon the ease of the retrofit and the boiler
following combustion [28]. Combustion modifi- design and operating characteristics, such as the
cation techniques will generally seek to lower furnace size, firing configuration, and the condi-
flame temperature or to control the flame stoichio- tion of existing equipment [124]. The remainder
metric ratio depending upon whether thermal of this section will provide an introduction to NOX
NOX or fuel NOX emissions are being targeted. control technologies.
For pulverized coal combustion, as discussed pre-
viously, the majority of the NOX emissions result
from the nitrogen in the fuel; hence, the most Low-Nox Burners
effective combustion modifications are those
which control the stoichiometric ratio during Low-NOX burners reduce NOX formation in the
coal devolatilization. Postcombustion techniques combustion process by delaying the mixing of
tend to rely on the use of additives to reduce the fuel and air in the flame [125]. One means of
NOX to molecular nitrogen, N2. Removal of NOX aerodynamically staging fuel and air mixing in a
by wet scrubbing is difficult due to the insolubility coal flame for a wall-fired boiler is illustrated in
of NO and poor solubility of NO2. Scrubbing Fig. 8. In this low-NOX burner design, the mixing
technologies that can be applied to industrial of fuel and air is controlled by dividing the com-
plants have not been widely applied to large pul- bustion air into multiple streams, with separate
verized coal-fired power plants due to high capital control over each stream. Coal entering the flame
and operating costs and the low initial NOX is initially burned in a fuel-rich core which grad-
concentrations [120]. ually becomes fuel lean as air is progressively
456 Pulverized Coal-Fired Boilers and Pollution Control

Pulverized Coal-Fired Outer

Boilers and Pollution Inner Secondary Air
Control, Fig. 8 Typical Secondary Air
low-NOX burner design
approach Coal and
Primary Air Flame Core

Very fuel Progressive air Final air addition

rich zone addition into into flame
core

mixed into the flame. Staging of the air into the equipment; hence, the trade-off between optimal
flame permits volatile nitrogen compounds burner performance and acceptable carbon-in-ash
released in the early portion of the flame to be levels is strongly tied to the furnace geometry.
processed in a reducing environment, resulting in NOX control levels for the application of low-NOX
lower NOX emissions. Staging of the flame also burners to wall-fired utility boilers typically range
reduces peak flame temperatures and, therefore, from 40% to 50%.
can reduce thermal NOX formation as well.
For wall-fired boilers, low-NOX burner designs
vary from manufacturer to manufacturer [126–128], Overfire Air
but all tend to have similar features, including sep-
arate control over the combustion air split into two Overfire air is a combustion modification technol-
or more passages, control over the degree of swirl ogy that, like low-NOX burners, stages the com-
imparted to the air streams, and a flame stabilizing bustion process to reduce NO emissions [130]. To
device. The inner and outer air split and swirl set- apply overfire air to a coal-fired boiler, combus-
tings are adjusted to generate a stable low-NOX tion air is diverted from the main combustion zone
emissions flame for a specific coal. These settings and is injected through ports located on the walls
can also be adjusted to ensure that the flames do not above the burners. In this process, the primary
impinge on the side or rear walls of the boiler. zone is operated slightly less fuel lean than
For tangentially fired boilers, low-NOX burner usual, and fuel and air mixing is delayed. This
designs typically consist of alternating passages of delay reduces the formation of fuel NO and also
air and fuel [129]. The fuel nozzles direct the fuel reduces peak flame temperatures, resulting in a
into the center of the furnace, while the air nozzles reduction in thermal NO formation as well. Over-
direct the air closer to the furnace walls. Offsetting fire air is added to complete combustion of
the air from the fuel generates a fuel-rich flame in the unburned fuel.
center of the furnace which leads to low-NOX emis- For pulverized coal-fired boilers, the perfor-
sions. The offset air also provides a measure of mance of overfire air is dependent upon the over-
protection along the wall from rich conditions that fire air system design and the burner/boiler
can lead to wall corrosion. Additional air staging is characteristics. Optimal NOX control for a partic-
achieved by adding a fraction of the combustion air ular system can be achieved provided that the
through air ports located on the top of the burner overfire air jets mix effectively, that good control
stack. The air added to the top of the burner is over the main flame can be maintained as com-
typically called “close-coupled overfire air” or bustion air is diverted to the overfire air system,
CCOFA. The performance of low-NOX burners in and that there is sufficient residence time in the
a particular application is very dependent upon the furnace for carbon burnout to occur. As with
boiler and coal characteristics. Low-NOX burners low-NOX burners, the NOX control performance
generally produce longer flames than the original of overfire air is limited by acceptable carbon-in-
Pulverized Coal-Fired Boilers and Pollution Control 457

ash. For pulverized coal-fired industrial boilers, Burnout Zone: In this final zone, combustion air is
NOX reductions between 15% and 30% are gen- added to complete the oxidation of the
erally expected with the application of overfire air. reburning fuel and to bring the boiler back to
a normal operating excess air level.

Reburning Studies of the reburning process have shown

that natural gas, fuel oil, and coal can be used as
Reburning is a combustion modification technology reburning fuels [132]. As shown in Fig. 10, hydro-
that removes NOX from combustion products by carbon fuels with low fuel-bound nitrogen, such
using fuel as the reducing agent [131]. As illustrated as propane or natural gas, provide the lowest NOX
in Fig. 9, application of reburning to a coal-fired emissions since nitrogen in the reburning fuel
boiler conceptually divides the furnace into three tends to contribute to the final NOX exiting the
zones (primary zone, reburning zone, and burnout process. In Fig. 10, the reburning zone stoichio-
zone) which are typically defined as follows: metric ratio is defined as the ratio of the total air
Primary Zone: The primary zone consists of the supplied to the primary and reburning zones to the
normal firing system and constitutes the bulk total stoichiometric air requirements of the pri-
(typically 80–90%) of the total heat release. mary and reburning fuels. The parameters which
This zone is operated fuel lean at an excess control the performance of the reburning process
air level which is close or slightly below nor- have been defined through an extensive series of
mal operation. NOX formed in this zone enters experimental studies [133]. For boiler applica-
the reburning zone. tions, the most critical parameters are the
Reburning Zone: The reburning fuel is injected reburning zone stoichiometric ratio and the initial
downstream of the primary zone to create a NOX level. As shown in Fig. 10, most fuels
slightly fuel-rich, NOX reduction zone. The exhibit a maximum NOX reduction (i.e., mini-
reburning fuel provides the remainder, nor- mum NOX emissions) at a reburning stoichiomet-
mally 10–20%, of the total heat input. In this ric ratio of about 0.9, which corresponds to a
zone, hydrocarbon radicals generated during reburning fuel input of approximately 18–19%
breakdown of the reburning fuel react with of the total heat input at a primary zone excess
NO molecules to form other reactive nitroge- air level of 10%. NOX reduction with reburning
nous species, such as HCN, which react with increases as the NOX level from the primary zone
other NO molecules to form molecular nitro- increases.
gen, N2. Similar to other combustion modification tech-
niques, the most significant factors which limit the
performance of reburning in full-scale applica-
tions are the design of the reburning system and
the boiler characteristics. The boiler geometry
Burnout Zone
must permit adequate mixing of the reburning
Overfire Air
fuel and overfire air to be achieved and must
Reburning Zone
Reburn Fuel provide adequate residence time for the process
to be implemented above the primary combustion
Main Burners Primary Zone zone. Reliable methodologies for scaling the
reburning process to boilers with varying charac-
teristics have been developed [134]. The use of
these methodologies has led to NOX control levels
between 50% and 70% for application of
Pulverized Coal-Fired Boilers and Pollution Control,
reburning to full-scale utility boilers, independent
Fig. 9 Application of reburning to a utility boiler of the firing configuration.
458 Pulverized Coal-Fired Boilers and Pollution Control

Pulverized Coal-Fired 700

Boilers and Pollution Reburn Fuel
Control, Fig. 10 Impacts Utah Coal
of reburning zone 600
stoichiometry and fuel on Colstrip Coal
reburning performance Beulah Coal

Exhaust NOx, ppm 0% O2)

500
Yallourn Coal

400 Propane

300

200

100

0
0.6 0.7 0.8 0.9 1.0 1.1 1.2
Reburn Zone SR

Selective Noncatalytic Reduction 100

Reagent
Selective noncatalytic reduction is a flue gas Ammonia

treatment process in which an amine-containing 80 Urea

agent, such as ammonia (NH3) or urea (CO(NH2)

2), is injected into combustion gases to react with
(NOx)out/(NOx)in

60
and reduce NO formed during the combustion
process [135, 136]. At the proper temperature,
NH2, generated from decomposition of the 40
injected reagent, reacts with NO via a complex
set of reactions to form N2. The chemistry of the
SNCR process is illustrated in the following 20
global reactions that are based upon the use of
ammonia as the reagent:
0
800 900 1,000 1,100 1,200
NH3 þ OH ! NH2 þ H2 O ð6Þ Temperature, °C

NH2 þ NO ! H þ OH þ N2 ð7Þ Pulverized Coal-Fired Boilers and Pollution Control,

Fig. 11 Impacts of gas temperature in SNCR performance
H þ OH ! H2 O ð8Þ
Figure 11 shows the impacts of gas tempera-
Initially, the injected ammonia is activated by ture on the NOX reduction achieved with urea or
OH to form a reactive amine, NH2, radical. This ammonia in tests performed in a pilot-scale com-
radical then selectivity reacts with NO to form bustion facility. As shown in this figure, at the
molecular nitrogen, N2, and H and OH radicals. proper flue gas temperature (approximately
The OH radical can help to activate additional 950–1000  C), the injected reagent selectively
ammonia or can react with H radicals to form reduces NO to molecular nitrogen, N2. As tem-
water, H2O. peratures increase, NOX reduction performance is
Pulverized Coal-Fired Boilers and Pollution Control 459

Pulverized Coal-Fired
Boilers and Pollution
Control,
Fig. 12 Operating range
for various SCR catalyst Zeolites
formulations

V-Ti Oxides

Pt-Pd Oxides

200 300 400 500 600

Catalyst Temperature, °C

reduced and a portion of the reagent can be oxi- reduction are generally achieved in boilers with
dized to NO. As temperatures decrease, a portion slower quench rates (temperature decay versus
of the reagent exits the process unconverted. The residence time) in the upper furnace and with
unconverted reagent exiting the process is referred good access to the optimal temperature window.
to as ammonia slip.
The performance of SNCR on utility boilers
depends upon the reagent used, the quantity of Selective Catalytic Reduction
reagent injected, and the boiler design and oper-
ating characteristics [137]. For maximum perfor- Selective catalytic reduction is a high-efficiency
mance, the injection system must be designed to flue gas treatment process in which ammonia is
provide effective mixing of the reagent within the added to the flue gas to react with NO over a
optimal temperature window. In addition, the catalyst [139]. The primary overall reactions for
boiler design should provide adequate residence this process can be expressed as:
time at the proper temperatures. For pulverized
coal-fired boilers, the optimal temperature win- Catalyst
dow generally resides within the convective heat 4NO þ 4NH3 þ O2 ! 4N2 þ 6H2 O ð9Þ
transfer sections, where access is limited and
Catalyst
quench rates are high [138]. Wide variations in 2NO2 þ 4NH3 ! 3N2 þ 6H2 O ð10Þ
temperature at the point of injection due to strat-
ification or to changes in load can also limit per- SCR catalysts are made of a solid ceramic
formance. Multiple injection locations can help to material that contains active catalytic compo-
reduce the negative impacts of temperature varia- nents. The ceramic may be in the form of a hon-
tions. Due to concerns over the handling and eycomb monolith or it may be wash coated onto a
storage of anhydrous ammonia, aqueous ammo- ceramic, metal, or fiber substrate [140]. The typ-
nia or urea is generally preferred for application of ical operating temperature range for available cat-
SNCR to utility boilers. alysts is shown in Fig. 12. Noble metal catalysts,
Application of SNCR to coal-fired boilers is such as platinum-palladium formulations, func-
expected to reduce NOX emissions by 30–60% tion at lower temperatures (150–290  C;
from uncontrolled levels. The higher levels of 300–550  F) than base metal and zeolite catalysts
460 Pulverized Coal-Fired Boilers and Pollution Control

but are susceptible to deactivation by sulfur influence system performance. Coal properties
oxides. Base metal catalysts, such as vanadium- are an important consideration for catalyst selec-
titanium formulations, function at temperatures in tion [143]. In general, since the SCR system is
the range of 260–490  C (500–900  F) and are separate from the combustion process and since
more resistant to sulfur deactivation. Zeolite cata- the extensive retrofit requirements permit the
lysts operate at a temperature range above that of design to be optimized, the overall control perfor-
the other catalysts (340–590  C, 650–1100  F) mance is less influenced by boiler characteristics
and have shown a good resistance to poisoning than the other NOX control technologies
from trace elements in coal such as arsenic. The described previously. On the other hand, the cost
majority of SCR technology applications to pul- of installing an SCR system is very sensitive to
verized coal-fired boilers use vanadiatitania cata- site-specific factors due to the extensive retrofit
lysts as they are widely available and have good requirements [144]. NOX reduction in the range of
resistance to deactivation by sulfur oxides in the 70–90% is expected for application of SCR to
flue gas. Catalyst formulations are specific to the coal-fired boilers.
catalyst manufacturer and may include the addi-
tion of other base metal oxides to increase the
catalyst activity, to increase poison resistance, to Future Directions
minimize ammonia slip, and to reduce SO2 oxi-
dation to SO3. Significant steps have been taken to reduce air
To meet the temperature requirements for pollutant emissions from coal-fired power plants
vanadiatitania catalysts, the catalyst needs to be throughout the developed countries in the world.
installed in the flue gas path between the boiler Promulgation of more stringent emission limits
economizer and air preheater. The arrangement in various counties will require further applica-
for both new and retrofit SCR systems depends tion of these technologies to smaller power plants
upon the boiler and balance of plant equipment and place constraints on new power plants. The
arrangement [141, 142]. Typically, one or two cost of the air pollution control technologies is a
layers of catalyst are incorporated into a catalyst challenge with both retrofit and new technologies
housing or reactor. Flue gas from the economizer and, hence, research and technology develop-
exit is either routed in a horizontal duct to the top ment is focused on reducing both the capital
of the reactor where it then flows downward over and operating costs of these technologies, on
the catalyst. Alternatively, the ductwork arrange- improving performance to meet stricter stan-
ment may need a vertical upward duct followed by dards, and on improving equipment reliability
a 180 transition into the top of the reactor. and availability. In addition, due to concerns
Ammonia is injected into the flue gas upstream over emissions of air toxic compounds from
of the SCR reactor using an ammonia injection coal-fired power plants, the potential positive
grid that typically consists of a series of injection and negative impacts of existing air pollution
lances spaced across the duct and designed to control equipment on these emissions is an area
provide good mixing of the ammonia with the of active investigation.
flue gas. Static mixers and flow conditioning
devices such as turning vanes are used to ensure
uniform distribution of the ammonia and flue gas
Bibliography
within the ductwork.
A number of factors need to be considered in 1. International Energy Agency (2010) World energy
designing an SCR system, including reactor loca- outlook 2010. International Energy Agency, Paris
tion and design, catalyst type and configuration, 2. World Coal Insitute (2009) The coal resource, a com-
prehensive overview of coal. World Coal Insitute,
and the location and design of the ammonia injec-
London
tion grid. The catalyst space velocity and unifor- 3. Freese B (2003) Coal: a human history. Perseus,
mity of ammonia and flue gas distribution all Cambridge, MA
Pulverized Coal-Fired Boilers and Pollution Control 461

4. Stuart R (1824) A descriptive history of the steam health criteria. EPA/600/R-08/071. Office of
engine. John Knight and Henry Lacey, London Research and Development, Research Triangle Park
5. Burn RS (1854) The steam engine. H. Ingram, London 24. Glass NR (1979) Environmental effects of increased
6. Thurston RH (1903) A history of the growth of the coal utilization: ecological effects of gaseous emis-
steam engine, 4th edn. D. Appelton, New York, sions from coal combustion, vol 33. U.S. National
Revised Institute of Environmental Health Sciences, Research
7. Tagliaferro L (2003) Thomas Edison: inventor of the Triangle Park, pp 249–272
age of electricity. Lerner, Minneapolis 25. Bennett DA, Goble RL, Linhurst RA (1985) The
8. Woodside M (2007) Thomas A. Edison: the man who acidic deposition phenomenon and its effects: critical
lit up the world. Sterling, New York assessment document. EPA/600/8–85/001. U.S.
9. Babcock & Wilcox Company (1992) Steam: its gen- Environmental Protection Agency, Office of
eration and use, Fortieth edition. Babcock & Wilcox, Research and Development, Research Triangle Park
Barberton 26. National Acid Precipitation Assessment Program
10. Combustion Engineering, Inc (1981) Combustion: (2005) National Acid Precipitation Assessment Pro-
fossil power systems, 3rd edn. Combustion Engineer- gram report to Congress: an integrated assessment.
ing, Windsor National Acid Precipitation Assessment Program
11. Heidorn KC (1978) A chronology of important 27. U.S. Environmental Protection Agency (1997) Nitro-
events in the history of air pollution meteorology to gen oxides: impacts on Public Health and the Envi-
1970. Bull Am Meteorol Soc, vol 59, issue 12. Amer- ronment. EPA-452-R-97-002. Office of Air and
ican Meteorological Society, Toronto, pp 1589–1597 Radiation, Research Triangle Park
12. Thorshiem P (2006) Inventing pollution: coal, 28. U.S. Environmental Protection Agency (1999) Nitro-
smoke, and culture in Britain since 1800. Ohio Uni- gen oxides (NOx), why and how they are controlled.
versity Press, Athens EPA-456/F-99-006R. Office of Air Quality Planning
13. Jacobs C, Kelly WJ (2008) Smogtown: the lung- and Standards, Research Triangle Park
burning history of pollution in Los Angeles. The 29. McConville A (1997) Emissions standards hand-
Overlook Press, New York book. IEACR/96. IEA Coal Research, London
14. Brimblecombe P (1987) The big smoke: a history of 30. U.S. Environmental Protection Agency (2000) Air
air pollution in London since medieval times. quality criteria for carbon monoxide. EPA 600/P-
Methuen, London 99/001F. Office of Research and Development,
15. Greater London Authority (2002) 50 years on: the Washington, DC
struggle for air quality in London since the great 31. U.S. Environmental Protection Agency (1998) Study of
smog of December 1952. Greater London Authority, hazardous air pollutant emissions from Electric Utility
London Steam Generating Units. Final report to congress, vols
16. Davis DL (2002) When smoke ran like water: tales of 1 and 2. EPA-453/R-98-004a, b. Office of Air Quality
environmental deception and the battle against pol- Planning and Standards, Research Triangle Park
lution. Basic Books, New York 32. U.S. Environmental Protection Agency (1997) Mer-
17. Reitze AW (2005) Stationary source air pollution cury study report to congress, vols I–VII. EPA-452/
law. Environmental Law Institute, Washington, DC R-97-003, 010. Office of Air Quality Planning and
18. Martineau RJ, Novello DP (2004) The clean air act Standards, Research Triangle Park
handbook. American Bar Association, Chicago 33. U.S. Environmental Protection Agency (2000) Regu-
19. U.S. Environmental Protection Agency latory finding on the emissions of hazardous air pol-
(2005) Review of the National Ambient Air Quality lutants from Electric Utility Steam Generating Units,
Standards for particulate matter. Policy assessment of vol 65(245). Federal Register, Research Triangle
scientific and technical information. OAQPS staff Park, pp 79825–79831
paper. EPA-425/R-05-005a. Office of Air Quality 34. National Research Council (2000) Toxicological
Planning and Standards, Research Triangle Park effects of methylmercury. National Academy Press,
20. U.S. Environmental Protection Agency (2009) Inte- Washington, DC
grated science assessment for particulate matter. 35. European Commission (2001) Ambient air pollution
Office of Research and Development, Research Tri- by mercury (Hg). Position paper. Office for Official
angle Park. EPA/600/R-08/139F Publications of the European Communities,
21. U.S. Environmental Protection Agency (2004) Air Luxembourg
quality criteria for particulate matter. EPA/600/P-99/ 36. Environment Canada (2010) Risk management strat-
002aF. Office of Research and Development, egy for mercury
Research Triangle Park 37. UNEP Chemicals Branch, DTIE (2010) The global
22. U.S. Environmental Protection Agency (2008) Inte- atmospheric mercury assessment: sources, emissions
grated science assessment for sulfur oxides – health and transport. United Nations Environment Pro-
criteria. EPA/600/R-08/047F. Office of Research and gramme, Geneva
Development, Research Triangle Park 38. US Environmental Protection Agency (2011) National
23. U.S. Environmental Protection Agency (2008) Inte- Emission Standards for Hazardous Air Pollutants
grated science assessment for oxides of nitrogen – from Coal and Oil-Fired Electric Utility Steam
462 Pulverized Coal-Fired Boilers and Pollution Control

Generating Units and Standards of Performance for 51. Kramlich JC, Chenevert B, Park J, Hoffman DA,
Fossil Fuel-Fired Electric Utility, Industrial- Butcher EK (1996) Suppression of fine ash formation
Commercial- Institutional, and Small Industrial- in pulverized coal flames. Final Technical Report
Commercial-Institutional Steam Generating Units, prepared for DOE Grant No. DE-FG22-
vol 76(85). Federal Register, Research Triangle Park, 92PC92548. U.S. Department of Energy, Pittsburgh
pp 24976–25147 Energy Technology Center, Pittsburgh
39. Mitchell JFB (1989) The “Greenhouse” effect and 52. Acurex Environmental Corporation (1993) Emission
climate change, vol 27, No 1. American Geophysical factor documentation for AP-42 Section 1.1 bituminous
Union, Washington, DC, pp 115–139 and subbituminous coal combustion. U.-
40. U.S. Environmental Protection Agency (1995) Com- S. Environmental Protection Agency, Office of Air
pilation of air pollutant emission factors, 5th edn. Quality Planning and Standards, Research Triangle Park
AP-42. Office of Air Quality Planning and Standards, 53. Attar A, Dupuis F (1979) Data on the distribution of
Research Triangle Park organic sulfur functional groups in coal. Preprints of
41. Eggleston HS, Buendia L, Miwa K, Ngara T, Tanabe the Papers of the Spring National Meeting of the
K (eds) (2006) 2006 IPCG guidelines for National Division of Fuel Chemistry, vol 24(1). American
Greenhouse Gas Inventories. National Greenhouse Chemical Society, Honolulu, pp 166–177
Gas Inventories Programme. Intergovernmental 54. Pershing DW, Silcox GD (1988) SOx fundamentals.
Panel on Climate Change, Hayama, Kanagawa Combustion of solid fuels – a one week intensive
42. Ciferno JD, Fout TE, Jones AP, Murphy JT course. International Flame Research Foundation,
(2009) Capture carbon from existing coal-fired Noordwijkerhout
power plants. Chemical engineering progress, Amer- 55. Davis K, Dissel A, Valentine J (2001) The evolution
ican Institute of Chemical Engineers, New York, of pyritic and organic sulfur from pulverized coal
pp 33–41 particles during combustion. The 2nd joint meeting
43. IRA Greenhouse R&D Programme (2007) CO2 cap- of the US Sections of The Combustion Institute. The
ture ready plants. Report No. 2007/4. International Combustion Institute, Oakland
Energy Agency, Cheltenham 56. Muller III CH, Schofield K, Steinberg M, Broida HP
44. Newton GH, Schieber C, Socha RG, Kramlich JC (1978) Sulfur chemistry in flames. In: Proceedings of
(1990) Mechanisms governing fine particulate emis- the seventeenth symposium (international) on combus-
sions from coal flames. DOE/PC/73743-T8. U.- tion. The Combustion Institute, Pittsburgh, pp 867–879
S. Department of Energy, Pittsburgh 57. Kramlich JC, Malte PC, Grosshandler WL
45. Gluskoter HJ (1978) An introduction to the occur- (1981) The reaction of fuel-sulfur in hydrocarbon
rence of mineral matter in Coal. [ed.] Richard combustion. In: Proceedings of the eighteenth sym-
W. Bryers. Ash deposits and corrosion due to impu- posium (international) on combustion. The Combus-
rities in combustion gases. Hemisphere Publishing tion Institute, Pittsburgh, pp 151–161
Department, Washington, DC, pp 3–19 58. Shao D, Hutchinson EJ, Heidbrink J, Pan W-P, Chou
46. Chen Y, Shah N, Huggins FE, Huffman GP, Linak C-L (1994) Behavior of sulfur during coal pyrolysis.
WP, Andrew Miller C (2004) Investigation of pri- J Anal Appl Pyrolysis 30:91–100. Elsevier,
mary fine particulate matter from coal combustion by Amsterdam
computer-controlled scanning electron microscopy, 59. Castaldini C, Angwin M (1977) Boiler design and
vol 85. Fuel Processing Technology, Elsevier, operating variables affecting uncontrolled sulfur
Amsterdam, pp 743–761 emissions from pulverized-coal-fired steam genera-
47. Folsom BA, Heap MP, Pohl JH (1986) State-of-the- tors. EPA-450/3–77-047. U.S. Environmental Pro-
art review summary and program plan. Effects of coal tection Agency, Research Triangle Park
quality on power plant performance and costs, 60. Folkedahl BC, Zygarlicke CJ (2004) Sulfur retention
vol 1. Report No. CS-4283. Electric Power Research in North Dakota lignite coal ash. Preprints of the
Institute, Palo Alto papers of the spring national meeting of the Division
48. Miller CA, Linak WP (2002) Primary particles gen- of Fuel Chemistry, vol 49(1). American Chemical
erated by the combustion of heavy fuel oil and coal. Society, Anaheim, pp 167–168
Review of research results from EPA’s National Risk 61. Blythe G, Dombrowski K (2004) SO3 mitigation
Management Research Laboratory. EPA-600/R-02- guide update. Report No. 1004168. Electric Power
093. U.S. Environmental Protection Agency, Office Research Institute, Palo Alto
of Research and Development, Research Triangle 62. Srivastava RK, Miller CA, Erickson C, Jambhekar
Park R (2004) Emissions of sulfur trioxide from coal-fired
49. Flagan RC, Seinfeld JH (1988) Fundamentals of air power plants. J Air Waste Manag Assoc 54:750–762.
pollution engineering. Prentice Hall, Englewood Air and Waste Management Association, Pittsburgh
Cliffs 63. Glassman I (1987) Combustion, 2nd edn. Academic,
50. Flagan RC (1978) Submicron particles from coal Orlando
combustion. In: Proceedings of the seventeenth sym- 64. Fenimore CP (1971) Formation of nitric oxide in
posium (international) on combustion. The Combus- premixed hydrocarbon flames. In: Proceedings of
tion Institute, Pittsburgh, pp 97–104 the thirteenth symposium (international) on
Pulverized Coal-Fired Boilers and Pollution Control 463

combustion. The Combustion Institute, Pittsburgh, control, 1970–1980. U.S. Environmental Protection
pp 373–389 Agency, Air and Energy Engineering Research Labo-
65. Pershing DW, Wendt JOL (1977) The influence of ratory, Research Triangle Park
flame temperature and coal composition on thermal 77. Chen SL, Heap MP, Pershing DW, Martin GB
and fuel NOx. In: Proceedings of the sixteenth sym- (1982) Influence of coal composition on the fate of
posium (international) on combustion. The Combus- volatile and char nitrogen during combustion. In:
tion Institute, Pittsburgh, pp 389–399 Proceedings of the nineteenth symposium
66. Sarofim AF, Beér JM (1990) The fate of fuel nitrogen (international) on combustion. The Combustion
and ash during combustion of pulverized coal, Institute, Pittsburgh, pp 1271–1280
Chapter 4. In: Lemieux PM, Air and Energy Engi- 78. Bartok W, Crawford AR, Piegari GJ (1971) System-
neering Research Laboratory (eds) Pulverized coal atic field study of NOx emission control methods for
combustion: pollutant formation and control, utility boilers. APTD-1163. U.S. Environmental Pro-
1970–1980. U.S. Environmental Protection Agency, tection Agency, Office of Air Programs, Research
Air and Energy Engineering Research Laboratory, Triangle Park
Research Triangle Park 79. Payne R, Heap MP, Pershing DW (1981) NOx for-
67. Hayhurst AN, Vince IM (1980) Nitric oxide forma- mation and control in pulverized-coal flames. In:
tion from N2 in flames. Prog Energy Combust Sci Proceedings of the low rank coal technology devel-
6:35–51. Elsevier, Amsterdam opment workshop, San Antonio, 17–18 June 1981
68. Kelemen SR, Gorbaty ML, Vaughn SN, Kwiatek PJ 80. Licht W (1980) Air pollution control engineering,
(1993) Quantification of nitrogen forms in argonne basic calculations for particulate collection. Marcel
premium coals. Preprints of the papers of the spring Dekker, New York
national meeting of the Division of Fuel Chemistry, 81. Strauss W (1975) Industrial gas cleaning, 2nd edn.
vol 38(2). American Chemical Society, Denver, Pergamon Press, New York
pp 384–392 82. Kinsey JS, Schliesser S, Englehart PJ (1985) Control
69. Mitra-Kirtley S, Mullins OC, Branthaver J, Van technology for sources of PM10. Draft report pre-
Elp J, Cramer SP (1993) Nitrogen XANES studies pared for EPA Contract No. 68-02-03891, work
of fossil fuels. Preprints of the papers of the fall assignment 4. U.S. Environmental Protection
national meeting of the Division of Fuel Chemistry, Agency, Office of Air Quality Planning and Stan-
vol 38(3). American Chemical Society, Chicago, dards, Research Triangle Park
pp 762–768 83. U.S. Environmental Protection Agency (1982) Con-
70. Solomon PR, Fletcher TH (1994) Impact of coal pyrol- trol techniques for particulate emissions from station-
ysis on combustion. In: Proceedings of the twenty-fifth ary sources, vol 1. EPA-450/3-81-005a. Emissions
symposium (international) on combustion. The Com- Standards and Engineering Division, Research Tri-
bustion Institute, Pittsburgh, pp 463–474 angle Park
71. Pohl JH, Sarofim AF (1977) Devolatilization and 84. White HJ (1977) Electrostatic precipitation of fly ash.
oxidization of coal nitrogen. In: Proceedings of the J Air Pollut Control Assoc 27(1):15–21. Air Pollu-
sixteenth symposium (international) on combustion. tion Control Association, Pittsburgh
The Combustion Institute, Pittsburgh, pp 491–501 85. U.S Environmental Protection Agency (2011) Elec-
72. Rees DP, Smoot LD, Hedman PO (1981) Nitrogen trostatic precipitator plan review. APTI Course
oxide formation inside a laboratory pulverized coal No. SI:412B. U.S. Environmental Protection
combustor. In: Proceedings of the eighteenth sympo- Agency, Research Triangle Park
sium (international) on combustion. The Combustion 86. White HJ (1977) Electrostatic precipitation of fly ash.
Institute, Pittsburgh, pp 1305–1311 Fly ash and furnace gas characteristics. J Air Pollut
73. Chen JC, Niksa S (1992) Suppressed nitrogen evo- Control Assoc 27(2):114–120. Air Pollution Control
lution from coal-derived soot and low-volatility coal Association, Pittsburgh
chars. In: Proceedings of the twenty-fourth sympo- 87. Turner JH, Lawless PA, Yamamoto T, Coy DW,
sium (international) on combustion. The Combustion Mckenna JD, Mycock JC, Nunn AB, Greiner GP,
Institute, Pittsburgh, pp 1269–1276 Vatavuk WM (2002) Electrostatic precipitators. In:
74. Zhang H (2001) Nitrogen evolution and soot forma- Mussatti DC (ed) EPA air pollution control cost man-
tion during secondary coal pyrolysis. PhD disserta- ual. U.S. Environmental Protection Agency, Office of
tion, Brigham Young University Air Quality Planning and Standards, Research Trian-
75. Pohl JH, Chen SL, Heap MP, Pershing DW gle Park. Chapter 3 in Section 6, Particulate matter
(1983) Correlation of NOx emissions with basic phys- controls
ical and chemical characteristics of coal. In: Proceed- 88. Szabo MF, Shah YM (1979) Inspection manual for
ings of the 1982 joint symposium on NOx control, evaluation of electrostatic precipitator performance.
vol 2. Electric Power Research Institute, Palo Alto EPA-340/1-79-007. U.S. Environmental Protection
76. Pershing DW, Heap MP, Chen SL (1990) Bench-scale Agency, Division of Stationary Source Enforcement,
experiments on the formation and control of NOx emis- Washington, DC
sions from pulverized coal combustion, Chapter 9. In: 89. White HJ (1977) Electrostatic precipitation of fly ash.
Pulverized coal combustion: pollutant formation and Precipitator design. J Air Pollut Control Assoc
464 Pulverized Coal-Fired Boilers and Pollution Control

27(3):206–217. Air Pollution Control Association, 102. Devitt T, Gestle R, Gibbs L, Hartman S, Klier
Pittsburgh R (1978) Flue gas desulfurization system capabilities
90. National Air Pollution Control Administration for coal-fired steam generators, vol II. Technical
(1969) Control techniques for particulate air pollut- report. U.S. Environmental Protection Agency,
ants. NAPCA Publication No. AP-51. National Air Office of Research and Development, Washington,
Pollution Control Administration, Washington, DC DC, EPA-600/7-78-032b
91. Turner JH, Lawless PA, Yamamoto T, Coy DW, 103. Wojciech J, Singer C, Srivastava RK, Tsirigotis PE
Mckenna JD, Mycock JC, Nunn AB, Greiner GP, (1999) Status of SO2 scrubbing technologies. EPRI-
Vatavuk WM (2002) Bag-houses and filters. In: DOE-EPA combined utility air pollution control
Mussatti DC (ed) EPA air pollution control cost man- symposium. TR-113187-V1. Electric Power
ual. U.S. Environmental Protection Agency, Office of Research Institute, Palo Alto
Air Quality Planning and Standards, Research Trian- 104. Srivastava RK (2000) Controlling SO2 emissions: a
gle Park. Chapter 1 in Section 6, Particulate matter review of technologies. EPA/600/R-00/093. U.S.
controls Environmental Protection Agency, Office of
92. Cushing KM, Merritt RL, Chang RL (1990) Operat- Research and Development, Washington, DC
ing history and current status of fabric filters in the 105. Kohl A, Riesenfeld F (1985) Gas purification,
utility industry. J Air Waste Manage Assoc 4th edn. Gulf, Houston
40(7):1051–1058. Air and Waste Management Asso- 106. Weiler H, Ellison W (1997) Wet gypsum-yielding
ciation, Pittsburgh FGD experience using quicklime reagent. EPRI-
93. Belba VH, Theron Grubb W, Chang R (1992) The DOE-EPA combined utility air pollutant control sym-
potential of pulse-jet baghouses for utility boilers. posium. The mega symposium. TR-108683-V2.
Part 1: a worldwide survey of users. J Air Waste Electric Power Research Institute, Palo Alto
Manage Assoc 42(2):209–217. Air and Waste Man- 107. United Engineers and Constructors, Inc (1991) Eco-
agement Association, Pittsburgh nomic evaluation of flue gas desulfurization systems.
94. Sloat DG, Gaikwad RP, Chang RL (1993) The poten- GS-7193, vol 1. Electric Power Research Institute,
tial of pulse-jet baghouses for utility boilers. Part 3: Palo Alto
comparative economics of pulse-jet baghouse pre- 108. United Engineers and Constructors, Inc (1992) Eco-
cipitators and reversegas baghouses. J Air Waste nomic evaluation of flue gas desulfurization systems.
Manage Assoc 43:120–128. Air and Waste Manage- GS-7193, vol 2. Electric Power Research Institute,
ment Association, Pittsburgh Palo Alto
95. EC/R Incorporated (1998) Stationary source control 109. United Engineers and Constructors, Inc (1995) Eco-
techniques document for fine particulate matter. nomic evaluation of flue gas desulfurization systems.
Report prepared for EPA Contract No. 68-D-98- GS-7193-V3, vol 3. Electric Power Research Insti-
026, Work Assignment No. 0-08. U.S. Environmental tute, Palo Alto
Protection Agency, Research Triangle Park 110. Fox MR, Hunt TG (1990) Flue gas desulfurization
96. Felix LG, Cushing KM, Grubb WT, Giovanni DV using dry sodium injection. Presented at the
(1988) Fabric filters for the electric utility industry. EPA/EPRI 1990 SO2 control symposium. s.n., New
In: Guildelines for fabrics and bags, vol 3, CS-5161. Orleans
Electric Power Research Institute, Palo Alto 111. Radojevic M (1991) Scrubbing of flue gases with sea-
97. Electric Power Research Institute (2005) Utility water. Presented at the AFRC/JFRC international
boiler bag-house update. Report No. 1010367. Elec- conference on environmental control of combustion
tric Power Research Institute, Palo Alto processes, 7–10 October 1991. s.n., Honolulu
98. Dennis R (1974) Collection efficiency as a function 112. Zhou W, Maly P, Brooks J, Nareddy S, Swanson L,
of particle size, shape and density: theory and expe- Moyeda D (2010) Design and test furnace sorbent
rience. J Air Pollut Control Assoc injection for SO2 removal in a tangentially fired
23(12):1156–1161. Air Pollution Control Associa- boiler. Environmental engineering science,
tion, Pittsburgh vol 27, Number 4. Mary Ann Liebert, New Rochelle,
99. Carr RC, Smith WB (1974) Fabric filter technology pp 337–345
for utility coal-fired power plants. J Air Pollut Con- 113. Nolan PS (1996) Emission control technologies for
trol Assoc 34(1):79–89. Air Pollution Control Asso- coal-fired power plants. Presented at the People's
ciation, Pittsburgh Republic of China Ministry of Electric Power semi-
100. Carr RC, Cushing KM, Gallaer CA, Smith WB nar, 22–25 April 1996, s.n., Beijing
(1992) Fabric filters for the electric utility industry. 114. Henzel DS, Laseke BA, Smith EO, Swenson DO
In: Guidelines for fabric filter design, vol 5, CS-5161. (1982) Handbook for flue gas desulfurization scrub-
Electric Power Research Institute, Palo Alto bing with limestone. Noyes Data Corporation, Park
101. Bustard CJ, Cushing KM, Pontius DH, Smith WB Ridge
(1988) Fabric filters for the electric utility industry. 115. Miller SF, Miller BG, Scaroni AW (1997) Limestone
In: General concepts, vol 1, CS-5161. Electric Power performance in a pilot-scale forced oxidation scrub-
Research Institute, Palo Alto ber. EPRI-DOE-EPA combined utility air pollutant
Pulverized Coal-Fired Boilers and Pollution Control 465

control symposium. The mega symposium. 1989 EPRI/EPA joint symposium on stationary com-
TR-108683-V2, Electric Power Research Institute, bustion NOx control, 6–9 Mar 1989. EPRI/EPA, San
Palo Alto Francisco
116. Brogen C, Klingspor JS (1997) Impact of limestone 128. Vatsky J, Sweeney TW (1991) Development of an
grind size on WFGD performance, August. EPRI- ultra-low NOx pulverizer coal burner. Presented at
DOE-EPA combined utility air pollutant control sym- the 1991 EPRI/EPA joint symposium on stationary
posium. The mega symposium. TR-108683-V2. combustion NOx control, 25–28 Mar 1991. EPRI/
Electric Power Research Institute, Palo Alto EPA, Washington, DC
117. Blythe G, Horton B, Rhudy R (1999) EPRI FGD 129. U.S. Department of Energy (1996) Reducing emis-
operating and maintenance cost survey. EPRI-DOE- sions of nitrogen oxides via low-NOx burner tech-
EPA combined utility air pollution control sympo- nologies. Topical report number 5. U.S. Department
sium. TR-113187-V1, Electric Power Research Insti- of Energy, Pittsburgh
tute, Palo Alto 130. Campbell LM, Stone DK, Shareef GS (1991) Source-
118. Masters K (1985) Spray drying handbook, 4th edn. book: NOx control technology data. U.S.
Wiley, New York Environmental Protection Agency, Research Trian-
119. Burnett TA, Anderson KD (1981) Technical review gle Park
of dry FGD systems and economic evaluation of 131. Wendt JOL, Sternling CV, Matovich MA
spray dryer FGD systems. EPA-600/7-81-014. U.S. (1973) Reduction of sulfur trioxide and nitrogen
Environmental Protection Agency, Office of oxides by secondary fuel injection. In: Proceedings
Research and Development, Washington, DC of the fourteenth symposium (international) on com-
120. Lachapelle DG, Brown JS, Stern RD (1974) Over- bustion. The Combustion Institute, Pittsburgh,
view of environmental protection agency’s NOx con- pp 897–904
trol technology for stationary combustion sources. 132. U.S. Environmental Protection Agency (1996) Con-
Presented at the 67th annual meeting American Insti- trol of NOx emissions by reburning. EPA/625/R-96/
tute of Chemical Engineers 001. U.S. Environmental Protection Agency,
121. Srivastava RK, Hall RE, Khan S, Culligan K, Lani Cincinnati
BW (2005) Nitrogen oxides emission control options 133. Chen SL, McCarthy JM, Clark WC, Heap MP,
for coal-fired electric utility boilers. J Air Waste Seeker WR, Pershing DW (1986) Bench and pilot
Manage Assoc 55:1367–1388. Air and Waste Man- scale process evaluation of reburning for in-furnace
agement Association, Pittsburgh NOx reduction. In: Proceedings of the twenty-first
122. Hjalmarsson A-K (1990) NOx control technologies symposium (international) on combustion. The Com-
for coal combustion. IEACR/24. IEA Coal Research, bustion Institute, Pittsburgh, pp 1159–1169
London 134. Payne R, Moyeda DK (1994) Scale up and modeling
123. U.S. Environmental Protection Agency (1992) Eval- of gas reburning. In: Moussa A, Presser C, Rini MJ,
uation and costing of NOx controls for existing utility Weber R, Woodward G, Gupta AK (eds) Combustion
boilers in the NESCAUM region. EPA-453/R-92- modeling, scaling and air toxins. ASME FACT,
010. Office of Air Quality Planning and Standards, vol 18. American Society of Mechanical Engineers,
Research Triangle Park New York, pp 115–122
124. U.S. Environmental Protection Agency (1994) Alter- 135. Lyon RK (1987) Thermal DeNOx, controlling nitro-
native control technologies document: NOx emis- gen oxides emissions by a noncatalytic process.
sions from utility boilers. EPA-453/R-94-023. Environ Sci Technol 21(3):231–236. American
Office of Air Quality Planning and Standards, Chemical Society, Washington, DC
Research Triangle Park 136. Jødal M, Nielsen C, Hulgaard T, Dam-Johansen
125. Heap MP, Folsom BA (1990) Optimization of K (1990) Pilotscale experiments with ammonia and
burner/combustion chamber design to minimize urea as reductants in selective non-catalytic reduction
NOx formation during pulverized coal combustion, of nitric oxide. In: Proceedings of the twenty-third
Chapter 10. In: Pulverized coal combustion: pollut- symposium (international) on combustion. The Com-
ant formation and control, 1970–1980. U.S. bustion Institute, Pittsburgh, pp 237–243
Environmental Protection Agency, Air and Energy 137. Muzio L, Quartucy G (1993) State-of-the-art assess-
Engineering Research Laboratory, Research Triangle ment of SNCR technology. Topical Report
Park No. TR-102414. Electric Power Research Institute,
126. Sommer TM, Jensen AD, Melick TA, Orban PC, Palto Alto
Christensen MS (1993) Applying European low 138. Berg M, Bering H, Payne R (1993) NOx reduction by
NOx burner technology to U.S. installations. Pre- urea injection in a coal fired utility boiler. Presented
sented at the 1993 EPRI/EPA joint symposium on at the 1993 EPRI/EPA joint symposium on stationary
stationary combustion NOx control, 24–27 May combustion NOx control, 24–27 May 1993. EPRI/
1993. EPRI/EPA, Miami Beach EPA, Miami Beach
127. LaRue AD (1989) The XCL burner – latest develop- 139. U.S. Department of Energy (2005) Selective catalytic
ments and operating experience. Presented at the reduction (SCR) technology for the control of
466 Pulverized Coal-Fired Boilers and Pollution Control

nitrogen oxide emissions from coal-fired boilers. Clean Air Companies ICAC forum’98, Durham,
Topical Report Number 23. U.S. Department of 18–29 Mar 1998
Energy, Pittsburgh 143. Khan S, Shroff G, Tarpara J, Srivastava R (1997)
140. Pereira CJ, Amiridis MD (1995) NOx control from SCR applications: addressing coal characteristics
stationary sources: overview of regulations, technol- concerns. EPRI-DOE-EPA combined utility air pol-
ogy, and research frontiers, Chapter 1. In: Umit S, lutant control symposium – the mega sumposium.
Agarwal SK, Marcelin Ozkan G (eds) Reduction of Technical report TR-108683-V1. Electric Power
nitrogen oxide emissions. American Chemical Soci- Research Institute, Palo Alto
ety, Washington, DC, pp 1–13 144. Foerter D, Jozewicz W (2001) Cost of selective cat-
141. Nischt W, Woolridge B, Bigalbal J (1999) Recent alytic reduction (SCR) application for nox control on
SCR retrofit experience in coal-fired boilers. Pre- coal-fired boilers. EPA/600/R-01/087. U.S. Environ-
sented at POWER-GEN international, New Orleans, mental Protection Agency, Office of Research and
30 Nov–2 Dec 1999 Development, Washington, DC
142. Tonn DP, Uysal TA (1998) 220 MW SCR installation
on new coal-fired project. Presented at the Institute of
 processing, making it an economical contributor
Mitigation of Airborne to the overall US energy portfolio. Inherent in coal
Pollutants in Coal combustion is the production of various airborne
Combustion: Use of trace pollutants including particulate matter. At
Simulation uncontrolled levels, these pollutants can create
environmental and health concerns. Pollution
Bradley R. Adams control technologies have been developed, com-
Department of Mechanical Engineering, Brigham mercialized, and deployed across the power gen-
Young University, Provo, UT, USA eration industry over the past 40 years resulting in
significant reductions in airborne emissions from
coal power plants. However, competitive markets
Article Outline and ever-tightening pollutant regulations require
increasingly sophisticated solutions to further
Glossary reduce emissions.
Definition of the Subject Because of the complex chemical processes
Introduction involved in coal combustion and pollutant forma-
Simulation Methodologies tion and the difficulties in making detailed mea-
Modeling Pollutant Formation and Control surements in coal boilers, it is beneficial to utilize
Future Work computer modeling to better understand the com-
Bibliography bustion process and to assess the performance and
impacts of pollutant control strategies. Develop-
Glossary ments over the past 30 years in combustion diag-
nostics, computational resources, and software
ACI Activated carbon injection development have led to the development and
APCD Air pollution control devices application of computational fluid dynamics
APH Air preheater (CFD) tools which can simulate combustion sys-
CFD Computational fluid dynamics tems. These tools have been used to gain insights
ESP Electrostatic precipitator into combustion behavior, evaluate new burner
FGD Flue gas desulfurization and furnace designs, optimize operating condi-
LNB Low-NOx burner tions, guide experiments, and identify pollutant
OFA Overfire air control impacts. As not all CFD software is capa-
PM Particulate matter ble of representing all combustion processes accu-
SCR Selective catalytic reduction rately, successful application of CFD modeling
SDA Spray dryer absorber requires an understanding of which tools are best
SNCR Selective non-catalytic reduction suited to specific pollutant control applications.
This chapter describes the development and
application of CFD models to simulate pollutant
Definition of the Subject formation and pollutant control technologies
associated with coal combustion. Focus will be
Combustion of coal is a significant source of on US EPA criteria air pollutants, namely, partic-
electrical power production in the United States, ulate matter, nitrogen oxides (NOx), carbon mon-
providing approximately one-third of US power oxide (CO), and sulfur dioxide (SO2), as well as
generation [1]. Coal provides a secure and sus- sulfur trioxide (SO3) and mercury (Hg). Technol-
tainable fuel source and utilizes existing infra- ogies related to carbon dioxide (CO2) emissions
structure for mining, transportation, and control are not covered here.
© Springer Science+Business Media, LLC, part of Springer Nature 2020 467
R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_959
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_959-1
468 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

Introduction past 40 years to significantly reduce airborne

emissions from coal power plants. Over the period
Coal Power and Pollutant Emissions 1990–2015, national emissions from all sources
In 2015, US power generation was approximately have decreased as follows [3]:
4 trillion kilowatt-hours. About two-thirds of this
generated electricity came from the combustion of • Carbon monoxide (CO), 62%
fossil fuels, with approximately one-third from • Nitrogen oxides (NOx), 51%
combustion of coal [1]. Coal is expected to be a • Direct particulate matter, 2.5 microns (PM2.5),
significant contributor to the future US power gen- 25%
eration portfolio as it provides a secure and econom- • Direct particulate matter, 10 microns (PM10),
ical source of fuel and utilizes existing infrastructure 19%
for mining, transportation, and power production. • Sulfur dioxide (SO2), 79%
Inherent in coal combustion is the production
of various airborne trace pollutants including Many of these reductions are attributable to
NOx, CO, SO2, SO3, Hg, and particulates. At decreased emissions from coal power plants. Spe-
uncontrolled levels, these pollutants can create cifically, national SO2 and NOx emissions from
environmental and health concerns. Emissions of coal power plants decreased by 86% and 79%,
NOx and SO2 can also contribute to other types of respectively, from 1990 to 2015 [4]. These signif-
air pollutants such as PM10 and PM2.5 and can icant reductions represent development and
impact ozone levels. These concerns have resulted deployment of advanced and innovative air pollut-
in the promulgation of a series of regional and ant control technologies. However, ever-tightening
national environmental regulations designed to pollutant regulations require increasingly sophisti-
improve air quality. The major US EPA regula- cated solutions to further reduce emissions, while
tions promulgated over the past 25 years to control market forces require increased efficiency and eco-
air pollutant emissions from coal power plants nomics from control technologies.
include: One approach used successfully to aid the devel-
opment and installed performance of new pollutant
• National Ambient Air Quality Standards control technologies is combustion simulation.
(NAAQS), which regulate emissions and ambi- Combustion models help guide technology devel-
ent concentrations of carbon monoxide (CO), opment and selection by providing insights into
lead, nitrogen dioxide (NO2), ozone (O3), fine performance and potential operational impacts. As
particulate pollution (PM10 and PM2.5), and pollutant regulations continue to tighten, increas-
sulfur dioxide (SO2). The NAAQS does not ingly sophisticated simulation tools are required to
cover emissions of particulates greater than develop and evaluate new pollutant control tech-
10 micron diameter, i.e., PM10 [2]. nologies as they are applied to the wide range of
• Title IV – Acid Deposition Control (Title IV of coal power plant designs and firing conditions.
the 1990 Amendments to the Clean Air Act),
which regulates SO2 and NOx emissions. Pollutant Formation in Coal Boilers
• Cross-State Air Pollution Rule (CSAPR), which Coal combustion is a complex process that
regulates regional SO2 and NOx emissions (this involves fluid-solids mixing, chemical reactions,
replaced the Clean Air Interstate Rule). and heat transfer. The formation of pollutants
• Mercury and Air Toxics Standards (MATS), during and after the combustion process is also
which regulates mercury emissions. complex. These processes are difficult to measure
because of the scale of the combustion system,
In response to these regulations and other mar- limited access through boiler equipment, and the
ket forces, pollution control technologies have turbulent, high temperature, particle-laden envi-
been developed, commercialized, and deployed ronment. Because of the complex interdependent
across the power generation industry over the processes involved in coal combustion and
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 469

pollutant formation and the difficulties in making and the stack. The primary purpose of the radiant
detailed measurements in coal systems, it is ben- section is to combust the coal and transfer the
eficial to utilize computer modeling to better resulting energy from the flame to the boiler water
understand the combustion process and to assess tubes. The combustion process liberates moisture,
the impacts of pollutant control strategies. carbon, hydrogen, oxygen, nitrogen, sulfur, and
A coal power plant is a highly integrated system mercury bound in the coal. Minerals remaining
that can be divided into three main functions: after combustion are called ash. Depending on
(1) fuel storage, preparation, and handling; (2) fuel coal properties and local temperatures and stoichi-
combustion, steam production, and flue gas ometry, varying amounts of CO, NOx, SO2, SO3,
cleanup; and (3) electricity generation via steam and mercury species will be formed. Controlling the
turbine and subsequent electricity distribution. The combustion process can mitigate formation of these
focus of this chapter is on fuel combustion and flue pollutants but can also impact boiler operation. For
gas cleanup, i.e., the processes which contribute to example, utilization of air staging can reduce NOx
coal combustion and pollutant formation and con- formation but may also increase deposition of
trol. Although coal power plants share similar pro- unburned carbon on boiler surfaces and increase
cesses, there are a wide variety of designs and firing CO concentrations in the boiler back pass.
conditions and fuels. These differences can impact The primary purpose of the convective section
the effectiveness of different control technologies; is to transfer heat from the combustion flue gas to
hence, power generators cannot select a one-size- the superheat and reheat steam tubes. The convec-
fits-all control technology for their plants. Air pol- tive section can also play a secondary role in
lution control technologies must be evaluated on a pollutant formation and control as CO and SO2
plant-by-plant basis for optimal performance. can be oxidized to CO2 and SO3, respectively, in
Figure 1 shows a schematic of a coal-fired boiler. this region, and NOx control technologies can be
For the purposes of this discussion, the boiler can be deployed here. This section can also experience
divided into a radiant heat transfer section, a con- some of the operational impacts of pollutant con-
vective heat transfer section (also called a back trol technologies such as increased ash deposition
pass), and a flue gas treatment section, which and increased flue gas temperatures due to
include everything between the convective section delayed combustion in the radiant section.

Convective Stack
Section /
Back Pass
Radiant
Furnace Wet
Scrubber
SCR ESP or
Baghouse
Combustion
Zone

Air Pre-
Heater

Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation, Fig. 1 Schematic of coal-fired boiler
470 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

Region of formation and control for each of the [5]. However, the modeling and control of CO2
focus pollutants is summarized as follows (see emissions is notably different than other air pollut-
Fig. 1): ants based on the quantity and nature of the emis-
Ash particles – small particles comprised of sions. Thus, technologies related to carbon dioxide
minerals in the coal that remain after combustion (CO2) emissions control are not covered here.
in the radiant furnace. Ash particles deposit on Subsequent sections in this chapter provide an
heating surfaces, drop out of the gas flow into overview of CFD modeling and describe how
hoppers, or are collected in a particulate collection modeling is used to understand pollutant formation
device such as an ESP or fabric filter baghouse. and to evaluate pollutant control technologies.
Sulfur dioxide (SO2) – formed as a product of
combustion in the radiant furnace. SO2 is captured
in a flue gas desulfurization unit (FGD), also called Simulation Methodologies
a wet scrubber or spray dryer absorber (SDA).
Sulfur trioxide (SO3) – formed in the radiant There are a variety of simulation tools that are used
furnace and convective back pass and in the selec- to predict pollutant formation and control. These
tive catalytic reduction (SCR) unit. SO3 is reduced range from simple correlations to complex compu-
in the air preheater, ESP and FGD. It can also be tational fluid dynamics (CFD) software. Where
controlled using sorbents injected into the flue gas practical, CFD software is preferable as it can incor-
ahead of the ESP or baghouse. porate many of the simpler models and provides the
Nitrogen oxides (NOx) – formed in the radiant most comprehensive simulation of pollutant forma-
furnace from a reaction of molecular or coal nitro- tion and control processes. The following sections
gen with oxygen at sufficiently high temperature. provide an overview of the key methodologies used
NOx can be reduced in the radiant furnace com- to model the chemistry and physics relevant to
bustion zone through the use of staged combus- pollutant formation and control.
tion, in the convective section with selective non-
catalytic reduction (SNCR), or after the back pass Correlations
with selective catalytic reduction (SCR). Correlations are equations developed from
Carbon monoxide (CO) – formed in the radiant observed or measured relations between vari-
furnace as a product of incomplete combustion. It ables. Correlations are often used when the
can be reduced in the upper furnace or early con- detailed relations between variables are not well
vective section by mixing with excess oxygen, understood or are too complex to represent with
while gas temperatures are sufficiently high. full physics or chemistry. An example of a corre-
Mercury (Hg) – released from coal during lation is the relation shown below for the change
combustion and forms gas-phase mercury (in the in CO concentration with time.
elemental or oxidized state) and a small amount of

particulate mercury. After the combustion zone, d½CO Ea
¼ A½COa H2 Ob ½O2 c exp :
elemental mercury is further oxidized across the dt RT
SCR and air preheater. Oxidized mercury can be
captured in air pollution control devices such as This correlation shows that the rate of change in
the ESP, baghouse, or FGD. Active control of gas- CO concentration is a function of CO, H2O, and
phase mercury is achieved through reaction with O2 mole fractions, gas temperature, and theoreti-
injected activated carbon particles which are sub- cally or experimentally determined constants. The
sequently collected in a baghouse. constants for this particular correlation were opti-
It should be noted that coal power plants, like mized over a specified temperature range by com-
other combustion processes that rely on fossil fuels, parison with a detailed chemical mechanism
produce significant amounts of carbon dioxide consisting of 277 reversible reactions and
(CO2) as a product of combustion. CO2 emissions 66 chemical species [6].
are considered to be a contributor to greenhouse Correlations are most commonly developed
gases and hence can be considered a pollutant based on theoretical relations (as illustrated
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 471

above) or experimental measurements. Correla- Code [8], STANJAN [9], Gaseq, HSC Chemistry,
tions are particularly useful when relating mea- and REACT. Many of these software packages are
surements from large combustion systems where updated frequently for new capabilities or com-
detailed measurements are difficult to make. puter systems, so the latest versions should be
Examples include sulfur dioxide removal or sulfur accessed online. Some are freeware whereas
trioxide formation across air pollution control others are offered as commercial products. All
devices at power plants. are capable of calculating reactions for gas-phase
Correlations are valuable because they use homogeneous coal combustion chemistry. Ver-
simple equations to relate the behavior of different sions of the NASA or STANJAN codes are fre-
variables and can be implemented into various quently included in CFD-based combustion
models ranging from CFD software to process software.
models to spreadsheets.
Chemical Kinetics
Chemistry Models Chemical kinetics refers to the calculation of chem-
Pollutant formation and reduction are inherently ical reactions with consideration of reaction rates.
chemistry-controlled processes. Chemistry models Such reactions are time dependent. Chemical kinet-
refer to approaches that calculate chemical reac- ics models require specification of individual reac-
tions for a given set of reactants and temperature, tion equations, each with an associated reaction
pressure, and time conditions. These approaches rate. The rate of combustion is controlled by chem-
employ algorithms for solving multiple interrelated ical reaction rates; however, for the main combus-
equations that describe the chemical reaction pro- tion reactions involving oxidation of coal volatiles
cess between different species. These models focus (carbon, hydrogen, and sulfur), these rates are so
only on the chemistry reactions and do not consider fast that they can be approximated with equilibrium
dimensionality or mixing of species due to fluid chemistry. Conversely, the formation and destruc-
flow behavior. These models may be used for tion of almost all trace pollutants are determined by
stand-alone calculations, simplified and presented chemical reaction rates [7]. Flame ignition and
as correlations, combined in networks of idealized extinction are also dependent on reaction rates. In
chemical reactors or included in computational coal combustion, the heterogeneous reactions of
fluid dynamics (CFD) models which include turbu- the coal particles (devolatilization of gases and
lent mixing and heat transfer effects. oxidation of coal char) are also kinetically con-
trolled [10]. Similarly, heterogeneous reactions of
Chemical Equilibrium sorbents with acid gases or mercury are kinetically
Chemical equilibrium is the state in which both controlled and require kinetic mechanisms and
reactants and products in a chemical reaction are rates to calculate extent of reactions.
at concentrations that do not change with time. Gas-phase kinetic mechanisms can include
Thermodynamics determines the extent to which hundreds to thousands of reaction rates and tens
the reactions occur. In combustion systems, chem- of species. Calculation of these mechanisms is
ical equilibrium assumes fuel and oxidizer react accomplished through the use of chemical kinetic
instantly to form equilibrium products as soon as software such as Chemkin [11] and Cantera
mixing occurs. This is a reasonable assumption [12]. These stand-alone codes calculate the
for the main combustion reactions of carbon, resulting reaction products for a specified set of
hydrogen, and sulfur oxidation. Chemical equilib- reactants, temperature, pressure, and residence
rium is commonly calculated using a Gibbs free time. A detailed set of chemical reaction relations
energy minimization approach [7]. There are (the chemical kinetic mechanism) must be also be
numerous software tools that may be used to specified or provided to the kinetic solver. Kinetic
calculate chemical equilibrium for combustion solvers are sometimes grouped together in net-
systems, including the NASA Chemical Equilib- works of idealized chemical reactors to represent
rium with Applications (CEA) code, which is a combustion process where different stoichiom-
based on the earlier NASA Chemical Equilibrium etries and temperatures are present [13].
472 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

Solution of detailed chemical kinetic mecha- combustion chamber are computed to determine
nisms is computationally intensive, so these cal- the convective and diffusive mixing of energy and
culations are not usually included in CFD solvers. relevant chemical species. The flow and chemical
Rather, simplified kinetic mechanisms can be reactions are influenced by the radiative heat
developed and included in CFD calculations at transfer that is generated by the combustion pro-
reasonable computational cost. One common cess. The radiative heat transfer influences the
approach for simplifying kinetic mechanisms is local gas temperature and thus the local fluid
the use of reduced mechanisms [14]. Reduced dynamics. Coupling between flow, reaction, and
mechanisms are comprised of a reduced number radiation mechanisms is accounted for in order to
of kinetic reactions (e.g., 10–20 reactions) identi- accurately simulate interaction between mecha-
fied from within a detailed mechanism as being nisms. In coal combustion, calculations must
the most relevant for the combustion conditions also include governing relations for turbulent par-
and reactions of interest. ticle transport, heterogeneous particle reactions,
Kinetic models for coal devolatilization and coupling of mass, species, momentum, and
include detailed information about coal structure energy transport between particle and gas phases.
and often require information from additional Further, calculation of formation and destruction
chemistry calculations [15, 16]. These complex of trace pollutant species such as NOx and CO
models are typically replaced with more compu- requires finite-rate chemistry.
tationally tractable one- or two-step Accurately representing these interdependent
devolatilization models for use in CFD combus- processes requires refined descriptions of the
tion models [10, 17, 18]. Similarly, kinetic models governing physics and chemistry and contempo-
for char oxidation include detailed information rary computational methods. Different combus-
about char structure [19] and are approximated tion applications require more or less
with simpler one-step oxidation models for use sophistication to represent the dominant processes
in CFD combustion models [10, 20]. for that application. Accordingly, different CFD
tools provide different levels of sophistication,
Computational Fluid Dynamics (CFD) based on the turbulence, chemistry, and heat trans-
Computational fluid dynamics (CFD) represents a fer submodels used [24, 25]. It is important that
general class of computer software that allows the the CFD practitioner become educated as to which
mathematical equations describing mass, momen- tool is best suited for their application.
tum, species, and energy transport to be approxi- Development of CFD for combustion applica-
mated with computers. Development and solution tions begun in the late 1960s, with significant
of these transport equations are summarized in work being done at Imperial College in Great
numerous computational fluid dynamics text- Britain. Early work focused on parabolic flows
books [21–23]. CFD simulations predict gas and for gas-phase combustion in laminar and turbulent
particle flows, heat transfer, and chemical reac- flame jets [26–28]. In the early 1970s, the stream-
tions in combustion systems. The goal of CFD function vorticity model was developed which
modeling is to represent real combustion systems allowed simulation of recirculating flows in two-
as accurately as feasible, based on computational dimensional combustion systems [29, 30]. These
resource and schedule limitations. This is accom- software programs also included early versions of
plished using submodels that accurately represent turbulence models based on kinetic energy. The
the turbulent mixing, heat transfer, and chemical stream-function vorticity model ultimately proved
reactions relevant to the combustion and pollutant unsuitable for three-dimensional flows, and in the
processes of interest. early to mid-1970s, alternate techniques such as
the SIMPLE algorithm allowed fluid flow and
Coal Combustion Modeling combustion calculations to be extended to three
In CFD combustion models, the fluid dynamical dimensions [31–33]. By the mid-1970s, the SIM-
properties of the gases flowing through a PLE algorithm for coupling velocity and pressure
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 473

had gained wide acceptance and was used in soft- predictions [64–69], and ash deposition
ware that included temperature- and pressure- [70–72]. Since that time, additional advances
dependent density and viscosity, the k-e turbu- have been made that enable predictions of many
lence model, the six-flux radiation model, and a pollutant formation processes and control
kinetically controlled global chemical reaction technologies.
model [34]. Also during the mid-1970s, research Due to the complex nature of the physics and
highlighted the importance of fluctuations and chemistry in the combustion process, current
chemical kinetics in combustion modeling [35], computational tools (computers and software)
and techniques such as the eddy breakup and pre- are still not able to solve exactly all of the equa-
sumed pdf methods were developed to account for tions used to describe all combustion and pollut-
the effects of turbulent fluctuations on species con- ant formation processes. In particular, current
centrations and combustion [36, 37]. combustion models utilize approximations to rep-
In the late 1970s and early 1980s, CFD devel- resent turbulent fluctuation impacts on chemical
opment focused on improved mesh generation reactions and radiative heat transfer and to repre-
techniques, improved submodels for radiative sent heterogeneous particle reactions. There are
heat transfer [38, 39], and turbulence impacts on current research efforts to investigate improved
chemical reaction rates [40–42]. CFD combustion methods of solving the governing equations for
applications also began to expand from simple jet turbulent combustion, including direct numerical
flows and gas turbines to gas and diesel engines, simulation (DNS) and large eddy simulations
power and industrial furnaces, fires, and explo- (LES) [73–76]. While these techniques hold
sions. Each of these applications utilized the promise for improved accuracy and have been
improving capabilities of the reacting flow models demonstrated in academic settings, the computa-
but was also limited by the lack of specialized tional cost is currently prohibitive for use with the
submodels specific to their physics and chemistry. geometrical complexities and multiple reaction
The early 1980s also brought about the first processes present in full-scale coal combustion
releases of commercial CFD software, specifically systems.
CHAM’s PHOENICS code released in 1981 and
Creare’s FLUENT code released in 1983. Usage Guidelines
Coal combustion modeling that begun in the Computational fluid dynamics is one of several
late 1970s and early 1980s was able to build on ways to model or approximate the physical behav-
the turbulence and chemical reaction techniques ior occurring in full-scale coal combustion sys-
of the early CFD researchers to add submodels for tems. Other modeling methods include pilot-scale
two-phase flows, including particle dispersion, test furnaces, scaled cold-flow physical mock-
reaction, and heat transfer. A number of coal com- ups, simple correlations, networks of idealized
bustion models were developed in the 1980s for reactors, and sophisticated artificial intelligence
axisymmetric and three-dimensional applications systems. Compared to these other methods, CFD
[43–52]. In particular, the texts by Smoot and modeling provides an excellent balance of accu-
Pratt [53] and Smoot and Smith [10] provide an rately representing the combustion system (full-
excellent summary of early submodels developed scale geometry, flow physics and chemistry) and
for coal combustion and reference contributions providing reasonable solution times at affordable
by several researchers in specific areas such as costs. CFD also has the flexibility to incorporate
coal particle devolatilization, particle reactions in correlations as simple or complex as needed to
turbulent flows, chemistry of coal reaction prod- represent specific processes such as heteroge-
ucts, and radiative heat transfer. Work on sub- neous surface chemistry reactions, surface deposit
model development continued through the 1990s behavior, thermal boundary conditions at steam
and 2000s, particularly on improved radiation tube surfaces, and tube surface corrosion.
modeling [39, 54–56], soot modeling [57–59], One of the key benefits in using combustion
turbulence-chemistry coupling [60–63], NOx simulations for coal boilers is the ability to assess
474 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

trade-offs between pollutant control, combustion codes may not have the refined GUI of commer-
efficiency (in terms of unburned carbon in ash), cial codes but have been specifically developed
and operational impacts (such as increased surface for one particular application or process, such as
deposition or corrosion). For example, a NOx coal combustion. These software programs incor-
control technology such as low-NOx burners can porate models for all important physics and chem-
be simulated to determine the amount of NOx istry occurring in the combustion system. CFD
reduction, the optimal design to achieve that users should be confident that the software
reduction, the impact on CO emissions and selected for modeling is appropriate for their
unburned carbon in ash, and operational impacts application. Where possible, software should be
such as increased radiant furnace slagging. The validated with experimental measurements or
relatively low cost (compared to full-scale testing) observations. Where not possible, software sub-
and completeness of process representation make models should be critically examined to ensure
CFD modeling an ideal candidate for “what-if” accurate representation of key processes.
studies. A number of different design and/or oper- User experience. In addition to having capable
ational changes can be assessed to determine opti- software, successful application of CFD requires
mal choices based on specific plants needs and qualified, experienced users or modelers. CFD
constraints. CFD models can be used to: users must understand how the software works
and what its capabilities and limitations are. Com-
• Better understand combustion system behavior puter resources today are capable of producing
• Evaluate impacts of proposed operational large amounts of data that can be displayed in
changes (e.g., fuel type, excess oxygen) dazzling color graphics. However, data quantity
• Evaluate the performance of new burners and attractive presentation do not guarantee accu-
• Evaluate the performance and impacts of racy. The modeler must be able to distinguish
potential pollutant control technologies accurate solutions based on software convergence
• Troubleshoot problematic performance areas (numerical accuracy) and comparison with obser-
and identify problem sources vations or experienced-based expectations. Even
• Guide boiler testing the best software can give erroneous results if
improperly used. In instances where modeling
Because pollutant formation is a complex pro- results are time critical, the CFD engineer may
cess, the CFD tools used to simulate combustion have to make trade-offs between simulation accu-
and pollutant formation are also complex. Suc- racy and completion time.
cessful application of CFD tools is dependent on Accurate problem description. Because CFD
several key factors as noted below. software is essentially a series of mathematical
CFD software selection. There is a variety of equations, the solution to these equations can
CFD software available, ranging from commer- only be as accurate as the information input to
cial all-purpose software to in-house application the equations (i.e., boundary conditions). CFD
specific software. The capabilities and ease of use software conforms to the maxim: “garbage in,
of these tools vary significantly and are continu- garbage out.” Accurate simulation of a combus-
ally evolving. Commercial software tends to focus tion process requires accurate specification of the
on relatively simple physics in complex geome- inputs (e.g., furnace and burner geometry, fuel and
tries and has a refined graphical user interface air properties, wall surface properties). An expe-
(GUI) making the software more intuitive to use rienced modeler will be able to, at least partially,
for end users in large industries such as aerospace screen the inputs for consistency and realism
and automotive. Thus these tools may not be as before beginning a simulation.
capable of modeling complex, specialized pro- Accurate calculations. The accuracy of the
cesses because they emphasize ease of use and CFD calculation is based on three major factors:
solution speed over more complex physics and (1) the accuracy of the information describing the
chemistry. Conversely, specialty combustion system that is entered into the program, (2) the
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 475

ability of the software to represent the physics and 100 mm [77, 78]. They may be deposited on boiler
chemistry governing the process, and (3) the accu- surfaces or carried out of the boiler with the flue
racy of the numerical methods used to approxi- gas. The size of these particulates is important as
mate the mathematical equations describing particles with aerodynamic diameters less than
combustion. 2.5 mm (PM2.5) may be a cause of increased
Accurate Results Evaluation. CFD results are health risks due to inhalation into human and
of little use if they cannot be understood and animal respiratory systems [79, 80]. Further, the
presented in a context relevant to the furnace collection efficiency of many particle collection
operator or engineer. The data resulting from devices such as ESPs decreases with decreasing
today’s simulations are so vast and complex that particle size. For example, one study has shown
they are best understood graphically. The manip- that the PM collection efficiency of an ESP typi-
ulation and presentation of this data is covered by cally decreases from 98% for PM10 (particulate
a related discipline – data visualization – from matter with an aerodynamic diameter less than
which CFD borrows graphical tools. The CFD 10 mm) to 96% for PM2.5 [81].
modeler uses graphical tools to understand the A second type of PM is fine particulates that
behavior of the combustion system and to sum- form from chemical processes or condensation
marize performance and operational impacts. The after combustion, e.g., sulfur trioxide and sulfuric
user then selects appropriate displays and summa- acid mist. These particulates are usually submi-
ries to describe to a broader audience what is cron in size and are not readily captured by the
happening in the furnace. usual ESP or fabric filter technologies [82]. The
One approach used to minimize mistakes lead- discussion here will focus on the first type of
ing to inaccurate solutions is to have the CFD particulates.
engineer work closely with an experienced plant Coal ash particulates are primarily comprised
or process engineer. This allows the CFD inputs of silicon oxide (SiO2), aluminum oxide (Al2O3),
and results to be reviewed for realism and accu- calcium oxide (CaO), and iron oxide (Fe2O3) but
racy from two different perspectives. The plant also contain additional minerals such as magne-
engineer can ensure that the process is being accu- sium, sodium, and potassium. Particulates vary in
rately represented by comparison to field observa- size and composition based on the original coal
tions and measurements. The CFD engineer can properties. Ash size determines whether the ash
ensure the inputs and results are numerically accu- particles drop into hoppers (bottom ash), are cap-
rate and consistent with previous experience. Note tured on heated surfaces (deposits), or are carried
that CFD modeling is not designed to make deci- along with the flue gas (fly ash). The composition
sions for engineers. Rather, the primary objective of ash determines the materials propensity to stick
of CFD modeling is to provide engineers and to boiler surfaces, processes known as slagging
operators with additional information to help (in the boiler radiant section) and fouling (in the
them make more informed decisions about plant boiler back pass). The majority of slagging and
technologies and processes. fouling comes from fly ash.
Particulate matter control is based on capturing
particulates using gravitational, centrifugal, iner-
Modeling Pollutant Formation and tial, interception, diffusional, or electrostatic
Control forces. Since larger ash particles will naturally
drop out of the flue gas flow due to gravity, the
Particulate Matter (PM) primary focus is capturing fly ash. The most com-
Particulate matter (PM) may be of two general mon devices for capturing fly ash in coal-fired
types. The most common type is comprised of boilers are electrostatic precipitators (ESPs) and
minerals from the original coal particles that fabric filters. ESPs use an electrical field to charge
remain after the coal is burned, i.e., ash. These and collect the ash particles on ESP surfaces. The
particles range in size from submicron to roughly particles are then mechanically removed from the
476 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

charged surfaces and dropped into hoppers for sub- by these elements condense on ash particles and
sequent removal. Fabric filters collect particulates heating surfaces, forming a sticky substance
on the filter as flue gas flows through the filters. which initiates deposition. The deposit grows as
Capture particulates are removed from fabric filters additional sticky particles impact the surface.
by mechanically shaking the bags, by reversing the Fouling processes can be modeled based on
flow through the filter, or by using pulsed air to impaction of ash particles on heating surfaces in
expand the bag such that particles fall off the fabric the boiler back pass using techniques similar to
into hoppers or are re-entrained into the flow. slagging deposition models. However, the predic-
tion of the volatile inorganic compounds and sub-
Slagging sequent condensation on ash particles are
Modeling of fly ash capture is done based on complex and to date have not been suitably
equations that describe the behavior of the collec- defined or included in CFD models. Rather, par-
tion process. In the radiant section of the boiler, ticle “stickiness” may be estimated based on coal
ash particle composition is used to determine the ash composition and the expected volatile inor-
particles “sticking” temperature. This is usually ganic concentrations in the flue gas. This
based on the T250 temperature of the particle, a approach is more qualitative than quantitative,
temperature at which the ash initially softens and and resulting correlations are based on measure-
becomes sticky. The T250 value may be deter- ments of specific coals. Although predictive
mined experimentally from ASTM methods [83, models for slagging and fouling are now widely
84] or calculated from correlations [85–88]. available, they are not capable of predicting ash
A T250 correlation or “sticking” temperature is impacts with sufficient reliability in all cases. The
then used in CFD codes along with calculations of prediction of ash impacts is more reliably
the particle temperature to determine if a particle achieved through a combination of predictive
will stick to surfaces, either through impaction or indices, models, historical information on coal,
attraction by van der Waals forces or Brownian and boiler performance and experience.
diffusion. Particle temperature is calculated based
on net radiative and convective heat transfer to the ESP
particle (assuming all moisture, volatiles, and fixed Fly ash particles that are not captured by sticking
carbon have been removed from the particle). to heating surfaces or dropped out of the flue gas
When an ash particle interacts with a surface and flow are captured directly from the gas stream
the particle temperature is greater than or equal to using either electrostatic precipitators (ESPs) or
the “sticking” temperature, the particle is assumed fabric filters. ESPs remove fine particulates by
to stick on the surface. Advanced models are avail- inducing an electric charge on the particulates,
able which predict ash deposit growth and then attracting them to downstream charged col-
sintering, which can be used to refine the thermal lection plates. Fabric filters, or baghouses,
resistance to heat transfer (via the thermal conduc- removed fine particulates by capturing them on
tivity and thickness) at the surface [70]. CFD the fabric filter as the flue gas is forced through the
models can use advanced ash submodels to refine material.
heating surface thermal properties. The classic equation (correlation) for calculating
collection efficiency of ESPs is the Deutsch-
Fouling Anderson equation, given as
Similar to slagging, ash particles may stick to  
superheater surfaces in the convective pass of Aw
 ¼ 1  exp 
the boiler. This is termed fouling. Fouling is typ- Q
ically caused by the vaporization of volatile inor-
ganic elements in the coal during combustion. where:  = collection efficiency of the precipitator
When flue gas temperatures decrease in the con- A = effective collecting plate area of the
vective section of the boiler, compounds formed precipitator (m2)
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 477

w = particle migration velocity (m/s) • The governing equations and interdependent

Q = gas flow through the precipitator (m3/s) parameters in them (gas velocity and pressure,
This correlation can be used to estimate the collec- particle velocity, electric field intensity, ion
tion efficiency of a precipitator based on the overall space charge density, particle charge) are
design and operating characteristics of the ESP. To solved by iterative solution until all properties
improve accuracy, an experimentally determined within the coupled equations are converged.
migration velocity can be used in the relation.
CFD modeling of ESPs is typically done using Several variations of this approach have been
one of two approaches. In the first approach, the flue used for CFD modeling of ESPs [91–93].
gas flow patterns and velocities are modeled using
standard CFD turbulent flow models where density Fabric Filters
is assumed constant (i.e., flow is isothermal). These Fabric filters collect particulates as flue gas flows
models may or may not include a solid-phase rep- through the filters. Particulates can either directly
resentation of the inert particulates. Flow calcula- impact fibers in the filter or be drawn onto the filter
tions are used to identify regions where by van der Waals forces. Fabric filters are com-
aerodynamic forces from high-velocity flow may monly configured in a baghouse. Unlike ESPs
be large enough to overcome the electrostatic forces which have collection efficiencies in the 96–98%
from the collection plates or regions where high range for small particles, a properly designed and
velocities may re-entrain particulates stored in hop- well-run baghouse will generally have an
pers below the collection plates. Alternately, regions extremely high particulate matter collection effi-
of low velocity or recirculated flow can be identified ciency, i.e., greater than 99.9%.
where maldistribution of flow leads to poor capture Baghouses are particularly effective for
efficiency. This approach does not calculate actual collecting small particles [94]. For example, tests
electrostatic forces but rather relies on correlations of baghouses on two utility boilers [95, 96]
and experience to identify flue gas flow conditions showed efficiencies of 99.8% for particles 10 mm
where ESP performance may be compromised [89]. in diameter and 99.6–99.9% for particles 2.5 mm
In the second approach, calculation of the flow in diameter. Because high-efficiency capture can
properties and turbulent particle dispersion is be assumed, the design and modeling of bag-
combined with calculation of the electric field houses focus on the flow distribution and pressure
surrounding the flow. The impact of the electrical drop through the baghouse.
field on the particulates is represented as source Modeling of pressure drop through a baghouse
terms in the gas and particle flow calculations. can occur at two levels of detail. A general CFD
Here again, the flow is assumed to be isothermal model of the flue gas flow through the baghouse
with constant density, so that only the mass, requires the baghouse duct geometry be defined
momentum, and electric field equations need be and the fabric filters or bags be represented with a
solved. One algorithm for solving these coupled flow resistance. Single-phase gas flow is assumed
equations is as follows [90]: (i.e., no particles), and the velocity profile of flue
gas entering the baghouse is specified. This
• Turbulent fluid flow is calculated by the time- approach has the advantage of using established
averaged Navier-Stokes equations closed by flow models and of representing the combined
the k-e turbulence model. effects of the baghouse inlets, outlets, and multiple
• The distribution of electric field and ion charge bags in the model. The model focuses on flow
density are obtained from the numerical solu- distribution through the baghouse and calculates
tion of a Poisson equation and the current con- flow velocities and pressure drops. This informa-
tinuity equation. tion can be used to determine if particles are being
• The process of particle charging is solved by uniformly distributed on bag surfaces and to iden-
Euler method, and the particle trajectories are tify regions of maldistribution. The disadvantage
computed by Lagrangian tracking method. of this approach is that pressure drops through the
478 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

bag filters must be specified by the user (as a func- These types of correlations can be
tion of flow resistance) rather than predicted by the implemented into CFD codes to model full bag-
model. This model also assumes steady-state houses if it assumed that all bags have the same
behavior so any dynamic system behavior such as pressure drop and the bag specific loss coefficients
bag cleaning must be modeled as two separate flow are specified. However, since so much informa-
resistance conditions. tion must be specified for the pressure drop corre-
The second type of model is a correlation- lation, it is sometimes preferable to simply specify
based model that focuses only on the pressure an overall flow resistance for the bags as a func-
drop across a single fabric bag. Several correla- tion of local gas velocity to represent the pressure
tions have been proposed for such models drop across the bags, similar to the first approach
[97–99]. All of these models calculate pressure described here.
drop as a function of filtration velocity, clean
fabric flow resistance, dust deposit flow resis- Carbon Monoxide (CO)
tance, and bag cleaning efficiency. The following Carbon monoxide is a product of incomplete coal
correlation for pressure drop across a pulse-jet combustion, where there is insufficient oxygen to
baghouse [100] illustrates one such model: locally oxidize CO produced during coal
devolatilization and/or char oxidation to CO2. Util-
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1  2 K2 ity boilers are typically operated with 10–15%
DP ¼ Ps þ K 1 V f  Ps  K 1 V f  4W o þ K v V 2f
2 K3 excess air (i.e., 10–15% more air than is required
for stoichiometric combustion) to help minimize
where: the amount of CO leaving the combustion zone.
Ps = maximum static pressure achieved in the Coal burners that rapidly mix fuel and air produce
bag during cleaning efficient combustion and low CO concentrations
K1 = clean fabric resistance but can also produce high NOx concentrations.
Vf = face velocity Conversely, low-NOx burners or staged firing sys-
K2 = dust deposit flow resistance tems gradually mix fuel and air to minimize com-
K3 = bag cleaning efficiency coefficient bustion temperature and NOx formation. This
Wo = areal density of the freshly deposited dust gradual mixing can potentially lead to poorly
Kv = loss coefficient for the venturi at the inlet mixed regions in the radiant furnace where pockets
to the bag of high CO concentrations exist (even with excess
The disadvantage of this equation, like that of all oxygen in other regions). Boilers equipped with
the correlations, is that the constants K1, K2, and K3 staged firing and overfire air (OFA) systems often
must be determined from laboratory measurements. use distributed injection of high-velocity air to
The most difficult to determine is the K3 value, better mix the flue gases and reduce the potential
which can only be found by making measurements for high CO pockets reaching the boiler back pass.
in a pilot-scale pulse-jet baghouse. A limitation of CO concentrations of several thousand parts
laboratory measurements is that actual filtration con- per million (ppm) are common in the lower fur-
ditions cannot always be adequately simulated. For nace, but these levels are greatly reduced by the
example, a recycled or redispersed dust may not time the flue gases mix, cool, and exit the boiler
have the same size distribution as the originally stack. Typical CO concentrations at the boiler
deposited dust, thereby yielding different values of stack are less than 100 ppm. US EPA has desig-
K1, K2, and K3 than would be measured in an nated CO as a criteria pollutant and has set
operating baghouse. While the form of the pressure National Ambient Air Quality Standards
drop correlations is applicable to different baghouse (NAAQS) for CO of 9 parts per million (ppm)
designs, the experimentally determined constants in measured over 8 h, and 35 ppm measured over 1 h
the correlations are applicable only to the specific [101]. Since this regulation applies to ambient air
bag fabric, design, and cleaning method tested and conditions and does not target single emitting
should not be used with different configurations. sources, there are no specific CO emission limits
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 479

for coal boilers. However, for emission, safety, Mitigation of Airborne Pollutants in Coal Combus-
and combustion efficiency reasons, boiler opera- tion: Use of Simulation, Table 1 CO oxidation kinetics
parameters (from [6])
tors need to minimize CO emissions.
CO oxidizes rapidly at high temperatures in the Optimized constants
presence of oxygen but does not oxidize as well at Parameter T < 1150 K T > 1150 K
the cooler temperatures or less mixed conditions. A 2.61E + 12 6.52E + 6
Ea 45,566 15,968
At temperatures greater than ~1150 K, CO oxida-
a 1.0 1.0
tion is rapid and can be modeled with chemical
b 0.5 0.5
equilibrium approaches. In the range from 975 to
c 0.25 0.25
1150 K, CO oxidation is slower and should be
Units: m3, kg-mol, sec, K, Kcal
modeled with finite-rate kinetics. Below 975 K,
CO oxidation does not occur even with oxygen
present [6]. the universal gas constant, T is the gas temperature
From a modeling perspective, CO reactions in in Kelvin, and a, b, and c are constants (see
the combustion zone (above 1150 K) can be Table 1).
modeled with equilibrium chemistry calculations. The use of the global CO kinetics in a CFD
This can be done with any chemical equilibrium model necessitates the incorporation and solution
code such as STANJAN or NASA CEA. For this of an additional transport equation for the local
type of calculation, the gas reactants, pressure, CO concentration. The assumed rate of CO oxi-
and temperature are specified, and the chemical dation depends on the local gas temperature and
equilibrium products are calculated. Combustion concentrations of CO, H2O, and O2. The CFD
CFD codes also calculate equilibrium CO concen- model can then calculate the effect of turbulence
trations as part of the combustion calculations, on this rate and the impact of the local CO oxida-
typically using the same techniques utilized in tion on other local species and gas temperature.
the stand-alone equilibrium codes. If a CFD option that incorporates finite-rate CO
As temperatures drop in post-combustion reactions is not available or desirable, then a
zones or in low-temperature combustion systems, hybrid approach may be used where a CFD
CO calculations require special treatment as the model is used to predict the equilibrium gas con-
equilibrium chemistry assumption becomes less centration and temperature properties at the exit of
accurate. Specifically, finite-rate reactions for CO the radiant furnace, and then a plug flow reactor
oxidation must be used. Finite-rate reactions can model can use these values as inputs to represent
be calculated from global mechanisms [102] or gas flow through the back pass of the boiler. In this
from detailed chemical kinetic mechanisms case, a temperature gradient must be prescribed in
[103]. A global mechanism uses only one equa- the plug flow reactor to represent the heat loss in
tion to describe the CO oxidation rate, whereas a the boiler convective section. This approach
detailed mechanism may use hundreds. allows CO oxidation behavior to be evaluated
The rate constants in the following global mech- for different CO inlet concentrations and back
anism were optimized over the temperature range pass configurations [6].
800–1600 K [6] by comparison with a detailed Reactions from detailed mechanisms can be
chemical mechanism consisting of 277 reversible calculated using stand-alone chemical kinetics
steps and 66 chemical species [104]. programs such as Chemkin and Cantera. Typi-

 cally the plug flow reactor option in these tools
d½CO a b c Ea is used to assess the extent of CO oxidation over a
¼ A½CO H2 O ½O2  exp
dt RT specified time period for a given set of tempera-
ture, pressure, and species concentrations. Current
where [CO], [H2O], and [O2] are the CO, H2O, computational limitations prohibit the inclusion of
and O2 mole fractions, respectively, A is the pre- full detailed kinetic mechanisms in CFD codes;
exponential factor, Ea is the activation energy, R is however, techniques have been developed by
480 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

which reduced kinetic mechanisms can be be less than 5% of the total NOx formed
included in CFD codes. Because CO reactions [109]. These reactions occur at slower rates than
can be accurately represented with an optimized the combustion reactions and must be represented
global mechanism, the development and use of with finite rate chemical kinetics rather than equi-
reduced mechanisms is typically done only for the librium reactions. Figure 3 provides an illustration
more complex chemical reactions such as NOx of the NOx chemistry pathways.
formation in coal boilers. Figure 2 shows pre- The amount of thermal NOx formed is based on
dicted CO concentration profiles for a coal-fired the local gas stoichiometry and temperature and
boiler with standard firing and two cases of staged the residence time or volume of gases in the fur-
firing with OFA. CO concentrations are much nace at formation conditions. Thermal NOx is
higher in the staged cases below the OFA ports formed under fuel lean conditions (i.e., excess
due to the fuel-rich stoichiometries in this region. oxygen) and at gas temperature above about
The differences in upper furnace CO in the staged 1800 K. As temperatures increase beyond this,
firing cases is due to different OFA designs. the rate of NOx formation increases. The principle
reactions for thermal NOx come from the
Nitrogen Oxides (NOx) extended Zeldovich mechanism [110].
NOx is a general term for nitrogen oxides NO Fuel NOx is formed from two primary paths
(nitric oxide) and NO2 (nitrogen dioxide). NOx [107, 108]. The first path involves the oxidation of
is formed by three mechanisms during the com- volatile nitrogen species during the initial stages
bustion process. First, molecular nitrogen in the of combustion. During the release of volatile
air combines with oxygen to form NO [105]. This nitrogen and before oxidation, nitrogen reacts to
is referred to as “thermal NOx.” Second, hydro- form several intermediaries which are then oxi-
carbons in the flame front react with molecular dized to NOx. If the volatiles are released into a
nitrogen during the initial stages of combustion fuel-rich (reducing) atmosphere, the nitrogen can
[106]. This is referred to as “prompt NOx.” Third, react to form N2 rather than NOx. The second path
fixed nitrogen species in the fuel are oxidized. involves reaction of the nitrogen bound in the coal
This is referred to as “fuel NOx.” For pulverized char. This char nitrogen reaction occurs much
coal combustion, the majority of NOx emissions more slowly than reaction of the volatile nitrogen.
result from fuel nitrogen rather than thermal NOx Typically about 20% of the char nitrogen is
formation [107, 108]. Prompt NOx is estimated to converted to fuel NOx with the remainder reduced

Mitigation of Airborne
Pollutants in Coal
Combustion: Use of
Simulation, Fig. 2 CO
concentration profiles for
unstaged (left) and staged
(center and right) firing
conditions (Figure courtesy
of Reaction Engineering
International)
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 481

Mitigation of Airborne Prompt NO

Pollutants in Coal CHi
Combustion: Use of NO Fuel NO
CO CN
Simulation, Fuel N2 O2
Fig. 3 Illustration of NOx Reburning
NO NO N2
CHi HCN NO
chemistry pathways
O2
Nvolatiles NHi NO
Coal
Reagent
(SNCR) O2
Nchar char
NO
O2
N2
Thermal NO

to N2 by reactions with the char. The amount of local gas properties are first calculated and then
fuel NOx formed is dependent on the amount of thermal and fuel NOx calculations are done using
nitrogen in the coal (typically 0.5–2% by weight), chemical kinetic mechanisms. This requires spec-
the local gas stoichiometry and temperature when ifying the local combustion environment, i.e., the
the volatiles are released, and the amount of char local stoichiometry, gas species, and gas temper-
nitrogen converted to NOx (based on coal type ature, the amount of fuel nitrogen devolatilizing
and local combustion conditions). The amount into the gas phase, the amount of char nitrogen
of char nitrogen converted to NOx can vary from converted to NOx, and the residence time for
near 0% to over 50% depending on the coal type, reactions at these conditions. To better represent
local oxygen concentration, NOx concentration, the varied stoichiometry and temperature condi-
and gas temperature [100]. tions in a boiler, a network of chemical reactors
NOx can be reduced by reaction with hydro- can be used to “map” the combustion conditions
carbon radicals (a process called reburning), by throughout a boiler and calculate NOx reactions in
homogeneous reaction with ammonia or urea each zone. This requires the user to have a good
(selective non-catalytic reduction (SNCR)), and understanding of the combustion conditions in the
by heterogeneous reaction with ammonia or urea boiler either from experience or from simulations.
and a catalyst (selective catalytic reduction These properties can be provided by a CFD sim-
(SCR)). Reburning processes may also include ulation of combustion in the boiler.
the introduction of a secondary fuel (e.g., natural The second approach utilizes a CFD combus-
gas) or reagent (e.g., urea) into the reburning zone tion model that includes NOx reactions. Here,
to augment the NOx destruction reactions. The equilibrium combustion calculations provide stoi-
chemical mechanisms for these reactions are chiometry, species, temperature, and particle
contained within the detailed chemical kinetics properties throughout the boiler. Finite rate NOx
used for NOx formation. reactions are then calculated throughout the boiler
Because of the multiple, complex reaction using transport equations for relevant NOx species
mechanisms controlling NOx formation and the (i.e., NO, HCN, NHi). This more tightly inte-
dependence on local stoichiometry and tempera- grated approach gives improved NOx predictions
ture, modeling NOx formation and destruction is due to the increased spatial refinement of gas
challenging. Full representation of NOx reactions properties and NOx calculations in the boiler.
in coal combustion requires calculation of thermal The most commonly used detailed chemical
and fuel NOx reactions (assuming prompt NOx is kinetic mechanisms for gas-phase NOx reactions
negligible) as well as the local combustion envi- in hydrocarbon combustion are GRI-Mech ver-
ronment. This can be done via two approaches. sions 2.11 and 3.0 [111]. GRI-Mech 3.0 is a com-
The first involves a piecewise approach where pilation of 325 elementary chemical reactions and
482 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

associated rate coefficient expressions and ther- species concentrations. NOx calculations can
mochemical parameters for the 53 species then be performed with this information using
involved in the mechanism. The use of this mech- the mechanisms described previously.
anism directly is done via chemical kinetic pro- Staged firing with overfire air (OFA) similarly
grams such as Chemkin and Cantera. These reduces NOx formation by staging or withholding
calculations predict NOx formation and destruc- some of the combustion air from the burners. This
tion in combustion gases. results in incomplete combustion and a fuel-rich
To facilitate inclusion of the NOx reactions cal- environment which inhibits NOx formation. In addi-
culations in CFD codes, simplified versions of the tion, the fuel-rich environment also serves to reduce
NOx reactions have been developed. These include existing NOx via a reburning process. After heat is
global mechanisms [112] for very simplified calcu- removed in the staged combustion zone, the
lations and more refined reduced chemical kinetic remaining air is added to complete combustion
mechanisms for greater accuracy. The reduced (at lower temperatures). The addition of OFA com-
mechanisms are comprised of a reduced number of pletes char oxidation in the coal particles and CO
reactions taken from a full detailed chemical kinetic oxidation in the flue gas. Again, these processes can
mechanism [104, 113–115]. These reduced mecha- be simulated with the use of CFD models which
nisms can be implemented in CFD software to allow predict the extent of mixing, gas and particle com-
NOx predictions as part of the combustion simula- bustion reactions, and the resulting temperatures,
tion. In many cases, the NOx calculations are done stoichiometry, and species concentrations. NOx cal-
after the main combustion calculations are com- culations can then be performed with this informa-
pleted as the NOx reactions do not impact the heat tion using the mechanisms described previously.
release, temperature, oxygen, or stoichiometry pro- Reduction of NOx via reburning can occur in
files in the furnace. Note that in regions where any fuel-rich region of the radiant furnace
reburning or reagent-based NOx reductions occur, [116]. Such regions occur naturally with staged
the reduction mechanisms must be formulated to firing but can also be created through the intro-
account for these reactions as well as the NOx duction of a secondary fuel such as natural gas or
formation reactions. micronized coal [117–121]. Reagents can also be
introduced into reburning zones to augment com-
Combustion Zone bustion [122]. Simulation of reburning process
NOx emissions are controlled via two main requires similar combustion zone properties and
approaches. First, formation of NOx in the com- reaction mechanisms as described previously.
bustion zone is suppressed using low-NOx Reburning should occur where gas temperatures
burners (LNB) and staging with overfire air are still high enough to support rapid chemical
(OFA). Second, existing NOx is reduced using reactions, i.e., before the upper furnace. In regions
reburning, selective non-catalytic reduction where reburning or reagent-based NOx reductions
(SNCR), and selective catalytic reduction (SCR). occur, NOx calculations should include reduction
LNB, OFA, and reburning are considered com- mechanisms formulated to account for these reac-
bustion zone controls, whereas SNCR and SCR tions as well as the NOx formation reactions.
are considered post-combustion controls. Figure 4 shows predicted NOx concentrations
Low-NOx burners reduce NOx formation by in a coal boiler for unstaged and staged firing
gradually mixing the fuel with the air oxidizer. conditions. Also shown are the predicted NOx
This produces lower flame temperatures by destruction rates in the boiler under staged firing,
spreading reactions and heat release over a larger highlighting regions in the fuel-rich zone where
region and extracting heat as the combustion pro- NOx is destroyed via reburning. Reburning stops
cess (gradually) occurs. This process is simulated once overfire air is added.
with the use of CFD models which predict the A significant benefit of calculating NOx forma-
extent of mixing, gas and particle reactions, and tion and reduction as part of a comprehensive
the resulting temperatures, stoichiometry, and CFD simulation is the ability to identify
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 483

Gas
NOx Destroyed
Tempering
in Fuel-Rich
Lower Furnace
NOx ppm
1000

100 NOx
OFA Destruction Rate
FGR (kg/s/m3)

Fuel Rich 0.0005

Zone

Cyclones Cyclones
0
Baseline Staged Staged

Mitigation of Airborne Pollutants in Coal Combus- destruction rates at same location under staged firing con-
tion: Use of Simulation, Fig. 4 Predicted NOx concen- ditions (right) (Figure courtesy Reaction Engineering
trations at boiler centerline a coal-fired boiler without (left) International)
and with (center) staged firing/OFA. Predicted NOx

operational impacts from NOx control technolo- reduced mechanism may be developed to calculate
gies. For example, simulating boiler combustion NOx reduction [126, 127].
and NOx concentrations first with existing burners In addition to correctly modeling the SNCR
and then with replacement low-NOx burners can- reaction kinetics, the distribution of the reagent
not only identify expected reduction in NOx levels droplets must be correctly predicted to assess NOx
but can also calculate changes in unburned carbon reduction. Reagent droplets are sprayed from noz-
in ash, CO concentrations, heating surface zles along the furnace walls or from lances inside
deposits, and heat flux profiles. Similarly, simu- the furnace. Droplet size distributions and veloc-
lating implementation of staged firing with OFA ity vectors must be specified in the CFD model
can identify changes in unburned carbon in ash, inputs and the droplet trajectories and evaporation
CO concentrations, surface deposits, heat flux calculated. As the droplets evaporate, the reagents
profiles, and furnace exit gas temperature, as are moved into the gas phase and the NOx chem-
well as NOx concentrations [123, 124]. istry is calculated. This approach is best done with
a CFD model coupled with SNCR chemistry since
Post-combustion Zone the evaporation of the droplets can affect the local
Selective non-catalytic reduction (SNCR) is accom- gas temperatures. It is also possible to trace
plished by injection ammonia or urea reagent into streamlines of droplets through an established
the combustion flue gas in the upper region (post- combustion property field and calculate the NOx
combustion zone) of the radiant furnace reactions along these streamlines using chemical
[125]. SNCR is most effective in a narrow temper- kinetics; however, this approach ignores two-way
ature window of approximately 1050–1350 K. At coupling between the droplets and the flue gas
temperatures higher than this, reagents thermally properties. Figure 5 shows predicted droplet dis-
decompose and, at even higher temperatures, may tributions from SNCR systems for two different
react with oxygen to form NOx. At temperatures boilers. The left figure shows trajectories from
lower than this, the reagent will incompletely react two injector types in the upper furnace of a wall-
with the NOx creating ammonia slip. The chemistry fired boiler. The right figure shows droplet trajec-
for SNCR is included in the detailed kinetic mech- tories in a tangentially fired boiler. The figure on
anisms. If only the SNCR section of a boiler is the right also shows the NOx concentration pro-
modeled, a simplified kinetics model or simplified files before and after the SNCR.
484 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

0.95
Droplet Water
Mass Fraction
0.0
Side Wall Injectors
135o flat fan
(50 mm SMD)
Front Wall Injectors
20o full cone
(300 mm SMD)

Mitigation of Airborne Pollutants in Coal Combus- boiler (left) and tangential-fired boiler (right)
tion: Use of Simulation, Fig. 5 Predicted SNCR droplet (Figures courtesy of Reaction Engineering International)
trajectories in the upper furnace of a wall-fired coal-fired

The performance of SNCR in utility boilers the most commonly used catalyst type in conven-
depends upon the type of reagent used, the quan- tional SCRs [130].
tity of reagent injected, and the boiler design and Chemistry of ammonia-based NOx reduction
operating characteristics [128]. For maximum across the SCR catalyst is represented by the
performance, the injection system must be reactions [130, 131]:
designed to provide effective mixing of the
reagent within the optimal temperature window. 4NH3 þ4NO þ O2 ! 4N2 þ6H2 O
CFD modeling provides an effective approach to 4NH3 þ2NO2 þO2 ! 3N2 þ6H2 O
optimize the injection system performance for 2NH3 þNO þ NO2 ! 2N2 þ3H2 O
different boiler operating conditions.
Selective catalytic reduction (SCR) utilizes Simulation of SCR systems can be divided into
ammonia (NH3) mixed into the flue gas stream three components:
to react with NOx over a catalyst [129]. The cat-
alyst is usually configured in a plate or honey- 1. A three-dimensional CFD duct flow model
comb geometry to provide a high surface area to used to calculate flue gas properties through
gas volume ratio. An outcome of this geometry is the ductwork before and after the SCR reactor
that gas flow through the catalyst channels is 2. A three-dimensional CFD model of the ammo-
essentially one dimensional. SCRs operate at nia injection grid (AIG) that uses a highly
much lower temperatures than SNCR systems refined computational mesh to calculate the
and are thus situated between the economizer ammonia injection and mixing in the AIG zone
and air preheater in a power plant. Ideal operating 3. A one-dimensional plug flow SCR catalyst
temperatures for an SCR depend on the catalyst submodel used to calculate the NH3 and NO
type, ranging from 420 K to 560 K for noble metal reactions across the SCR catalyst
catalysts to 600–860 K for zeolite catalysts. Metal
oxide catalysts such as vanadium or titanium Three-dimensional CFD models calculate the
oxide operate in the 570–670 K range and are velocity profiles, pressure drop, temperature
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 485

profiles, NH3 injection and mixing, and species with flue gas profiles and resulting impact of
concentration profiles based on the system geom- AIG tuning.
etry and incoming flue gas properties. Ammonia • Determine the sensitivity of NOx reduction and
injection is modeled as jets exiting the injection ammonia slip to NH3 and NOx concentration
lances. The flow, temperature, and species con- profiles.
centration profiles entering the SCR catalyst are • Guide the amount of NH3 needed to reach a
geometrically divided into groups representing NOx reduction goal.
inlet conditions for different sections of the cata- • Predict how much ammonia slip will occur for
lyst. NOx chemistry across the SCR catalyst is a given design.
calculated using a one-dimensional catalyst sub- • Guide catalyst quantity and type needed for
model that accounts for catalyst performance as a NOx reduction goal.
function of catalyst design, temperature, NOx con- • Select best measurement locations for sam-
centration, NH3 concentration, flue gas composi- pling, testing, and analyzing.
tion, and residence time. Key submodel
capabilities include one-dimensional flow with One useful feature from CFD simulations is the
heterogeneous kinetics for NOx reduction by ability to “back track” from a predicted region of
NH3 over vanadia/titania catalysts, gas transport, poor NOx reduction at the catalyst to the region
and surface reactions solved along the catalyst where the ammonia is injected. Flow streamlines
length, number of assumed “active sites” for can be used to trace the ammonia species concen-
chemical reactions calibrated based on catalyst trations from the catalyst reactor to the injection
manufacturer’s design data [132]. zone and to identify locations where ammonia
Heterogeneous kinetics and catalyst active injection rates should be adjusted to give a more
sites should be adjusted depending on the catalyst efficient NH3 to NOx distribution at the reactor.
used. If detailed information about the catalyst is Figure 6 shows an example of streamlines tracked
not available, the CFD simulations of flue gas backward from the duct exit, through the catalyst,
flow velocities and NOx and ammonia concentra- to the ammonia injection grid.
tions (particularly the NH3/NOx ratio) can be used
to determine the potential system performance
Sulfur Dioxide (SO2)
based on velocity distributions (implying resi-
Sulfur is present in coal in inorganic and organic
dence time in catalyst) and NH3/NO ratio profiles
(implying correct matching of NH3 to NOx con- forms. During the combustion process, the inor-
ganic and organic sulfur is released and converted
centrations for optimal NOx reduction).
to sulfur dioxide, SO2. For bituminous and subbi-
There are many additional insights that can be
gained from CFD modeling of SCR systems. For tuminous coals, nearly all of the sulfur in the coal
(90–95%) is released as sulfur oxides, both SO2
example, simulations can be used to:
and sulfur trioxide, SO3, with the remainder ending
• Improve understanding of flue gas properties up as sulfates in the ash. For pulverized coal-fired
boilers, emissions of SO2 are independent of the
through the ammonia injection grid (AIG),
boiler design and operating conditions [133]. For
SCR catalyst, and ductwork.
• Evaluate the impact of flue gas profiles and lignitic and subbituminous coals containing ash
with high-alkali (calcium and sodium) content, a
flow rates on SCR performance.
higher fraction of the sulfur is retained in the ash.
• Identify potential operation/design problems such
as nonuniform velocity profiles, high-pressure Sulfur oxide emissions from lignites can be corre-
lated with the sulfur, sodium oxide (Na2O), and
drop zones, or high-ammonia slip zones.
silica (SiO2) content of the ash [134].
• Suggest duct designs to provide more uniform
flow to SCR reactor. The primary mechanisms for removal of a gas-
eous pollutant from a gas stream involve (1) absorp-
• Improve design and operation of the AIG
tion into a liquid, (2) gas/solid reaction, and
including guidance for matching AIG flows
486 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

Mitigation of Airborne
Pollutants in Coal
Combustion: Use of
Simulation,
Fig. 6 Predicted flow
traces through an SCR
installation illustrating the
correlation between NH3
profiles exiting the AIG,
NH3/NO ratios entering the
catalyst, and NO profiles
exiting the catalyst
(Figure courtesy of
Reaction Engineering
International)

(3) adsorption onto a solid. The first two mecha- Clean Gas
nisms are the primary methods for removal of sulfur to Stack
dioxide from coal-fired boiler flue gas. The primary
technologies for removing sulfur dioxide from coal-
fired power plants are wet limestone (CaCO3) and Mist Eliminator

dry lime (Ca(OH)2) scrubbing [135–137], typically Limestone

Sprays Slurry
referred to as wet and dry flue gas desulfurization
(FGD). A wide variety of wet and dry FGD pro-
cesses exist, but the basic processes are similar Absorber
Tower
[138–140]. A less commonly used alternative or
Flue Gas
enhancement to scrubber technologies is dry sorbent (post PM)
injection, where dry sorbents such as calcium car-
bonate or sodium carbonate (e.g., trona) are used to
decrease sulfur removal costs or improve capture Effluent
efficiency [141–143].
Mitigation of Airborne Pollutants in Coal Combus-
tion: Use of Simulation, Fig. 7 Schematic of wet FGD
Wet Flue Gas Desulfurization absorber tower
Figure 7 shows a schematic of a wet FGD
absorber tower. In a wet FGD, flue gas from the The resulting slurry is captured in an effluent
particulate matter control device (ESP or bag- holding tank. The cleaned flue gas exits the
house) enters the bottom of an absorber tower absorber after passing through a mist eliminator
(typically a spray tower) and flows upward. An designed to remove entrained droplets from the
alkaline solution or slurry containing partially gas. The primary steps involved in the process are
dissolved limestone (CaCO3) is sprayed into the gas/liquid mass transfer of the SO2, dissolution of
gas stream. The SO2 is absorbed into the solution the limestone, oxidation of SO3 to SO4 in solution,
where it reacts with the limestone to form calcium and crystallization of CaSO3 and CaSO4.
sulfite (CaSO3) and calcium sulfate (CaSO4). The Simulation of wet FGD absorber tanks are typi-
detailed chemistry of the process is complex and cally of three types. First, CFD flow models are used
depends upon the composition of the slurry and to simulate flue gas flow and slurry spray patterns.
extent of dissolution of the limestone [136, 144]. These models evaluate the effectiveness of the flue
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 487

gas-spray mixing or contacting. Two-phase gas and determine the effectiveness of SO2 capture given
droplet flows are represented with an Eulerian- specific residence times, temperatures, and spe-
Lagrangian or Eulerian-Eulerian framework cies concentrations. The overall chemistry pro-
depending on the model selected. Eulerian- cess for a forced oxidation absorber is
Lagrangian approaches allow more detailed repre-
sentation of droplet dispersion and mixing at greater SO2 þCaCO3 ðsÞþO2 þH2 O ðlÞ
computational cost. Eulerian-Eulerian formulations ! CaSO4 2H2 O ðsÞþCO2
are more commonly used when droplet and gas-
phase chemical reactions are included in the CFD This reaction can be further broken down into
model. In either approach, two-way mass, momen- four key chemical processes in the absorber: SO2
tum, and energy coupling between gas and droplets absorption, reagent (limestone) dissolution, sulfite
are included. Standard k-e or k-o technique is used oxidation, and solids precipitation. The reactions
to account for gas turbulence and turbulent droplet for these processes are shown below.
dispersion. For this first model type, no chemical
reactions are considered. For isothermal flows, tem- SO2 SO2 þH2 O $ H2 SO3
peratures are assumed constant and droplet evapo- absorption: H2 SO3 $ Hþ þHSO3 
ration is not considered. For non-isothermal flows, Limestone CaCO3+H2O $ Ca+++HCO3–
the effects of droplet evaporation and local gas dissolution: +OH–
cooling and composition change are considered via Oxidation: HSO3+O2 $ H++SO4=
a local energy balance between the gas and droplets. Precipitation: Ca+++SO4=+2H2O $ CaSO4 
The spray headers are represented as a distribu- 2H2O
tion of solid tubes, ending in injection nozzles. The
mist eliminator is commonly modeled as a porous
media with a suitable pressure drop. Droplets are The complex combination of gas-solid-liquid
represented by spherical solid particles of specified reactions make this process challenging to model.
size distribution (e.g., Rosin-Rammler distribution) Some chemistry models focus only on the SO2
emanating from spray header nozzles. Droplet size absorption, whereas other research has shown that
is a key parameter in determining absorber perfor- all four rate-based chemical processes should be
mance [144]. Droplets are introduced with an initial included [145]. Absorption of SO2 has been modeled
velocity vector to represent the spray pattern (e.g., using dual-film theory and appropriate correlations.
hollow cone). Input data for modeling injectors is The model for aqueous phase chemistry considered
acquired from nozzle manufacturers, who have instantaneous equilibrium reactions of eight
conducted extensive testing to quantify nozzle char- dissolved species into a slurry droplet, namely: SO2,
+  2 
acteristics. Given the large number of nozzles at aq, CO2,aq, H , OH , HSO3, SO3 , HCO3 , and
2
multiple spray header elevations, the resolution of CO3 [146]. A more complete model has been
the spray header geometry, mist eliminator geome- developed which accounts for interaction between
try, and number of droplet trajectories tracked must gas, liquid, and solid phases as well as a detailed
be balanced against the available computational liquid chemistry with the main acid-base equilibria
resources. More droplet trajectories will produce a and nonideal thermodynamic behavior [147]. All
more uniform and well-distributed representation of major rate-controlling steps are modeled: limestone
the sprays at higher computational cost. A wall dissolution, sulfites oxidation, and precipitation of
rebound and impingement model available in com- gypsum. Sometimes more complex chemistry
mercial CFD codes is used to represent droplet models can be used to develop a simplified represen-
interaction with surfaces. Modeling is simplified tation of the SO2 absorption chemistry for a limited
by assuming droplets do not interact with each other. range of conditions. This simpler representation of
Second, detailed chemistry models are used to the chemistry can then be more readily implemented
assess the reactions of the SO2 in the flue gas with into a CFD code to combine the gas-droplet mixing
the alkaline spray solutions. These models and chemistry effects.
488 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

Third, CFD models which include chemistry Mitigation of Airborne Pollutants in Coal Combus-
submodels are also used to guide scrubber designs tion: Use of Simulation, Table 2 CFD model charac-
teristics for a wet FGD absorber [144]
and estimate overall scrubber performance. These
closely coupled models may use simplified chemis- Parameter Modeling method
try relations, often based on more complex chemis- Modeling Eulerian-Lagrangian
approach
try mechanisms and/or laboratory-scale testing,
Phase Two-way coupling
which can be more readily coupled into the CFD interaction
model. Equations for gas-droplet mass, momentum, Turbulence k-e model
and energy transfer are solved simultaneously with model
chemical reactions. Alternately, a loosely coupled Nozzle type Dual flow hollow cone point
approach may be used to link CFD and chemistry injectors
calculations. In this approach, CFD calculations are Drag force Liu dynamic drag coefficient model
performed to determine local gas temperature, spe- Droplet TAB distortion model
distortion
cies concentration, and droplet property conditions.
Mist eliminator Porous media with suitable pressure
Using these inputs, chemical reactions are calculated drop
to update the gas and droplet properties. This infor- Droplet-wall Rebound and impingement model
mation is then fed back into the CFD model, typi- interaction
cally as source terms in the mass, momentum, Droplet size Rosin-Rammler distribution
energy, and species equations, and the process is distribution
Domain Polyhedral and prism layer cells with
iterated until a converged solution is reached. This
discretization surface remesher
approach has the advantage of using complex chem- Evaporation Quasi-steady-state droplet
istry models and commercial CFD software without evaporation model
having to input the chemistry calculations into the
CFD code. This approach also allows a range of
chemistry models of different fidelity, potentially
customized, to be used based on user needs. droplet evaporates leaving a fine powder that can
A number of three-dimensional models of full- be collected from the spray dryer surfaces and in a
scale wet FGD systems have been developed subsequent particulate collection device.
[144–149]. These models include some level of After evaporation, collection and retention of
chemical calculations as well as gas and droplet sorbent solids on a fabric filter permit additional
calculations. They have been used to evaluate SO2 removal to occur resulting in increased sor-
system performance (e.g., droplet dispersion, bent utilization. The primary processes involved
liquid-to-gas ratio, gas-droplet contacting time) in spray drying consist of atomization of the alka-
and to guide spray header positioning, nozzle line slurry into a spray of droplets, contacting of
design, flue gas inlet location and velocity, and the spray with the flue gas, absorption of SO2 into
gas to slurry mass flow ratios. Table 2 lists the the droplets, reaction with the suspended alkali
modeling techniques used in one such model material, drying of the spray by evaporation of
[144] that does not include gas-liquid chemistry. the moisture in the droplets, and, finally, separa-
tion of the solids from the gas [150].
Dry Flue Gas Desulfurization Similar to modeling of the absorber tower in the
Dry FGD is usually implemented via a spray dryer wet FGD process, simulation of a spray dryer
absorber (SDA). In an SDA system, hot flue gas absorber in the dry FGD process can range from
from the air preheater enters the top of an absorber simple CFD flow models to complex chemistry
tower where an atomizer sprays an alkaline solu- models to combinations of the two. CFD simulation
tion containing hydrated lime (Ca(OH)2) into the of the flue gas and solution droplets is similar to the
flue gas. SO2 is absorbed into the spray droplets techniques used for the wet FGD absorber. Spray
where it reacts with the alkali material to form dryers may have fewer injection nozzles allowing
solid calcium sulfite (CaSO3). The water in the better resolution of the droplet sprays. Droplet
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 489

injection properties are specified in the same way as liquid mixing, temperatures, and evaporation
wet FGD nozzles. During their lifetime, the sprayed rates. CFD models can also include simplified
droplets simultaneously evaporate and absorb SO2. SO2 absorption models or be loosely coupled
Essentially, the desulfurization efficiency in a spray with single-droplet models to assess SO2 absorp-
dryer is the result of the competition between the tion and droplet evaporation.
SO2 absorption rate in the slurry droplets and the
water evaporation rate from the droplets. SDA
Dry Sorbent Injection
models solve mass, momentum, energy transport
Dry sorbent injection (DSI) refers to the practice of
equations for the two-phase flows, and chemistry
injecting a dry alkaline mineral into a flue gas stream
for the SO2 absorption process.
in an effort to capture acid gases such as SO2, SO3,
Several researchers have developed single
and HCl. The most common sorbents are calcium
droplet models which model the SO2 absorption
and sodium based. CFD models are used to calculate
process [151–158]. The process can be described
flue gas and injected sorbent flow patterns. The
in the following steps:
models evaluate the effectiveness of the sorbent dis-
persion and flue gas-sorbent mixing. Two-phase gas
1. Gas-phase diffusion of SO2 from the gas bulk
and particle flows are represented with an Eulerian-
to the droplets surface
Lagrangian framework, which allows more detailed
2. Dissolution of SO2 at the droplet surface
representation of sorbent dispersion and mixing than
3. Formation of sulfurous acid and dissociation
less commonly used Eulerian-Eulerian formulations.
into ionic sulfur species following the scheme:
In either approach, two-way mass, momentum, and
energy coupling between gas and particles are
SO2ðaqÞ þH2 O ! H2 SO3ðaqÞ
included. Standard k-e or k-o technique is used to
H2 SO3ðaqÞ $ Hþ þHSO3  ðaqÞ
 account for gas turbulence and turbulent particle
HSO3  ðaqÞ $ Hþ þSO3 2 ðaqÞ
dispersion. Sorbent particle size distribution and
injection nozzle velocity vectors are specified. CFD
4. Liquid-phase diffusion of sulfur species
models can be used to calculate gas-particle mixing
toward the droplet center
only or to also calculate SO2 absorption where
5. Parallel dissolution of calcium hydroxide
appropriate chemistry is included [159, 160].
particles
The calcium reagent most widely utilized for
6. Liquid-phase diffusion of alkaline species
DSI is hydrated lime or calcium hydroxide (Ca
toward the droplet surface
(OH)2). The basic reaction for SO2 capture [161] is
7. Neutralization by reaction between acid and
alkaline species
CaðOHÞ2 þSO2 þ :5O2 ! CaSO4 þH2 O
The single-droplet models can be coupled with
The most common sodium reagents are trona
other chemistry and/or flow models to study the
and sodium bicarbonate. The basic trona reaction
relevance of the different resistances to SO2
for SO2 capture [160–162] is
absorption and to predict the influence of the
main operating variables on the spray-dryer desul-
2ðNa2 CO3  NaHCO3  H2 OÞþ3SO2
furization performance. For example, simulations
! 3Na2 SO3 þ4CO2 þ5H2 O
using a steady-state one-dimensional spray-dryer
model showed that the operating variables that
A secondary sulfate capture reaction is
exerted the largest influence on the overall desul-
furization efficiency were the Ca/S molar feed
3Na2 SO3 þ 1:5O2 ! 3Na2 SO4
ratio, the mean initial droplet size, and the mean
lime particle size [151].
CFD models of spray dryer absorbers calculate Single particle models with sorbent chemistry
gas and droplet flows in order to evaluate gas- have been developed which include chemical
490 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

reaction rates for the sulfation reaction and pore sulfur trioxide (SO3) emissions. Increased SO3
diffusion and boundary layer diffusion effects. emissions often result in an increase in stack
Often a shrinking core model is used to represent opacity, a phenomenon referred to as the “blue
the sorbent particles as the porosity of the particles plume” effect. Plume opacity is caused when SO3
decreases as sulfation degree increases. Many of is converted into vapor-phase sulfuric acid
these models are computationally simple enough (H2SO4). When this H2SO4 cools, it forms fine
that they can be coupled with CFD models via sub- particles of sulfuric acid aerosol. In sufficient con-
routines or user-defined functions to evaluate the centration, these fine particulates scatter light and
mixing and reaction processes simultaneously [163]. create a visible plume that is typically bluish or
DSI chemistry and overall SO2 removal effi- brownish orange in color. The plume color and
ciency are determined by three factors: flue gas extent of opacity are dependent on the aerosol
properties, sorbent properties, and the DSI system concentration, aerosol size, sun angle, gas temper-
configuration. Gas temperatures and acid gas spe- ature, and atmospheric conditions. In most cases,
cies concentrations affect the reaction kinetics of a plume will become visible when H2SO4 aerosol
sorbents with SO2 and SO3. Sorbent reactivity, concentrations in the flue gas exceed 10–20 ppm.
surface area, pore volume, and particle size impact The higher the plume concentration, the greater is
removal efficiency. The DSI system configuration the discoloration and the longer the discoloration
determines the residence time and sorbent to gas remains in the atmosphere. In severe conditions,
mixing characteristics for the system. CFD the sulfuric acid mist may settle to the ground
modeling coupled with sorbent chemistry models creating a health hazard.
can account for each of these variables. Of the factors affecting plume opacity, only the
From a design and analysis perspective, CFD SO3 concentration, which leads to H2SO4 aero-
modeling of a sorbent injection system can be used sols, can be controlled by a power plant. SO3
to ensure uniform sorbent distribution, maximize concentrations depend on a number of factors
sorbent residence times, determine optimal sorbent and vary significantly from plant to plant. Differ-
feed rates and injection location, optimize duct ences between plants are a result of differences in
layout, and ensure proper mixing within the flue fuel, design, and operational characteristics. SO3
gas stream for maximum adsorption and capture levels in coal-fired plants are affected by several
efficiency. plant components, including the radiant furnace
and back pass, selective catalytic reduction sys-
Sulfur Trioxide (SO3) tem (SCR), air preheater (APH), electrostatic pre-
The addition of air pollution control devices cipitator (ESP), and wet flue gas desulfurization
(APCDs) to coal-fired power plants has led to unit (wet FGD). Differences within a plant are a
significant reductions in airborne emissions such result of different chemical reactions that form
as NOx and SO2 but in some instances has also SO3 during combustion and flue gas conditioning.
produced an unexpected side effect of increased Figure 8 illustrates the plant components that can

Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation, Fig. 8 Schematic illustrating SO3
behavior in a coal boiler and APCD components
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 491

have the greatest impacts on SO3 levels. SO3 composition. Reductions typically range from
behavior in each of these components is discussed 10% to 50%. For 30 ppm SO3 in the flue gas, a
below. 25% reduction would result in an 8 ppm decrease
Inside the boiler, SO3 is formed by conversion to approximately 22 ppm SO3 levels. From an
from SO2. The extent of conversion from SO2 to operational perspective, if the amount of SO3
SO3 depends on the fuel composition, flue gas transferred to the fly ash is too high (due to high
temperature, oxygen concentration (excess air), SO3 concentrations in the flue gas), ash resistivity
and ash composition (tube deposits) [164]. Con- may decrease to the point that particle collection
version from SO2 to SO3 in the radiant section and efficiency is reduced at low gas temperatures.
convective back pass is generally 0.5–1.5% of the SO3 is reduced in wet FGDs and typically
total sulfur in the coal [165], meaning that a fur- converted to H2SO4 mist (vapor-phase sulfuric
nace burning a 2.5% sulfur coal that produces acid) because the temperatures in an FGD are
around 2000 ppm of SO2 in the flue gas would below the H2SO4 dew point. The extent of reduc-
in turn produce around 20 ppm SO3 in the flue gas tion depends on the FGD design, i.e., the temper-
leaving the boiler back pass. ature profile and spray contact in the scrubber.
SO3 emissions can be affected by the presence Reduction is generally in the 30–40% range,
of an SCR. The vanadia-based catalysts used to meaning a 35% reduction of 22 ppm SO3 in the
reduce NOx in SCRs also oxidize SO2 to SO3. The flue gas would result in approximately 14 ppm
extent of conversion of SO2 to SO3 depends on the SO3 levels. FGDs are not efficient at removing
catalyst type and flue gas temperature (i.e., higher H2SO4 aerosols, so most of the SO3 converted to
temperatures produce higher conversions). Con- H2SO4 in the FGD will exit the stack. It should be
version levels are generally 0.5–1.5%, meaning a noted that it is very hard to differentiate between
flue gas with 2000 ppm SO2 that experienced a SO3 and H2SO4 in stack measurements; thus, the
1% conversion would yield an additional 20 ppm two are often referred to interchangeably. Once
SO3 in the flue gas at the SCR outlet. SO3 or the vapor-phase H2SO4 is emitted to the
Air preheaters (APH) have the opposite effect atmosphere and cools, fine particles of sulfuric
of SCRs, i.e., they can reduce SO3 levels. The acid aerosol are formed, which in sufficient quan-
extent of reduction varies greatly from unit to tity create visible plumes. The chemistry of SO3
unit and is dependent on flue gas temperature formation in boilers and SCRs and the reduction
and type of air preheater (i.e., regenerative versus of SO3 across air preheaters and APCDs are com-
tubular). The faster the temperature quench and plex and have been summarized in the literature
lower the APH outlet temperature, the greater the [164, 166].
SO3 reduction. Reductions typically range from The most common methods of SO3/H2SO4
10% to 50%. For example, a 25% reduction will control in existing boilers are wet ESPs or the
lower SO3 emissions from a baseline of 40 ppm to injection of a sorbent upstream of the particulate
30 ppm. This reduction is tempered by the fact collection device. Wet ESPs are best suited to use
that lower outlet temperatures can increase the where greater than 90% SO3 capture is required
risk of H2SO4 formation, particularly below and a new wet FGD is planned for the unit. Wet
530 K. In addition, SO3 can also combine with ESPs are significantly more expensive and require
NH3 slip (from SCR or SNCR systems) at lower more space than sorbent injection systems. Sor-
APH temperatures to form ammonia bisulfate, bent injection systems can be fitted into existing
which can deposit on APH surfaces and cause flue gas ductwork. Sorbents adsorb the SO3 gas
plugging and increased pressure drop. and are in turn captured in the particulate collec-
Electrostatic precipitators (ESP) can also tion device, most often a fabric filter. Sorbents that
reduce SO3 levels. SO3 is captured on fly ash may be used for control of SO3 emissions include,
particles and then removed with the particles in but are not limited to, ammonia, hydrated lime,
ESP. The extent of reduction primarily depends on magnesium hydroxide, sodium bicarbonate,
the flue gas temperature and the fly ash trona, and sodium bisulfite (SBS) solution. Note
492 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

that the lime and sodium compounds are also used droplet evaporation before impaction on struc-
to capture other acid gases such as SO2 and HCl tures, and the residence time required for sorbents
from the flue gas. to react with flue gas can also be studied. CFD
The main reactions for adsorption of SO3 modeling can be done for SO3 control in the
inside a hydrated lime sorbent are furnace (magnesium hydroxide) or in the duct-
work before or after the air preheater (lime and
SO3 þ CaðOHÞ2 ! CaSO4 þH2 O sodium injection).
H2 SO4 þ CaðOHÞ2 ! CaSO4 þ2H2 O
Mercury (Hg)
The main reaction for sodium compounds is Mercury is liberated from coal during the com-
Na2CO3+SO3 ! Na2SO4+CO2 bustion process and exists in three forms – ele-
These reactions will proceed in conditions mental mercury (Hg0), oxidized mercury (Hg2+),
where the SO3 in the flue gas comes in contact and particulate-bound mercury (Hgp). Mercury
with the sorbent for temperatures and residence exists in elemental form in the high-temperature
times typical of that in ductwork between the air regions of coal-fired boilers. As the flue gas cools,
preheater and the baghouse. The challenge is to a series of complex reactions begin to convert the
ensure that sorbents are sufficiently distributed Hg0 to gaseous oxidized (Hg2+) and particulate-
and mixed with the flue gas. The distribution and bound (Hgp) mercury. In the presence of chlorine,
mixing depend on the sorbent feed rates and the gas-phase equilibrium conditions favor the forma-
injection system design. CFD modeling is valu- tion of HgCl2 at flue gas cleaning temperatures.
able in guiding the design and assessing the per- However, the oxidation of Hg0 is kinetically lim-
formance of dry sorbent injection (DSI) systems. ited by the gas-phase (homogeneous) and gas-
Effective modeling can be done with two-phase solid (heterogeneous) reaction rates [167,
models of flue gas with inert particles (which pre- 168]. As a result, mercury enters the flue gas
dicts sorbent distribution only) or with reacting cleaning device(s) as a mixture of Hg0, Hg2+,
particles (which predicts actual capture). and Hgp. This partitioning of mercury into three
SO3 sorbent injection systems can be modeled forms is known as mercury speciation. In general,
with CFD software using similar approach to DSI the majority of gaseous mercury in bituminous
modeling for SO2 systems (for dry sorbent) or coal-fired boilers is Hg2+. Conversely, the major-
slurry modeling of SNCR reagents for NOx con- ity of gaseous mercury in subbituminous- and
trol. Typical turbulent flow models can be used lignite-fired boilers is Hg0 [169]. The forms or
with two-phase, Eulerian-Lagrangian models to speciation of mercury can have considerable
represent dispersion and mixing of the sorbent influence on the selection of mercury control
droplets/particles. For slurry systems, droplet technologies.
evaporation can also be modeled. Sorbent size The extent of conversion of Hg0 to Hg2+ and
distribution must be specified along with injection Hgp depends on the coal and flue gas composition,
velocity vectors for each nozzle. Nozzle spray the amount and properties of fly ash, and the flue
characteristics can be obtained from the nozzle gas quench rate. Data on mercury speciation from
manufacturers. full-scale utility boilers provide insight into mer-
Detailed CFD modeling of injector sprays and cury chemistry in flue gas. Utility boiler field-test
droplet dispersion can be used to lower perfor- data show that the halogen (e.g., chlorine) content
mance risk associated with different designs. Spe- of the coal has some impact on the speciation of
cific design and operating considerations such as mercury at the inlet to the air pollution control
injector locations, droplet size distribution, and devices (APCDs) [170–172]. Additionally, mer-
slurry flow rates at different furnace loads can be cury speciation is impacted by APCDs. In some
modeled to aid the design process. Key perfor- instances, Hg0 is oxidized by the vanadium-based
mance parameters such as the extent of sorbent catalysts used in selective catalytic reduction
dispersion throughout the flue gas, the extent of (SCR) units. Measurements of mercury speciation
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 493

across full-scale catalysts [173–176] show that SO3 upstream of the ACI or by using advanced
substantial oxidation of elemental mercury can activated carbon formulations which are designed
take place across the catalyst. Mercury can also for SO3 tolerance [162].
be oxidized across air preheaters [167]. CFD modeling has been used to simulate mer-
Mercury can also be captured by APCDs. For cury control using activated carbon injection
example, data on mercury speciation in full-scale [179]. Two-phase models are used to calculate
coal boilers taken from EPA’s Information Collec- the injection, dispersion, mixing, and residence
tion Request (ICR) showed that the average mer- time of ACI in the flue gas. For these applications,
cury removal across wet FGD scrubbers is well duct geometries, injector geometries, nozzle loca-
over 50% for bituminous coals but only around tion, sorbent size distribution, injection angle and
30% for subbituminous coals [171, 172]. These velocity, and flue gas inlet velocity profiles must
differences are attributable to the differences in be specified. Calculations provide information on
mercury speciation for bituminous and subbitumi- the coverage and residence time of the sorbents in
nous coals. The particulate matter (PM) control the flue gas. Capture performance is inferred from
device captures Hgp, and the wet FGD scrubber the degree of coverage and residence time.
captures soluble compounds of Hg2+. Mercury There is limited understanding of the mecha-
can be captured more effectively by combinations nisms of mercury oxidation and adsorption, which
of ESPs or fabric filters (for Hgp capture) and limits the development and application of chemi-
spray dryer absorbers (SDA) or wet FGDs (for cal kinetics models for mercury reduction and
Hg2+ capture). Capture rates up to 98% have been capture, particularly related to ACI technologies
reported for these combinations with bituminous [180, 181]. This is a relatively new research field
coals [169]. Since SDAs and wet FGDs effec- wherein a number of kinetic mechanisms have
tively remove oxidized mercury but not elemental been proposed [180, 182], but no unifying models
mercury, maximizing the amount of Hg2+ in the have been identified.
flue gas upstream of the SDA or wet FGD One kinetics model that has shown promise in
improves Hg removal. predicting mercury speciation and capture
Mercury can also be controlled by sorbent includes both homogeneous and heterogeneous
injection. The most common form of this is acti- kinetic mechanisms for modeling of mercury
vated carbon injection (ACI). ACI systems inject behavior in practical coal combustion systems
fine activated carbon particles into the flue gas [167]. The homogenous kinetic model includes
stream after the air preheater and before the par- 385 reactions involving 110 species. Mechanisms
ticulate matter (PM) control device. The activated are included for hydrocarbon chemistry, NOx
carbon captures gaseous mercury through reac- chemistry, SO2 chemistry, chlorine chemistry,
tions on the surface of the carbon particles and mercury chemistry. The mercury mechanism
[177]. The carbon particles may be treated with consists of eight reactions of Hg and HgCl with
bromine to improve performance in systems that chlorine species to form HgCl2. The homogenous
fire coals with lower native levels of halogens model is dependent on accurate prediction of
(i.e., low chlorine coals) [170]. Effectiveness of chlorine chemistry to account for the effects of
ACI systems is a function of coal type, mercury quench rate on mercury oxidation. The chlorine
speciation, flue gas temperature, activated carbon speciation is predicted using a mechanism that
injection rate, gas-sorbent mixing (i.e., coverage), includes 29 reactions and 6 species and is depen-
and residence time. Depending on flue gas com- dent on the presence of O, OH, and H radicals
position, mercury removal is relatively effective at [183, 184]. The homogeneous model was derived
temperatures below 450 K. from previous work in the literature [183, 184]
It has been noted that high levels of SO3 can and has been shown to accurately predict mercury
interfere with activated carbon mercury control speciation in gas-phase homogeneous reactors in
effectiveness [177, 178]. These impacts can be the presence of chlorine [185]. The heterogeneous
mitigated by using dry sorbent injection to control reactions of mercury in coal include pathways for
494 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

mercury oxidation [167, 186]. This model allows control. CFD is best used to study flow and tem-
for sorption/desorption of Hg on/off the solid perature profiles and dispersion of ACI sorbents in
phase, particularly carbon, and promotion of oxi- specific sections or APCD components of the
dation of Hg on the surface, where the oxidized plant, with some limited application of chemical
products are desorbed. The combined process of reactions. As an alternative to CFD models, plant-
oxidation/desorption is strongly determined by wide process models have been developed that
the HCl and SO2 present and trace amounts of track mercury speciation across the full coal
NO2 and SO3. plant from fuel input to stack outlet.
The kinetic model can be used to predict mer- These models rely on kinetic mechanisms and
cury speciation across the air preheater and SCR. experimentally based correlations for assessing
The inputs to the kinetic model are the coal com- mercury behavior across full coal plants, includ-
position, excess O2 at the economizer exit, ing impacts of SO3 and ACI [167, 187, 188]. The
unburned carbon in fly ash, air heater/SCR inlet, models use coal properties and descriptions of
and exit temperatures. CFD models can be used to APCD operating parameters to estimate mercury
predict gas temperature, species distribution, and speciation and removal for different power plant
particle properties. Predicting mercury removal configurations. This approach is different from
across ESPs is more challenging due to the more CFD models which typically focus only on the
complex mercury interactions with HCl, SO2, furnace or one specific APCD at a time.
SO3, NO2, and fly ash. In the absence of compre- An example of mercury removal predictions
hensive kinetic models, test data taken from from one such mercury process model [167] is
power plants is used to develop general correla- shown in Fig. 9. Results compare Hg removal
tions for removal across ESPs. Similarly, predic- for different plant equipment configurations and
tion of mercury removal across wet FGDs is control technologies relative to a baseline config-
limited to the use of correlations based on coal uration of the boiler that includes an air preheater
properties and flue gas properties [167]. (APH) and ESP. Mercury removal was calculated
Because mercury emissions are dependent on based on the addition of different control technol-
boiler components from furnace to stack and ogies including an SCR and wet FGD (co-benefits
because detailed descriptions of all mercury reac- or CB), a boiler-based additives (BBA), activated
tions are not well established, CFD modeling has carbon injection (ACI), brominated ACI (BrACI),
limited applications for modeling mercury fabric filter (FF), and various combinations of

Mitigation of Airborne
Pollutants in Coal
Combustion: Use of
Simulation,
Fig. 9 Predicted mercury
removal for various
mercury control scenarios
using a kinetics-based
process model [188]
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 495

these technologies. In this example for a boiler near-pure oxygen for oxidation of coal in pressur-
firing bituminous coal, configurations which ized and/or dense-phase environments, resulting
included BBA, ACI, and ESP produced the in high CO2 concentrations in flue gas. Existing
highest mercury removals (99%). An alternate submodels for devolatilization, char oxidation,
combination of BBA with co-benefits also pro- soot formation, and radiative heat transfer have
duced high removal (95%). not been fully developed or tested for these non-
traditional conditions. Ongoing research funded
by the US Department of Energy is supporting
Future Work development of pilot-scale test facilities and sub-
model development to address some of these
The quantity and quality of CFD simulations of issues [189, 190]. Development of fundamentally
pollutant control in coal combustion systems con- based CFD models are critical to development of
tinue to grow, aided by developments and appli- new combustion technologies as models will be
cations in other combustion industries. This work relied on to guide scale up from bench-scale and
is expected to improve mesh generation, numeri- pilot-scale systems to full-scale designs. Models
cal methods, and design optimization techniques. are also critical to optimizing design and opera-
However, continued improvements are needed in tion of next-generation systems.
three areas specific to coal combustion. First, con-
tinued development is needed for advanced chem-
istry models of pollutants and control
Bibliography
technologies and for techniques for coupling this
chemistry with existing turbulent flow and heat 1. (2016) US Energy Information Administration. Elec-
transfer models. These developments range from tric power monthly
improved detailed and reduced chemical kinetic 2. https://www.epa.gov/naaqs
mechanisms to advanced turbulence models to 3. https://gispub.epa.gov/air/trendsreport/2016/
4. Power plant emission trends. https://www3.epa.gov/
improved turbulence-chemistry coupling tech- airmarkets/progress/datatrends/index.html
niques. Research in large eddy simulations 5. Mitchell JFB (1989) The “Greenhouse” effect and
(LES) and reduced chemical mechanisms are climate change. Am Geophys Union 27(1):115–139.
helping to address some of these issues. Washington, DC
6. Adams B, Cremer M, Wang D (2000) Modeling non-
Second, improved characterization of coal ash equilibrium CO oxidation in combustion systems. In:
deposits is needed to improve prediction of heat Proceedings of the ASME heat transfer division
transfer and temperature profiles within coal HTD-366-5:29–34
boilers. Radiant furnace and back pass deposits 7. Turns S (1996) An introduction to combustion.
McGraw-Hill, New York
impact not only heat transfer but also NOx forma- 8. Gordon S, McBride BJ (1994) Computer program for
tion, CO oxidation, and SO3 formation and cap- calculation of complex chemical equilibrium compo-
ture via temperature. Accurate prediction of sitions and applications: I Analysis, vol 1311. NASA
deposit profiles, quantities, and properties is Reference Publication, Washington, DC
9. Reynolds W (1986) The element-potential method
needed to improve boiler operation and mitigation for chemical equilibrium analysis: implementation
of pollutant control technology impacts. Improve- in the interactive program STANJAN. Stanford Uni-
ment of deposit models likely requires additional versity, Palo Alto
testing to better characterize deposit properties as 10. Smoot L, Smith P (1985) Coal combustion and gas-
ification. Plenum Press, New York
a function of different operating environments. 11. Kee R, Rupley F, Miller J (1989) Chemkin-II: a
Third, there is a need for more fundamentally fortran chemical kinetics package for the analysis of
based and experimentally validated submodels for gas-phase chemical kinetics. Technical report
advanced combustion systems such as pressurized SAND-89-8009. Sandia National Labs
12. Goodwin D (2001) Cantera user’s guide fortran ver-
oxy-coal combustion and chemical looping. sion. California Institute of Technology, Pasadena
These next-generation combustion systems are 13. Meeks E, Chou C, Garratt T (2013) Package equiv-
designed to maximize CO2 capture by utilizing alent reactor networks as reduced order models for
496 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

use with CAPEOPEN compliant simulations – final 31. Caretto LS, Curr RM, Spalding DB (1972) Two
report. US DOE report DOE/FE0001074-4. numerical methods for three-dimensional boundary
14. Chen JY (1988) A general procedure for constructing layers. Comput Methods Appl Math Eng 1(1):39–57
reduced reaction mechanisms with given indepen- 32. Caretto LS, Gosman AD, Patankar SV, Spalding DB
dent relations. Combust Sci Technol 57(1–3):89–94 (1973) Two calculation procedures for steady, three-
15. Grant D, Pugmire R, Fletcher T, Kerstein A (1989) dimensional flows with recirculation. Proc, 3rd Int
Chemical model of coal devolatilization using perco- Conf on Numerical Methods in Fluid Mechanics,
lation lattice statistics. Energy Fuel 3:175–186 Springer Verlag
16. Solomon P, Hamblen D, Carangelo R, Serio M, 33. Zuber I (1972) A mathematical model of the com-
Deshpande G (1988) General model of coal bustion chamber. Staatliches Forschungsinstitut fuer
devolatilization. Energy Fuel 2:405–422 Maschinenbau, Bechovice
17. Ubhayakar SK, Stickler DB, Von Rosenberg C, 34. Patankar SV, Spalding DB (1974) Simultaneous pre-
Gannon, RE (1976) Rapid devolatilization of pulver- dictions of flow patterns and radiation for three-
ized coal in hot combustion gases. In: 16th interna- dimensional flames. In: Afgan NH, Beer JM (eds)
tional symposium on combustion, Pittsburgh, Heat transfer in flames. Wiley, New York
pp 427–436 35. Serag-El-Din MA (1977) The numerical prediction
18. Kobayashi H, Howard JB, Sarofim AF (1977) Coal of the flow and combustion processes in a three-
devolatilization at high temperatures. Symp Combust dimensional combustion chamber. London Univer-
16(1):411–425 sity PhD thesis, Imperial College
19. Hurt R, Sun J, Lunden M (1998) A kinetic model of 36. Spalding DB (1971) Mixing and chemical reaction in
carbon burnout in pulverized coal combustion. Com- confined turbulent flames. In: 13th international sym-
bust Flame 113:181–197 posium on combustion, Combustion Institute, Pitts-
20. Baxter LL (1987) Particle phase behaviour in com- burgh, pp 649–657
bustion environments. Ph.D. Dissertation, Chemical 37. Spalding DB (1971) Concentration fluctuations in a
Engineering Department. Brigham Young Univer- round turbulent free jet. J Chem Eng Sci 26:95
sity. Provo, Utah 38. Lockwood FC, Shah NG (1981) A new radiation
21. Pletcher RH, Tannehill JC, Anderson D (2012) Com- solution method for incorporation in general combus-
putational fluid mechanics and heat transfer, 3rd edn. tion prediction procedures. In: 18th symposium
CRC Press, Boca Raton (international) on combustion. Elsevier, Amsterdam
22. Patankar S (1980) Numerical heat transfer and fluid 39. Fiveland WA (1984) Discrete-ordinates solutions of
flow. Hemisphere, Washington, DC the radiative-transport equation for rectangular
23. Versteeg HK, Malalasekera W (2007) An introduc- enclosures. J Heat Transf 106:699–706
tion to computational fluid dynamics, the finite vol- 40. Magnussen BF, Hjertager BH (1976) On mathemat-
ume method, 2nd edn. Pearson Education Limited, ical modelling of turbulent combustion with special
England emphasis on soot formation and combustion. In: 16th
24. Eaton AM, Smoot LD, Hill SC, Eatough CN symposium (international) on combustion, Combus-
(1999) Components, formulations, solutions, evalu- tion Institute, Pittsburg, pp 719–729
ation, and application of comprehensive combustion 41. Bray KNC (1980) Topics in applied physics.
models. Prog Energy Combust Sci 25:387–436 Springer, New York
25. Williams A, Backreedy R, Habib R, Jones JM, 42. Pope SB (1982) An improved turbulent mixing
Pourkashanian M (2002) Modelling coal combus- model. Combust Sci Technol 28:131
tion: the current position. Fuel 81(5):605–618 43. Abbas AS, Lockwood FC (1986) Prediction of power
26. Patankar SV, Spalding DB (1970) Heat and mass station combustors. In: 21st symposium
transfer in boundary layers, 2nd edn. Morgan- (international) on combustion, Elsevier, Amsterdam,
Grampian, London pp 285–292
27. Spalding DB (1977) GENMIX; a general computer 44. Boyd RK, Kent JH (1986) Three-dimensional fur-
program for two-dimensional parabolic phenomena. nace computer modelling. In: 21st Symposium
Pergamon Press, Oxford (international) on combustion, Elsevier, Amsterdam,
28. Spalding DB, Stephenson PL (1971) Laminar flame pp 265–274
propagation in hydrogen & bromine mixtures. Proc 45. Cetegen BM, Richter W (1987) Heat transfer model-
Roy Soc A 324:315–337 ing of a large coal-fired utility boiler and comparisons
29. Gosman GD, Pun WM, Runchal AK, Spalding DB, with field data. In: Proceedings of the 2nd ASME/
Wolfshtein M (1969) Heat and mass transfer in JSME Thermal Engineering Joint Conference,
recirculating flows. Academic Press, London vol 1. New York, pp 225–234
30. Pun WM, Spalding DB (1968) A procedure for pre- 46. De Michele G, Ghiribelli L, Pasini S, Tozzi A (1989)
dicting the velocity and temperature distributions in a A 3-D code for predicting radiative and convective
confined, steady, turbulent, gaseous diffusion flame. heat transfer in boilers. In: Shah, RK (ed) Heat trans-
In: Proceedings of 18th International Aeronautical fer phenomena in radiation, combustion, and fires,
Congress, Belgrade. Pergamon Press, London vol. 106, HTD – ASME, New York, pp 275–286
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 497

47. Fiveland WA, Wessel RA (1986) FURMO: 64. Falcitellia M, Pasinib S, Tognotti L (2002) Modelling
A numerical model for predicting performance of practical combustion systems and predicting NOx
three-dimensional pulverized-fuel fired furnaces. emissions with an integrated CFD based approach.
ASME Paper 86-HT-35. ASME, New York Comput Chem Eng 26(9):1171–1183
48. Gillis PA, Smith PJ (1990) An evaluation of three- 65. Cremer M, Valentine J, Shim H, Davis K, Adams B,
dimensional computational combustion and fluid- Letcavits J, Vierstra S (2003) CFD-based development,
dynamics for industrial furnace geometries. In: 23rd design, and installation of cost-effective NOx control
symposium (international) on combustion, Elsevier, strategies for coal-fired boilers. In: The mega sympo-
Amsterdam, pp 981–991 sium: EPRI-DOE-EPA combined utility air pollutant
49. Lockwood FC, Mahmud T (1988) The prediction of control symposium, AWMA, Washington, DC
swirl burner pulverized coal flames. In: 22nd sympo- 66. Cremer M, Adams B, Valentine J, Letcavits J,
sium (international) on combustion, Elsevier, Vierstra S (2002) Use of CFD modeling to guide
Amsterdam, pp 165–173 design and implementation of overfire air for NOx
50. Smith PJ, Fletcher TH, Smoot LD (1981) Model for control in coal-fired boilers. In: Proceedings of nine-
pulverized coal fired reactors. In: 18th symposium teenth annual international Pittsburgh coal confer-
(international) on combustion, Elsevier, Amsterdam, ence, Pittsburgh
pp 1285–1293 67. Adams B, Cremer M, Valentine J, Bhamidipati V,
51. Smoot LD (1984) Modeling of coal-combustion pro- O’Connor D, Letcavits J, Vierstra S (2002) Use of
cesses. Prog Energy Combust Sci 10:229–272 CFD modeling for design of NOx reduction systems
52. Truelove JS (1984) The modelling of flow and com- in utility boilers. In: International joint power gener-
bustion in swirled, pulverized-coal burners. In: 20th ation conference, Phoenix
symposium (international) on combustion, Elsevier, 68. Cremer M, Wang D, Montgomery C, Adams
Amsterdam, pp 523–530 B (2001) Utilization of reduced mechanism methods
53. Smoot L, Pratt DT (1979) Pulverized coal combus- in CFD simulations for improved NOx predictions in
tion and gasification. Plenum Press, New York utility boilers and furnaces. In: Joint AFRC/JFRC/
54. Adams B, Smith PJ (1993) Three dimensional dis- IEA international combustion symposium, Kauai
crete ordinates modelling of radiative transfer in a 69. Adams B, Wang DH, Cremer M, Frizzell K, Conn
geometrically complex furnace. Combust Sci S (2001) Modeling NOx reduction from fuel lean gas
Technol 88(5–6):293–308 reburning and selective non-catalytic reduction com-
55. Viskanta R, Menguc MP (1987) Radiation heat trans- bined with overfire air at OMU’s Smith Unit 1. US
fer in combustion systems. Prog Energy Combust Sci EPA/DOE/EPRI combined power plant air pollutant
13:97–160 control symposium: The mega symposium, paper
56. Fiveland W, Jessee P (1995) Comparison of discrete 147, Chicago, AWMA, Washington, DC
ordinates formulations for radiative heat transfer in 70. Wang H, Harb J (2007) Modeling of ash deposit
multidimensional geometries. J Thermophys Heat growth and sintering in PC-fired boilers. Impact of
Transfer 9:1 mineral impurities in solid fuel combustion, Gupta R,
57. Fletcher TH, Ma J, Rigby JR, Brown AL, Webb BW Wall T, Baxter L (Eds), Berlin Springer Science &
(1997) Soot in coal combustion systems. Prog Business Media 697—708
Energy Combust Sci 23:283–301 71. Lee1 FCC, Lockwood FC (1999) Modelling ash
58. Adams BR, Smith PJ (1995) Modeling effects of soot deposition in pulverized coal-fired applications.
and turbulence-radiation coupling on radiative trans- Prog Energy Combust Sci 25(2):117–132
fer in turbulent gaseous combustion. Combust Sci 72. Rushdi A, Gupta R, Sharma A, Holcombe D (2005)
Technol 109:121–140 Mechanistic prediction of ash deposition in a pilot-
59. Leung KM, Lindstedt P, Jones WP (1991) A simplified scale test facility. Fuel 84(10):1246–1258
reaction mechanism for soot formation in nonpremixed 73. Pedel J, Thornock JN, Smith PJ (2012) Large eddy
flames. Combust Flame 87:289–305 simulation of pulverized coal jet flame ignition using
60. Magnussen B, Hjertager B (1977) On mathematical the direct quadrature method of moments. Energy
modeling of turbulent combustion with special Fuel 26(11):6686–6694
emphasis on soot formation and combustion. Symp 74. Pedel J, Thornock JN, Smith PJ (2013) Ignition of
Combust 16(1):719–729 co-axial turbulent diffusion oxy-coal jet flames:
61. Peters N (1988) Laminar flamelet concepts in turbu- experiments and simulations collaboration. Combust
lent combustion. Symp Combust 21(1):1231–1250 Flame 160(6):112–1128
62. Christo F, Dally B (2005) Modeling turbulent 75. Warzecha P, Boguslawski A (2014) LES and RANS
reacting jets issuing into a hot and diluted coflow. modeling of pulverized coal combustion in swirl
Combust Flame 142(1–2):117–129 burner for air and oxy-combustion technologies.
63. Gran I, Magnussen B (1996) A numerical study of a Energy 66:732–743
bluff-body stabilized diffusion flame. Part 1: influ- 76. Stein OT, Olenik G, Kronenburg A et al
ence of turbulence modeling and boundary condi- (2013) Towards comprehensive coal combustion
tions. Combust Sci Technol 119(1–6):171–190 modelling for LES. Flow Turbul Combust 90:859
498 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

77. Kulaots I, Hurt RH, Suuberg EM (2004) Size distri- 93. Soldati A (2000) On the effects of electro-
bution of unburned carbon in coal fly ash and its hydrodynamic flows and turbulence on aerosol trans-
implications. Fuel 83:223–230 port and collection in wire-plate electrostatic
78. Wang C, Seames WS, Gadgil M, Hrdlicka J, Fix precipitators. J Aerosol Sci 31(3):293–305
G (2007) Comparison of coal ash particle size distri- 94. Turner J, McKenna J, Mycock J, Nunn A, Vatavuk
butions from Berner and Dekati low pressure impac- W (1998) Baghouses and filters. EPA/452/B-02-001,
tors. Aerosol Sci Technol 41:1049–1062 OAQPS, US EPA, Research Triangle Park, NC
79. Seaton A, Godden D, MacNee W, Donaldson K, 95. Broadway RM, Cass RW (1975) Fractional effi-
Gooden D (1995) Particulate air pollution and acute ciency of a utility boiler baghouse: Nucla generating
health effects. Lancet 345:176–178 plant. EPA-600/2-75-013-a, Office of Research and
80. U.S. Environmental Protection Agency (1998) Partic- Development, US EPA, Washington, DC
ulate matter research needs for human health risk 96. Cass RW, Broadway RM (1976) Fractional effi-
assessment to support future reviews of the national ciency of a utility boiler baghouse: Sunbury Steam
ambient air quality standards for particulate matter. Electric Station. EPA-600/2-76-077a, Office of
National Center for Environmental Assessment, Research and Development, US EPA, Research Tri-
Research Triangle Park. EPA/600/R-97/132F angle Park, NC
81. U.S. Environmental Protection Agency (1998) Station- 97. Dennis R, Klemm HA (1980) Modeling concepts for
ary source control techniques document for fine partic- pulse jet filtration. JAPCA 30(l):38–43
ulate matter. Research Triangle Park: National Center 98. Leith D, Ellenbecker MJ (1980) Theory for pressure
for Environmental Assessment. EPA 68-D-98-026 drop in a pulse-jet cleaned fabric filter. Atmos Envi-
82. Wilder J, Pilat M (1983) Calculated droplet size ron 14:845–852
distributions and opacities of condensed sulfuric 99. Koehler JL, Leith D (1983) Model calibration for
acid aerosols. APCA J 33(9) pressure drop in a pulse-jet cleaned fabric filter.
83. (2016) ASTM D1857/D1857M-16 standard test Atmos Environ 17(10):1909–1913
method for fusibility of coal and coke ash. ASTM 100. Spinti JP, Pershing DW (2003) The fate of char-N at
International, West Conshohocken pulverized coal conditions. Combust Flame 135:299–313
84. Walsh PM, Sayre AN, Loehden DO, Monroe LS, 101. https://www.epa.gov/sites/production/files/2016-07/
Beér JM, Sarofim AF (1990) Deposition of bitumi- documents/cofactsheetaugust12v4.pdf
nous coal ash on an isolated heat exchanger tube: 102. Dryer G (1973) High temperature oxidation of CO
effects of coal properties on deposit growth. Prog and CH4. In: 14 symbosium (international) on com-
Energy Combust Sci 16(4):327–345 bustion, The Combustion Institute, Pittsburgh,
85. Miller B (1985) Clean coal engineering technology, pp 987–1003
2nd edn. Butterworth-Heinemann Elsevier Ltd, 103. Miller B (1989) Mechanism and modeling of nitro-
Oxford gen chemistry in combustion. Prog Energy Combust
86. Raask E (1985) Mineral impurities in coal combus- Sci 15:287–338
tion: behavior, problems, and remedial measures. 104. Chen JY (1988) A general procedure for constructing
Hemisphere Publishing Corp, Washington, DC reduced reaction mechanisms with given indepen-
87. López C, Unterberger S, Maier J, Hein KRG dent relations. Combust Sci Technol 57(1–3):9–94
(2003) Overview of actual methods for characteriza- 105. Glassman I (1987) Combustion, 2nd edn. Academic,
tion of ash deposition. In: Heat exchanger fouling Orlando
and cleaning: fundamentals and applications, engi- 106. Fenimore CP (1971) Formation of nitric oxide in
neering conferences international, New York, NY premixed hydrocarbon flames. In: Proceedings of
88. Bryers RW (1996) Fireside slagging, fouling, and the 13th symposium (international) on combustion,
high-temperature corrosion of heat-transfer surface Elsevier, Amsterdam, pp 373–389
due to impurities in steam-raising fuels. Prog Energy 107. Pershing DW, Wendt JOL (1977) The influence of
Combust Sci 22(1):29–120 flame temperature and coal composition on thermal
89. Dumont BJ, Mudry RG (2003) Computational fluid and fuel NOx. In: Proceedings of the 16th symposium
dynamic modeling of electrostatic precipitators. In: (international) on combustion, pp 389–399
Electric power conference, Houston, TX 108. Sarofim AF, Beér JM (1990) The fate of fuel nitrogen
90. Choi BS, Fletcher CAJ (1997) Computation of parti- and ash during combustion of pulverized coal, chap-
cle transport in an electrostatic precipitator. ter 4. In: Lemieux PM (ed) Air and Energy Engineer-
J Electrost 40–41:413–418 ing Research Laboratory (eds) Pulverized coal
91. Guo B, Yu A, Guo J (2015) Numerical modelling of combustion: pollutant formation and control. U.-
ESP for design optimization. The 7th World Con- S. Environmental Protection Agency/Air and Energy
gress on Particle Technology (WCPT7). Procedia Engineering Research Laboratory, Research Triangle
Eng 102:1366–1372 Park, pp 1970–1980
92. Liu Q, Zhang S, Chen J (2015) Numerical analysis of 109. Hayhurst AN, Vince IM (1980) Nitric oxide forma-
charged particle collection in wire-plate ESP. tion from N2 in flames. Prog Energy Combust Sci
J Electrost 74:56–65 6:35–51. Elsevier, Amsterdam
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 499

110. Zel’dovich YB (1946) The oxidation of nitrogen in 125. Lyon R (1987) Thermal DeNOx: controlling nitrogen
combustion explosions. Acta Physicochim USSR oxides emissions by a noncatalytic process. Environ
21:577–628 Sci Technol 21(3):231
111. http://combustion.berkeley.edu/gri-mech/ 126. Blejchař T, Dolníčková D (2013) Numerical simula-
112. Perry S, Fletcher TH, Pugmire RJ, Solum MS tion of SNCR technology with simplified chemical
(2000) A global free-radical mechanism for light kinetics model. In: EPJ web of conferences,
gas nitrogen release from coal during 45, 01015, EDP Sciences
devolatilization. Energy Fuel 14:1094–1102 127. Javed M, Ahmed Z, Ibrahim M, Irfan N (2008)
113. Sung CJ, Law CK, Chen JY (2001) Augmented A comparative kinetic study of SNCR process using
reduced mechanisms for NO emission in methane ammonia. Braz J Chem Eng 25(1):109–117
oxidation. Combust Flame 125(1–2):906–919 128. Muzio L, Quartucy G (1993) State-of-the-art assess-
114. Lu T, Law CK (2008) A criterion based on computa- ment of SNCR technology. Topical report
tional singular perturbation for the identification of no. TR-102414. Electric Power Research Institute,
quasi steady state species: a reduced mechanism for Palo Alto
methane oxidation with NO chemistry. Combust 129. U.S. Department of Energy (2005) Selective catalytic
Flame 154(4):761–774 reduction (SCR) technology for the control of nitro-
115. Massias A, Diamantis D, Mastorakos E, Goussis gen oxide emissions from coal-fired boilers. Topical
D (1999) Global reduced mechanisms for methane report no 23. U.S. Department of Energy, Pittsburgh
and hydrogen combustion with nitric oxide formation 130. Forzatti P (2001) Present status and perspectives in
constructed with CSP data. Combust Theor Model de-NOx SCR catalysis. Appl Catal A Gen
3(2):233–257 222(1–2):221–236
116. Meadows ML (1997) Summary report control of 131. Yun BK, Kim MY (2013) Modeling the selective
NOx emissions by reburning. U.S. Environmental catalytic reduction of NOx by ammonia over a
Protection Agency, Washington, DC. EPA/625/R- Vanadia-based catalyst from heavy duty diesel
96/001 (NTIS 97-208201) exhaust gases. Appl Therm Eng 50:152–158
117. Wendt J, Sternling C, Matovich M (1973) Reduction 132. Dumesic JA, Topsøe NY, Topsøe H, Chen Y, Slabiak
of sulfur trioxide and nitrogen oxides by secondary T (1996) Kinetics of selective catalytic reduction of
fuel injection. In: 14th symposium (international) on nitric oxide by ammonia over Vanadia/Titania.
combustion, Elsevier, Amsterdam, pp 897–904 J Catal 163(2):409–417
118. Yagiela A, Maringo G, Newell R, Farzan H (1992) 133. Castaldini C, Angwin M (1977) Boiler design and
Demonstration of coal reburning for cyclone boiler operating variables affecting uncontrolled sulfur
NOx control. In: First annual clean coal technology emissions from pulverized-coal-fired steam genera-
conference, Cleveland tors. EPA-450/3-77-047. U.S. Environmental Protec-
119. Folsom B, Payne R, Sommer T, Engelhardt D, Ritz tion Agency, Research Triangle Park
H (1995) Demonstration of gas reburning-low NOx 134. Folkedahl BC, Zygarlicke CJ (2004) Sulfur retention
burner technology for cost-effective NOx emission in North Dakota lignite coal ash. Preprints of the
control. In: Fourth annual clean coal technology con- papers of the spring national meeting of the Division
ference, Denver of Fuel Chemistry 49(1):167–168. American Chem-
120. Folsom B, Sommer T, Englehardt D, O’Dea D, ical Society, Anaheim
Hunsicker S, Watts (1997) Coal reburning for cost- 135. Srivastava RK (2000) Controlling SO2 emissions: a
effective NOx compliance. In: 5th annual clean coal review of technologies. EPA/600/R-00/093. U.-
technology conference, Tampa S. Environmental Protection Agency, Office of
121. Savichky W, Gaufillet G, Mahlmeister M, Research and Development, Washington, DC
Englehardt D, Mereb J, Watts J (1998) Micronized 136. Kohl A, Riesenfeld F (1985) Gas purification,
coal reburning demonstration of NOx control. In: 6th 4th edn. Gulf, Houston
annual clean coal technology conference, Reno 137. Weiler H, Ellison W (1997) Wet gypsum-yielding
122. Cremer MA, Adams BR, O’Connor DC, FGD experience using quicklime reagent.
Bhamidipati VN, Broderick RG (2001) Design and TR-108683-V2. Electric Power Research Institute,
demonstration of rich reagent injection (RRI) for Palo Alto
NOx reduction at Conectiv’s B.L. England Station. 138. United Engineers and Constructors, Inc (1991) Eco-
US EPA/DOE/EPRI combined power plant air pol- nomic evaluation of flue gas desulfurization systems.
lutant control symposium: the Mega symposium, GS-7193, vol 1. Electric Power Research Institute,
Chicago, AWMA, Washington, DC Palo Alto
123. Wan HP, Yang CS, Adams BR, Chen SL (2008) Con- 139. United Engineers and Constructors, Inc (1992) Eco-
trolling LOI from coal reburning in a coal-fired nomic evaluation of flue gas desulfurization systems.
boiler. Fuel 87:290–296 GS-7193, vol 2. Electric Power Research Institute,
124. Wu KT, Lee HT, Juch CI, Wan HP, Shim HS, Adams Palo Alto
BR, Chen SL (2004) Study of syngas co-firing and 140. United Engineers and Constructors, Inc (1995) Eco-
reburning in a coal fired boiler. Fuel 83:1991–2000 nomic evaluation of flue gas desulfurization systems.
500 Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation

GS-7193-V3, vol 3. Electric Power Research Insti- absorption into a calcium hydroxide slurry in a
tute, Palo Alto spray dryer. Sep Sci Technol 25:1843–1855
141. Fox MR, Hunt TG (1990) Flue gas desulfurization 157. Ramachandran PA, Sharma MM (1969) Absorption
using dry sodium injection. Presented at the with fast reaction in a slurry containing sparingly
EPA/EPRI 1990 SO2 control symposium, New soluble fine particles. Chem Eng Sci 24:1681–1686
Orleans 158. Scala F, D’Ascenzo M (2002) Absorption with
142. Zhou W, Maly P, Brooks J, Nareddy S, Swanson L, instantaneous reaction in a droplet with sparingly
Moyeda D (2010) Design and test furnace sorbent soluble fines. AICHE J 48:1719–1726
injection for SO2 removal in a tangentially fired 159. Marocco L, Mora A (2013) CFD modeling of the
boiler. Environ Eng Sci 27(4):337–345. Mary Ann Dry-Sorbent-Injection process for flue gas desulfuri-
Liebert, New Rochelle zation using hydrated lime. Sep Purif Technol
143. Stouffer M, Withum J, Rosenhoover W, Maskew 108:205–214
J (1994) Advanced in-duct sorbent injection for 160. Schantz M, Sewell M (2013) The growth of dry
SO2 control. Final technical report, DOE/PC/90360- sorbent injection (DSI) and the impact on coal com-
49 bustion residue. In: 2013 World of Coal Ash
144. Arif A, Stephen C, Branken D, Everson R, (WOCA) conference, Center for Applied Energy
Neomagus H, Piketh S (2015) Modeling wet flue Research, Lexington, KY
gas desulfurization. In: Conference of the National 161. Muziak J (2005) Successful mitigation of SO3 by dry
Association for Clean Air (NACA 2015), Bloemfon- sorbent injection of trona upstream of the ESP. 2005
tein, South Africa EPA/DOE/EPRI mega symposium, Air and Waste
145. Neveux T, le Moullec Y (2011) Wet industrial flue Management Association, Pittsburgh, PA
gas desulfurization unit: model development and val- 162. (2016) Dry sorbent injection for acid gas control:
idation on industrial data. Ind Eng Chem Res process chemistry, waste disposal and plant opera-
50:7579–7592 tional impacts. White paper, Institute of Clean Air
146. Gómeza A, Fueyoa N, Tomás A (2007) Detailed Companies. Arlington
modelling of a flue-gas desulfurisation plant. Comput 163. Shi L, Liu G, Higgins BS, Benson L (2011) Compu-
Chem Eng 31(11):1419–1431 tational modeling of furnace sorbent injection for
147. Maroccoa L, Inzoli F (2009) Multiphase SO2 removal from coal-fired utility boilers. Fuel Pro-
Euler–Lagrange CFD simulation applied to wet flue cess Technol 92(3):372–378
gas desulphurisation technology. Int J Multiphase 164. Srivastava RK, Miller CA, Erickson C, Jambhekar
Flow 35(2):185–194 R (2004) Emissions of sulfur trioxide from coal-fired
148. Dou BL, Pan WG, Jin Q, Wang WH, Li Y (2009) power plants. J Air Waste Manag Assoc 54:750–762.
Prediction of SO2 removal efficiency for wet flue gas Air and Waste Management Association, Pittsburgh
desulfurization. Energy Convers Manag 50:2547–2553 165. Blythe G, Dombrowski K (2004) SO3 mitigation
149. Zhong Y, Gao X, Huo W, Luo ZY, Ni MJ, Cen KF guide update. Report no. 1004168. Electric Power
(2008) A model for performance optimization of wet Research Institute, Palo Alto
flue gas desulfurization systems of power plants. Fuel 166. Walsh PM, McCain JD, Cushing KM (2006) Evalua-
Process Technol 89:1025–1032 tion and mitigation of visible acidic aerosol plumes
150. Masters K (1985) Spray drying handbook, 4th edn. from coal fired power boilers. EPA/600/R-06/156
Wiley, New York 167. Senior C, Fry A, Montgomery C, Sarofim A, Wendt
151. Scala F, D'Ascenzo M, Lancia A (2004) Modeling J (2006) Modeling tool for evaluation of utility mer-
flue gas desulfurization by spray-dry absorption. Sep cury control strategies. Paper #23, DOE-EPRI-EPA-
Purif Technol 34(1–3):143–153 A&WMA Power plant air pollutant control “mega”
152. Partridge GP, Davis WT, Counce RM, Reed GD symposium, Washington, DC
(1990) A mechanistically based mathematical 168. Senior CL, Sarofim AF, Zeng T, Helble JJ, Mamani-
model of sulfur dioxide absorption into a calcium Paco R (2000) Gas-phase transformations of mercury
hydroxide slurry in a spray dryer. Chem Eng in coal-fired power plants. Fuel Process Technol
Commun 96:97–112 63(2–3):197–213
153. Newton GH, Kramlich J, Payne R (1990) Modeling 169. (2004) Control of mercury emissions from coal-fired
the SO2-slurry droplet reaction. AICHE electric utility boilers. Air Pollution Prevention and
J 36:1865–1872 Control Division, National Risk Management
154. Hill FF, Zank J (2000) Flue gas desulphurization by Research Laboratory, Office of Research and Devel-
spray dry absorption. Chem Eng Process 39:45–52 opment, U.S. Environmental Protection Agency,
155. Karlsson HT, Klingspor J (1987) Tentative modelling Research Triangle Park
of spray-dry scrubbing of SO2. Chem Eng Technol 170. (2014) Mercury emissions capture efficiency with
10:104–112 activated carbon injection at a Russian coal-fired
156. Dantuluri SR, Davis WT, Counce RM, Reed GD thermal power plant. Scientific report, EPA600/R-
(1990) Mathematical model of sulfur dioxide 14/29, US EPA, Washington, DC
Mitigation of Airborne Pollutants in Coal Combustion: Use of Simulation 501

171. Afonso RF, Senior CL (2001) Assessment of mer- native fly ash carbons in coal-fired power plants.
cury emissions from full scale power plants. Paper Prog Energy Combust Sci 36(4):510–529
presented at the EPRI-EPA-DOE-AWMA mega 181. Padak B, Wilcox J (2009) Understanding mercury
symposium and mercury conference, Chicago, binding on activated carbon. Carbon
AWMA, Washington, DC 47(12):2855–2864
172. Kilgroe JD, Sedman CB, Srivastava RK, Ryan JV, 182. Wilcox J, Rupp E, Ying SC, Lim DH, Suarez
Lee CW, Thornloe SA (2002) Control of mercury Negreira A, Kirchofer A, Feng F, Lee K (2012) Mer-
emissions from coal-fired electric utility boilers: cury adsorption and oxidation in coal combustion
interim report. EPA –600/R-01-109. National Risk and gasification processes. Int J Coal Geol
Management Laboratory, US EPA, Research Trian- 90–91:4–20
gle Park, NC 183. Niksa S, Helble JJ, Fujiwara N (2001) Kinetic model-
173. Laudal D (2002) Effect of selective catalytic reduc- ing of homogeneous mercury oxidation: the impor-
tion on mercury. 2002 field studies update, EPRI, tance of NO and H2O in predicting oxidation in coal-
Palo Alto. Product ID 1005558. derived systems. Environ Sci Technol
174. Chu P, Laudal D, Brickett L, Lee CW (2003) Power 35(18):3701–3706
plant evaluation of the effect of SCR technology on 184. Qiu J, Sterling RO, Helble JJ (2003) In development
mercury. Presented at the DOE-EPRI-U.S. EPA of an improved model for determining the effects of
-A&WMA combined power plant air pollutant control SO2 on homogeneous mercury oxidation. In: 28th
symposium – The mega symposium, Washington, DC international technical conference on coal utilization
175. La Marca C, Cioni M, Pintus N, Rossi N, Malloggi S, & fuel systems, Clearwater
Barbieri A (2003) Macro and micro-pollutant emis- 185. Senior CL, Sadler B, Sarofim AF (2005) Modeling
sion reduction in coal-fired power plant. In: Pre- mercury behavior in practical combustion systems.
sented at 7th international conference on energy for In: 229th American Chemical Society National
a clean environment (Clean Air 2003), Lisbon Meeting, San Diego
176. Gutberlet H, Schlüter A, Licata A (2000) SCR 186. Senior CL (2006) Oxidation of mercury across SCR
impacts on mercury emissions from coal-fired catalysts in coal-fired power plants. J Air Waste
boilers. Presented at EPRI SCR workshop, memphis Manag Assoc 56:23–31
177. Sjostrom S, Durham M, Bustard CJ, Martin C (2010) 187. Understanding mercury chemistry via the reaction
Activated carbon injection for mercury control: over- engineering international (REI) ProMerc™ model.
view. Fuel 89(6):1320–1322 EPRI report 1014893
178. Looney B, Meeks N, Cecil J, Huston R, Wong J, 188. Adams B, Van Otten B (2014) Evaluation of mercury
Johnson E (2013) Advanced activated carbons for control strategies in the presence of SO3. Paper #18,
efficient solutions. Presented at EUEC presentation power plant pollutant control “mega” symposium,
179. Cremer M, Senior C, Chiodo A, Wang D, Valentine Baltimore
J (2004) CFD modeling of activated carbon injection 189. Ciferno JD, Fout TE, Jones AP, Murphy JT
for mercury control in coal fired power plants. In: (2009) Capture carbon from existing coal-fired
Joint EPRI DOE EPA combined utility air pollution power plants, chemical engineering progress. Amer-
control symposium, the mega symposium, AWMA, ican Institute of Chemical Engineers, New York,
Washington, DC pp 33–41
180. Hower JC, Senior CL, Suuberg EM, Hurt RH, 190. https://www.netl.doe.gov/research/coal/energy-syste
Wilcox JL, Olson ES (2010) Mercury capture by ms/advanced-combustion/project-portfolio
 oxygen-rich or pure oxygen environment for
CO2 Capture and generating high-concentration CO2 in flue gas.
Sequestration Precombustion capture Separation of CO2
from multicomponent gas streams that are the
Abhoyjit S. Bhown1, Grant Bromhal2 and products of fuel conversion (e.g., gasification,
Gabriel Barki3 methanation, or fermentation) before combus-
1
Electric Power Research Institute, Palo Alto, tion, commonly through physical solvents.
CA, USA Post-combustion capture Separation of CO2
2
US DOE National Energy Technology from multicomponent gas streams that are the
Labaratory, Pittsburgh, PA, USA products of hydrocarbon combustion, com-
3
NETL’s Mission Execution and Strategic monly through chemical solvents, sorbents, or
Analysis Site Support Contract, KeyLogic selective membranes.
Systems, Pittsburgh, PA, USA

Definition of Subject
Article Outline
Carbon capture and storage (CCS) sometimes
Glossary
called carbon capture and sequestration is the
Definition of Subject
long-term isolation of carbon dioxide from the
Introduction
atmosphere through physical, chemical, biologi-
Carbon Capture
cal, or engineered processes. The purpose is to
Bibliography
reduce anthropogenic CO2 emissions, primarily
from fossil fuel usage. CCS can be a cost-effective
Glossary
partial solution to climate change, accounting for
10–15% of the needed abatement to limit global
Carbon capture and storage (CCS) sometimes
warming to less than 2  C above pre-industrial
referred to as carbon capture and
levels [1]. CCS should be considered as part of a
sequestration The long-term isolation of car-
portfolio of solutions which include efficiency,
bon dioxide from the atmosphere through
conservation, renewable energy, nuclear power,
physical, chemical, biological, or engineered
and other options.
processes.
CO2 capture The separation and concentration
of CO2 from multicomponent gas streams.
CO2-enhanced oil recovery (EOR) Injection of Introduction
CO2 into depleted oil fields for the purpose of
increasing oil production. Depending on geologic In the broadest sense, CCS includes a range of
formations, this can result in long-term storage approaches including soil carbon sequestration
of CO2. (e.g., through no-till farming), terrestrial biomass
Geologic storage The long-term physical or sequestration (e.g., through planting forests),
chemical storage of CO2 in deep geological direct ocean injection of CO2 either onto the
formations to isolate it indefinitely from the deep seafloor or into the intermediate depths,
atmosphere. injection into deep geological formations, or
Oxyfired combustion The burning of a hydro- even direct conversion of CO2 to carbonate min-
carbon fuel (coal, oil, gas, biofuel) in an erals. Some of these approaches are considered

© Springer Science+Business Media, LLC, part of Springer Nature 2020 503

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_106
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2019, https://doi.org/10.1007/978-1-4939-2493-6_106-3
504 CO2 Capture and Sequestration

geoengineering (see the appropriate entry herein). coupled. The first step, commonly called “carbon
All are considered in the 2005 special report by the capture,” is the separation of carbon dioxide from
Intergovernmental Panel on Climate Change [2]. industrial flue streams, including power plants. The
Of the range of options available, CCS most second step is compression of the separated CO2 to
commonly entails the capture of CO2 from power a pressure required to transport, use, or otherwise
and industrial plants followed by injection into store the CO2. In practice, capture and compression
deep geological formation (geological carbon are often discussed together since both processes
storage or sequestration, or GCS). In many are needed in order to render the CO2 in the state
ways, this form of CCS appears to be a critical it’s ultimately needed. The third step is the trans-
option for major greenhouse gas reduction in the port of the CO2 via pipelines, ships, or other means
next few years. The basis for this interest includes: to the final site. In some cases, transport may not be
needed if the final site is the same location as the
• There is no obvious immediate technical bar- carbon capture process. The fourth and final step is
rier to deployment. Systems to capture and the utilization or storage of CO2. For CO2 storage
concentrate CO2 are well-understood and in deep geological formations, CO2 is injected as a
widely used commercially [3, 5], and the oil dense, supercritical (liquid-like) phase. This step is
and gas industry has injected CO2 for enhanced commonly called geological carbon sequestration
oil recovery for over 45 years. or GCS. If CO2 is utilized in another process,
• Most countries across the globe with high CO2 instead of stored, the CCS chain is sometimes
emissions have published GCS storage esti- written as CCU or CCUS. Utilization options,
mates much higher than what is needed to however, are small relative to anthropogenic CO2
store expected future emissions [4]. emissions [9], though it could provide an important
market opportunity for the first deployment of car-
Testing of large-scale GCS is feasible and has bon capture technologies. Enhanced oil recovery
begun. In many industrialized countries, large CO2 (EOR) is one of the larger utilization options that
sources like power plants and refineries lie near has been key to deploying some of largest carbon
prospective storage sites. These plants could be capture plants thus far [10].
retrofit today, and injection begun (while bearing Typically, CCS requires the separation of CO2
in mind scientific uncertainties and unknowns). from industrial flue streams and concentration to
Part of this interest comes from several key CO2 purities of 95% or greater [5, 6]. This limits
documents written in the last few years that pro- compression costs and makes effective and effi-
vide information on the status, economics, tech- cient use of available sequestration resource
nology, and impact of CCS. (subsurface pore volume). Currently, several tech-
When coupled with improvements in energy nology pathways exist for CO2 capture and sepa-
efficiency, renewable energy supplies, fuel ration (Fig. 1):
switching, and nuclear power, CCS helps dramati-
cally reduce current and future emissions [5–7]. If • Post-combustion capture: This involves sepa-
CCS is not available as a carbon management ration of CO2 from nitrogen and other gases in
option, it will be much more difficult and much air, commonly with chemical sorbents (e.g.,
more expensive to stabilize atmospheric CO2 emis- aqueous monoethanolamine (MEA)).
sions. Estimates of the cost of carbon abatement to • Pre-combustion capture: This involves con-
450 ppm CO2-eq without CCS are 138% higher version of fuel feedstocks (e.g., coal) into syn-
than if CCS were to be available [8]. gas via gasification, steam reformation, or
partial oxidation, then shifting the syngas
chemically to hydrogen and CO2, and then
Carbon Capture separating the H2 from CO2. Currently, this
last step is commonly done with solvents.
CCS is a series of sequential steps. While each step • Oxy-combustion: This involves combustion of
may be considered separately, they are serially fuels in a pure oxygen or O2-CO2-rich
CO2 Capture and Sequestration 505

N2
O2
Coal
Post combustion Gas CO2
Power & Heat
Separation
Biomass
Air CO2

Coal
Air/O2
Gas CO2
Steam
Biomass

Pre combustion H2
Gasification Reformer N2 O2
+CO2 sep Power &
Gas, Oil Heat CO2
Compression
Air & Dehydration
Coal CO2
Oxyfuel Gas Power & Heat
Biomass
O2
N2
Air Air Separation
Coal Gas CO2
By-product Biomass
Chemical SNG, liquids,
Gasification
Process methanol, NH3

CO2 Capture and Sequestration, Fig. 1 Schematic diagram of the four main CO2 capture pathways and the main
processes involved (After Thambimuthu et al. [12])

environment such that almost no nitrogen is Each of these approaches requires varying
present in the flue gas. Separation of O2 from energy and cost to operate [5, 12, 13].
air (N2) is required and is the main cost Industry has experience with each of these
element. technology pathways, chiefly from operation of
• Chemical looping combustion. This involves hydrogen plants, fertilizer plants, refineries, and
oxidizing a metal in air in one reactor, trans- natural gas processing facilities. CO2 has been
porting it to another reactor and reducing it separated from industrial flue streams at scales
with a hydrocarbon fuel to release CO2 and much greater than 1 Mt. CO2/year (270,000 t
water. The reduced metal is then sent back to C/year). Similarly, CO2 has been separated from
the first reactor. Both reactors are operated as small-scale power plants for decades, and some of
fluidized bed reactors, and the exothermic air these have been scaled to 100–150 MWe
reactor provides high temperature heat that can [10]. Large pipelines transport millions of tons
be converted to work [11]. of CO2 across hundreds of kilometers, and mil-
• Direct capture of high-purity streams: In these lions of tons of CO2 and other acid gases are
cases, CO2 is produced or generated in a compressed and injected into geological forma-
chemical or industrial process and already tions every year. Thus, a great deal is known about
at or above 95% purity. This requires carbon capture, separation, and transportation,
only compression and transportation before and many countries have regulatory frameworks
sequestration. in place to accommodate the permitting of sepa-
• Direct air capture: This involves separation of ration facilities and pipelines.
CO2 directly from the atmosphere. The very While small-scale carbon capture processes are
low concentrations of CO2 in air (400 ppm) commercial for utilization, the scale of anthropo-
make this approach highly energy intensive. genic emissions is orders of magnitude larger.
506 CO2 Capture and Sequestration

Hence, the lack of a market or regulatory driver • Solid sorbents: In this process, selective solid
remains an important barrier to wider commercial materials adsorb CO2 from mixed gas streams.
deployment for CCS [5, 14]. However, as the These are released under different physical or
concepts for geological carbon sequestration are chemical conditions, typically through heating
proven to be reliable for current power plant tech- the adsorbent or altering the pressure of the gas
nology, improved power plant designs are surrounding it.
expected to be able to bring down CCS costs • Phase separation: At sufficiently low temper-
dramatically. Several large pilot projects are test- atures, CO2 can condense from a gas mixture
ing pre-, post-, and oxyfired combustion tests at and can thus be physically separated from the
the 2–30 MW scale, a necessary precursor to gas mixture. Examples include cryogenic
broad commercial deployment. These are methods and supersonic gas expansion [18].
expected to bring CCS costs down from a current
estimate of ~$70/t to ~$40/t and lower. Further The cost and performance of different systems
R&D is needed to pursue such pathways and vary as a function of loading, energy costs, capital
possibly reduce costs even further. costs, vapor pressure, process efficiency, and
other issues. While some processes appear to be
Post-combustion Capture highly efficient, they may have cost or volumetric
As the name suggests, post-combustion capture issues which limit deployment. Many prospective
(PCC) involves capture and separation of CO2 processes remain at the bench scale and have not
from flue gas streams after combustion [5, 12]. It been tested in large-scale pilots or commercial
can also apply to capture from high- or low- plants.
concentration streams from industrial sources Today, most post-combustion capture is
(e.g., cement manufacturing). Most PCC studies performed by liquid solvents, chiefly aqueous
have focused on coal and gas power plants, both amines. This is in large part due to the familiarity
because of their central role in global power pro- of the technology patented in 1930 [19] and the
duction and power sector is one of the largest CO2 conventional equipment set used, such as gas
emitters. Typically, coal-fired power plants have stripper towers and thermal regeneration. In addi-
~12–15% CO2, while gas-fired power plants have tion, improvements in both amine chemistry and
~3–4% CO2. The balance of the gas mixture is process development have led to lowering of costs
typically nitrogen, oxygen, argon, and other over the past few decades [3].
minor components. Aqueous amines have thus become the de facto
Typically, CO2 capture involves one of these standard against which other newer capture tech-
separation processes: nologies are judged. Two large-scale amine cap-
ture systems have been deployed at commercial
• Chemical solvents: This process operates when power plants [20], and a host of other capture
flue gas contacts liquid solvents and CO2 selec- technologies is currently under development at
tively absorbs into the liquid. The CO2-rich various scales.
solvent is typically heated to release the CO2
and reconstitute the solvent, which is then
recirculated. Pre-combustion Capture
• Membranes: This process separates CO2 from Precombustion capture involves the removal of
mixed gases as CO2 preferentially transports CO2 after a hydrocarbon fuel such as coal or
across a selective membrane which rejects natural gas is converted into syngas but before
other gases. Membranes can be made of poly- combustion in a turbine or boiler. Typically, the
mers [15], ceramics [16], or more exotic mate- first step is to oxidize the carbon and make syngas
rials (e.g., [17]). Costs and viability are a and some CO2. This CO2 may be separated at this
function of gas selectivity, permeance, and stage for partial capture (typically around 20% of
membrane lifetime, among other factors. carbon content). To achieve higher fractional
CO2 Capture and Sequestration 507

capture, a water-gas shift reactor converts carbon has focused on alternative separation technolo-
monoxide and steam to CO2 and hydrogen. This gies, including hydrate formation, ceramic mem-
increases the concentration of CO2 which is then branes, or exotic materials. Alternative
removed using either a chemical or a physical gasification technologies such as molten metal
solvent, such as Selexol™ or Rectisol™, and is gasifiers [27] or underground coal gasification
compressed. This approach is used widely where [28] remain potential avenues to large cost reduc-
coal is used for chemical feedstock. tions for precombustion capture-based plant
For power production applications, the rem- designs.
nant hydrogen burns in a turbine to generate elec-
tricity [5]. The likeliest configuration for these Oxy-combustion
systems is an integrated gasification combined The burning of a hydrocarbon fuel (coal, oil, gas,
cycle (IGCC) power plant. While both IGCC biofuel) in an oxygen-rich or pure oxygen envi-
and precombustion CO2 capture technologies are ronment for the purpose of CO2 concentration or
available today, the costs of the power generation capture is called oxy-combustion.
block limit commercial deployment. About One of the potential challenges with oxy-
10,000 MWth of syngas, derived predominantly combustion is that many hydrocarbon fuels will
from coal, are being used for power production at burn extremely hot in a pure oxygen environment,
the end of 2016 [21, 22]. making retrofits difficult and new reactors expen-
Overall, because CO2 is generated at higher sive. One common approach to this problem is to
pressure in the power plant, CCS on IGCC plants mix oxygen with CO2 at near atmospheric oxygen
have lower energy penalties than other kinds of concentrations (e.g., 25% oxygen and 75% CO2)
plants with carbon capture [24]. The commercial creating a “synthetic air.” The CO2 is recirculated
availability of gasifiers, water-gas shift reactors, after separation and acts to moderate the temper-
solvent towers, and gas turbines also provides ature and kinetics of the fuel combustion in the
some commercial advantages. However, the total reactor. This approach is considered promising for
capital costs for IGCC systems with CCS remain retrofits, both in the power sector (subcritical and
higher than conventional plants with post- supercritical pulverized coal boilers) and in the
combustion carbon capture. industrial sector (e.g., in catalytic crackers com-
Other precombustion plant designs show mon in hydrocarbon refining).
promise. For example, hydrogen and syngas An alternative approach is to reduce the
from gasification can also be used to make amount of fuel in the oxygen by staging the com-
chemicals such as ammonia, urea, or olefins. bustion. In this process, the fuel is introduced in
Coal-to-chemical plants can use some hydrogen stages, such that the temperature is controlled at
like an IGCC plant and generate power as well. each stage. Each stage is also operated at elevated
These polygeneration plants show some advan- pressure which partially offsets the amount of
tages in terms of economic return (e.g., [25]). CO2 compression after the last stage.
Several US projects, including the Texas Clean The primary cost of oxyfired combustion is the
Energy Project [26] and HECA project [23], are cost of oxygen separation from air. The air sepa-
attempts to manage costs and create economic ration unit (ASU) requires substantial capital and
return for precombustion plants through operating costs. There are also some specific tech-
polygeneration. nical considerations associated with the operating
Given the high efficiency of physical solvents pressure of the combustion unit. If the reactor
for precombustion, little research has focused on burns its fuel at lower than atmospheric pressure
solvent-based precombustion technology (a common configuration for boilers), then there is
improvement. Much research has focused on a risk of air leaking into the reactor and contami-
improving efficiency or reducing costs of gasifi- nating the stream with nitrogen.
cation or on better plant integration to improve Chemical looping [11] is considered an
overall efficiency and cost. Additional research oxyfiring variant, although in most ways the
508 CO2 Capture and Sequestration

technology is radically different from synthetic air of CCS projects across the globe, including over
combustion. Instead, a metal is oxidized, creating 16 million metric tons in the United States as part
heat which can be used to run a steam cycle. Most of DOE’s Clean Coal Research, Development,
commonly, the process is configured with two and Demonstration Programs [4]. The most prom-
interconnected fluidized bed reactors: an air reac- ising reservoirs are porous and permeable rock
tor and a fuel reactor. The solid oxygen carrier is bodies at depth (Figs. 2 and 3).
circulated between the air and fuel reactors, and
the metal oxide is reduced in the second stage. • Saline formations contain brine in their pore
This process remains promising, in large part due volumes, commonly with salinities greater
to the very high theoretical conversion efficiencies than 10,000 ppm. Because CO2 is buoyant in
(as high as 80%). However, no reactors larger than most geological settings, target saline forma-
benchtop have proven viable yet, and questions tions require a cap rock or sealing unit above
remain about the operation, longevity, and stabil- the main injection zone.
ity of the metal oxide carriers. • Depleted oil and gas fields have some combi-
nation of water and hydrocarbons in their pore
volumes. In some cases, economic gains can be
Direct Air Capture (DAC)
achieved through enhanced oil recovery or
The key technical concern for direct air capture is
enhanced gas recovery [8, 30, 31]. Substantial
the low partial pressure of CO2 in air due to its low
CO2-enhanced oil recovery already occurs in
concentration (~400 ppm). Any appreciable
the United States with both natural and anthro-
removal of CO2 thus requires the movement of
pogenic CO2. These fields provide much of the
very large volumes of air through a capture unit,
knowledge base available about the potential
and the dilute concentration requires high energy
issues related to CO2 sequestration.
to execute the separation. The American Physical
• Deep coal seams, often called unmineable coal
Society Panel on Public Affairs [29] completed a
seams, are composed of organic minerals with
fairly comprehensive assessment of the 2009 state
brines and gases in their pore and fracture vol-
of the art. The key messages of the report include
umes that can preferentially adsorb and bind
three conclusions:
CO2 as well as store it in pores and minor
fractures. These targets present some challenges
• DAC is not an economically viable approach.
in that coals are relatively low permeability
• Generally, low-carbon power and energy are
units, presenting challenges to injection.
best used directly to minimize losses rather
than used for DAC.
Because of their large storage potential and
• Eventually, DAC could play a role in capture
broad distribution, it is likely that most geological
and storage from decentralized sources of CO2,
sequestration will occur in saline formations.
such as vehicles, ships, or planes.
However, initial projects have been proposed for
depleted oil and gas fields, accompanying
While the situation may change with CO2 enhanced oil recovery, due to the high density
demand and price, the consensus today is that
and quality of subsurface data and the potential
the economics and engineering of DAC remain for economic return. Although there remains
formidable and difficult to overcome relative to
some economic potential for enhanced coal bed
removing CO2 from more concentrated streams.
methane recovery, much less is known about this
style of sequestration [2, 32–34]. Even less is
Geological Sequestration known about sequestration in basalts. As such,
A number of geological reservoirs appear to have many are not convinced of the economic viability
the potential to store many hundreds to thousands of sequestration projects in coal, basalts, or oil
of Gt of CO2 [2]. In the past 15 years, tens of shales given today’s technology and
millions of tons of CO2 have been injected as part understanding [5].
CO2 Capture and Sequestration 509

Geolog
Geologiica
log ical Stor
ical Stor
torage
age Op
Optio
tions
tio nss for
for
o CO2 Produced oil or gas

1 Depleted oil and gas reservoirs injected CO2

2 Use of CO2 in enhanced oil recovery Stored CO2
3 Deep unused saline water-saturated reservoir rocks
4 Deep unmineable coal seams
5 Use of CO2 in enhanced coal bed methane recovery 6
6 Other suggested options (basalts, oil shales, cavities) 1 4 5

3
2

1 km

2 km

CO2 Capture and Sequestration, Fig. 2 Options for storing CO2 in underground geological formations (After [30])

CO2 Capture and Sequestration, Fig. 3 Schematic dia- whereas the lower unit is homogeneous. Central insets
gram of large injection at 10 years time illustrating the main show CO2 as a mobile phase (lower) and as a trapped
storage mechanisms. All CO2 plumes (yellow) are trapped residual phase (upper). Right insets show CO2 dissolution
beneath impermeable shales (not shown). The upper unit is (upper) and CO2 mineralization (lower) (After [5])
heterogeneous with a low net percent usable porosity,
510 CO2 Capture and Sequestration

Storage of large CO2 volumes in geological dissolution of minerals generally traps CO2 as an
formations requires that the CO2 be relatively ionic species (usually bicarbonate) in the pore
dense, so that storage capacity is efficiently used. fluid.
Given typical geothermal gradients and hydro- Although work remains to characterize and
static loads, CO2 is likely to be in a supercritical quantify these mechanisms, the current level of
state at most target sites greater than 800 m depth understanding can be used today to develop esti-
(e.g., [35]). At the likely range of injection pres- mates of the percentage of CO2 that can be stored
sures and temperatures for most projects, CO2 over some period of time. Confidence in these
would be buoyant relative to the in situ brine. estimates is bolstered by studies of hydrocarbon
Consequently, trapping mechanisms are systems, natural gas storage operations, hazard-
needed to store CO2 effectively. For depleted oil ous waste injection, and CO2-enhanced oil recov-
and gas fields or for saline formations, CO2 stor- ery (CO2-EOR). Current evaluations of CCS
age mechanisms are reasonably well defined and effectiveness based on the current understanding
understood. CO2 sequestration targets will require of trapping mechanisms estimate that more than
physical barriers to CO2 migration out of the crust 99.9% of injected CO2 can be reliably stored over
to the surface. These barriers will commonly take 100 years, and it is likely that 99% of CO2 can be
the form of impermeable layers (e.g., shales, evap- reliably stored for 1,000 years [2]. These esti-
orites) overlying the sequestration target. This mates assume careful siting, due diligence before
storage mechanism is highly or directly analogous injection, and appropriate management of injec-
to that of hydrocarbon trapping, natural gas stor- tion and reflect the view that the crust contains
age, and natural CO2 accumulations. At the pore many sites that are generally well configured to
scale, capillary forces can immobilize a substan- store CO2 effectively.
tial fraction of a dispersed CO2 bubble, commonly
measured to be between 5% and 25% of the CO2- Large-Scale Commercial Deployment
bearing pore volume. The volume of CO2 trapped Many countries across the world have invested
as a residual phase is highly sensitive to pore substantial funding in technologies to ensure
geometry and consequently is difficult to predict; safe and effective carbon storage. The first large-
however, standard techniques can measure resid- scale carbon storage operation was the Sleipner
ual phase trapping directly in the laboratory with field, where Equinor (formerly Statoil) has
rock samples. injected approximately one million tons of CO2
Once in the pore volume, the CO2 will dissolve per year into the Utsira saline formation below the
into other pore fluids, including hydrocarbon spe- North Sea since 1996 [36]. The US Department of
cies (oil and gas) or brines. Depending on the fluid Energy’s Office of Fossil Energy and the National
composition and reservoir condition, this may Energy Technology Laboratory have supported a
occur rapidly (seconds to minutes) or over a variety of research programs, including large-
period of tens to hundreds of years. Once scale commercial deployment initiatives, with a
dissolved, the CO2-bearing brines are denser goal of enabling widespread commercial deploy-
than the original brines, and so the strong buoyant ment by 2025–2035 [37]. The EU’s New
forces of free phase gas are replaced by small Entrants’ Reserve (NER) 300 program is a two-
downward forces. Over longer time scales phase research and development initiative meant
(hundreds to thousands of years), the dissolved to catalyze CCS demonstration projects [38]. In
CO2 may react with minerals in the rock volume the past decade, a number of commercial-scale
to dissolve or precipitate new carbonate minerals. GCS projects have been initiated. In order to
For the majority of the rock volume and major achieve substantial GHG reductions, geological
minerals, this process is slow and may take hun- storage deployment has several requirements:
dreds to thousands of years to achieve substantial
storage volumes. Precipitation of carbonate min- • Projects must be large in scale, roughly in the
erals permanently binds CO2 in the subsurface; order of 1 Mt/year CO2 or more.
CO2 Capture and Sequestration 511

• There must be minimal leakage from the of the Illinois Industrial Carbon Capture and Stor-
underground storage reservoirs back to the age (Illinois ICCS) project, the effort offers an
atmosphere. opportunity to collect scientific and engineering
• There must be minimal impact on other uses of data of large-scale CO2 storage in saline forma-
the subsurface environment and the resources tions, including the injection of approximately
it contains. one million tons of CO2 annually [41].

The issue of scale dominates deployment of GCS Boundary Dam

[5, 39]. These volumes would have geological car- In 2014 SaskPower’s Boundary Dam Carbon
bon sequestration, providing 25–75 Gt C over Capture Project became the first power station in
50 years or 15–43% of emissions reduction needed the world to effectively use CCS technology at
to stabilize atmospheric CO2 levels at 550 ppm [39]. scale. Located in Canada, the coal-burning
Below are summaries of three large-scale Boundary Dam Unit #3 CCS project captures up
injection projects, some of which have an ambi- to 1.3 million tons of CO2 per year, reducing SO2
tious scientific program that includes monitoring emissions by up to 100% and CO2 by 90%
and verification. [42]. The captured carbon dioxide is injected
both in oil fields in the region and a saline-filled
sandstone aquifer through the Petroleum Technol-
Petra Nova Parish Holdings, LLC
ogy Research Centre’s Aquistore Project [42].
The Petra Nova Parish Holdings project offers the
potential to significantly reduce CO2 emissions
Worldwide Large-Scale CO2 Sequestration
from existing coal-fueled electric power genera-
Projects
tion by demonstrating a combined commercial-
Two well-maintained databases house and update
scale post-combustion carbon capture technology
project information and statistics of large-scale
and EOR from the WA Parish Generating Sta-
CO2 sequestration projects from around the
tion’s coal-fired plant to the West Ranch oil field.
world. The US Department of Energy’s National
The project’s specific objectives include
Energy Technology Laboratory maintains the
(1) remove CO2 from treated flue gas from an
Carbon Capture and Storage Database, which
existing coal-fired electrical generating station
provides information on active, proposed, and
and (2) compress and transport pipeline-quality
terminated CCS projects worldwide, and main-
CO2 to a storage site for use in EOR [40].
tains a listing of large-scale projects at various
The proposed benefits include the capture and
stages of deployment [31]. The database can be
storage of up to 1.4 million metric tons of CO2 per
found at https://www.netl.doe.gov/coal/carbon-
year from a flue gas stream, increased oil production
storage/worldwide-ccs-database. Table 1 provides
from CO2-enhanced oil recovery, and the incorpo-
a subset of some current and pending international
ration of CO2 capture technologies within existing
CO2 injection projects. The Global CCS Institute
coal-fueled power generation plants, which allows
maintains a similar database, the CCS Facilities
them to continue producing energy while meeting
Database (https://www.globalccsinstitute.com/re
environmental sustainability goals [40]. The project
sources/ccs-database-public/) [43].
has been operating since January 2017 [40].
Today there are over a dozen large-scale injec-
tion projects, some of which have an ambitious
Archer Daniels Midland Company scientific program that includes monitoring and
The Archer Daniels Midland Company is testing verification. Each project has injected CO2 at the
large-scale industrial CCS technologies through rate of 1 Mt CO2/year (280,000 t C/year).
the development and demonstration of an inte- Each project has had a substantial supporting sci-
grated system in the processing and transportation ence program or anticipates one.
of CO2 from an ethanol plant to saline reservoir These projects have sampled a wide array of
within the Mt. Simon Sandstone formation. Part geology with varying trapping mechanisms,
512 CO2 Capture and Sequestration

CO2 Capture and Sequestration, Table 1 Some current and pending large CO2 injection projects
Overall Plant Amount of Project
Project name status status Country location State location CO2 stored Unit start date
Weyburn-Midale Project Completed Existing Canada Saskatchewan 40,000,000 tons 10/1/2000
PEMEX CCS Project Active Existing Mexico 8,459,550 tons 1/1/2004
SECARB Development Completed Existing United States Mississippi 4,743,898 tons 4/1/2009
Phase – Cranfield Project
Bent County IGCC Plant Hold Planned United States Colorado 4,200,000 tons 1/1/2016
Big Sky Development Active United States Montana 1000,000 tons 7/25/2011
Phase – Kevin Dome
Project
MRCSP Development Active Existing United States Michigan 1,000,000 tons 10/1/2005
Phase – Michigan Basin
Project
SWP Development Phase – Active Existing United States Texas 1,000,000 tons 10/1/2007
Farnsworth Unit Ochiltree
Project
MGSC Development Completed Existing United States Illinois 999,215 tons 1/1/2010
Phase – Illinois Basin
Decatur Project
Purdy, Sho-Vel-Tum EOR Active Existing United States Oklahoma 675,000 tons 1/1/2003
Project
SECARB Validation Completed United States Mississippi 627,744 tons 7/15/2008
Phase – Stacked Storage
Test
Salt Creek, Monell, Sussex Active United States Wyoming 5,480 tons/ 1/1/2004
Unit EOR day
In Salah Gas Storage Hold Hold Algeria 3,288 tons/ 8/1/2004
Project day
Boundary Dam Integrated Active Existing Canada Saskatchewan 2,740 tons/ 1/1/2014
CCS Project day
Sleipner Project Active Existing Norway 2,740 tons/ 8/1/1996
day

injection depths, reservoir types, and injectivity. with long-established Class II regulations, while
Each of these projects appears to have ample saline projects operate under the recent and more
injectivity and capacity for success, and none stringent monitoring requirements for Class VI
has detected CO2 leakage of any significance. In wells [32].
addition to the sequestration projects, many indus- It is worth noting that many of these projects
trial applications have injected large volumes of have come online recently in China, which has the
CO2 into the subsurface. EOR operations in West potential to ramp up projects very rapidly
Texas, New Mexico, Colorado, Wyoming, Okla- [45–47]. It should also be noted that all of the
homa, Mississippi, Trinidad, Canada, and Turkey first projects do not involve capture from a
have individual injection programs as large as power plant but rather from industrial facilities
3 Mt CO2/year (820,000 t C/year) and cumula- where CO2 is available at a relatively low cost.
tive anthropogenic emission injections of 10 Mt
CO2/year (2.7 Mt C/t) [44]. It should be said that Science and Technology Status
the monitoring and verification program at each As discussed above, the knowledge of trapping
site varies substantially [5]. In the United States, mechanisms and large projects provides substan-
EOR projects are typically monitored consistent tial information. These are augmented by studies
CO2 Capture and Sequestration 513

of naturally occurring CO2 systems, [48], natural effectively execute key tasks around single large-
gas storage facilities, hazardous waste disposal, scale injection projects:
acid gas injection, and CO2-EOR [30]. This
knowledge provides a firm foundation for com- • Characterize a site.
mercial action and a nascent foundation for the • Design and operate the project.
development of regulation, standards [49], and • Monitor the CO2 injection.
legal frameworks for sequestration. GCS itself, • Mitigate problems that might arise.
however, drives study into specific technical and • Close and abandon the project.
scientific challenges associated with the central
elements of site characterization, selection, oper- Although this knowledge is currently being
ation, and monitoring [50]. Forward investigation brought to bear on specific injection projects
around these topics will enhance the technical and around the world, greater work is needed to codify
operational understanding of commercial GCS. and develop tools, regulations, and standards for
deployment of multiple million ton injections in
thousands of wells nationwide and worldwide
Monitoring and Verification (M&V)
across a range of geological settings. This affects
Monitoring and verification must detect and track
both sides of the deployment rubric over the pro-
CO2 in the deep subsurface near injection targets,
ject life cycle. Potential operators must execute a
in the shallow subsurface, and above ground.
set of tasks to prepare for and implement permit-
Monitoring and verification studies are a chief
ting and injection operations. Similarly, potential
focus of many applied research efforts. The US
regulators, investors, insurers, and public stake-
Department of Energy has defined M&V technol-
holders require information to make decisions.
ogy development, testing, and deployment as a
Part of the challenge is to provide a technical
key element to their technology roadmap [51]).
basis for each set of actors to make decisions
The National Energy Technology Laboratory has
concerning the minimal amount of information
published several versions of a best practice man-
needed to serve all stakeholders.
ual on Monitoring, Verification, and Accounting
While many possible goals and terms may be
(MVA) for Geologic Storage Projects, which
pursued in site characterization, it is difficult to
discussed tools, approaches, regulatory regimes,
imagine the success of a large-scale injection pro-
and lessons learned from application in the
ject without knowledge of three parameters.
field [52].
These are injectivity, capacity, and effectiveness
The European Union CO2ReMoVe effort was
[54]. In general terms, injectivity and capacity
dedicated to monitoring and verification, and the
may be estimated by conventional means, such
industry-led CO2 Capture Project continues to
as core analysis, regional and local structural and
study monitoring in commercial settings [53].
stratigraphic mapping, and simple multiphase
Some form of M&V will be required at commer-
fluid flow simulations. However, there are explicit
cial sites, but the extent of monitoring required by
standard measures of effectiveness. Ultimately,
regulators, operators, or financiers remains uncer-
characterizations must rely on estimates of
tain. Many geophysical and geochemical methods
geomechanical integrity, hydrodynamic stability,
are sufficiently well understood for them to be
and seal continuity for the rock system, fault sys-
used to make reasonable performance predictions
tem, and well system [55–59].
at candidate storage sites [50, 52].
Given this complexity, it has been difficult to
get broadly acceptable guidelines on what consti-
Deployment Challenges tutes effective storage. The World Resources
Despite the current gaps in sequestration science Institute [50] proposed a set of guidelines to
and technology, commercial projects have begun potential operators based on a working group of
and are ready to proceed with confidence in their over 40 experts ranging from geoscientists to law-
success. Today, enough is known to safely and yers and regulators. More recently, an
514 CO2 Capture and Sequestration

international group has been brought together to probability (frequency) and magnitude (severity)
establish standards for carbon storage based on [34]. Several risk assessment factors include
the knowledge from several large and small stor- (1) environment, (2) health and safety, (3) techni-
age projects to date, including monitoring and cal, (4) cost, (5) reputation, and (6) schedule [62].
verification, risk management, and regulatory An often-cited approach to help understand and
guidelines [60]. categorize the risks associated with GCS projects
is the features, events, and processes, or FEPs,
approach [62, 63].
Hazards Assessment and Risk Management
In the past 15 years, a great deal has been
Supercritical CO2 is buoyant and will seek the
learned about assessing and managing risks from
Earth’s surface; therefore, CO2 injection carries
carbon storage projects, bringing the state of the art
the possibility of leakage. Importantly, CO2 leak-
from qualitative risk assessment to semiquantita-
age risk will not be uniform across all sites; thus
tive and quantitative approaches, some of which
CO2 storage sites will have to demonstrate mini-
have been applied at commercial projects [64]. For
mal risk potential in their site characterization
example, the US DOE has funded the National
plans [50, 61]. Based on analogous experience in
Risk Assessment Partnership (NRAP) to develop
CO2 injection such as acid gas disposal and
tools and approaches to quantify risks and uncer-
enhanced oil recovery, these risks appear to be
tainties associated with fluid migration and ground
less than those of current oil and gas
motion [45]. Table 2, taken from Best Practices:
operations [32].
Risk Management and Simulation for Geologic
The direct hazards associated with geologic
Storage Projects [62], provides information on
sequestration fall into three distinct categories:
tools related to geologic risk assessment.
• Hazards associated with the release of the car-
Water Issues
bon dioxide to the Earth surface
A number of issues arise from CCS regarding
• Hazards associated with release into ground-
water use and quality. These include additional
water and subsequent degradation
water use due to process steam and additional
• Hazards associated with Earth movement
hear requirements in the plants, displacement of
caused by the injection process itself
in situ brines, and potential risks to groundwater
quality from unplanned CO2 leakage [66]. Addi-
The hazards themselves in turn are associated tional water needs for surface facilities are highly
with failure mechanisms and triggers. Potential trig- sensitive to capture process and plant type, rang-
gering of events associated with these hazards could ing from 20% to 110% additional water consump-
lead to undesired consequences. As such, it is an tion [67]. CO2 injected into saline formations
important goal to identify and understand these necessarily interacts with water and brines there.
hazards in order to avoid triggering hazard events. This interaction includes displacement of the
Identification and characterization of these hazards brines there, local drying of the reservoir, and
are the critical first steps to managing the risks at a dissolution of CO2 into brine forming carbon
site. They also serve as the basis for a quantitative acid. Early research focused on the potential
probabilistic risk assessment (PRA). A robust PRA risks and impacts of acid brine formation, ulti-
is still challenging due to substantial scientific and mately concluding that these risks are small.
technical gaps. However, the hazards associated Injection also creates a pressure transient in the
with a site can be identified, mapped, characterized, reservoir which increases through time. In the
and parameterized sufficiently to avoid failure or right context, this pressure can drive water
(alternatively) avoid selecting a poor site. co-production (enhanced water recovery) and
The management of risks involves analysis and even drive surface desalination processes [68,
evaluation of risk factors, as well as ranking and 69]. Based on preliminary calculations, it appears
prioritizing risks based on the assessment of that water co-production can cut the water
CO2 Capture and Sequestration 515

CO2 Capture and Sequestration, Table 2 A summary Bibliography

of geologic carbon storage risk assessment tools
Tool Methodology family 1. International Energy Agency (2018) Sustainable
Carbon Storage Scenario Qualitative, scenario-based development scenario [Online]. https://www.iea.org/
Identification Framework weo/weomodel/sds/
(CASSIF), TNO 2. IPCC (2005) Intergovernmental panel on climate
change. IPCC special report on carbon dioxide capture
Vulnerability Evaluation Qualitative and storage. Interlachen. https://www.ipcc.ch/
Framework (VEF), US 3. Rochelle GT (2009) Amine scrubbing for CO2 cap-
EPA ture. Science 325(5948):1652–1654
Screening and Ranking Qualitative, expert-elicited 4. Consoli CP, Wildgust N (2017) Current status of
Framework (SRF), LBNL probabilities global storage sites. Energy Procedia 114:4623–4628
CO2QUALSTORE Qualitative/ 5. MIT (2007) Future of coal in a carbon constrained
guideline, DNV semiquantitative, with world. MIT Press. http://coal.mit.edu/
“panel” inputs 6. US CCTP (2005) US climate change technology pro-
Quintessa FEP database Semiquantitative, FEPs gram strategic plan. Washington, DC, 256 p. http://
screened by experts www.climatetechnology.gov
TNO risk assessment Semiquantitative, expert- 7. Edmonds J, Clarke J, Dooley J, Kim SH, Smith SJ
methodology elicited probability and (2004) Stabilization of CO2 in a B2 world: insights on
consequence matrices the roles of carbon capture and disposal, hydrogen,
and transportation technologies. Energy Econ
Risk Identification and Semiquantitative, expert-
26(4):517–537, Special Issue EMF 19 Alternative
Strategy using elicited probability and
technology strategies for climate change policy, John
Quantitative Evaluation consequence matrices
P. Weyant, ed
(RISQUE), URS
8. Intergovernmental Panel on Climate Change (2014)
CarbonWorkFlow Semiquantitative, FEPs Climate change 2014: mitigation of climate change.
process for long-term ranked through expert Cambridge University Press, New York
CO2 storage elicitation using a risk 9. Bhown AS, Freeman BC (2011) Analysis and status of
matrix approach post-combustion carbon dioxide capture technologies.
Performance Assessment Quantitative, evidence- Environ Sci Technol 45:8624
(PA), Quintessa support (three-valued) 10. Large Scale CCS Facilities (2018) Global CCS Insti-
logic (ESL) Distinguishes tute. https://www.globalccsinstitute.com/projects/large-
cases of poor-quality data scale-ccs-projects. Accessed 2 July 2018
from uncertain data 11. Hossain MM, de Lasa HI (2008) Chemical-looping
CarbonSCORE software Quantitative, all evaluated combustion (CLC) for inherent CO2 separations – a
to pre-assess potential criteria are weighted, review. Chem Eng Sci 63:4433–4451
CO2 storage sites jointly evaluated, and 12. Thambimuthu K, Soltanieh M, Abanades JC
summarized (2005) Chapter 3: Capture. In: IPCC special report
Oxand Performance & Quantitative, risk matrix on carbon dioxide capture and storage. Intergovern-
Risk (P&RTM) evaluation mental Panel on Climate Change, Interlachen, pp 3–1,
Methodology 3–114. https://www.ipcc.ch
13. Rao AB, Rubin ES, Keith DW, Morgan MG
CO2-PENS, LANL Quantitative, hybrid
(2006) Evaluation of potential cost reductions from
system-process model
improved amine-based CO2 capture systems. Energy
NRAP-IAM-CS Quantitative, hybrid Policy 34:3765–3772
system-process model 14. Al-Juaied M, Whitmore A (2009) Realistic costs of
evolved from CO2-PENS carbon capture. Belfer center discussion paper 2009-
Certification Framework Quantitative, system-level 08. Harvard University, 73 p. http://belfercenter.ksg.
(CF), LBNL model, probabilities partly harvard.edu/files/2009_AlJuaied_Whitmore_Realistic_
calculated using fuzzy Costs_of_Carbon_Capture_web.pdf
logic 15. Hiroshi M, Hiroshi H (2000) Development of CO2
separation membranes: (1) polymer membrane.
Sumitomo Electric Tech Rev 157:22–26
16. Meinema HA, Dirrix RWJ, Brinkman HW, Terpstra
RA, Jekerle J, Kosters PH (2005) Ceramic membranes
demands for an integrated CCS plant by 50% or for gas separations – recent developments and state of
produce freshwater for agricultural or civil use. the art. Interceram 54:86–91
More work is needed to understand the viability of 17. Holt JK, Park G, Wang Y, Staderman M, Artyukhin
AB, Grigoropoulos CP, Noy A, Bakajin O (2006) Fast
this technical approach.
516 CO2 Capture and Sequestration

mass transport through sub-2-nanometer carbon nano- 32. Jarrell PM, Fox CE, Stein MH, Webb SL (2002) Prac-
tubes. Science 312:1034–1036 tical aspects of CO2 flooding. Monograph 22. Society
18. Berger AH, Wang Y, Bhown AS, Castrogiovai A, of Petroleum Engineers, Richardson
Kield R, Balepin V (2017) Thermodynamic analysis 33. Oldenburg CM, Stevens SH, Benson SM (2004) Eco-
of post-combustion inertial CO2 extraction system. nomic feasibility of carbon sequestration with enhanced
Energy Procedia 114:7 gas recovery (CSEGR). Energy 29:1413–1422
19. Bottoms RR (1930) Separating acid gases. U.S. Patent 34. Stevens S (1999) Sequestration of CO2 in depleted oil
1783901 and gas fields: barriers to overcome in implementation of
20. The Global Status of CCS: 2017 (2017) Global CCS CO2 capture and storage (disused oil and gas fields). IEA
Institute Greenhouse Gas R&D Programme. IEA/CON/98/31
21. Engineering-economic evaluations of advanced fossil 35. NPC (2007) Facing hard truths about energy: a com-
fuel power plants with carbon capture and storage. prehensive view to 2030 of global oil and gas.
EPRI, Palo Alto 2018. 3002013869 National Petroleum Council, Washington, DC,
22. Scientific American (2008) “Clean” coal power plant 442 pp. https://www.npc.org
cancelled – hydrogen economy, too. http://www.scientifi 36. Bachu S (2000) Sequestration of CO2 in geological
camerican.com/article.cfm?id=clean-coal-power-plant- media: criteria and approach for site selection in
canceled-hydrogen-economy-too response to climate change. Energy Convers Manag
23. MIT (2011) GreenGen fact sheet: carbon dioxide cap- 41:953–970
ture and storage project. MIT Energy Initiative. http:// 37. US DOE (2007) Basic Research needs for
sequestration.mit.edu/tools/projects/greengen.html geosciences: facilitating 21st century energy systems.
24. DOE-NETL (2007) U.S. Department of Energy, Department of Energy Office of Basic Energy Sci-
National Energy Technology Laboratory, Cost and ences, Washington, DC, 287 p. http://www.sc.doe.
performance baseline for fossil energy plants, gov/bes/reports/list.html
vol 1. Bituminous coal and natural gas to electricity, 38. UK CCS Commercial Scale Demonstration Programme
Revision 1, Aug 2007 (2010) Delivering projects 2–4. http://decarboni.se/sites/
25. Kreutz T, Williams R, Consonni S, Chiesa P (2005) default/files/publications/159703/UK-carbon-capture-
Co-production of hydrogen, electricity, and CO2 from storage-commercial-scale-demonstration-programme-
coal with commercially ready technology, part B: eco- delivering-projects-2-4.pdf
nomic analysis. Int J Hydrog Energy 30:769–784 39. Pacala S, Socolow R (2004) Stabilization wedges:
26. US DOE (2011) Innovative Texas clean coal project solving the climate problem for the next 50 years
takes major step forward as DOE issues record of using current technologies. Science 305:986–999
decision, press release. http://www.fe.doe.gov/news/ 40. U. S. DOE National Energy Technology Laboratory
techlines/2011/11053-Texas_Clean_Coal_Project_Mo (2019) Petra Nova Parish holdings: W.A. Parish Post-
ves_For.html Combustion CO2 capture and sequestration project.
27. Redman E, Fennerty K, Fowler M (2009) Mobilizing https://www.netl.doe.gov/sites/default/files/netl-file/
the next generation coal gasification technology for FE0003311.pdf
carbon capture and sequestration. In: Coal without 41. U. S. DOE National Energy Technology Laboratory
carbon, Clean Air Task Force, pp 1–16. https://www. (2019) Archer Daniels Midland Company: CO2 cap-
coaltransition.org ture from biofuels production and storage into the
28. Friedmann SJ (2009) Accelerating development of Mt. Simon Sandstone. https://www.netl.doe.gov/
underground coal gasification: priorities and chal- sites/default/files/netl-file/FE0001547-Factsheet.pdf
lenges for U.S. Research and Development. In: Coal 42. Plains CO2 Reduction Partnership (2019) SaskPower
without carbon, Clean Air Task Force, pp 1–1625. Boundary Dam Carbon Capture Project. http://www.
https://www.coaltransition.org. Solomon S et al undeerc.org/PCOR/CO2SequestrationProjects/SaskPo
(2009) Irreversible climate change due to carbon diox- wer.aspx?AspxAutoDetectCookieSupport=1
ide emissions. Proc Natl Acad Sci 106(6):1704–1709 43. Global CCS Institute (2018) CCS facilities database.
29. American Physical Society (2010) Direct air capture https://www.globalccsinstitute.com/resources/ccs-data
of CO2 with chemicals. APS, 119 p. http://www.aps. base-public/
org/policy/reports/popa-reports/loader.cfm?csModule= 44. Kuuskraa VA, DiPietro P, Koperna GJ (2006) CO2
security/getfile&PageID=244407 storage capacity in depleted and near-depleted US oil
30. Benson SM, Cook P (2005) Chapter 5: Underground and gas reservoirs. In: NETL 5th annual conference on
geological storage. In: IPCC special report on carbon carbon sequestration. ExchangeMonitor, Alexandria
dioxide capture and storage. Intergovernmental Panel 45. Asia Society (2009) Roadmap for US-China collabo-
on Climate Change, Interlachen, pp 5–1 to 5–134. ration on carbon capture and sequestration. Asia Soci-
https://www.ipcc.ch ety Press, 41 p. http://asiasociety.org/files/pdf/AS_
31. U.S. DOE National Energy Technology Laboratory CCS_TaskForceReport.pdf
(2019) Carbon capture and storage database. https:// 46. Dalhousie R, Li X, Davidson C, Wei N, Dooley
www.netl.doe.gov/coal/carbon-storage/worldwide-ccs- J (2009) Establishing China’s potential for large
database scale, cost effective, deployment of carbon dioxide
CO2 Capture and Sequestration 517

capture and storage. Pacific Northwest National labo- 58. Hovorka SD, Benson SM et al (2006) Measuring per-
ratory report, PNNL-SA-XXXX, 6 p manence of CO2 storage in saline formations – the Frio
47. Liu H, Gallagher K-S (2010) Catalyzing strategic experiment. Environ Geosci 13:103–119
transformation to a low-carbon economy: a CCS 59. Streit JE, Hillis RR (2004) Estimating fault stability
roadmap for China. Energy Policy 38:59–74 and sustainable fluid pressures for underground stor-
48. IEA GHG (2005) A review of natural CO2 emissions age of CO2 in porous rock. Energy 29:1445–1456
and releases and their relevance to CO2 storage. Inter- 60. Dixon T (2017) ISO CO2 storage ISO TC 265. CTS
national Energy Agency greenhouse gas R&D pro- workshop. EIA, Paris
gramme, report 2005/8. http://www.ieagreen.org.uk/ 61. Bradshaw J, Boreham C, la Pedalina F (2004) Storage
49. Carpenter S, Koperna G (2014) Development of the retention time of CO2 in sedimentary basins; examples
first internationally accepted standard for the geologic from petroleum systems. Paper presented at the
storage of carbon dioxide utilizing Enhanced Oil GHGT-7 conference, Vancouver
Recovery (EOR) under the International Standards 62. National Energy Technology Laboratory (2017) Best
Organization (ISO) Technical Committee TC-265. practices: RIsk management and simulation for Geo-
Energy Procedia 63:6716–6729 logic Storage Projects. https://www.netl.doe.gov/coal/
50. WRI (2008) Guidelines for carbon capture and seques- carbon-storage/strategic-program-support/best-practices-
tration, major contributing author. World Resources manuals
Institute, Washington, DC, 103 pp. https://www.wri.org 63. Yamaguchi K, Takizawa K, Shiragaki O, Xue Z,
51. National Energy Technology Laboratory (2017) Car- Komaki H, Metcalfe R, Yamaguchi M, Kato H, Ueta
bon Storage Program Fact Sheet. https://www.netl. S (2013) Features events and processes (FEPs) and
doe.gov/sites/default/files/2017-11/Program-116.pdf scenario analysis in the field of CO2 storage. Energy
52. National Energy Technology Laboratory (2017) Best Procedia 37:4833–4842
practices: monitoring, verification, and accounting 64. Pawar RJ, Bromhal GS, Carey JW, Foxall W, Korre A,
(MVA) for Geologic Storage Projects. https://www.netl. Ringrose PS, Tucker O, Watson MN, White JA
doe.gov/coal/carbon-storage/strategic-program-support/ (2015) Recent advances in risk assessment and risk
best-practices-manuals management of geologic CO2 storage. Int
53. CO2 Capture Project (2009) A technical basis for CO2 J Greenhouse Gas Control 40:292. https://doi.org/
storage. In: Cooper C (ed) CO2 capture project. CPL 10.1016/j.ijggc.2015.06.014
Press, UK. ISBN: 978-1-872691-48-0 65. Carroll S, Bromhal G, Richard T (eds) (2016) Interna-
54. Friedmann SJ (2006) The ICE framework for CO2 stor- tional Journal of Greenhouse Gas Control. Virtual
age site characterization. In: NETL 5th Annual Confer- Special Issue: NRAP
ence on Carbon Sequestration. ExchangeMonitor 66. Newmark RA, Friedmann SJ, Carroll SJ (2010) Water
Publications, Alexandria challenges for carbon capture and sequestration. Envi-
55. Burton EA, Myhre R, Myer LR, Birkinshaw K (2007) ron Manag 45:651. https://doi.org/10.1007/s00267-
Geologic carbon sequestration strategies for Califor- 010-9434-1
nia, The assembly bill 1925 report to the California 67. DOE-NETL (2007) U.S. Department of Energy,
legislature. California Energy Commission, Systems National Energy Technology Laboratory, Power plant
Office. CEC-500-2007-100-SD. http://www.energy.ca. water usage and loss study, May 2007 revision
gov/2007_energypolicy/documents/index.html#100107 68. Aines RD, Wolery TJ, Bourcier WL, Wolfe T,
56. Chiaramonte L, Zoback M, Friedmann S, Stamp Hausmann C (2010) Fresh water generation from
V (2006) CO2 sequestration, fault stability and seal aquifer-pressured carbon storage: feasibility of
integrity at teapot dome, Wyoming. In: NETL 5th treating saline formation waters. Energy Procedia
annual conference on carbon sequestration. 4:2269–2276
ExchangeMonitor Publications, Alexandria 69. Buscheck TA, Sun Y, Hao Y, Aines RD, Wolery TJ,
57. Cugini A, DePaolo D, Fox M, Friedmann SJ, Tompson AFB, Jones ED, Friedmann SJ (2010) Com-
Guthrie G, Virden J, (2010) US-DOE’s National risk bining brine extraction, desalination, and residual-
assessment program: bridging the gaps to provide the brine injection with CO2 storage in saline formations:
science base to ensure successful CO2 storage. Inter- implications for pressure management, capacity, and
national Conference on Greenhouse Gas Technolo- risk management. Energy Procedia 4:4283–4290
gies, Amsterdam, September 2010
 equivalents) in order to create a disincentive
Policy Instruments for for pollution emissions.
Mitigating Carbon Dioxide Economic efficiency Obtaining the maximum
Emissions outcome from the use of limited resources, as
in the maximum economic output given a fixed
Adam Rose1 and Brandt Stevens2 amount of inputs by a producer and the maxi-
1
Price School of Public Policy, and Faculty mum utility from a given combination of goods
Fellow, Schwarzenegger Institute for State and and services by a consumer.
Global Policy, University of Southern California, Economic policy instruments The institutional
Los Angeles, CA, USA means of implementing environmental goals.
2
BK Stevens Consulting, Sacramento, CA, Emission allowance auction Selling emission
USA allowances to the highest bidder, as opposed
to allocating the allowances free of charge.
Environmental regulation Specific require-
Article Outline ments pertaining to pollution emissions that
are backed up by the force of government and
Glossary penalties. This includes setting a bound on
Definition of the Subject pollution emissions, limiting the use of certain
Introduction fuels, or requiring certain production or pollu-
Evaluative Criteria tion control technology.
Environmental Goal Setting Equity Fairness in the distribution of income or
Regulation the distribution of impacts of a given policy more
Taxes and Subsidies generally. Note that there is no universal consen-
Tradable Property Rights (“Cap and Trade”) sus on the best definition. One prevalent equity
Operational Examples concept in environmental policy is vertical
Future Directions for Research equity – seeking to narrow the disparities
Conclusion between rich and poor, as in allocating relatively
Bibliography more greenhouse gas emission allowances to
poorer countries. Another is horizontal equity –
Glossary treating everyone in the same category equally,
as in requiring the same percentage emission
Benefit-cost analysis The systematic appraisal reductions from all polluters.
of all benefits and all costs of a specific course Evaluative criteria Norms by which to judge
of action. the effectiveness of policies.
Cap and trade The policy instrument of Renewable portfolio standard A requirement
assigning a limited number of property rights that a given percentage of electricity be gener-
to the environment (typically in the form of ated by renewable resources.
emission allowances) and then letting recipi-
ents of those rights to trade them in the mar-
ketplace to achieve the least-cost mitigation of Definition of the Subject
a given pollutant.
Carbon tax Imposing a fee per unit of carbon Following the establishment of an environmental
(or other greenhouse gas measured in carbon policy goal, attention must turn to attaining

© Springer Science+Business Media, LLC, part of Springer Nature 2020 519

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3_953
Originally published in
R. A. Meyers (ed.), Encyclopedia of Sustainability Science and Technology, © Springer Science+Business Media, LLC,
part of Springer Nature 2017, https://doi.org/10.1007/978-1-4939-2493-6_953-1
520 Policy Instruments for Mitigating Carbon Dioxide Emissions

it. Policy goals are often codified in legislation or penalties for non-compliance. Over the last cen-
administrative orders, initially without specifics tury, economists have devised various types of
about how they are to be implemented. Economic incentive and disincentive systems to attain desir-
policy instruments are the means of implementing able environmental and economic outcomes,
environmental goals, such as greenhouse gas mit- though policymakers have been slow to adopt
igation. Several alternatives exist from which to them until the last three decades. These are
choose, with nearly all of them having a rich intended to offer polluters more choice in their
history of both success and failure in the behavior, which is considered a superior approach
environmental and other domains. Moreover, to ensuring compliance, and doing so at a mini-
care must be taken to minimize unintended nega- mum cost of pollution mitigation. A carbon tax,
tive side-effects in their design and application. and a cap and trade system are two prime exam-
ples of incentive-based approaches.
The section “Evaluative Criteria” presents some
Introduction criteria for the evaluation of alternative pollution
control policy instruments. The section “Environ-
Following the establishment of an environmental mental Goal Setting” presents a benefit-cost analysis
policy goal, attention must turn to attaining (BCA) framework for overall goal setting. The sec-
it. Policy goals are often codified in legislation tions “Regulation” through “Tradable Property
or administrative orders, initially without specifics Rights (“Cap and Trade”)” represent a primer on
about how they are to be implemented. Economic economic policy instruments in the environmental
policy instruments are the means of implementing arena. The intent is to help the reader understand the
environmental goals, such as greenhouse gas mit- underlying rationale of policy instruments in general
igation. Several alternatives exist from which to and their strengths and weaknesses. The
choose, with nearly all of them having a rich section “Operational Examples” provides an assess-
history of both success and failure in the environ- ment of real-world examples of these policy instru-
mental and other domains. Moreover, care must ments applied to mitigation of carbon dioxide and
be taken to minimize unintended negative side- other greenhouse gases. The section “Future Direc-
effects in their design and application. tions for Research” discusses possible directions for
This entry provides an introduction to the major additional research, and the section “Conclusion”
economic policy instruments used in environmen- summarizes this entry.
tal policymaking. We focus on three major sets of Although the carbon tax and cap and trade instru-
instruments: (1) regulations (sometimes referred to ments have gained the most attention in recent years,
as standards or “command and control”), (2) taxes not all mitigation and sequestration measures
(sometimes referred to as charges or fees) and respond to price signals, and, in other cases, these
subsidies, and (3) tradable property rights two policy instruments are unwieldy. Therefore, we
(sometimes referred to as “cap and trade”). This also present a discussion of the regulation approach,
entry begins with a discussion of criteria by which which manifests itself in such ways as renewable
to evaluate the alternative instruments. We then portfolio standards (RPS), land use planning, and
explain the workings of each instrument, including strictures against deforestation.
strengths and limitations, and how each one stacks
up against the various criteria. Finally, we present
examples of design and application in Evaluative Criteria
policymaking related to greenhouse gas mitigation.
The design and implementation of policy Policy should start with the establishment of
instruments to reduce various types of pollutants criteria by which to judge the effectiveness of
have a long history of research and practice. Until environmental policy instruments in attaining
recently, this field was dominated by a regulatory their objectives and also whether the policy instru-
approach that was backed up by various types of ment does so in an efficacious manner and with a
Policy Instruments for Mitigating Carbon Dioxide Emissions 521

minimum of unintended negative side-effects. In Dynamic equity refers to fairness over time and
some cases, these criteria serve as intended sup- is sometimes referred to as intergenerational
plementary objectives of the policy. The follow- equity for policies with long time horizons, such
ing criteria are most often applied in this setting: as those related to climate change. It poses special
Static economic efficiency typically refers to problems, because future generations are not pre-
being able to attain the objective at the minimum sent during the policymaking process to represent
cost at a given point in time (the static aspect). themselves. Discounting of future benefits from
There is general consensus among economists, as greenhouse gases (GHG) reductions poses such
well as most policymakers, on its basic definition sufficient challenges that have been discussed in
and the worthiness of this criterion, as it seeks to several IPCC assessment reports (IPCC [1]
reduce any waste of resources in achieving the 228:232). The US EPA also convened a special
policy goal. panel in 2011 seeking advice on discounting in
Dynamic economic efficiency can have several BCA (Arrow [2]). The benefits from GHG reduc-
interpretations, but, in general, it refers to mini- tions will occur over at least the next two or three
mizing the cost of compliance over the anticipated centuries. For example, $1,000,000 of benefits
time horizon of the policy. A related aspect would 100 years from now, discounted at 4%, has a
be to evaluate whether the policy provides incen- present discounted value of $19,800. Some
tives to reduce direct costs of compliance scholars think that this result strongly implies a
over time. GHG policy decision taken today inappropriately
Flexibility pertains to the ability to adjust to diminishes the welfare of distant future
changing conditions. Obviously lack of flexibility generations [3].
would have an adverse effect on dynamic effi- There are, of course, several other criteria that
ciency, but it does merit separate attention. could be applied, depending on the particulars of
Enforceability addresses several consider- the situation, policymaker and stakeholder prefer-
ations, including whether the policy instrument ences, and existing laws. It is important, however,
has sufficient strength itself or provides incentives to present and justify any criteria at the outset, so
(making it self-enforcing) to attain the policy goal one is not accused of making up the rules while
or whether the government agency that estab- the game is being played.
lishes and/or administers the policy can enforce it. Also, in comparing policy instruments, it is
Administrative cost refers to the cost of imple- important to do so in an even-handed fashion.
mentation by the administrative agency, rather than That is, policies should be compared either on a
the cost of those entities that must comply with theoretical or a real-world plane, but not across
policy. Policy instruments will score high on this planes. A not uncommon tactic in policy debates
criterion if they require a minimum of such costs. is to present one’s own favored policy instrument
Static equity generally refers to the fairness under ideal circumstances and to compare it with
aspects of the policy instrument and its implemen- alternatives in the real world, warts and all.
tation at a given point in time. Unlike economic
efficiency, there is no consensus on the best defi-
nition of fairness. Several alternative equity prin- Environmental Goal Setting
ciples exist, including ability to pay and vertical
equity (which give preference to those lower on Ideally, any environmental goal would be
the scale of such metrics as income, profits, or established on the basis of sound reasoning and
size), or horizontal equity (which promotes equal careful justification. In the economic realm, this
access or outcomes). The criteria are often applied typically involves the application of benefit-cost
not just to those that need to directly comply with analysis (BCA), which is the systematic enumer-
the policy but also to all those affected by it, which ation of all benefits and all costs of a specific
often translates into the entirety of the broader course of action. (In some instances, a BCA is
community. not conducted as part of the policymaking
522 Policy Instruments for Mitigating Carbon Dioxide Emissions

process, due to either the legislative origin of the BCA in terms of marginal benefits and marginal
policy goal or sufficient uncertainty about either costs. The former represents the additional ben-
benefits or costs or both that a BCA is inadequate efits from increasing the level or scale of the
to guide policymakers about the optimal scale of activity being implemented. It is the first deriva-
the policy. Evaluating an established policy goal tive of the total benefit function. Analogously,
with respect to cost-effectiveness – is the policy the marginal cost is the additional cost that must
goal achieved at the lowest cost to society – is an be incurred in increasing the level or scale of the
alternative to BCA that allows analysts to sort activity. The shape of the MC curve in Fig. 1
through the merits of different policy instruments depicts the concept of diminishing returns
that could be used to achieve the policy goal.) The (or alternatively that costs increase at an increas-
“all costs” refers to not only out-of-pocket costs to ing rate). The MB curve is an approximation of
those that need to directly comply with the policy the demand curve for a good or service, while the
but also any implicit or opportunity costs, as well marginal cost curve is in effect the short-run
as any negative spillover effects. An example of a supply curve for that activity.
negative spillover effect is the generation of huge The intersection of the marginal benefit and
amounts of sludge, which requires a relatively marginal cost curves represents the optimal level
costly disposal activity, in the application of
“scrubbers” to electric power plant stack gases.
It is also possible that there would be positive
spillover effects, which would then be subtracted
to determine net costs. One example would be if a
by-product of pollution control could be sold at a
significant price. The “all benefits” refers to not
only gains that have a market price but also to
those for which a market may not exist, which
then requires “non-market valuation.” BCA is
often required because the market does not exist
and the benefits, analogous to the demand for the
policy outcome, must be estimated. In effect a
BCA represents a “surrogate market.”
Figure 1 presents two depictions of BCA. The
first partition illustrates the standard shape of the
total benefits and total cost curves. That is, once
benefits are identified, they are ranked in
descending order in relation to increasing policy
stringency, for example, greater reductions in
GHG emissions and thus increase at a decreasing
rate. In contrast, once costs are identified, they
are ranked in a descending order in relation to
increasing policy stringency and thus increase at
an increasing rate. A policy is worthy of adoption
if its total benefits exceed its total costs. If there
are alternative levels of the policy under consid-
eration, for example, alternative levels of pollu-
tion control, then the optimal scale is at the level
that maximizes the net benefits or the greatest
Policy Instruments for Mitigating Carbon Dioxide
distance between the total benefit and total cost Emissions, Fig. 1 Benefit-cost analysis of environmental
curves. The second partition of Fig. 1 illustrates regulation
Policy Instruments for Mitigating Carbon Dioxide Emissions 523

of the policy action. That level is consistent with specified reduction. Strictly speaking, marginal
the point at which net benefits are maximized. benefits would be conceived as infinite below
This result stems from mathematical optimization the standard and zero beyond it, but again this is
theory where the first-order condition for an opti- intended only as an approximation, and the inter-
mum is to set the partial derivative of net benefits pretation should not be taken literally as the policy
(TB – TC) equal to zero, or, in this case, (MB assessment. Note also that the standard takes pri-
MC) equal to zero, or MB = MC. Alternatively, it macy, and its intersection with the marginal cost
is an approximation of the intersection of supply curve would appear to be of secondary impor-
and demand. tance. For example, the Clean Air Act, as origi-
nally written in the early 1970s, made no reference
to cost, i.e., costs were not to be considered as part
Regulation of the standard setting process [4]. However, in
actual environment policymaking, an outcry
Basic Considerations and Workings regarding excessive costs from those upon which
This regulatory policy instrument is the traditional the regulation was directly imposed sometimes
approach in environmental policy. It establishes had the effect of reducing the stringency of the
strict requirements related to pollution emission standard.
limits or reductions or the means of achieving The next consideration in most environmental
them. Prior to the enactment of the National Envi- policy is how to distribute the burden of compli-
ronmental Policy Act (NEPA) in 1969, there were ance among relevant entities within the political
limited examples of environmental policymaking, jurisdiction enacting the regulation. It is not atyp-
but they were dominated by the regulatory ical to translate a region-wide percentage standard
approach. The “thou shalt” or “thou shalt not” reduction to that same percentage for each
tenor of this policy instrument is analogous to polluter.
the way laws and codes in other domains are We also note some special cases and more
stipulated. To some extent, the strict proscriptive sophisticated aspects of regulations. First, prohi-
nature of this policy instrument had the advantage bition is a special case and is warranted where the
of unequivocally requiring action to comply with marginal benefits exceed the marginal costs at
the policy. Some say that such an approach was 100% emission/effluent reduction. However, pro-
necessary to get the attention of polluters and to hibition is often invoked when the damages are
win over supporters of NEPA. especially serious or widespread or where a mar-
Definitely in early applications of the regula- gin of safety is warranted. Classic examples relate
tion approach, and still to a great extent today, the to the emissions of radiation from nuclear power
level of air pollution emission or water pollution plants or to the discharge of especially dangerous
effluent reduction was not based on full knowl- carcinogens. However, when there are choices
edge of the marginal benefit function. Instead, a about alternative levels of the policy, economic
health-based standard was invoked to begin the analysis suggests it is a rare situation when no
policy process. This could have the form of a pollution control or eliminating the pollutant
threshold beyond which there was a serious threat entirely is optimal.
of chronic lung disease, cancer, or damage to an Another important distinction is between
ecosystem. The standard was then translated imposing regulation in the form of procedural
through a “rollback” formula to determine the requirements or performance standards. The for-
emission/effluent reduction needed to comply mer requires a specific action and limits choices
with the standard. This usually translated into a that regulated entities could implement and,
percentage reduction in emissions/effluents in hence, may also eliminate lower cost options.
relation to a baseline or translated into some max- The classic example was EPA’s Best Available
imum level of discharge. In effect, this represents Control Technology (BACT) requirements as
a perfectly vertical marginal benefit curve at the applied to electric utility sulfur dioxide emissions
524 Policy Instruments for Mitigating Carbon Dioxide Emissions

in the 1970s. In contrast, performance standards which the regulation applied directly, or to the
simply require that some emission reduction goal population of the affected jurisdiction indirectly,
be met but without restricting the means of doing as the cost of regulation is passed forward to
so. In this case, electric utilities would have the business customers and final consumers in the
choice of various control technologies, as well as form of higher prices.
using lower sulfur coal. The BACT rules were
sometimes justified on the grounds that they
represented a type of “technology forcing” that Taxes and Subsidies
would inspire the development of better and
lower-cost technology alternatives and therefore Basic Considerations and Workings
might limit static economic efficiency but inspire Taxes and subsidies represent two examples of
dynamic economic efficiency, as we have defined what economists refer to as “incentive-based”
it above. Analysis to support this justification is policy instruments. They provide a price signal
not often undertaken, in part, because of difficulty to regulated entities with the intent of prompting a
in predicting the future composition and implica- desired action. Taxes represent a disincentive to
tions of new technology. pollute, while subsidies provide an incentive not
to pollute. Taxes and subsidies represent two sides
Assessment of the same coin. Polluters are faced with a strong
How does the regulation policy instrument stack economic incentive to recognize the cost of the
up against the various criteria that we presented at pollution they generate, because, even in the sub-
the outset? Clearly most regulations are not flexi- sidy case, foregoing a subsidy represents a real
ble. On the other hand, they are definite, and this opportunity cost of their action. In the discussion
adds to enforceability. Enforcement is a function of below, we will focus on the tax approach and
several variables, a full discussion of which is simply note here that the explanation of the sub-
beyond the scope of this entry. These include the sidy approach is symmetric without going into
ability to monitor the pollution discharge and the much detail. However, when we perform our
severity of the penalty. All of these factors, how- assessment against evaluative criteria, we will
ever, also bear on the extent of administrative costs, note some important distinctions between the
which can be significant for some pollutants. two policy instruments.
In terms of static economic efficiency, the reg- In the ideal emission/effluent tax situation, the
ulation/standards approach misses the mark in two tax rate would be determined by the intersection
ways. First, the marginal benefit function is some- of the marginal benefit and marginal cost curves,
times not considered at all or is at best a rough as in the lower partition of Fig. 1. Figure 2 shows
approximation, so that it is unlikely that the overall this tax rate and the amount of pollution control
net benefit optimum will be achieved. Second, that individual firms would undertake. However,
when applied in a uniform percentage manner for policymakers cannot know aggregate marginal
all matters, this approach does not allow for the benefits or costs with certainty. When marginal
utilization of the optimal (least-cost) mix of pollu- benefits are not known, then utilizing a standard is
tion control options. Dynamic economic efficiency invoked, with the same likely deviation from the
in general is sacrificed, though there is some incen- overall efficient level of pollution control com-
tive to reduce costs over time. pared to regulation. Even if marginal benefits are
In terms of equity, this policy instrument would known with certainty, a tax can result in more or
seem to be consistent with horizontal equity when less pollution control than regulation or cap and
uniform pollution reduction requirements are trade if there is uncertainty about aggregate or
applied to all emitters. However, the reduction regional marginal costs. For example, in Fig. 2,
requirement is not the only “reference base” by if both firms’ actual marginal costs are lower than
which to gauge equity. Others would include the what regulators have estimated (marginal costs
relative health (profitability) of businesses to are shifted to the right), then actual pollution
Policy Instruments for Mitigating Carbon Dioxide Emissions 525

Policy Instruments for Mitigating Carbon Dioxide Emissions, Fig. 2 Pollution tax

control from both firms will be greater than emitters with relatively lower costs are penalized
required to meet the environmental policy goal. by controlling a higher percentage of their pollu-
When marginal costs are higher, actual pollution tion. However, mitigation costs are only part of
control is less than required. Regulators may the story. By inspecting Fig. 2, one can readily see
therefore prefer regulation or cap and trade policy that the lower-cost polluters will have lower tax
instruments than taxes as they assure attainment payments than the higher-cost polluters. The
of the environmental goal, though polluters may direct hit on the emitters is a combination of
prefer the price certainty of taxes. However, mitigation costs and tax payments, and the
uncertain marginal costs means uncertain revenue lower-cost emitters will therefore have a lower
collected either via taxes or by sale of cap and overall cost burden.
trade allowances. Note also one major distinction between the
The tax is usually applied to all emitters at the tax/subsidy approach and that of regulation.
same rate, e.g., a tax of $20 per metric ton of Under regulation, the quantity of emissions/efflu-
carbon emitted. Each polluter will choose to mit- ents that are in excess of the pollution control
igate pollution up to the point where the marginal requirement have no cost to polluters, while in
cost of doing so is lower than the tax. However, the tax/subsidy approach, it is generally the case
once the marginal cost exceeds the tax rate, then it that no emission/effluents are free. While environ-
will be cheaper to just pay the tax. Although the mental taxes are usually applied to all emissions,
tax rate is uniform, the response rate will differ there is no theoretical reason why some level of
among emitters. In general, emitters with rela- emissions could not be tax-free, similar to income
tively steep marginal cost curves will reduce emis- tax structures that allow some low level of income
sions less than those with relatively flat marginal to be earned with no tax obligation.
cost curves. The resulting aggregate level of emis-
sions is the least-cost combination of pollution Assessment
control across all regulated sources within the How does the tax/subsidy policy instrument stack
jurisdiction. Note that it may at first seem as if up against our evaluative criteria? Again, if the
526 Policy Instruments for Mitigating Carbon Dioxide Emissions

complete marginal benefit function is not known, Tradable Property Rights (“Cap and
then an initial inefficiency results in a manner Trade”)
similar to that under regulation. However, because
the tax results in differentiated responses among Basic Considerations and Workings
polluters according to the shape of their marginal One of the most unique policy design innovations
cost curve, it attains the least-cost overall response in all of economics over the last 50 years has been
to meet the policy goal. In terms of dynamic the development of the “property rights”
efficiency, there is in fact an incentive to reduce approach. This approach, pioneered by Nobel
pollution mitigation costs over time in order to Laureate Ronald Coase [5] and refined by several
reduce pollution control outlays and to reduce tax theorists and policy analysts, posits that external-
payments. In fact, one might consider this incen- ities, such as pollution, exist because of an
tive to be stronger than the case of regulation, absence of property rights. In essence, most envi-
which allows some pollution to be generated ronmental media, such as the oceans and the
for free. atmosphere, are not owned by anyone, and there-
The tax/subsidy approach is considered to be fore no one takes custodial responsibility for
relatively flexible because it allows choice in the them. Even if countries claim airspace or the
response. It is also considered to be relatively oceans as their territory, the “public good” nature
strong in terms of enforcement because it provides of many environmental resources are such that
incentives, which some characterize as making it ownership claims cannot be enforced, because
“self-enforcing.” Administrative costs are not others cannot be excluded from using or abusing
necessarily any higher than that of regulation on them. (A public good is said to have two charac-
first glance. Moreover, tax revenues can be used to teristics: (1) jointness in consumption, meaning
offset these costs. Of course, the opposite is true that two or more people can consume the good
for subsidies, where this variant of policy instru- simultaneously without detracting from the quan-
ment necessitates an additional burden on the tity available, and (2) non-exclude ability, mean-
government treasury. ing that it is technologically impossible,
As we noted above, there are many alternative economically impractical, or socially unaccept-
concepts of equity that are typically applicable to able to exclude anyone. For many years, it was
each of the policy instruments. This instrument assumed that a public good was best provided by
would appear to be consistent with horizontal government. However, the property rights
equity because, in typical applications, all pol- approach represents an alternative approach by
luters face the same tax rate. However, obviously establishing a market for the public good, e.g.,
the direct burden of the imposition of the tax either as the environment or pollution control.
varies but this instrument ranks well in terms of The latter is usually considered a “pure” public
some equity concepts in that it favors those pol- good, meaning that the jointness condition holds
luters that have relatively lower costs. At the to the extreme, while the former can be subject to
same time, the policy instrument does not distin- deterioration issues stemming from overuse,
guish between the relative economic health of the hence, rendering it less than a pure public good.)
entities to which it is applied. However, any Essentially, the problem is that of the “tragedy of
inequities here can be offset by the redistribution the commons” [6], whereby communal resources
of tax revenues. Another perspective on equity is are often overused, sometimes to the extent of
that dollars flow from polluters (typically busi- extinguishing them in quantity or quality.
nesses) to the general treasury (i.e., the public at The property rights solution is to assign rights
large). And, of course, lower-income people do where none previously existed and then to make
not generally own businesses, resulting in a high these rights tradable. This establishes a market
score in terms of ability to pay or vertical equity. where one did not previously exist and facilitates
The subsidy approach works in the opposite a market-based solution to the externality prob-
direction. lem. The application of this approach to
Policy Instruments for Mitigating Carbon Dioxide Emissions 527

environmental issues, such as GHG emissions, is The workings of the cap and trade approach are
to establish a “cap” on emissions, assign rights to illustrated in Fig. 3, which depicts two polluters
emit, and then allow those rights to be “traded.” that can represent either countries or firms. When
The cap sets the total supply of rights, and market a polluter’s marginal cost of CO2 mitigation is
participants determine the price at which the rights higher than the market price of the permits, the
(permits or allowances) will be traded. Alterna- polluter will purchase permits up to the point
tively, the rights can be auctioned to the highest where the marginal cost equals the price
bidder to achieve the same market outcome. Each (MC = P), thereby saving some mitigation costs.
of these alternatives represents examples of an Those polluters whose marginal costs of CO2
“incentive-based” system. mitigation are lower than the market price will
In the absence of environmental policy, the sell permits and make a profit by doing so.
polluter has two choices: (1) emit pollution to
the atmosphere at zero cost or (2) clean up the Assessment
pollution before emitting it but at some cost. The In evaluating the property rights approach, it is
profit motive dictates that the first option will be best to begin with a statement of what is known as
taken. The key is to make it profitable not to the Coase Theorem: externalities can be elimi-
pollute. By granting the polluter (or the general nated if property rights can be assigned and are
citizenry, as will be discussed below) rights to allowed to be traded, regardless of the initial dis-
pollute, it gives the entity receiving the rights a tribution of the rights (assuming no significant
stake in the environment. This makes the polluter income effects or transaction costs). Below, we
take into account the opportunity cost of his/her begin our discussion invoking the assumptions in
actions, because every metric ton of CO2 emitted the parentheses at the end of the stated theorem.
represents a lost opportunity to sell a permit. This The trading guarantees static economic efficiency
has the effect of “internalizing” to the polluter the will be attained by virtue of typical gains from
externality cost imposed on others. trade (a win-win outcome) of competitive markets

Policy Instruments for Mitigating Carbon Dioxide Emissions, Fig. 3 Cap and trade
528 Policy Instruments for Mitigating Carbon Dioxide Emissions

in general. Dynamic efficiency, as we have and then distributing the permits accordingly, a
defined it, also scores high, in that both buying major logjam was broken by the now annual Con-
and selling entities have an incentive to reduce ference of the Parties (COP) to the United Nations
their mitigation costs over time to reduce the Framework Conference on Climate Change in
number of permits they need to purchase and to Paris in 2015 (COP 21), through a “bottom-up”
make higher profits on permit sales, respectively. approach where each country made its own
The static equity aspect of the Coase Theorem Intended Nationally Determined Contributions
is rather unique, however, as it says that the effi- (INDCs), essentially conditional and uncondi-
cient outcome will arise regardless of to whom the tional pledges to reduce GHGs in the coming
rights are distributed. Thus, the outcome is the decades (UNFCCC [10]). The cap and trade
same if polluters are given property rights or if approach also scores well on dynamic
the general citizenry were given the property (intergenerational) equity, an especially serious
rights and polluters would have to buy these rights issue in the context of climate change because it
from them. The more typical case is to distribute holds the potential of ominous long-term impacts
the rights to polluting entities and allow them to that may affect many future generations.
do the trading. The important thing is that this As to other criteria, the cap and trade system is
means a very standard concern in environmental flexible, in that the cap can be easily adjusted over
policymaking, and policymaking in general will time. Enforcement can be problematic, but not
not arise – that of the efficiency-equity trade-off. necessarily more so than in the case of the other
Economists typically give primacy to the effi- two policy instruments previously discussed. In
ciency criterion and in many cases point to the all three, monitoring of emissions is necessary to
fact that the pursuit of equity will undercut it ensure compliance but in the climate change
(a prime example is the concern that unemploy- arena, monitoring is facilitated greatly by the
ment insurance benefits will provide a disincen- fact that the majority of GHG emissions can be
tive for people to find jobs). Thus, equity can be traced to some tangible and straightforward indi-
addressed head on in the application of this policy cator, rather than the need for sophisticated equip-
instrument by changing the initial distribution of ment being placed on smokestacks. The primary
rights without undercutting an economically effi- example is the fact that each type of fossil fuel has
cient market outcome. Moreover, in many appli- associated with it an “emissions factor,” which
cations of cap and trade, such as in the indicates the CO2 generated per unit of coal, oil,
international domain, where there is no suprana- and gas. Hence the emission estimates can be
tional authority to force an agreement on GHG made fairly accurately using this indirect method.
emission reduction, voluntary cooperation is nec- Additionally, of course, market considerations
essary to obtain an agreement. Rose [7] and others help make this policy instrument self-enforcing.
have pointed to the fact that equity might be Administrative costs can potentially be a major
paramount in this situation. In effect, no country consideration – recall the parenthetical expression
will join an agreement in which it believes it is in the Coase Theorem regarding “transaction
being treated unfairly, and many countries would costs” undercutting the effectiveness of this policy
also be concerned about the equitable treatment of instrument. Transaction costs might be considered
others. Unfortunately, there is no single best def- a subset of administrative costs, which would
inition of equity, and the policy design of and include not only the cost of information gathering
actual negotiations regarding international treaties and the establishment of some institutional mech-
to mitigate GHG emissions have brought forth anism for exchange but also the cost of monitoring
literally dozens of alternative equity principles or and enforcement. This is likely to be relatively
combinations (see, e.g., [1, 8, 9]). After many modest in the case where there are a limited num-
years of the countries of the world attempting to ber of entities to which the property rights are
come to a truly global agreement through a top- assigned. There is typically some flexibility in
down approach of agreeing on an equity principle the point in the production chain at which this is
Policy Instruments for Mitigating Carbon Dioxide Emissions 529

best applied in light of this concern. For example, renewable energy sources, for example, hydro,
rather than assigning emission rights to every geothermal, solar, and wind. There are several
driver in an air basin, the rights could simply be components to the planned GHG reductions to
assigned “upstream” to oil refineries. meet the goals in AB 32, California’s legislation
Much research has been done comparing cap initiating a reduction in GHGs by 2020 to the
and trade and carbon tax systems in particular and 1990 level of emissions. The RPS program is
also comparing these two policy instruments in expected to account for 11.4 million metric tons
general (see, e.g., [11, 12]), with the general (MMT) of CO2e reductions by 2020, about 14%
assessment that they are very similar on a concep- of the GHG reductions. California’s reductions in
tual level. However care must be taken in making the WCI cap and trade program, discussed below,
these comparisons. For example, Hahn [13] dem- are expected to account for 32.1 MMT of CO2e
onstrates that cap and trade’s economic efficiency reductions or about 23%. California’s RPS pro-
may be adversely affected if the market has one or gram is summarized below.
two large polluters. Hahn and Hester [14] identify California’s Renewables Portfolio Standard
specific transaction costs that can affect cap and program was initiated by SB 1178 passed in
trade markets. Kolstad and Wolak [15] conclude 2002, followed by SB 107 in 2006, SB X1–2 in
that the nitrogen oxide (NOx) market in Southern 2011, and SB 350 in 2015. SB 1178 required all
California allowed some generators to exercise retail sellers of electricity in California to procure
market power in wholesale electricity markets a minimum of 20% of their retail electricity sales
and increase bid prices. In addition, this equiva- from eligible renewable resource generation.
lency only holds when we consider the auction Retail sellers were to increase their procurement
form of cap and trade [16]. Under this arrange- by 1% per year until they had reached the 20%
ment, there are no free emissions (in contrast to minimum. SB 107 accelerated the requirement
what is known as “grandfathering” or free alloca- such that retail sellers must have met the 20%
tion of permits), just as there are no free emissions requirement by 2010 rather than 2017. SB X1–2
in the context of carbon taxes. Also, in both cases, increased the standard to 33% by 2020, and SB
equity can be adjusted depending on how the 350 further increased the standard to 50% by
auction revenues and tax revenues are eventually 2030. SB 584, introduced in the 2017 legislative
redistributed or spent. session, proposes a standard of 100% to be
Some major differences have been pointed out achieved by 2045. Compared to other state RPS
at both the conceptual and real-world implemen- programs, California’s goals are ambitious.
tation levels. For example, there are typically Responsibility for the program is divided among
more institutional obstacles in changing the tax the California Public Utilities Commission
rate as opposed to the cap. On the other hand, (CPUC) and the California Energy Commission
there are some advantages of the tax approach (CEC), with a small role for the California Air
under conditions of uncertainty [17]. These issues Resources Board (CARB). The CPUC’s primary
will be discussed in greater detail in the context of responsibility is to oversee procurement and
actual applications below. related activities by investor-owned utilities, i.e.,
Southern California Edison (SCE), San Diego
Gas & Electric (SDG&E), and Pacific Gas and
Operational Examples Electric (PG&E), along with developing and
implementing RPS compliance rules for all retail
Regulation electricity sellers in California. The CEC’s pri-
Nearly 30 states have a renewable portfolio stan- mary responsibilities are certifying RPS-eligible
dard in place. Although the programs differ in generation facilities, tracking electricity purchases
their details, each state’s regulation promulgates from such facilities by retail sellers in California
a standard that requires a minimum percentage of (and other states), and overseeing publicly owned
total electricity sold in the state be generated by utility procurement and compliance with the RPS.
530 Policy Instruments for Mitigating Carbon Dioxide Emissions

Noncompliant POUs are referred to CARB by Procurement plans for 2020 for each utility would,
CEC. Carl et al. [18] recommend replacing this if all planned projects are completed on time, result
fragmented bureaucratic structure with a lead in more than 40% sales, exceeding the 33% stan-
agency that would develop an implementation dard. While the banking provisions are restrictive,
road map. these utilities, serving slightly less than 70% of
Eligible RPS technologies include, for exam- electricity load in California, have the opportunity
ple, solar thermal, solar photovoltaics, small to carry over excess compliance if needed or possi-
hydro, geothermal, and wind. The tracking system bly to sell excess RECs. However, the CPUC’s
is known as Western Renewable Energy Genera- Biennial RPS Program Update ([19]; Table 1)
tion Information System (WREGIS) and covers shows that some of the remaining retail sellers are
California and other states that are part of the strong candidates to employ excess compliance and
Western Electricity Coordinating Council historic carryover. Also, a review of 6 of the 44 POU
(WECC). Renewable energy credits (RECs) are procurement plans show that 4 of these utilities plan
tracked via WREGIS. One REC represents to use banked RECs from 2011 onward, historic
1 MWh of electricity sold. carryover, and unbundled RECs to the maximum
In addition to either procurement via contract extent possible through 2030. A review of the ver-
with an RPS resource within the WECC or own- ification reports for those same POU’s showed that
ership of an eligible facility, retail sellers have five retired unbundled RECs for the first compliance
other options for compliance. First, banking is period, four retired the maximum number of
allowed by carrying forward “excess” unbundled RECs allowed, and three had substantial
compliance – RECs from retail sales in excess of historic carryover RECs that were not used in first
the RPS beginning in 2011 can be applied in later compliance period [20].
periods by the same entity; however, such RECs While there have been numerous analyses of
must be submitted for compliance within 3 years RPS programs, to date, there has been no formal
of their creation. Second, banking is also allowed BCA for any of the state programs, including
for historic carryover, i.e., any excess compliance California’s RPS. In the absence of such analysis,
from 2004 to 2010 can be applied for compliance policymakers do not know whether any of the
in later years through 2030. Third, unbundled required RPS of 20, 33, or 50% are too low or
RECs, RECs that have been sold separately from too high. While a BCA would be useful, though
the generated electricity, may also be submitted difficult to conduct at least in part due to highly
for compliance purposes. Additional restrictions imperfect information regarding the benefits of
exist in regulatory distinctions between three port- GHG reduction, the cost-effectiveness of a policy
folio categories: portfolio category 1 is electricity is an important second best question. As noted
from a renewable resource that is delivered earlier in this entry, a uniform standard that
directly or is interconnected with the California applies to all regulated entities is not likely to be
grid and the associated REC and portfolio cate- cost-effective if there are differences in marginal
gory 3 is primarily unbundled RECs. Beginning costs of complying with the standard, and larger
in 2011 with the first compliance period, compli- differences in marginal costs imply more costly
ance requires baseload RPS must be at least 50% compliance compared to a cost-effective policy.
of the entity’s retail sales and unbundled RECs Clearly, there are differences in compliance costs
can be no more than 25%; by the third compliance across the various retail sellers, some of whom
period, concluding in 2020, the requirements are own generation facilities (with very different costs
no less than 75% and 10%, respectively. Begin- characteristics as shown by bids into the whole-
ning in 2021, no unbundled RECs may be used for sale electricity market), while other sellers do not
compliance with the RPS. and there are substantial differences in climate and
According to the CPUC, in 2015, SCE, SDG&E, weather in the territories served by different
and PG&E retail sales from eligible RPS sources sellers and these are evidenced in differences in
were 24.3, 35.2, and 29.5%, respectively. retail rates and rate structures. As discussed
Policy Instruments for Mitigating Carbon Dioxide Emissions 531

above, different entities have quite different com- Another criterion discussed earlier in this entry
pliance strategies, driven by differences in com- is equity of the policy. While no analysis of equity
pliance costs. For example, the Port of Stockton of RPS policies are available, California’s RPS
[21] initially considered adding a small RPS gen- program is likely regressive. It is well-known
eration facility but eventually concluded that it that low-income households spend a larger frac-
was cheaper to buy RECs to the extent allowed tion of their income on energy – primarily trans-
by the regulations. portation fuels and household energy use – than
While the program offers some flexibility, the high-income households. Most of the smaller util-
restrictions with respect to the minimum amount ities in the state, for example, Lassen Municipal
of RPS resources and the maximum amount of Utility District, Modesto Irrigation District, and
unbundled RECs for compliance purposes are a Redding Electric Utility, serve areas with lower
barrier to a cost-effective policy. GHGs are a than state average household incomes and have
global pollutant and since any REC registered in some of the state’s poorest households. Based on
WREGIS represents a MWh generated by some the review of POU procurement plans and the
RPS resource within the WECC, an open and verification reports, along with CPUC data show-
transparent market would allow entities to choose ing progress toward the 2020 RPS, these are reg-
to procure RPS resources so long as that is less ulated entities that appear most likely to employ
than or equal to the price of a REC. For some retail banking, historical carryover, and unbundled
sellers, that might have resulted in more or less RECs for compliance purposes.
than 50% of RPS electricity sales from baseload
renewable generation in 2011–2013. Limiting British Columbia’s Carbon Tax
banking during the 2011–2020 compliance To date, only a few jurisdictions have considered
periods also reduces cost-effectiveness (see [22] or implemented a carbon tax. British Columbia
for analysis in the context of global GHG per- (BC) has had a carbon tax since July 2008. Prime
mits), and, as Schmalensee [23] notes, discour- Minister Trudeau’s government has announced
ages early investment in RPS resources, a likely that all Canadian provinces must implement either
unintended consequence of limiting REC bank- a carbon tax or a cap and trade program by
ing. His analysis also includes an examination of January 2018. Washington state voters rejected a
several REC markets. He concludes that price carbon tax ballot measure by a 58–42 margin in
transparency – meaning market participants are November 2016. In the USA, former Democratic
well aware of market prices via brokers and/or Party presidential candidate Bernie Sanders has
published information on prices and market advocated for a carbon tax, and prominent Repub-
transactions – is poor. For those REC markets lican Party members George Schultz and James
that were analyzed, price dispersion is large, mar- Baker are also advocates for this approach to
kets are thin, and bid/ask spreads are unusually GHG reduction [24]. A brief summary of British
large compared to municipal and corporate bond Columbia’s carbon tax follows.
markets, indicating poor market liquidity. This is British Columbia (BC) passed legislation to
due in part to limitations on banking, which makes implement a carbon tax on all fossil fuels burned
different vintages of RECs imperfect substitutes. in BC as reported in Environment Canada’s
While California’s REC market (or even the National Inventory Report, beginning in July
WECC) was not included in this examination, 2008, coinciding with the province’s fiscal year.
there is no systematic publication of REC trans- The tax would start at $10 (Canadian) per metric
actions and market prices from any source. In the ton of CO2e, increasing by $5 each year until
absence of this information, it would be surprising 2012, leveling out at $30 per ton [25]. According
if California’s REC market did not exhibit the to [26], the $30 (Canadian) tax would be about US
same problems. Carl et al. [18] note that public $24 per metric ton of CO2e. The increase in the tax
release of program cost information would assist rate was designed to give those affected by the
California’s development of renewable resources. carbon tax time to adjust their behavior and to
532 Policy Instruments for Mitigating Carbon Dioxide Emissions

provide certainty with respect to the tax level tax. In addition, personal income tax rates for the
during and after the initial implementation period. bottom two tax brackets were reduced by 2% in
The tax is collected by fuel suppliers at the whole- 2008 and 5% in 2009 and thereafter. The general
sale level based on the carbon content of the fuel. corporate tax rate was reduced to 11% in 2008,
The carbon tax for natural gas sold to retail con- followed by additional cuts in 2010 and 2011 to
sumers and propane are collected at the retail 10.5% and 10%, respectively. The small business
level. Collecting the tax is thus administratively tax rate was also reduced in steps to 3.5% and
simple and easily enforceable, as it relies on sales subsequently to 3% and 2.5% on a similar schedule
information that is already required for fuel excise to general corporate tax rate. Over 60% of the pro-
taxes. The fuel tax rates for gasoline and diesel jected tax cuts funded by carbon revenues are in
fuel are different, with diesel having a slightly reductions to personal income taxes. For example,
higher tax rate per liter. Similarly, high-heat coal total estimated tax recycling in the 2010/2011 fiscal
and low-heat coal have different fuel tax rates. BC year is $880 million, $547 million of which is some
released a second carbon reduction plan in 2016 type of personal income tax reduction.
[27], but the plan does not propose a further series Actual experience with revenue recycling has
of increases in the carbon tax rate. Some observers been somewhat different. From the inception of
think that additional increases in the carbon tax the program, carbon tax revenues were less than
would help the province meet its carbon reduction revenues recycled. Murray and Rivers’ [26] anal-
target. Provincial officials may prefer to wait until ysis of provincial budget documents from 2008 to
the Canadian federal plan takes effect as BC is 2015 shows carbon tax revenues of $6.1 billion
well ahead of other Canadian provinces. (Canadian) compared to about $7.1 billion in rev-
All carbon tax revenue is legally required to be enue recycled. Only in the 2013–2014 fiscal year
recycled each fiscal year via reductions in other were carbon tax revenues and revenue recycled
taxes levied by the province. The carbon tax is equal to each other. Also, in contrast to the initial
intended to be “revenue neutral,” and thus the tax plan with larger personal income tax reductions
revenues cannot be spent on government programs. relative to business tax reductions, by the
The carbon tax revenues are used to offset price- 2012–2013 fiscal year, business tax reductions
distorting taxes on labor and capital (income and were equal to or larger than revenue recycling to
corporate taxes), which lead to inefficiencies in the BC taxpayers and are projected to continue to
economy. Thus, a carbon tax can result in a “double- grow relative to personal income tax reductions.
dividend,” with the first dividend being the reduc- The primary reason is that targeted tax reductions
tion in carbon emissions and the second being an were implemented beginning in 2013, mostly for
increase in overall economic activity due to the agricultural sector, with additional tax reduc-
improved economic efficiency (see [28] for further tions planned for the motion picture industry.
discussion). Such revenue recycling could also be Finally, Murray and Rivers summarized several
employed with cap and trade. The provincial gov- studies of distributional impacts of the carbon tax.
ernment presents a plan each year in conjunction These analyses indicated small distributional
with the provincial budget that shows how the car- impacts of less than 1% of income for low-income
bon tax revenue is recycled to taxpayers. In 2008, earners, and in analyses where the models can
the revenue recycling plan included provisions for account for changes in expenditures and sources
both reductions in personal income and business of income, high-income households bear most of
taxes. For example, fuel taxes, whether excise the carbon tax incidence.
taxes or carbon taxes, fall more heavily on low-
income earners because fuel expenditures are a pro- Western Climate Initiative Cap and Trade
portionally larger share of their income. A Low- Program
Income Climate Action Tax Credit provided a larger Several operational cap and trade programs for
share of the carbon tax revenue to low-income reducing greenhouse gas (GHG) emissions are in
earners to offset the regressive nature of the carbon use in various jurisdictions, including the
Policy Instruments for Mitigating Carbon Dioxide Emissions 533

European Union Emissions Trading System allocations relative to other sectors of the state
(EU ETS), the Regional Greenhouse Gas Initia- economy (see Table 8-1 of the Regulation
tive (RGGI) that covers nine states in the North- Section 95870). Quarterly auctions are used to
eastern United States, and the Western Climate dispose of the remaining allowances in the annual
Initiative (WCI) joint cap and trade program, cap. The primary purpose of the auctions is to
started by California, with Québec and Ontario promote price discovery, that is, to allow all mar-
as jurisdictional partners. A brief summary of the ket participants to collectively determine the value
WCI cap and trade program follows. of an allowance relative to the cost of mitigating
The California Legislature passed the Global GHG emissions. The regulation allows for a sec-
Warming Solutions Act (AB 32) in 2006, requiring ondary market in allowances with provisions in
a reduction in greenhouse gas emissions to 1990 the regulation that require reporting of a change in
levels by 2020. A further reduction to 40% below ownership of allowances held. Future contracts
1990 levels by 2030 is contained in SB 32, passed in are reportable when the contract expires and the
2016. The legislature clearly believed that the ben- buyer wants to take physical delivery of the allow-
efits of the GHG reductions in both acts outweigh ances because the option price is less than the
the costs, though no formal BCA was conducted. current market price. Transactions are conducted
However, the legislative text makes clear the inten- both on the Intercontinental Exchange and in
tion for cost-effective GHG reductions. The Califor- bilateral transactions between covered entities.
nia Air Resources Board (CARB) was given Price data from the secondary market is available
administrative responsibility to implement the pro- from several sources.
visions in AB 32, including the legislative directive California and Québec provisions with respect
to develop a market-based program. to enforcement are similar. Both jurisdictions
The cap and trade program initiated by CARB require emissions reporting by covered entities
covers nearly all sectors of the state economy, for and use reported emissions to determine covered
example, electricity generation, industrial entities allowance surrender obligation (Any cap
sources, transportation fuels, and natural gas dis- and trade program or a carbon tax requires accu-
tribution. Any source of GHG emissions exceed- rate knowledge of emissions by covered entities to
ing 25,000 metric tons annually (CO2 or CO2 be able to determine how many allowance-
equivalent) is a covered entity and has a compli- covered entities need to surrender or the quantity
ance obligation. The California Regulation of emissions subject to the GHG tax. In Califor-
established an initial GHG emissions cap, begin- nia, such reporting started in 2008, 5 years prior to
ning in 2013, with an annual decline to 2020 the initial compliance period, and is governed by
([29]) (See Table 6–1 of the Regulation, the Mandatory Reporting Regulation (MRR).
Section 95841. There is an increase in the cap in Emissions are reported annually to CARB and
2015 due to the inclusion of transportation fuels are subject to verification by third parties
and natural gas distributors as covered sectors.). ([30]).). Entities are allowed to bank allowances,
Allowances are allocated annually at no charge to i.e., allowances can be held after their calendar
electricity generators, industrial sources, and nat- year vintage if not needed to satisfy a compliance
ural gas distributors (The legislature was obligation [22]. Both programs allow for the cre-
concerned that implementation of AB 32 might ation of GHG offsets, for example, via forestry
increase prices of some goods produced in Cali- projects; offsets are issued by the jurisdiction to
fornia relative to the same products produced out- project operators when the project meets condi-
side California, making imports of those goods to tions codified in the regulations. Project operators
California attractive to state consumers. AB can sell offsets after issuance, and covered entities
32 requires that the regulation not unduly disad- can surrender offsets for compliance purposes,
vantage sectors that are determined to be highly though no more than 8% of their emissions obli-
sensitive to increased import competition. These gation in any compliance period. If for some rea-
sectors are protected by larger allowance son, the offsets are later determined to be invalid,
534 Policy Instruments for Mitigating Carbon Dioxide Emissions

the jurisdictions can withdraw them without com- floor was $10, and in 2017, it is $13.57
pensating the owner of record. With offsets typi- (California also has a soft price cap via the
cally sold shortly after their issuance, there is Allowance Price Containment Reserve
some risk to purchasers, primarily covered enti- (APCR). About 122 million allowances
ties, and thus offsets are sold at a discount relative have been set aside from the initial
to allowances. While both jurisdictions use the 2012–2020 budget and can be sold at a fixed
same compliance periods through 2020 price of $50 in 2013 (now $57.04), adjusted
(2013–2014, followed by a second compliance by inflation annually. However, once the
period of 2015–2017, and finally, 2018–2020), APCR is exhausted, market prices can con-
California requires a surrender of allowances tinue to rise above the APCR price unless the
equal to 30% of a covered entity’s emission obli- governor suspends the program per AB 32.).
gation for years preceding the final year in the 8. Participants register in advance and submit a
compliance period, with the remainder due the bid guarantee prior to the auction; bids sub-
final year. Québec entities surrender allowances mitted by participants are limited to a total
due in the final year of the compliance period with cost of no more than the amount of the bid
no interim obligation. There is a substantial pen- guarantee
alty for noncompliance, as entities that fail to 9. The auction is settled by selling allowances to
surrender the required allowances owe the juris- bidders in descending order of bid price
diction three times the unfulfilled emissions obli- until all allowances have been sold or until
gation. This penalty appears to be a strong there are no more bids (auction is under-
economic incentive to motivate compliance by subscribed); the last bid price at which all
covered entities. Reports issued by the jurisdic- allowances are sold is the price paid by all
tions indicate that compliance has been nearly successful bidders
100%, with only two California-registered entities 10. Results are published by jurisdictions within
failing to comply for the 2013–2014 period [31]. 5 business days.
The provisions for the quarterly auctions are as
follows: In 2012, the legislature passed AB 1532 and
SB 535, both of which broadly guide the expen-
1. Auctions are held each quarter. diture of auction revenues. AB 1532 requires that
2. Québec began participating jointly in the revenues be spent on programs and projects
November 2014, and in 2018 Ontario plans to reduce GHG emissions, with several second-
to participate. ary objectives, including directing expenditures
3. The auction is a first-price sealed-bid auction, to disadvantaged communities and households in
as recommended by economists that advised California, as well as expenditures to reduce the
CARB on market design. impacts of climate change on California. SB
4. Bids are submitted electronically to a website/ 535 further requires California Environmental
auction platform constructed and operated by Protection Agency (CalEPA) to identify disad-
a WCI contractor for the participating vantaged communities in California, with at least
jurisdictions. 25% of the auction revenues to be spent on pro-
5. Each quarter there is a current auction for grams that would benefit disadvantaged commu-
calendar year vintage allowances and an nities and 10% of the funds to be spent directly in
advance auction for allowances with a vin- those communities. AB 1532 requires a 3-year
tage 3 years after the current year. investment plan to determine program priorities,
6. Covered entities and other registered entities with actual expenditures by state agencies
can purchase in the current auction; the funded through the state’s annual budget process.
advance auction is for covered entities only. An initial investment plan was released in 2013,
7. There is a price floor for both auctions that is and a second investment plan, covering expendi-
adjusted annually for inflation; the 2012 price tures through the 2018–2019 fiscal year, was
Policy Instruments for Mitigating Carbon Dioxide Emissions 535

submitted to the legislature in January 2016 [32, Recycling revenue is a positive step, even more
33]. The California Climate Investments Annual so if combined with some form of income tax
Report provides details about specific programs reduction or an income tax rebate. However, elim-
and expenditures to date [34]. Spending has been ination of banking would increase price volatility
divided primarily to three categories: transporta- since it would reduce market liquidity and it
tion and sustainable communities, clean energy would make compliance more costly without
and energy efficiency, and natural resources and any change in GHG reductions by covered entities
waste diversion. Expenditures have been on a in the program. Numerous observers have
wide variety of projects, for example, high- expressed concern about the fact that auction and
speed rail, low-carbon transportation, dairy secondary market prices have hovered near the
methane reductions, water and energy efficiency price floor since May 2016. Making allowances
improvements, and energy efficiency in public ineligible for post-2020 compliance will continue
buildings. Through the 2017–2018 fiscal year, to depress allowance prices through 2020 as any
about $3.4 billion has been appropriated by the excess allowances will be quite literally worth-
legislature. less. With clear knowledge that the more stringent
Along with SB 584, SB 775 has been intro- emission reduction targets post-2020 mean a
duced in the 2017 legislative session, though the smaller supply of allowances, the price of pre-
legislation is still under consideration and may 2020 allowances would increase over the next
change considerably before passage or may not few years in the absence of this provision.
pass at all. SB 775 proposes significant changes to
California’s implementation of the cap and trade
program beginning in 2021. Briefly, SB 775 pro- Future Directions for Research
poses the following:
Two areas of additional research are worth men-
1. Most of the revenue would be recycled to tioning. First, this entry’s analysis of California’s
California taxpayers RPS program omitted potential problems in
2. Annual compliance instead of 3-year compli- wholesale and retail electricity markets from the
ance periods; also implies no allowance increase in California’s RPS. In January,
banking February, and March 2017, California paid Ari-
3. Allowances from the current program, those zona utilities to take its solar power to avoid
with a vintage prior to 2021, would be ineligi- overloading the California grid, i.e., electricity
ble for surrender for compliance purposes prices were negative on several days at specific
beginning in 2021. times. Negative pricing also occurred in 2014,
4. A stronger price ceiling mechanism that would 2015, and 2016, though with less frequency than
allow the sale of an unlimited number of allow- in 2017. Whether this phenomenon is simply the
ances at a predetermined price; if such a sale result of inadequate integration of solar into
were to occur, the allowances sold could California’s mix of electricity supplies, too rapid
undercut the GHG reduction goals in SB 32. an increase in the RPS (which is not binding for
5. Potentially stronger protection against carbon the state’s large utilities) or some more fundamen-
leakage (producers move production outside of tal issue in the design of the RPS program is not
California to avoid the need for allowances and well understood.
then import products to California) with a bor- Second, there have been few analyses of the
der tax adjustment for imports into California interaction between regulation and either taxes
with the tax varying according to the carbon and subsidies or cap and trade. As noted earlier
intensity of the imported good (legal issues in this entry, AB 32’s GHG reductions are to be
with the interstate commerce clause unknown, met with various regulatory programs, including
also issues with World Trade Organization). the RPS, the low-carbon fuel standard, various
6. No carbon offsets would be allowed. energy efficiency programs, as well as cap and
536 Policy Instruments for Mitigating Carbon Dioxide Emissions

trade. Bushnell et al.’s [35, 36] prospective anal- Conclusion

ysis of California’s cap and trade program and
Bushnell et al. [37] indicate that cap and trade Economic policy instruments are the means of
could lead to little change in GHG emissions implementing environmental goals, such as
from the electricity sector. When renewable elec- GHG mitigation. Several alternatives exist from
tricity resources outside of California exceed which to choose, with nearly all of them having a
California’s demand, generators may be able to rich history of both success and failure in the
meet their compliance obligations by changing environmental and other domains. This entry has
the source of electricity imports to California provided an introduction to and illustration of the
from coal-fired or other fossil fuel generation to major economic policy instruments used in envi-
renewable resources or other less carbon- ronmental policymaking: (1) regulations,
intensive resources. Referred to as “reshuffling” (2) taxes and subsidies, and (3) tradable property
(a special case of leakage), there may be no rights. It has illustrated their real-world applica-
change in electricity output or GHG emissions tion to California’s Renewables Portfolio Stan-
(in the WECC), although GHG emissions by dard, the British Columbia Carbon Tax, and the
California-covered entities would decline. Western Climate Initiative cap and trade program.
Reshuffling, if it does occur, does nothing to To solve difficult environmental problems such as
reduce emissions globally. The 2007 paper dem- climate change may take a combination of policy
onstrates that renewable capacity, including antic- instruments.
ipated load growth, was likely more than These policy instruments have been analyzed
sufficient relative to California demand and and refined for more than 50 years and have been
GHG reductions required for the electricity sector. reasonably successful in achieving their goals in
Even if this were not true, the additional increase relation to air, water, solid waste, and other types
in the RPS from 20% in 2010 (before cap and of pollution. The conceptual aspects are well-
trade took effect) to 33% in 2020, the fact that at developed for the established approaches, but fur-
least the investor-owned utilities were ahead of ther innovations, such as for the property rights
schedule in meeting the 2020 requirement, and the approach underpinning cap and trade, would be
growth in renewable capacity in the WECC would welcomed. The design and actual implementation
make it more likely that reshuffling would occur. of these policy instruments would benefit from
Section 95852(b)(2) of the California Regulation further research and experience. Another chal-
prohibits reshuffling, though the regulation pro- lenge is the idiosyncrasies of each individual
vides a safe harbor for (an exemption from application, which limit the knowledge transfer
enforcement of the reshuffling provision) electric- across individual applications. With more experi-
ity purchases to satisfy the purchaser’s compli- ence, however, one should expect that many of the
ance with the California RPS. design and implementation problems in recent
While reshuffling in the electricity market has years will be overcome.
been recognized by California regulators, it Yet another concern is the politics of design
remains a rare instance in which interaction and implementation. For example, in the area of
between regulation and cap and trade has been climate change negotiations, several valuable pol-
analyzed. For the foreseeable future, GHG taxes icy instruments exist and have had some success
or cap and trade programs will coexist with regu- in their implementation. The lack of progress in
lation, either current regulations or new regula- this realm, however, relates to the inability to
tions thought to be complementary to market- negotiate agreements within and across political
based GHG policy instruments. More research jurisdictions. This issue is even more pronounced
on interaction between different policy instru- at the international level, where there is no supra-
ments would improve understanding of possible national authority that can force in agreement. In
unintended and undesirable consequences of that context, the major stumbling block is the
coexisting policy instruments. burden-sharing of the costs of mitigating and
Policy Instruments for Mitigating Carbon Dioxide Emissions 537

sequestering GHGs. Fortunately, progress is 18. Carl J et al. (2013) Renewable and distributed power in
being made in lowering these costs and in the California – simplifying the regulatory maze. Hoover
Institution/Stanford University, Stanford
design of policy instruments that can reflect con- 19. California Public Utilities Commission (2016) Bien-
sensus on fairness. nial RPS program update
20. California Energy Commission (2017) Publicly
owned utility verification reports – compliance
period 1
Bibliography 21. Port of Stockton (2016) Renewable portfolio standard
procurement plan update
22. Stevens B, Rose A (2002) A dynamic analysis of the
1. (IPCC) Intergovernmental Panel on Climate Change
marketable permits approach to global warming pol-
(2014) Climate change 2014: mitigation of climate
icy: a comparison of spatial and temporal flexibility.
change. IPCC, Genf
J Environ Econ Manag 44:45–69
2. Arrow K (2013) Determining benefits and costs for
23. Schmalensee R (2012) Evaluating policies to increase
future generations. Science 341:349–350
electricity generation from renewable energy. Rev
3. Stern N et al. (2007) Stern review: the economics of
Environ Econ Policy 6:45–64
climate change. Cambridge University Press,
24. Schultz G, Baker III J (2017) A conservative answer to
Cambridge
climate change. Wall Street Journal, New York City
4. Ackerman B, Hassler W (1981) Clean coal, dirty air.
25. British Columbia (2008) Climate action plan. British
Yale University Press, New Haven
Columbia Ministry of Environment, Victoria
5. Coase R (1960) The problem of social cost. J Law
26. Murray B, Rivers N (2015) British Columbia’s
Econ 3:1–44
revenue-neutral carbon tax: a review of the latest
6. Hardin G (1968) The tragedy of the commons. Sci
“grand experiment”. Energ Policy 86:674–683
New Ser 162(3859):1243–1248
27. British Columbia (2016) Climate leadership plan. Brit-
7. Rose A (1990) Reducing conflict in global warming
ish Columbia, Victoria
policy: the potential of equity as a unifying principle.
28. Bovenberg A, Goulder L (2002) Environmental taxa-
Energ Policy 18:927–935
tion and regulation. In: Auerbach A, Feldstein M (eds)
8. Kverndokk S, Rose A (2008) Equity and justice in
Handbook of public economics. North Holland,
global warning policy. Int Rev Environ Resour Econ
Amsterdam, pp 1471–1545
2:135–176
29. California Air Resources Board (2010–2016) Cap-
9. Rose A, Stevens B, Edmonds J, Wise M (1998) Inter-
and-Trade regulation
national equity and differentiation in global warming
30. California Air Resources Board (2008–2016) Manda-
policy. Environ Resour Econ 12:25–51
tory reporting regulation
10. (UNFCCC) United Nations Framework Convention
31. California Air Resources Board (2016) 2013–2014
on Climate Change (2015) Historic Paris Agreement
Compliance obligation report
on climate change
32. California Air Resources Board (2013) Cap-and-Trade
11. Baumol W, Oates W (1975) The theory of environ-
auction proceeds investment plan: fiscal years
mental policy. Cambridge University Press,
2013–14 through 2015–16
Cambridge
33. California Air Resources Board (2016) Cap-and-Trade
12. Weitzman ML (1974) Prices vs. quantities. Rev Econ
auction proceeds second investment plan: fiscal years
Stud 41:477–491
2016–17 through 2018–19
13. Hahn R (1984) Market power and transferable prop-
34. California Air Resources Board (2017) Annual report
erty rights. Q J Econ 99:753–765
to the legislature on California climate investments
14. Hahn R, Hester G (1989) Where did all the markets
using Cap-and-Trade auction proceeds
go? An analysis of EPA’s emissions trading program.
35. Bushnell J et al. (2007) California’s greenhouse gas
Yale J Regul 6:109–153
policies: local solutions to a global problem? Center
15. Kolstad J, Wolak F (2008) Using environmental emis-
for the Study of Energy Markets, University of Cali-
sions permit prices to increase electricity prices: evi-
fornia Berkeley http://www.escholarship.org/uc/item/
dence from the California electricity market. Stanford
6m3355kj
University
36. Bushnell J et al. (2008) Local solutions to global
16. Pezzey J (1992) The symmetry between controlling
problems: climate change policy and regulatory juris-
pollution by price and controlling it by quantity. Can
diction. Rev Environ Econ Policy 2:175–193
J Econ 25:983–991
37. Bushnell J et al. (2014) Downstream regulation of CO2
17. Watson W, Ridker R (1984) Losses from effluent taxes
emissions in California’s electricity sector. Energ Pol-
and quotas under uncertainty. J Environ Econ Manag
icy 64:313–323
11:310–326
 Index

A Bioleaching, 333
Abandoned mine, 333 Biomining, 333
Abandoned mine methane (AMM), 251 and bioprocessing, 350–351
Acid (rock/mine) drainage, 333 Bioremediation, 351–352
irrigation with lime-treated, 348 Biotic and abiotic metal recovery, 352–353
Acidizing, 23 Bitumen/bitumoid, 47, 56, 101
Acidophile, 333 recovery from oil sand deposits, 37
Administrative cost, 521, 528 Bituminous coal, 309, 439
Air blown autothermal reforming, 215 Boiler exit gas temperature, 430
Airborne trace pollutants, coal combustion, 467 Boundary Dam Carbon Capture Project, 511
Air quality and dust, 381 Brake torque, 137
Air ratio, 430 Brayton cycle, 249, 262–266
Alaska gas hydrate production potential, 194 Brayton-Rankine combined cycle efficiency, 272
Alaska North Slope oil field developments, 187 British thermal unit (Btu), 203
Alkylation, 109–111
Ammonia
commercial activities, 244–245 C
process description, 241 California Air Resources Board (CARB), 533
production, 240–241 California Council on Science and Technology
reaction pathway, 241–242 (CCST), 87, 93
reactors and catalysis for synthesis, 242–244 California Environmental Protection Agency
Ammonification, 333 (CalEPA), 534
Amorphous, 333 California Public Utilities Commission (CPUC), 529
Anthracite, 309, 439 Calorific value, 309, 439
Archer Daniels Midland Company, 511 Cap and trade, 519, 527
Ash, 309, 357, 439, 444 Western Climate Initiative Program, 532–535
particles, 470 Carbon capture and sequestration (CCS), 4, 429
Associated natural gas, 22 Carbon capture and storage
Autothermal reforming (ATR), 203, 206, 211 Archer Daniels Midland Company, 511
Boundary Dam Carbon Capture Project, 511
carbon capture, 504–506
B CO2 injection projects, 512
Baghouse, 439 database, 511
Bakken formation, 12 deployment, 513–514
Balance of plant (BOP), 278 direct air capture, 508
Barnett Shale, 28 factors, 504
Barrel of oil equivalent (BOE), 203 geological sequestration, 508–510
Barrels (of oil) (bbl), 203 geological storage deployment, 510
Barrow Gas Field (BGF), 186, 187 geologic risk assessment tools, 514
Basin-centered gas, 10 hazards assessment, 514
Benefit-cost analysis, 519 monitoring and verification, 513
Best Available Control Technology (BACT), 523 oxy-combustion, 507
Billion barrels of oil (BBO) post-combustion capture, 506
Biogenic processes, 8 precombustion capture, 506–507

© Springer Science+Business Media, LLC, part of Springer Nature 2020 539

R. Malhotra (ed.), Fossil Energy,
https://doi.org/10.1007/978-1-4939-9763-3
540 Index

Carbon capture and storage (cont.) modeling pollutant formation and control, 475–478
risk management, 514 particulate emissions, 442–444
technologies, 299 particulate matter (PM), 475–476
technology pathways, 504, 505 pollutant formation, coal boilers, 468–470
water issues, 514 simulation methodologies, 470–475
Carbon dioxide (CO2), 439 slagging, 476
augmented waterflooding, 26 Coal gasification combined gas turbine-steam cycle
capture and compression in PC and IGCC Plant, (IGCC), 432–433
433–435 Coalification, 309
indirect coal liquefaction (ICL), 418–420 Coal liquefaction
injection projects, 512 direct, 387, 388, 393–401
Carbon dioxide emissions, policy instruments indirect, 387, 389, 401–420
design and implementation, 520 pyrolysis processes, 392
environmental goal setting, 521–523 survey of specific, 391–401
evaluative criteria, 520–521 Coal mine methane (CMM), 11, 251
operational examples, 529–535 Coal mining
regulation, 523–524 air pollution, 339–341
taxes and subsidies, 524–526 dust impact, 341
tradable property rights, 526–529 excavation process, 339
Carboniferous, 309 growth rate, 337
Carbon monoxide (CO), 439, 478–480 pollution from, 339
Carbon tax, 519 process of coal excavation, 339
in British Columbia, 531–532 water pollution, 341–342
Carnot cycle, 249 See also Mining operations
efficiency, 143, 264 Coal power and pollutant emissions, 468
Casinghead gasoline, 22 Coal preparation, 309
Catalyst beds and reactors, 225–227 air quality and dust, 381
Catalytic membrane reactor (CMR), 203, 217–218 beneficial impacts, 370–376
Catalytic partial oxidation (CPOx), 203 centrifugal dewatering, 367
reforming, 215–217 challenges for, 376–381
Catalytic reforming, 111–112 classifying cyclones, 362
Catalytic two-stage liquefaction (CTSL) process, 396 cleaning characteristics assessment, 358–359
CFBC with Oxy/FGR, 435–436 cleaning of fine particles, 376–377
Charge heating, 141 coal dewatering, unit operations, 366–367
Chemical looping, 507 coarse waste disposal, 378
combustion, 505 conversion processes, 382–383
Christmas tree, 19 definition, 357
Clathrates, see Gas hydrates dense medium separators, 363–364
Clean Air Act, 523 dewatering of fine particles, 377
Cleats, 11 dry processing, 381–382
Close-coupled integrated two-stage liquefaction environmental issues, 378
(CC-ITSL), 397 filtration dewatering, 367–368
Coal, 439 froth flotation separators, 365–366
abundance of, 440 on-line analysis and control, 378
disadvantages, 429 pollution abatement, 372–376
properties, 443 processing operations, 359–370
Coal bed methane (CBM), 5, 11, 251 process water quality, 379–380
Coal combustion reserves, 370–371
airborne trace pollutants, 467 screens and sieves, 361
air pollutant emissions from, 441–442 sizing of fine particles, 377–378
chemical equilibrium, 471 slurry disposal and storage, 378–379
chemistry models, 471 thermal dewatering, 368
computational fluid dynamics (CFD), 472–475 thickening and clarification, 369
correlations, 470–471 transportation, 371
electrostatic precipitators (ESPs), 476–477 unit operations, 363
fabric filters, 477–478 utility performance, 371–372
fouling, 476 waste disposal and storage, 369–370
gas-phase kinetic mechanisms, 471 water-based separators, 364–365
inherent in, 468 Coal resources, 3
Index 541

Coal seam gas (CSG), 11 hydrocracking, 107–109

Coal-to-liquids (CTL), 203, 387 hydrotreating, 107–109
definition, 388 ignitability-flammability, 125
processes, 393 isomerization, 112
Coarsewaste disposal, 378 leaching, 124
Coase Theorem, 528 liquid effluents, 120–121
Cobalt, 409 lower-boiling constituents, 125
Cogeneration, 249 and natural gas (see Oil and natural gas)
Cold flame oxidation, 246 polymerization, 109–111
Cold heavy-oil production with sand (CHOPS), 41 reactivity, 125
Cold-water process, 36 recovery program, 17
Combined cycle power plant, 249 refinery, 102
Combined heat and power (CHP), 249 refinery emissions, 115–116
Combustion refinery processes, 102–103
efficiency, 143 release into environment, 122–123
natural gas, 286–294 sedimentation/adsorption, 124
oil sand exploration, 38–39 solid effluents, 121
petroleum exploration, 26, 27 spreading, 124
processes, 38–40 storage and handling, 121–122
stroke, 136 total petroleum hydrocarbons (TPH), 127–128
Combustor, 249 toxicity, 125
Community, 333 vis breaking and coking, 105–106
Compliant tower platform, 30 waste cleanup, 129
Compression waste management, 128–129
ignition combustion, 159 waste treatment and disposal, 129
stroke, 136 wastewater, 128
Compressor, 249 wind, 124
Computational fluid dynamics (CFD) tools, 467 Crystalline, 334
accurate problem description, 474 Cut point, 357
calculations, 474 Cut size, 357
coal combustion modeling, 472–473 Cyclic steam injection
results evaluation, 475 oil sand exploration, 38
software selection, 474 petroleum exploration, 26, 27
usage guidelines, 473–475
user experience, 474
Contaminated land/water, 333 D
Conventional oil and gas, 5 Deasphalting, 112–113
Coprocessing (of coal), 387 Denitrification, 334
Corrosivity, 125 Dense medium, 357
Cretaceous, 309 separators, 363–364
Cross-State Air Pollution Rule (CSAPR), 468 Density functional theory (DFT), 232
Crude oil (petroleum), 101 Department of Energy (DOE), 203, 387
alkylation, 109–111 Desalting, 103–104
catalytic reforming, 111–112 Devonian, 309
corrosivity, 125 Devonian–Mississippian black shales, 11
deasphalting, 112–113 Dewatering
desalting, 103–104 centrifugal, 367
dewaxing, 113 coal, unit operations, 366–367
dispersion, 123 filtration, 367–368
dissolution, 123–124 of fine particles, 377
distillation, 104–105 thermal, 368
emulsification, 124 Dewaxing, 113
entry into environment, 121 Diffusion heat release, 160
environmental regulations and pollutants, 113–115 Dimethylether (DME), 203, 391, 402, 414–418
evaporation, 124 commercial activities in synthesis, 236–237
fluid catalytic cracking, 106–107 gas phase production, 237
gaseous emissions, 116–120 vs. methanol, 236
hazardous chemicals, 125–126 reaction pathway, 236
higher-boiling constituents, 126–127 slurry phase production, 237–238
542 Index

Direct air capture (DAC), 505, 508 Electric power, 427

Direct capture of high-purity streams, 505 Electrostatic precipitator (ESP), 439, 446–449, 476–477
Direct coal liquefaction (DCL), 387, 388 Electrowinning, 334
advantages, 401 Emissions, 249
block flow diagram (BFD), 394, 395 allowance auction, 519
catalytic two-stage liquefaction (CTSL) process, 396 natural gas, 286–294
in China, 400 Emulsification, 124
coprocessing, 401 Energy
disadvantages, 401 definition, 310
in Europe, 398 resources, 87, 310
H-Coal, 395, 396 Energy Information Administration (EIA) data, 317, 427
H/C ratio, 393 Enforceability, 521
hydrogen, 394 Engine downsizing, 170
in Japan, 399 Enhanced oil recovery (EOR), 37, 504
operating conditions, 394 chemical flood recovery methods, 26
Wilsonville Facility in United States, 397–398 gas flood recovery methods, 26
Direct line-drive pattern, 25 methods, 25–28
Direct retorting, 47 thermal recovery methods, 25–27
Discrete fracture network (DFN), 91, 92 Environmental aspects, of hydraulic fracturing, 93
Dispersion, 123 ground-water contamination, 93–94
Dissimilatory reduction, 334 methane emissions, 94–96
Dissolution, 123–124 water usage, 94
Dissolved gas drive, 22 Environmental Impact Statement (EIS), 337
Distillate Environmental issues, 378
light, 203, 387 Environmental regulation, 519
middle, 203, 387 benefit-cost analysis for, 522
Distillation, 104–105 and pollutants, 113–115
Drilling depths, oil and natural gas, 10 Equity, 519
Drilling, petroleum exploration dynamic, 521
directional drilling, 20–21 static, 521
seismic or electric energy, 18 Evaluative criteria, 519
without depth control, 18 Evaporation, 124
Dry flue gas desulfurization, 439, 454–455 Exergy analysis, engine system, 152–156
See also Gas desulfurization Exhaust stroke, 136
Dry processing, 381–382 Expansion stroke, 136
Dry sorbent injection (DSI), 489–490 Exploration, 101, 309
Ex situ retorting, 47
Extractant, 334
E Extraction, 337
East Barrow field material balance study, 198–199 Extra heavy oil, 32
Economic efficiency, 519 Exxon Donor Solvent (EDS) process, 398
dynamic, 521 ExxonMobil, 239
static, 521
Economic factors, 89, 98
Economic policy instruments, 519 F
design and implementation, 520 Fabric filter, 439, 477–478
environmental goal setting, 521–523 baghouse, 449–451
evaluative criteria, 520–521 Field scale reservoir simulation study, 199–200
operational examples, 529–535 Fine particles, coal
regulation, 523–524 cleaning of, 376–377
taxes and subsidies, 524–526 Firing temperature, 249
tradable property rights, 526–529 Fischer assay yields, 64
Efficiency, 249 Fischer-Tropsch (F-T) reaction, 203, 387, 390
Effective stimulated volume (ESV), 92 ASF distribution, 224
Electrical submersible pump (ESP), 23 catalyst beds and reactors, 225–227
Electricity, 3 catalysts, 224–225
generation by fuel, 1970–2020, 428 chain initiation and propagation, 224
generation, world net, 253 chain termination, 224
prosperity and, 428 commercial activities in, 227
Index 543

C1 species formation, 220 resource potential, 191

elements, 221–223 resource pyramid, 190
indirect coal liquefaction (ICL), 403 in South Korea, 192
product upgrading, 227 stability modeling, BGF pools, 197–198
reaction pathway, 220 in United States, 193
synthesis, 219–220 world distribution, 189
Five-spot pattern, 24 Gasification (of coal), 387
Flexibility, 521 Gas-phase kinetic mechanisms, 471
Floating production platform, 30–31 Gas to liquids (GTL), 203
Floating tension leg platform, 21 chemical conversion, 205
Flue gas desulfurization (FGD), 439 definition, 204
Flue gas recirculation (FGR), 435 location and transportation, 205
Fluid catalytic cracking, 106–107 modern synthesis technology, 204
Flush stage, 21 survey of, 206
Formaldehyde, 240 Gas turbine, 249
Fossil energy, world production, 311 Brayton cycle, 262
Fossil fuels, 440 F-Class, 257
Fouling, 476 heavy-duty industrial, 262
Four-spot pattern, 24 history of, 257–261
Four-stroke cycle engines, 135 thermodynamics, 261–277
Fracture-assisted steam technology (FAST), 38 Generator, 250
Fracture characterization, 89–91 Geological carbon sequestration (GCS), 504
Friction work, 138 Geological sequestration, 508–510
Froth flotation separators, 365–366 Geologic carbon storage risk assessment tools, 515
Fuel-air cycle analysis, 145, 146, 148 Geomechanics, of hydraulic fracturing, 90–93
Fuel conversion efficiency, 142 Geophysical borehole logging, 18
Fuel knock, 159, 168 Global warming, 250
Glow plugs, 159
Gravity base complex, 30
G Gravity drive, 23
Gas cap, 20 Greenhouse effect, 250
drive, 22 Greenhouse gas (GHG) emissions, 532
Gas desulfurization Ground-water contamination, 93–94
dry flue, 439, 488–489 Gusher, 21
wet flue, 486–488 Gypsiferous, 342
Gaseous emissions, 116–120 Gypsum, 334
Gas flood, 23 precipitation, 348
Gas-heated reforming/heat exchange reforming
(GHR/HER), 203, 206, 213
Gas hydrates, 2, 5, 12 H
Alaska gas hydrate production potential, 194 Hazards, 514
in Canada, 193 Hazardous chemicals, 125–126
in China, 192 H-Coal, 395
definition, 185 Heap leaching, 334
East Barrow field material balance study, 198–199 Heat engine, 143
exploration in Alaska, 194 Heat exchange reforming (HER), 203, 206, 211
field scale reservoir simulation study, 199–200 Heating value, 250
global projects, 192–193 Heat rate, 250
history of, 185–186 Heat recovery steam generator (HRSG), 250, 432
Hot Ice No. 1, 194–195 High-acid crude oil, 32
in India, 193 Higher-boiling constituents, 126–127
in Japan, 192 High temperature Fischer-Tropsch (HTFT)
Mt. Elbert-01, 195–196 technology, 390
in New Zealand, 193 Homogeneous charge compression ignition (HCCI)
North Slope Borough Methane Hydrate Project engines, 176
Review, 196 Hot Ice No. 1, 194–195
production and commercialization, 189 Hot-water process, 35–36, 101
research on Alaskan gas hydrates, 186 Hubbert curve, 7
resource, global distribution of, 188–191 Huff and puff, see Cyclic steam injection
544 Index

Hydrate Energy International (HEI), 189 Intended Nationally Determined Contributions

Hydraulic fracturing/fracking (HF), 23, 28–30, 99 (INDCs), 528
definition, 86–87 Internal combustion (IC) engines, 254
economics of, 97–98 compression ignition combustion, 159–163
environmental impact of (see Environmental aspects, definition, 133
of hydraulic fracturing) diesel vs. gasoline engine, 167–170
fracture characterization, 89–91 emission formation and exhaust pollution, 163–166
geomechanics of, 90–93 future research, 170
global oil and gas resources, assessment of, 97 history, 134
and horizontal drilling, 87, 88 operating cycle, 134–136
induced seismicity, 96–97 power, 137
operational aspects of, 98 spark ignition combustion, 156–159
shale production, explosion of, 88 thermodynamic analysis, 142
Hydrocarbons synthesis, 206 types, 133
Hydrocracking, 107–109 uses, 134
Hydrocracking heavy paraffins (Wax), 227–229 International Institute for Applied Systems Analysis
Hydrogen, 394 (IIASA), 189
Hydrotreating, 107–109 Inverted seven-spot pattern, 25
light paraffins, 229–230 Isomerization, 112

J
I Jurassic period, 309
Ignitability-flammability, 125
Immiscible carbon dioxide displacement, 26
Immiscible nonhydrocarbon gas displacement, 26 K
Incentive-based policy instruments, 524 Kerogen, 47
Indirect coal liquefaction (ICL), 387, 389 bitumen/bitumoid, 56
carbon dioxide, 418–420 chemical structure, 57
cobalt, 409 classification, 53
dimethylether (DME), 402 decomposition rates, 61
downstream product upgrading, 413–414 degradative analysis, 55
entrained flow gasifiers, 405 destruction, 61
environmental factors, 418 elemental compositions, 53
Fischer-Tropsch (F-T) reaction, 403, 407 insolubility, 53
fluidized bed gasifiers, 404–405 isolation, 53–54
gasification, 403–404 laboratory techniques, 55
HTFT reaction, 410 macromolecular structure, 52–53
LTFT reaction, 410 mechanisms and kinetics, 68–69
methanol and dimethylether, 414–418 nondestructive analysis, 55–56
moving bed gasifiers, 404 during pyrolysis, 67–68
parameters and product distribution, 407 separation methods, 54
secunda plant block flow diagram, 411 types, 54
steps for, 402 Kerr-McGee critical solvent de-ashing (CSD)
utility water, 420 technology, 397
water-gas shift reaction (WGSR), 412 Knuckle joint, 21
Indirect retorting, 47
Induced seismicity, 87, 89, 93, 96–97 L
Inert gas injection (IGI), 40 Leaching, 124
Injectors, 24, 25 Lenoir engine, 135
In-line four-cylinder engine, 136 Lignite, 439, 442, 445
In-place resources Limestone reservoir rock, 21
in hydrate-bearing sands, 190 Limestone scrubbing process
and recoverable resource, 186 dry, 454
unconventional gas hydrate, 186 wet, 453
In situ conversion, 101 Linear elastic fracture mechanics (LEFM) model, 91
In situ retorting, 47 Liquefaction energy, 207
Intake stroke, 136 Liquefied natural gas (LNG), 7, 203
Integrated gasification CC (IGCC) power plants, 299 air blown autothermal reforming, 215
Integrated two-stage liquefaction (ITSL) mode, 397 applications, 207
Index 545

autothermal reforming (ATR), 211 Microbial enhanced oil recovery (MEOR), 42

catalytic membrane reactors (CMR), 217–218 Microemulsion flooding, 26
catalytic partial oxidation (CPOx) reforming, 215–217 Micro-seismic data, 91, 92
compact reforming, 211 Miller concept, 172
gas-heated reforming/heat exchange reforming Mining life cycle, 334, 336–337
(GHR/HER), 213 Mining operations
heat exchange reforming, 211 and beneficiation, 335
liquefaction energy, 207 coal, 337–342
methane reforming (see Methane reforming) of mineral resources, 336
partial oxidation, 215 recognition and acceptance, 334
pipelines construction, 207 Miscible fluid displacement, 26
plant considerations, 207–206 Modified in situ extraction processes, 38
postreforming, 218 Modular Formation Dynamics Tester
prereforming, 218 (MDT™), 186
production of, 207 Monitoring and verification (M&V), 513
propane prechilling, 208 Mt. Elbert-01, 195–196
propylene compressor, 208
refrigeration system in, 208
shell side combustion zone, 211 N
soot-free operation, 212 National Ambient Air Quality Standards
steam methane reforming (SMR), 209–211 (NAAQS), 468
thermal efficiency, 207 National Environmental Policy Act (NEPA), 523
Liquefied petroleum gas (LPG), 388 Natural gas, 3
Liquid effluents, 120–121 advantages of, 250
Liquids from Coal (LFC™), 392 associated gas, 8
Lower-boiling constituents, 125 bulk of, 205
Low-NOX burners (LNB), 439 coal bed methane, 11
Low temperature combustion (LTC), 176 coal mine gas, 251
LTFT naphtha and blends, 230 cogeneration, 280–286
®
Lurgi MegaMethanol process, 235 combustion and emissions, 286–297
composition and heating values, 252
in conventional fields, 7–10
M dual-fuel and, 255
Macarels, 309 energy content, 252
Macrophyte, 334 forming, 251
Marcellus Shale, 28 on future electrical production, 250
Mercury (Hg), 492–495 future shortage predictions, 7–8
Mercury and Air Toxics Standards (MATS), 468 gas hydrates, 12
Mesophile, 334 gasoline, 22
Methane (CH4), 251 history of, 256–261
coal bed methane, 11 to liquids, direct conversion of, 245
content, natural gas, 204 methane content, 204
Methane reforming methanol, selective oxidation of, 245–246
in F-T LNG, 209 nonassociated gas, 8
for methanol/DME synthesis, 218–219 production and distribution companies, 252
Methanogenesis, 334 production history, 7–10
Methanol, 414–418 shale gas, 12
activities in synthesis, 233–234 tight gas, 10
aging time for selected catalysts, 233 Natural gas liquid (NGL), 9, 203
to chemicals, 239–240 Natural gasoline, 22
density functional theory (DFT), 232 New Energy and Industrial Technology Development
dimethylether vs., 236 Organization (NEDO), 389
process description, 230–231 Nine-spot pattern, 24
production of, 206 Nitric oxide (NO), 440
reaction pathway, 231–233 Nitrogen dioxide (NO2), 440
reduction mechanism of CO, 232 Nitrogen oxides (NOx), 440, 480–485
removal, 232 combustion zone, 482–483
selective oxidation of natural gas, 245–246 emissions control technologies, 455
Methanol to gasoline (MTG) process, 238–239 formation, 445–446
546 Index

Nitrogen oxides (NOx) (cont.) oleophilic sieve process, 36

low-NOx burners, 455–456 opportunity crude oils, 32
post-combustion zone, 483–485 sand-reduction process, 36
Nonintegrated two-stage liquefaction (NTSL), 397 solvent extraction process, 36
Non-mining methods, oil sand spherical agglomeration process, 36
bitumen from oil sand deposits, 37 Oil shale, 5, 11–12, 47
cold heavy-oil production with sand, 41 assay methods, 62
combustion processes, 38–40 average oil yield, 50
enhanced oil recovery, 37 characteristics, 49–57
inert gas injection, 40 chemical aspects, oil recovery by retorting, 66–67
for long horizontal wells, 40 chemical nature, 49
microbial enhanced oil recovery, 42 composition, 51
modified in situ extraction processes, 38 conversion processes, 58–60
pressure-pulsing technologies, 41 definition, 48, 49
steam-assisted gravity drainage, 40–41 direct combustion, 58–59
steam-based processes, 38 ex situ retorting technologies, 75
thermal recovery methods, 37–38 Fischer assay yields, 64
vapor-assisted petroleum extraction, 41 gas atmosphere on retorting, 65–66
North Slope Borough Methane Hydrate Project heating rate effects, 65
Review, 196 heating value, 50
heat transfer medium, 76
high-temperature pyrolysis and gasification, 59–60
kerogen, 52–53, 67–68 (see also Kerogen)
O low-temperature pyrolysis, 59
Offshore production, petroleum mineral matter, 51
compliant tower platform, 30 organic and inorganic phases, 48
floating production platform, 30–31 organic matter, 52
gravity base complex, 30 pressure effects on retorting yields, 65
semisubmersible platform, 31 process temperatures, 60–61
shallow water complex, 30 production, 60–72
SPAR platform, 31 pyrolysis phenomena, 67
subsea production systems, 31 recovery, technological aspects, 72–77
tension leg platform, 31 in situ technologies, 76–77
Oil, 2 thermal effects during retorting, 62
mobility ratio, 24 thermochemical-based conversion technologies, 60
Oil and natural gas thermochemical conversion, 64
coal bed methane, 11 trace elements, 62
in conventional fields, 7–10 Oleophilic sieve process, 36
estimated volumes of, 12–14 On-line analysis and control, 378
future shortage predictions, 7–8 Onshore production, petroleum, 19–21
gas hydrates, 12 adverse effects, 21, 22
oil sands, 10–11 drilling a well, 20–21
oil shale and shale oil, 11–12 enhanced oil recovery methods, 25–28
production history, 5–7 hydraulic fracturing, 28–30
shale gas, 12 injection and producing wells, 20
tight gas, 10 primary recovery methods, 18, 21–23
Oil mining, 32–33, 101 secondary recovery methods, 18, 23–25
reservoir properties, 34 Opportunity crude oils, 32
Oil sand, 5, 10–11 Ordovician, 309
cold-water process, 36 Ore, 334
direct heating, 36–37 Organic efficiency, 357
extra heavy oil, 32 Organization of the Petroleum Exporting Countries
high-acid crude oil, 32 (OPEC), 6
hot-water process, 35–36 Original oil-in-place (OOIP), 21
improved mining, 37 Otto cycle, 143, 144
mining, 34–35 Overfire air (OFA), 440, 456–457
non-mining methods (see Non-mining methods, Oxidative coupling to light hydrocarbons, 245
oil sand) Oxy-combustion, 507
oil mining, 32–33 Oxyfiring combustion, 504
Index 547

P Processing operations, 359–370

PAH-containing soot aerosols, 3 Process water quality, 379–380
Partial oxidation (POx), 203, 206 Propane, prechilling, 208
liquefied natural gas, 215 Property rights, 526–529
Particulate emissions, 442–444 Proppants, 21, 29
Particulate matter (PM), 475–476 Propylene, compressor, 208
control technologies, 446 Pulse jet baghouse, 450
PC/Oxy/USC, 435 Pulverized coal in ultra-supercritical steam cycle
Peat, 310 (PC/USC), 430
formation of, 311–313 Pyrolysis (of coal), 47, 387, 392
occurrences, 313–314
principle mechanism, 313
production, 323–326 Q
resource and reserve classifications, 314–317 Quasi-static effect, 141
resources in world, 317–320
in world, 320–323 R
Perforating gun, 19, 28 Ram effect, 142
Permafrost, 185 Rankine cycle, 174, 250
Permian, 309 efficiency, 430
Petra Nova Parish Holdings project, 511–512 Raton Basins, 11
Petroleum, 310 Raw coal oil, 388
exploration, 17–19 Reactivity, 125
offshore production (see Offshore production, Reburning, 440, 457–458
petroleum) Reclamation, 334
onshore production (see Onshore production, Reconfigured integrated two-stage liquefaction
petroleum) (RITSL), 397
production rates, 17 Recovery, 101
site specificity, 17 Refinery, 102
See also Crude oil (petroleum) emissions, 115–116
Petrophysical properties, 185 processes, 102–103
Phytotechnology, 348–350 Remediation, 334
Piston/cylinder reciprocating engine Renewables Portfolio Standard program, 519, 529
four-stroke cycle, 136 Reserves, 5, 309
geometrical considerations, 134 (see also Internal coal, 370–371
combustion (IC) engine) growth, 9
mean effective pressure, 138 Reservoir, 8
power/torque/speed curve, 139, 140 geometry, 17
pressure-crankangle diagram, 137 rock, 21
P-V diagram, 138 Resources, 5, 185, 309
thermodynamic work, 137 Restoration, 334
torque, 137 process, 342–344
volumetric efficiency curve, 141 Retort gas, 47
Pollution abatement, 372–376 Retorting, 47
Pollution tax, 525 residue/spent shale, 47
Polyaromatic hydrocarbons, 165 Retort water, 47
Polymer-augmented waterflooding, 26 Reverse air flow baghouse, 450
Polymerization, 109–111 Risk assessment factors, 514
Polyphenylene sulfide (PPS), 451
Polytetrafluoroethylene (PTFE), 451
Porous rock, 18 S
Post-combustion capture (PCC), 504, 506 Sand-reduction process, 36
Postreforming, 218 Sandstones, 10
Powder River Basin, 11 San Juan Basins, 11
Power stroke, 136 Sasol Slurry Phase Distillate (SSPD) process, 390
Precombustion capture, 504, 506–507 Sedimentation/adsorption, 124
Prereforming, 204, 218 Selective catalytic reduction (SCR), 440, 459–460, 484
Pressure-pulsing technologies (PPT), 41 Selective noncatalytic reduction (SNCR), 440, 458–459
Probabilistic risk assessment (PRA), 514 Semisubmersible platform, 31
Process development unit (PDU), 389 Seven-spot pattern, 24
548 Index

Shale gas, 5, 12 Thermogenic processes, 8

Shale oil, 11–12, 47 Three-pressure, reheat (3PRH) combined cycle
Shale resources, 88, 89, 94, 97 diagram, 278
Shallow water complex, 30 Tight gas, 5, 10
Silurian, 309 Tight oil, 48
Slagging, 476 Toe-to-heel air injection (THAI), 39
Slurry disposal and storage, 378–379 Topsøe’s Integrated Gasoline Synthesis (TIGAS)
Solid effluents, 121 process, 239
Solution gas drive, 22 Total petroleum hydrocarbons (TPH), 127–128
Solvent extraction process, 36 Toxicity, 125
Spark ignition combustion, 156 Transaction costs, 528
SPAR platform, 31 Trap, 17
Spherical agglomeration process, 36 Triassic, 309
Staggered line-drive pattern, 25 Trillion cubic feet (TCF), 13
Static equity, 521 Tubing, 19
Steam-assisted gravity drainage (SAGD), 40–41 Turbocharging, 171
Steam-based processes, for oil recovery, 38
Steam drive injection
oil sand exploration, 38 U
petroleum exploration, 25–27 Ultra-supercritical (USC) steam, 431
Steam methane reforming (SMR), 204, 206, 209–211 Unconventional oil and gas, 5
Steam soak, see Cyclic steam injection US Geological Survey (USGS), 9, 13
Steam turbine, 250
Storage and handling, 121–122
Stress field, 91
Stripping ratio, 309 V
Subbituminous coal, 440 Vapor-assisted petroleum extraction (VAPEX), 41
Subsea production systems, 31 Visbreaking and coking, 105–106
Subsurface fluid injection and production (SFIP), 92 Volatile matter, 310, 440
Sulfur dioxide (SO2), 440, 485–490
emissions control technologies, 451–452
Sulfur oxides, 440 W
formation, 444–445 Washability, 357
Sulfur trioxide (SO3), 490–492 Waste heat recovery, 173–174
Supercharging, 170–172 Waste management, 128–129
Surfactant flooding, 26 cleanup, 129
Surfactant-polymer flooding, 26 disposal and storage, 369–370
Sustainability and mining, 344–347, 354 treatment and disposal, 129
Synthesis gas (syngas), 204, 206, 387 Waste rock, 334
generation technologies from natural gas, 210 Wastewater, 128
intermediacy, 206 Water-based separators, 364–365
Water-cooled surface condenser, 278
Water drive, 22
T Water-flooding process, 23–24
Tailings, 334 Water-gas shift reaction (WGSR), 204, 387, 412
Tar sand mining, 101 Water quality and acid mine drainage, 353
Taxes and subsidies, economic policy instruments, Water use and quality, 514
524–526 Western Renewable Energy Generation Information
Tension leg platform, 31 System (WREGIS), 530
Terminal attemperators, 277 Wet flue gas desulfurization, 440, 452–454
Tertiary, 309 See also Gas desulfurization
Thermal barrier coatings (TBC), 254 Whipstock, 21
Thermal decoupling, 277 Wilsonville Facility in United States, 397–398
Thermal efficiency, 207 Wind, 124
Thermal recovery methods World Energy Council (WEC), 317
oil sand exploration, 37–38
petroleum exploration, 25–27
Thermobitumen, 47 Y
nature and role of, 69–72 Yield, 357