Jean-François NAUROY

IFP Energies nouvelles

GEOMECHANICS
APPLIED TO
THE PETROLEUM
https://telegram.me/Geologybooks

INDUSTRY

2011

Editions TECHNIP 25 rue Ginoux, 75015 PARIS, FRANCE

 FROM THE SAME PUBLISHER

• Heavy Crude Oils

From Geology to Upgrading. An Overview
A.Y. HUC

• CO2 Capture
Technologies to Reduce Greenhouse Gas Emissions
F. LECOMTE, P. BROUTIN, E. LEBAS

• Corrosion and Degradation of Metallic Materials

Understanding of the Phenomena and Applications in Petroleum and Process Industries
F. ROPITAL

• Multiphase Production
Pipeline Transport, Pumping and Metering
J. FALCIMAIGNE, S. DECARRE

• A Geoscientist’s Guide to Petrophysics

B. ZINSZNER, F.M. PERRIN

• Acido-Basic Catalysis (2 vols.)

https://telegram.me/Geologybooks

Application to Refining and Petrochemistry

C. MARCILLY

• Petroleum Microbiology (2 vols.)

Concepts. Environmental Implications. Industrial Applications
J.P. VANDECASTEELE

• Physico-Chemical Analysis of Industrial Catalysts

A Practical Guide to Characterisation
J. LYNCH

• Chemical Reactors
From Design to Operation
P. TRAMBOUZE, J.P. EUZEN

• Petrochemical Processes (2 vols.)

Technical and Economic Characteristics
A. CHAUVEL, G. LEFEBVRE

• The Technology of Catalytic Oxidations (2 vols.)

P. ARPENTINIER, F. CAVANI, F. TRIFIRO

• Marine Oil Spills and Soils Contaminated by Hydrocarbons

C. BOCARD

All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopy, recording, or any information storage and retrieval system, without the
prior written permission of the publisher.

© Editions Technip, Paris, 2011.

Printed in France
ISBN 978-2-7108-0932-6
 Preface

According to great organisations involved in energy prospective studies as IEA (Interna-

https://telegram.me/Geologybooks

tional Energy Agency) and DOE (US Department of Energy), world energy demand will
still grow in the coming years and fossil energies will contribute in the range of 60 up to
80% to the energy supply, oil and gas contributing to about 50% of the energy supply. This
has two major implications. On one hand, oil and gas will have to be produced from more
and more difficult fields; on the other hand, more CO2 will be emitted and will have to be
captured and stored.
Both fluid production and CO2 storage involve fluid flow within porous rocks. This
induces pressure changes, saturation changes, temperature changes with water injection or
thermal recovery, even fluid interaction with the rock for CO2 injection. All these phenom-
ena have an impact on stresses and strains, or on the mechanical behaviour of rocks.
Rock mechanics in the oil and gas industry was traditionally used for drilling, wellbore
stability and fracturing. Reservoir application came with the observed consequences of sub-
sidence and the associated reservoir compaction. It should be stressed here that to satisfy the
increase in energy demand, fluids will be produced from complex environments where the
mechanical behaviour of the rock will be more and more important, for instance for wellbore
stability or faults behaviour.
A new area is emerging with CO2 management and more specifically CO2 sequestration.
Massive storage will occur by injecting in deep saline aquifers. Along with the description
of the reservoir behaviour, the knowledge of the overburden behaviour from a mechanical
standpoint will be essential to assess the storage integrity.
This book covers all these aspects and aims at providing a large range of potential users
with a simple approach of a broad field of knowledge. The reader will enjoy the tables of
reference values for the needed parameters such as wave velocity, elastic properties given
for both static and dynamic properties. Furthermore, the Figures are easy to understand and
illustrate very accurately the theoretical development. The book is application oriented, thus
closed forms, or empirical formulae are provided to give first estimates of the problem. For
instance, the reader will find indicators of drilling efficiency, closed forms for hydraulic or
thermal fracturing or for subsidence prediction. The practical consideration is always
 nearby. Hence, sections are devoted to lab determination of properties, in situ determination
of stresses, sand production prevention and finally geomechanical monitoring of reservoirs.
While rock mechanics is of importance in many fields of application such as drilling,
production or reservoir engineering, it is still thought of as a complex field restricted to a
few experts. I hope and even guess that, through this book, any engineer involved in those
fields will understand the basics and the practical applications of rock mechanics.

Maurice Boutéca
Director
Resources Business Unit
IFP Energies nouvelles
https://telegram.me/Geologybooks
 Table of Contents
https://telegram.me/Geologybooks

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII

List of Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XV

Main Symbols. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XVII

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XIX

Chapter 1
ELEMENTS OF ROCK MECHANICS FUNDAMENTALS

1.1 Basic notions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1.1 Rock characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1.1 Minerals and rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1.2 Rock classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1.3 Sedimentary rock and porous medium . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.1.4 Properties of heterogeneous media. . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.2 Stresses and strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.2.1 Stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.2.2 Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.1.2.3 Effective stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.1.2.4 State Boundary Surface (SBS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.1.2.5 Elastic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.1.2.6 Dynamic moduli of elasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.1.2.7 Anisotropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.2 Geomechanical characterisation of rocks in the laboratory. . . . . . . . . . . . . . 15
1.2.1 Laboratory tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.2.2 Tests without interstitial fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
 X Table of Contents

1.2.2.1 Uniaxial compression test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

1.2.2.2 Uniaxial tension test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.2.2.3 Brazilian test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.2.3 Uniaxial strain (or oedometric) test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.2.4 Conventional triaxial test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.2.5 Effective stresses and drained triaxial test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.2.6 Rock behaviour during compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.2.6.1 Elastic behaviour and measurement of elastic parameters . . . . . . . . 23
1.2.6.2 Sandstones (damage, failure). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.2.6.3 Carbonates (plasticity, failure) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.2.6.4 Shales. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
1.2.7 Simulating depletion in the laboratory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.3 Modelling rock behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.3.1 Representation of the porous medium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.3.2 Fundamental assumptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.3.3 Equilibrium equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
https://telegram.me/Geologybooks

1.3.3.1 Conservation of mass. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

1.3.3.2 Equation of motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.3.4 Fluid constitutive equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.3.5 Clausius-Duhem inequality and thermal equation . . . . . . . . . . . . . . . . . . . . . . . 35
1.3.6 Conduction laws. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.3.6.1 Fluid conduction law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.3.6.2 Heat conduction law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
1.3.7 Constitutive equations of the skeleton and the porous medium . . . . . . . . . . . . 37
1.3.8 Linear poroelastic behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.3.8.1 Skeleton constitutive equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.3.8.2 Porous medium constitutive equations . . . . . . . . . . . . . . . . . . . . . . . 39
1.3.8.3 Equation governing the variation in elementary Eulerian porosity . . 40
1.3.9 Nonlinear poroelastic behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
1.3.10 Poroelastoplastic behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.3.10.1 Plastic strains, plastic porosity and trapped energy . . . . . . . . . . . . . . 45
1.3.10.2 Cam-Clay model constitutive equations . . . . . . . . . . . . . . . . . . . . . . 47
1.3.10.3 Plasticity criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
1.3.10.4 Flow rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.3.10.5 Strain-hardening equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
1.3.10.6 Incremental expression of the flow rule . . . . . . . . . . . . . . . . . . . . . . 52
1.4 Determination of in situ stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
1.4.1 Determination from in situ measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
1.4.1.1 Horizontal stress directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
1.4.1.2 Stress amplitude. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
1.4.2 Determination from core measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
1.4.3 How to constrain the stress tensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1.5 Geomechanical modelling elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1.5.1 Geology and griding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
1.5.2 Geomechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
1.5.3 Boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
 Table of Contents XI

Chapter 2
GEOMECHANICS, DRILLING AND PRODUCTION

2.1 Drilling performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

2.1.1 Empirical drilling efficiency: weight on bit, rate of penetration and torque. . . . 71
2.1.1.1 Axial kinematics of the bit: normalised rate of penetration . . . . . . . . 71
2.1.1.2 Circumferential response and representation in the plane RD – TD . 74
2.1.1.3 Cutter wear, a major concern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.1.2 Drillability and detailed analysis of failure mechanisms at the working face . . 78
2.1.2.1 Rock cutting mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
2.1.2.2 Shear cutting regimes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
2.1.2.3 Consequences and applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2.1.2.4 Punching models for bit behaviour. . . . . . . . . . . . . . . . . . . . . . . . . . . 83
2.1.3 Drilling performance including evacuation of cuttings and bit dynamics . . . . . 84
2.1.3.1 Modelling of jamming during drilling . . . . . . . . . . . . . . . . . . . . . . . . 85
2.1.3.2 Drilling efficiency and vibrations of the drilling system . . . . . . . . . . 86
https://telegram.me/Geologybooks

2.2 Borehole stability during drilling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

2.2.1 Instabilities and drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
2.2.1.1 Description of drilling modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
2.2.1.2 Borehole instability mechanisms while drilling . . . . . . . . . . . . . . . . . 88
2.2.2 Instabilities by formation type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
2.2.2.1 Shale formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
2.2.2.2 Unconsolidated formations (erosion) . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.2.2.3 Fractured/cracked formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
2.2.2.4 Creeping formations (salt or plastic shales) . . . . . . . . . . . . . . . . . . . . 95
2.2.2.5 Formations in tectonic zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
2.2.3 Stability calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
2.2.3.1 Analytical solution to the elasticity problem . . . . . . . . . . . . . . . . . . . 96
2.2.3.2 Wall instability model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
2.3 Sand production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
2.3.1 Types of sand production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.3.1.1 Transient solid production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.3.1.2 Continuous production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.3.1.3 Catastrophic solid production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.3.2 Mechanisms involved . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
2.3.2.1 Initiation of sand production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
2.3.2.2 Special case of CHOPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
2.3.3 Sand production prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
2.3.3.1 Parameters affecting solid production . . . . . . . . . . . . . . . . . . . . . . . . 105
2.3.3.2 Empirical methods from in situ measurements . . . . . . . . . . . . . . . . . 106
2.3.3.3 Rock failure around a wellbore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
2.3.3.4 Numerical modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
2.3.3.5 Experimental validation of sand production prediction . . . . . . . . . . . 111
2.3.4 Sand production prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.3.4.1 Reducing the effective stresses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.3.4.2 Immobilising the material. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.3.4.3 Filters used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
2.3.4.4 Frac-pack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
 XII Table of Contents

2.3.4.5 Increasing the strength of the material . . . . . . . . . . . . . . . . . . . . . . . 116

2.3.4.6 Sand management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.4 Well stimulation by hydraulic fracturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
2.4.1 Description of the hydraulic fracturing process. . . . . . . . . . . . . . . . . . . . . . . . . 118
2.4.2 Opening a fracture around a vertical wellbore. . . . . . . . . . . . . . . . . . . . . . . . . . 119
2.4.3 Description of fractures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
2.4.4 Modelling of hydraulic fracture propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
2.4.4.1 Simplified 2D model: Carter’s model . . . . . . . . . . . . . . . . . . . . . . . . 123
2.4.4.2 GDK (Geertsma and de Klerk) type 2D models . . . . . . . . . . . . . . . . 123
2.4.4.3 PKN (Perkins, Kern, Nordgren) model . . . . . . . . . . . . . . . . . . . . . . . 125
2.4.4.4 Comments on 2D models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
2.4.4.5 Pseudo-3D models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
2.4.4.6 Fully-3D models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
2.4.5 Fracture monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
2.4.5.1 Pressure analysis during fracturing . . . . . . . . . . . . . . . . . . . . . . . . . . 127
2.4.5.2 Geometry of the hydraulic fracture . . . . . . . . . . . . . . . . . . . . . . . . . . 128
https://telegram.me/Geologybooks

2.5 Induced thermal fracturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

2.5.1 Description of fracturing induced by injection of cold fluid . . . . . . . . . . . . . . . 131
2.5.2 Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

Chapter 3
GEOMECHANICS AND RESERVOIR

3.1 Modelling the geomechanical and reservoir coupling . . . . . . . . . . . . . . . . . . . 136

3.1.1 Flow and permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
3.1.1.1 Hydraulic diffusivity equation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
3.1.1.2 Reservoir rock compressibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
3.1.1.3 Intrinsic permeability variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
3.1.2 Coupling between geomechanical and reservoir models. . . . . . . . . . . . . . . . . . 140
3.1.2.1 Conventional reservoir approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.1.2.2 Limitations of the conventional reservoir approach . . . . . . . . . . . . . 141
3.1.2.3 Geomechanical approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.1.2.4 Coupling between fluid flow and geomechanics. . . . . . . . . . . . . . . . 143
3.1.2.5 Example of external coupling in porosity on a highly compactable
reservoir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
3.1.2.6 Example of coupling in permeability . . . . . . . . . . . . . . . . . . . . . . . . 149
3.1.3 Stress path followed by the reservoir during depletion . . . . . . . . . . . . . . . . . . . 151
3.1.3.1 In situ measurements and assumption of uniaxial compaction . . . . . 152
3.1.3.2 Factors influencing reservoir behaviour during depletion. . . . . . . . . 153
3.2 Environmental aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.2.1 Subsidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.2.1.1 Introduction and famous examples . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.2.1.2 Analysis of subsidence related to the production of underground fluids 159
3.2.1.3 Subsidence measurement and prediction. . . . . . . . . . . . . . . . . . . . . . 160
3.2.2 CO2 sequestration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
3.2.2.1 Short-term geomechanical risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
3.2.2.2 Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
 Table of Contents XIII

3.2.2.3 Medium-term geomechanical risks . . . . . . . . . . . . . . . . . . . . . . . . . . 164

3.2.3 Well abandonment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.2.3.1 Abandonment procedure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
3.2.3.2 Behaviour of the well plug after abandonment . . . . . . . . . . . . . . . . . 166
3.3 Geomechanical monitoring of reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
3.3.1 Measurement of surface deformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
3.3.2 Seismic monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
3.3.3 Passive seismic monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
3.3.3.1 Temporary or permanent passive seismic monitoring . . . . . . . . . . . . 174
3.3.3.2 Surface PSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
3.3.3.3 Downhole PSM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
https://telegram.me/Geologybooks
 1

Elements of Rock Mechanics

Fundamentals
https://telegram.me/Geologybooks

1.1 BASIC NOTIONS

This chapter provides a general introduction to rock mechanics dedicated to petroleum

applications. Some simplifications have been required to respect the objective set. For more
details, refer to more specific books on petroleum geology and fundamental rock mechanics.

1.1.1 Rock characteristics

1.1.1.1 Minerals and rocks

The earth’s crust accessible to geological observation consists of rocks. Almost all rocks are
aggregates of minerals, forming grains, visible under the naked eye or the microscope. Min-
erals are natural chemical species, generally consisting of crystalline solids, characterised by
a three-dimensional periodic arrangement of atoms or ions. More than 4,000 minerals are
known. Silica and silicates are the most abundant [Guillemot, 1986].
The physical, chemical and mechanical properties of rocks depend on the physical,
chemical and mechanical properties of the constituent minerals, the volume fraction of each
mineral and the distribution and arrangement of the minerals with respect to each other. The
microstructure of this assembly is rarely homogeneous: arrangement of minerals, mineralog-
ical variability, distribution of grain sizes, spatial distribution of the grains, degree of frac-
turing, etc. The rock heterogeneity depends on the scale of observation, especially since its
physical properties depend on the scale of measurement: centrimetric for laboratory samples
to metric, or even kilometric, for blocks in situ.

1.1.1.2 Rock classification

There are three broad classes of rock: magmatic, sedimentary and metamorphic.
 2 Chapter 1 • Elements of Rock Mechanics. Fundamentals

1.1.1.2.1 Magmatic rocks

Magmatic rocks are mainly silicates. They result from the solidification of a magma (liquid
rock, possibly containing crystals in suspension). When the magma cooled very slowly deep
within the lithosphere, the grains (crystals) are well developed: this is the case of plutonic
(or intrusive) rocks, such as granite. When the magma cooled quickly after irrupting on the
surface or at the bottom of the oceans, the grains are generally very small, apart from the
occasional macrocrystals: volcanic (or extrusive) rocks, such as basalt.

1.1.1.2.2 Sedimentary rocks

Although magmatic rocks represent about 95% of the earth’s crust (by volume), 75% of the
area of the continents and virtually all the ocean floors are covered with sedimentary rocks
resulting from the accumulation and compaction of mineral debris – in this case we speak of
detrital rock (degradation of other rocks), organic debris (plant or animal remains, fossils) –,
or chemical precipitation.
https://telegram.me/Geologybooks

When the sediments are deposited in lacustrine and marine environments, they are soft
and generally water saturated. They are converted into coherent rocks under the effect of
diagenesis, a term referring to several phenomena that can occur to varying extents depend-
ing on the deposition conditions and the type of sediment concerned (compaction, cementa-
tion, recrystallisation). Sedimentary rocks are brought to the surface during the large tec-
tonic movements which take place in the earth’s crust. When the ocean or continental plates
collide, the sediments are folded, deformed and uplifted to form valleys and high mountain
belts.
During these tectonic movements, the rock may have been subjected locally to very high
pressures and temperatures.
Sedimentary rocks are classified into four groups:
– Sandstones are detrital sedimentary rocks composed of grains of sand (quartz or
silica) bound together by a siliceous, calcareous or ferruginous cement which deter-
mines the strength of the sandstone. Quartz and feldspars are the main components.
Sandstone elements belong to the granulometric class of arenites (2 mm to 0.05 mm).
– Shales are defined as detrital rocks of fine grain size distribution (grain diameter less
than 2 mm). Phyllites (chlorite, montmorillonite, muscovite and kaolinite) are the
major component minerals, but quartz, feldspar and calcite are also present. Shales
have very complex properties.
– Carbonate rocks are composed mainly of the minerals calcite, aragonite and dolo-
mite, of origin either organic (marine organisms) or inorganic (chemical precipita-
tion). Calcite is the dominant mineral.
– Evaporites: these rocks are composed of elements precipitated from water. As the
water evaporates, the species in solution become more and more concentrated, up to
oversaturation then precipitation of the components. The main minerals precipitated
are gypsum, anhydrite and halite.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 3

1.1.1.2.3 Metamorphic rocks

Metamorphic rocks are formed from mineralogical transformations of pre-existing rocks
(sedimentary or magmatic) without fusion, mainly through recrystallisations due to tempera-
ture and pressure increases.

1.1.1.3 Sedimentary rock and porous medium

Sedimentary rocks are porous media composed of a deformable skeleton and interstitial
voids, which are extremely important for oil companies. In terms of geomechanics, these
voids account for the major differences in behaviour observed between the compact solids
(e.g. metals) and the porous media, complex assemblies where the presence of fluids, even
in small quantities, makes the rock even more complex. For further information on the
mechanics of porous media, refer to the books by Coussy (1995) and Bourbié et al. (1986).
The total porosity of a given rock is the ratio of the non-solid volume (pore volume) to
https://telegram.me/Geologybooks

the total volume

Vp V − Vs
φ= = (1.1.1)
V V
where Vp is the porous volume,
Vs the solid volume,
V the total volume.
This porosity is also called Eulerian porosity, to distinguish it from Lagrangian porosity
Vp
used in modelling F � , with respect to the initial total volume V0.
V0
There are numerous ways of measuring porosity but the measurement result may depend
on the method used [Bourbié et al., 1986].
This petrophysical quantity provides no information on the pore dimensions, distribution
or degree of connection.
The porosity of a sediment varies during burial. The sediment changes progressively
from a quasi-liquid state to a compact solid state. The initial intergranular porosity may be
very high, 70% to 80% for a clay sediment. During burial of the sediment, the confining
pressure and the temperature increase. Various diagenetic processes (mechanical, chemical
and biochemical) change the porosity by consolidation and compaction of the rock, dissolu-
tion and recrystallisation of the minerals, etc.
Extensive theoretical and experimental studies have been conducted on the flow of fluids
through the porous media. The permeability of a porous medium is a property which reflects
the ability of the rock to allow a fluid to flow through its pores. The absolute permeability is
an intrinsic characteristic of the material, independent of the fluid. This permeability k was
defined by Darcy, by demonstrating that the discharge rate Q through a porous medium of
cross-sectional area A was proportional to the pressure drop ΔP, divided by the dynamic vis-
cosity m of the saturating fluid and the length L over which the pressure drop is taking place:
 4 Chapter 1 • Elements of Rock Mechanics. Fundamentals

k ΔP
Q = −A (1.1.2)
μ L
The negative sign is needed since fluids flow from high to low pressure.
When the following units are used: Q in m3.s–1, A in m2, DP/L in Pa.m–1 and m in Poises,
the permeability k is expressed in m2 and in Darcy, with 1 Darcy = 0.97.10–12 m2. The
Darcy, denoted D, is widely used in petroleum engineering and petrophysics. The m2 is used
more by material physicists.
The hydraulic conductivity K, or permeability coefficient, is defined as the ratio of per-
meability to fluid viscosity:
k
K= (1.1.3)
μ
K is expressed in m.s–1. So, while the permeability k is a characteristic of the material
https://telegram.me/Geologybooks

that is highly dependent on the porosity, the hydraulic conductivity K is characteristic of the
conditions of flow in a given material for a given fluid. To a certain extent, the hydraulic
conductivity K takes into account the physical interactions between the fluid and the rock.
The permeability depends on the structure of the porous medium. Two rocks with the
same porosity may have quite different permeabilities. When the porous space is filled by
several fluids, the capillary effects act in addition to the dynamic head losses and, in this
case, permeabilities relative to the fluid considered can be determined.

1.1.1.4 Properties of heterogeneous media

Application of the macroscopic equations of mechanics to porous media assumes that these
media are continuous, in other words that physical values (porosity, permeability, saturation,
etc.) can be defined at each point as functions differentiable at the point considered. At
microscopic scale, however, a point of the rock is either in the solid or in the pores, thereby
forming a discontinuity (if we take the variable “porosity” as an example). The notion of
“Representative Elementary Volume” (REV) is related to the need to describe the properties
of a porous medium at a scale where the structural discontinuities have no influence on these
properties. Consequently, the dimensions of the elementary volumes required to include the
effect of the discontinuities may vary in particular for a given medium, depending on the
parameter considered.
When a macroscopic constitutive equation is defined for a given medium, these various
minimum homogenisation volumes must be taken into account. The mechanical experi-
ments will only be relevant “macroscopically” if the sample is much larger than the mini-
mum homogenisation volume of the parameter considered. When dealing with cracked
media, this volume is sometimes highly underestimated.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 5

1.1.2 Stresses and strains

Analysis of stresses and strains is essential whenever studying rock mechanics. Since these
notions have been studied extensively in numerous books [e.g. Jaeger and Cook, 1969], only
a few details will be given here.

1.1.2.1 Stresses
For the simple case of an axially loaded body, e.g., a cylindrical bar subjected to tension or
compression by a force passing through its centre, the stress s, or intensity of the distribu-
tion of internal forces, can be obtained by dividing the total tensile or compressive force
amplitude F by the cross-sectional area A over which it is acting. In this case the stress is a
scalar that represents an average stress savg over the area, meaning that the stress in the
cross section is uniformly distributed. We therefore obtain:
https://telegram.me/Geologybooks

F
σavg =
A
In general, however, the stress is not uniformly distributed over a cross section of a mate-
rial body, and consequently the stress at a point is different from the average stress over the
entire area. The local stress at a specific point in the body is defined as being the limiting
value of ΔF/ΔA when ΔA tends to zero.
If the force is not applied perpendicular to the cross-section A, the stress can be decom-
posed into a normal component called the normal stress and a tangential component called
the shear stress.

• Stress tensor
To provide a complete description of the stress state at a given point, we must be able to
identify the stresses relative to the surfaces in three orthogonal directions. There are 9 com-
ponents of the stress at a point, given by the tensor expression:

⎛σ τ τ ⎞
⎜ x xy xz ⎟
σ = ⎜ τ yx σ y τ yz ⎟
⎜ ⎟
⎜τ τ σ ⎟
⎝ zx zy z ⎠

• Invariants
Invariants remain unchanged during a change of coordinate system:
The mean normal stress is equal to the trace of the stress tensor matrix:

( )
σmean = σ x + σ y + σ z / 3 (1.1.4)
 6 Chapter 1 • Elements of Rock Mechanics. Fundamentals

The other invariants commonly used are:

( )
I2 = − σ xσ y + σ yσ z + σ zσ x + τ xy
2 + τ2 + τ2
yz zx
(1.1.5)
I3 = σ xσ yσ z + 2τ xy τ yz τ zx − σ x τ 2yz − σ y τ zx
2 − σ τ2
z xy

• Deviatoric stress
The average stress defined previously expresses a uniform compression or extension. Dis-
tortions are expressed by the deviatoric stress tensor. Deviatoric stresses can be obtained by
subtracting the mean normal stress components from the stress tensor.

⎛s sxy sxz ⎞ ⎛ σ − σ τ xy τ xz ⎞
x
⎜ ⎟ ⎜ x mean ⎟
⎜ s s s ⎟ = ⎜τ σ − σ τ ⎟
⎜ yx y yz ⎟ ⎜ yx y mean yz ⎟
https://telegram.me/Geologybooks

⎜ s s s ⎟ ⎜τ τ zy σ z − σmean ⎟⎠
⎝ zx zy z ⎠ ⎝ zx

• Principal stresses
At any point of a body subjected to a system of forces, three planes can be defined on which
there are only normal stresses and no shear stresses. These planes are called the principal
planes and the stresses the principal stresses. In this coordinate system, the stress tensor has
a simple expression.

⎡σ 0 0⎤
⎢ 1 ⎥
⎢ 0 σ2 0⎥
⎢ ⎥
⎢⎣ 0 0 σ3 ⎥⎦

The three principal stresses are known as the major, intermediate and minor principal
stresses, in descending order of magnitude.

• Mohr circle
In 2D, knowing the principal stresses s1 and s2 (s1 > s2), the normal and shear stresses can
be calculated in any other plane inclined at an angle q in the xy plane (Figure 1.1.1).
Solving the equilibrium equations (sum of forces equal to zero) implies:

σθ =
1
2 1
( 1
) (
σ + σ2 + σ1 − σ2 cos 2θ
2
)
(1.1.6)
1
(
τθ = − σ1 − σ2 sin22θ
2
)
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 7

Direction of s2
sn
q

q
s1
Direction of s1
P

tn

s2
https://telegram.me/Geologybooks

Figure 1.1.1
State of stress at a point.

The stress state (sq, tq) in a plane at an angle q to the horizontal describes a circle in the

plane (s, t), called Mohr’s circle. Figure 1.1.2 shows a semi-circle of radius
1
(σ − σ2
2 1
)
1
(
and centre at σ1 + σ2 .
2
)

s1 – s2
2
tq

q 2q
s
s2 s 1 + s2 sq s1
2

Figure 1.1.2
Mohr circle.
 8 Chapter 1 • Elements of Rock Mechanics. Fundamentals

• p-q diagram and stress path

Successive stress states applied to a rock element are often described in geomechanics. One
way of doing this is to draw a series of Mohr circles. A more satisfactory way is to plot a
series of points and connect these points together with a line or a curve. These points are
called stress paths [Lambe and Whitman, 1969]. Due to the easy generalisation offered, the
p-q diagram (Figure 1.1.3) is currently widely used, where p is the mean principal stress and
q is the deviatoric stress:

(σ1 − σ2 )2 + (σ2 − σ3 )2 + (σ1 − σ3 )2

q= (1.1.7)
2
σ1 + σ2 + σ3
p=
3

tq q
https://telegram.me/Geologybooks

E Stress path
E
D
D
C
C
B
B
A sq p
A

(a) (b)

Figure 1.1.3
Representation of successive stress states, (example when s1 increases with s2 = s3
constant), (a) Mohr circles and (b) p-q diagram [Lambe and Whitman, 1969].

1.1.2.2 Strains
The material is deformed under the effect of stresses. The strains are calculated assuming
there has been a change between two states: the initial state and the final state. The strain of
a body can be measured by calculating the change in length of a line on the body or the angle
between two lines on this body. The change in length of a line, an elongation or a contrac-
tion, can be denoted as dl. The strain e, a dimensionless ratio, is defined as the ratio of elon-
gation with respect to the original length:

d l l - l0
e� � (1.1.8)
l0 l0
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 9

with
e strain in the direction measured,
l0 initial length of the line,
l current length of the line.
The local strain in the given direction can be expressed as:

dl
e � lim
l

when l tends to zero.

Angular strain g is defined by the amount of deformation Dl perpendicular to a given line
rather than parallel to it. The angular deformation is also called the shear strain (measured in
radians):
Dl
https://telegram.me/Geologybooks

� tgg � g
l

• Strain tensor and invariants

Assuming infinitesimal strains, if a point undergoes a displacement by a vector u, the math-

ematical expression of the strains is given by:

1 ⎛ ∂ui ∂u j ⎞
εij = ⎜ + ⎟ (1.1.9)
2 ⎜⎝ ∂x j ∂xi ⎟⎠

The strain tensor is written:

⎛ε ε ε ⎞
⎜ xx xy xz ⎟
ε = ⎜ ε yx ε yy ε yz ⎟
⎜ ⎟
⎜ε ε ε ⎟
⎝ zx zy zz ⎠

The tensor trace εv = εxx + ε yy + εzz expresses the volume strain. This value, independent
of the coordinate system chosen, is an invariant. Other strain invariants can be defined:

( )
J2 = − εx ε y + ε y εz + εz εx + εxy
2 + ε2 + ε2
yz zx
(1.1.10)
J3 = εx ε y εz + 2εxy ε yz εzx − εx ε2yz − ε y εzx
2 − ε ε2
z xy

As for the stresses, the strain tensor can be written with respect to its principal planes.
 10 Chapter 1 • Elements of Rock Mechanics. Fundamentals

1.1.2.3 Effective stress

The pore pressure plays a fundamental role studied through the notion of effective stress,
used to take into account the twofold effect of the stress (sij) and the pore pressure (pp) via
an expression, which is a linear combination of these two variables:

σij' = σij − α pp δij or σ ' = σ − α pp 1 (1.1.11)

where dij is the Kronecker tensor.

Since this concept is applied to problems involving distinct mechanisms – constitutive
equations, failure criterion and rock characteristics – the corresponding expressions of the
effective stress do not necessarily coincide. When the coefficient a is equal to 1, we speak of
Terzaghi’s effective stress, named after the inventor of the concept for soil mechanics.
For more in-depth information on this topic, refer to [Guéguen and Boutéca, 1999].
https://telegram.me/Geologybooks

1.1.2.4 State Boundary Surface (SBS)

Using the critical state concept, soil mechanics scientists have developed the notion of stress
path. Stress paths plotted in the 3-dimensional space p'-q-v (p': mean effective stress, q:
deviatoric stress, v: volume variation) determine a unique surface known as the State
Boundary Surface (SBS) (Figure 1.1.4). This surface separates the area of the possible

q'
p' Critical state line

Hvorslev surface

Roscoe surface

Normal consolidation
line
Tension failure

Figure 1.1.4
State boundary surface [Schofield and Wroth, 1968].
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 11

“stress-volume” states from the area where “stress-volume” states cannot exist. The states
presented by the parameters (p', q, v) always lie inside or on the SBS. The response of soils
inside the SBS is purely elastic (reversible but not necessarily linear) and becomes plastic
when the soil state reaches the state boundary surface. This notion has been extended to rock
mechanics. The SBS is also called yield surface.

1.1.2.5 Elastic properties

1.1.2.5.1 Static moduli of elasticity

The theory of elasticity describes situations where there are linear relations between the
stresses applied and the resulting strains. We consider a cylindrical sample of length l and
cross-sectional area A to which a force F is applied at its two ends. The stress applied to the
F
sample is σ z = . The sample will undergo an axial strain of εz = Δl l and two radial
A
https://telegram.me/Geologybooks

strains εx and ε y ( εx = ε y for a cylindrical sample). The strains in this type of experiment
are related to the stresses by the following linear relation:
σz
εz = relation known as Hooke’s Law. (1.1.12)
E
E, called the modulus of elasticity or Young’s modulus, characterises the stiffness of the
sample, in other words its resistance to an axial stress. The radial deformability can also be
characterised using the Poisson’s ratio:
εx
ν=− (1.1.13)
εz
For isotropic materials, whose response is independent of the direction of the stresses
applied, the relations between the stresses and strains are given by the generalised Hooke’s Law:

εij =
1⎡
E⎣
( )
1 + ν σij − νδijσαα ⎤
⎦
or (1.1.14)
σij = λδij εαα + 2Gεij

where l and G, the Lamé parameters, are also elastic moduli. G is called the shear modulus,
since it characterises the resistance of the sample to shear. The shear modulus is the ratio of
shear stress to shear strain.
The bulk modulus K defines the ratio of an isotropic stress to a volume strain. In an iso-
σp
tropic stress state sx = sy = sz, the other components of the stress tensor are zero: K = .
εv
K is the sample resistance to an isotropic compression. The reciprocal 1/K, called compress-
ibility, is written Cp. The various elastic constants and their definitions are given in
 12 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Table 1.1.1. Note that for porous media, we must introduce Kd and Ku, bulk moduli, respec-
tively in drained and undrained conditions, Ed and Eu, drained and undrained Young’s mod-
uli, and nd and nu, drained and undrained Poisson’s ratios.

Table 1.1.1 Definitions of elastic constants [Guéguen and Placiauskas, 1992].

K E n G
Isotropic Uniaxial Uniaxial Shear sxy
Experiment compression sp compression sz compression sz

sp sz ex s xy
Definition K� E� n�- G�
ev ez ez exy
https://telegram.me/Geologybooks

The relations between the various elastic constants K, E, G, l and n are listed in
Table 1.1.2.

Table 1.1.2 Relations between the various elastic constants of an isotropic material [Birch, 1961].

K E n G l

2 3l � 2G l
___ ___
l� G
3
G
l �G �
2 l �G �
K -l l
___ 9K
3K - l 3K - l
3
2
�
K -l � ___

9 KG 3 K - 2G 2
___ ___
3K � G �
2 3K � G � K-
3
G

EG E E - 2G
___ ___ G
�
3 3G - E � 2G
-1
� 3G - E )�
___ ___ 3K - E 3 KE 3K - E
3K
6K 9K - E 9K - E

l
1� n
l
�1 � n ��1 - 2n � ___ l
1 - 2n ___
3n n 2n

G
�
2 1� n � �
2G 1 � n � ___ ___ G
2n
�
3 1 - 2n � 1 - 2n

___ �
3 K 1 - 2n � ___ 3K
1 - 2n
2 � 2n
3K
n
1� n
E E En
___ ___
�
3 1 - 2n � 2 � 2n �
1 � n 1 - 2n�� �
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 13

1.1.2.6 Dynamic moduli of elasticity

The moduli of elasticity listed above are determined by experiments on rock samples sub-
jected to tension – compression (static strain). They are called the static moduli of elasticity.
These moduli of elasticity can also be determined by measuring the velocities of propaga-
tion of seismic (or acoustic) waves (dynamic strain). In this case, they are called the
dynamic moduli of elasticity.
The compression Vp and shear Vs wave velocities are given by the following expres-
sions:
4
Kdyn + Gdyn
VP = 3 (1.1.15)
ρ

Gdyn
https://telegram.me/Geologybooks

VS = (1.1.16)
ρ

When the P and S wave velocities are available, the following dynamic moduli of elas-
ticity can be determined:

4 2
Gdyn = ρVS2 and Kdyn = ρVP2 − ρV (1.1.17)
3 S
The dynamic values of Young’s modulus and the Poisson’s ratio are given respectively by:

Edyn =
(
ρVS2 3VP2 − 4VS2 ) and νdyn =
VP2 − 2VS2
(1.1.18)
(VP2 − VS2 ) (
2 VP2 − VS2 )
The static moduli of rocks are generally lower than the dynamic moduli, due to the
amplitudes of the strains considered. In oil exploration, seismic data, Vertical Seismic Pro-
file (VSP) data and P- and S-wave logging data provide measurements of propagation veloc-
ities for rocks around and between the wells, both in the reservoir and in the surrounding
strata. These measurements are not recorded at the same frequency (Table 1.1.3), but data of
velocity dispersion as a function of frequency is interesting for petroacousticians.

Table 1.1.3 Frequency of acoustic waves.

Acquisition frequency
Exploration seismic 100-300 Hz
VSP 100-300 Hz
Sonic logging 100 kHz
Measurement on samples in laboratory 0.5 to 1 MHz
 14 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Seismic velocities are sensitive to lithology, porosity, fluid content, temperature and
stress field. Orders of magnitudes of the velocities Vp and Vs for some rocks are provided for
information in Table 1.1.4.

Table 1.1.4 Order of magnitude of P and S wave velocities and density for different rock types [Lavergne, 1986].

P wave velocity S wave velocity

Rock type Density (g/cm3)
(m/s) (m/s)
Scree, organic soil 300-700 100-300 1.7-2.4
Dry sands 400-1,200 100-500 1.5-1.7
Wet sands 1,500-2,000 400-600 1.9-2.1
Saturated clays 1,100-2,500 200-800 2.0-2.4
Marls 2,000-3,000 750-1,500 2.1-2.6
Saturated argillaceous-sandstone series 1,500-2,200 500-750 2.1-2.4
https://telegram.me/Geologybooks

Porous and saturated sandstones 2,000-3,500 800-1,800 2.1-2.4

Limestones 3,500-6,000 2,000-3,300 2.4-2.7
Chalk 2,300-2,600 1,100-1,300 1.8-2.3
Salt 4,500-5,500 2,500-3,100 2.1-2.3
Anhydrite 4,000-5,500 2,200-3,100 2.9-3.0
Dolomite 3,500-6,500 1,900-3,600 2.5-2.9
Granite 4,500-6,000 2,500-3,300 2.5-2.7
Basalt 5,000-6,000 2,800-3,400 2.7-3.1

1.1.2.7 Anisotropy
If the elastic response of the material depends on its orientation for a given stress field, the
material is said to be anisotropic. In this configuration, the moduli of elasticity vary depend-
ing on the direction in which they are evaluated.
Anisotropy is defined as heterogeneities at a smaller scale than that of the volume
observed. At large scale, the development of geological strata (foliations, lineations, etc.)
results in a certain degree of anisotropy, while at small scale this anisotropy exists due to the
presence of the anisotropic minerals forming the rocks.
For anisotropic medium, a generalised form of Hooke’s law is written:
σij = Cijkl εkl or σ = C ε (1.1.19)

C , elastic stiffness tensor, is a 9x9 tensor. The symmetry of the stresses and strains
reduces the number of independent components to 36. In addition, the existence of a unique
energy potential reduces the number of independent components to 21, the maximum
number of elastic constants that a medium can possess.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 15

Most sedimentary rocks (sandstone, mudstone, shale, etc.) exhibit anisotropic character-
istics in terms of strength and deformational properties (differences between elastic moduli
measured parallel and perpendicular to bedding), and are generally considered to be trans-
versely isotropic materials, symmetric about an axis normal to a plane of isotropy. The num-
ber of independent elastic constants is reduced to 5 (2 Young’s moduli, 2 Poisson’s ratios
and one shear modulus). Some rocks may have three distinct directions of anisotropy. These
materials are orthotropic and have 9 independent elastic constants (3 Young’s moduli,
3 Poisson’s ratios and 3 shear moduli).

1.2 GEOMECHANICAL CHARACTERISATION OF ROCKS

IN THE LABORATORY

Apart from its structural and mineralogical variability at microscopic scale, at macroscopic
https://telegram.me/Geologybooks

scale a rock mass may exhibit pronounced stratification, cracks, faults, etc. Generally there-
fore, it is not homogeneous. The fundamental hydromechanical properties of a rock are con-
sidered to be those of a sample big enough to contain a large number of component grains,
but small enough to exclude any major structural discontinuity. More precisely, it is recom-
mended that the diameter of a sample should be more than 10 times the diameter of the
largest rock grain [Kovari et al., 1983]. Samples of a few centimetres generally correspond to the
mesoscopic scale so defined and offer the advantage of being easily tested in the laboratory.
The aim of laboratory tests is to characterise the hydromechanical properties of a rock
using relatively simple experiments. The properties determined are based on a certain num-
ber of assumptions (homogeneous stress and strain states in the sample, isotropic behaviour,
etc.). It is also essential to ensure that they are actually representative of the behaviour of the
rock, not of the experimental device. Relations between stresses and strains can generally be
deduced from the measurements taken. These relations are then extrapolated to deal with
more complex situations at a scale different from that of the tests, this scale change requiring
additional assumptions.

1.2.1 Laboratory tests

In situ, a rock is subjected to a triaxial stress state and to the pressure exerted by the intersti-
tial fluid(s). The stresses are generally assumed to be oriented in the vertical direction and
two orthogonal horizontal directions, the vertical stress being determined by the overburden
(see Chapter 1.4).
The time scale of the strains may vary considerably depending on the problem consid-
ered: for instance, the strain rates associated with progressive depletion of a reservoir are
much slower than those encountered when drilling a well. The strength of a rock generally
increases with strain rate. In the laboratory, strain rates between 10–6 and 10–1 s–1 are gener-
ally applied. Within this interval (not necessarily representative of the problem studied), the
properties of most rocks can be considered as globally independent of the strain rate.
 16 Chapter 1 • Elements of Rock Mechanics. Fundamentals

However, with very poorly permeable rock, such as shale, the strain rate chosen must be low
enough to prevent excess pore pressure generation.
The various types of test used to study the hydromechanical characteristics of a rock, in
particular its failure properties, are mainly characterised by the specific stress states imposed
and the accessible properties. The associated cost and ease of implementation must also be
taken into consideration when selecting the tests to be conducted during a characterisation
campaign.
This chapter outlines the main types of test on rock and the properties that can be deter-
mined. We will then focus on the compressive behaviour of a rock, using the results of the
main rock mechanics test: the conventional triaxial test.
This document does not discuss hollow cylinder tests that can be used to study well wall
behaviour [Fjaer et al., 1992] [Jaeger and Cook, 1979].

1.2.2 Tests without interstitial fluid

https://telegram.me/Geologybooks

1.2.2.1 Uniaxial compression test

The uniaxial compression test (also called unconfined compression test) is one of the oldest
and easiest to implement. It consists in compressing a generally cylindrical rock sample
between the two platens of a load frame. An increasing axial compressive force is applied to
the sample, resulting in its gradual shortening and eventual failure (Figure 1.2.1). The axial
force applied and the associated axial strain are measured. The radial strain can also be mea-
sured. The axial stress is equal to the force measured divided by the initial cross-section of
the sample. With a highly deformable rock, it may be worthwhile evaluating the current
cross-section of the sample from the measurement of radial strain and calculating a “true”
axial stress.
The Unconfined Compressive Strength (UCS) of the rock (value of the axial stress peak
associated with the failure) and estimations of its Young’s modulus and its Poisson’s ratio
are deduced from this test. Assuming linear elastic behaviour, these parameters correspond
respectively to the slope of the axial stress-axial strain curve and to the opposite of the slope
of the radial strain-axial strain curve.
A representative value of unconfined compressive strength for a given rock can be diffi-
cult to obtain as it can vary widely within the same rock mass. Considering this large vari-
ability, a UCS obtained by testing just a few samples cannot be considered as totally reliable.
Various phenomena are involved. Firstly, the friction at the area of contact between the
platens generates a non-uniform distribution of the stresses in the sample. To reduce these
edge effects, the sample height must be at least twice the diameter. The moisture content
may also have a significant effect on the measurement, due to the presence of capillary
bridges. Lastly, this type of test is highly sensitive to the presence of the slightest heteroge-
neity or crack in the sample. Note that it would be more appropriate to measure UCS using a
triaxial device on a saturated sample subjected to very low confinement. Table 1.2.1 gives
the orders of magnitude of unconfined compressive strengths for rocks frequently encoun-
tered in petroleum engineering.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 17

35

30

25
Axial stress (MPa)

20 Increase
in pore
15 pressure
ea
er
10

0
-0.1 0.0 0.1 0.2 0.3
Strain (%)
https://telegram.me/Geologybooks

Figure 1.2.1
Unconfined compression test on a sample of Berea sandstone with 24%
porosity.

Table 1.2.1 Order of magnitude of the static elastic properties of rocks with different UCSs
[partially from Le Tirant and Gay, 1972 and Fjaer et al., 1992].

Rock type UCS (MPa) E (GPa) u

Poorly consolidated sandstone < 10 <1 0.4
Poorly cemented sandstone 10 to 50 5 to 10 0.3 – 0.35
Consolidated sandstone 50 to 250 10 to 30 0.1 to 0.35
Highly consolidated sandstone > 100 30 to 50 0.15 to 0.25
Hard and very hard limestone 100 to 200 50 to 100 0.2 to 0.3
Soft limestone marl 2.5 to 5 6 to 10 0.4 to 0.45
Soft shale 0.5 to 5 1 to 5 0.3 to 0.4
Highly consolidated shale 5 to 100 5 to 70 0.2 to 0.3
Salt 10 to 30 15 to 40 0.3 to 0.4

1.2.2.2 Uniaxial tension test

Tensile strength describes the ability of the rock to resist tensile stress. The direct measurement
of tensile strength can be performed by gluing the opposite ends of a cylindrical rock sample to
metal jaws which are fixed to the platens of a test machine. This type of test is rarely carried
out, being difficult to implement and interpret. The tensile strength measured in this way is
highly dependent on the sample size and the stress concentrations generated by the techniques
used to attach the rock to the platens of the press. Consequently, to estimate the tensile strength
of a rock the samples must be chosen carefully and a large number of tests conducted.
 18 Chapter 1 • Elements of Rock Mechanics. Fundamentals

1.2.2.3 Brazilian test

The Brazilian test is an indirect experimental method to measure the tensile strength of rock.
Although it is more frequently used than the uniaxial tension test, being much simpler to
implement, it has the disadvantage of introducing non-homogeneous stresses. It is carried out
by compressing a cylindrical rock sample of height/diameter ratio less than 1 diametrically
between the two platens of a load frame (Figure 1.2.2). The failure occurs under this load
along a stress crack, locally roughly in the diametral loading plane. The tensile strengths mea-
sured are reproducible and consistent with the values obtained in uniaxial tension.

W
h

D
https://telegram.me/Geologybooks

Figure 1.2.2
Principle of the Brazilian test.

In practice, to estimate the tensile strength, the rule is to take a ratio of 8 to 10 with UCS.

1.2.3 Uniaxial strain (or oedometric) test

The oedometric test is used to determine the uniaxial compressibility of a soil- or rock-type
material.
The oedometric cell is a “non-deformable” cylindrical mould that can contain a thin sam-
ple (typically 20 mm high by 40 mm diameter), placed between two highly porous discs and
compressed with a piston (Figure 1.2.3). The horizontal strains being prevented, the sample
is in a state of uniaxial strain. The axial strain is then exactly equal to the volumetric strain.
The porous discs placed at each end of the sample ensure uniform distribution of the
upstream and downstream pore pressures. These pressures can be controlled independently.
A differential pressure and therefore a flow can be applied between the upstream and down-
stream ends of the sample, to evaluate in particular its vertical permeability.
The oedometric test consists in applying an increasing axial load on the sample, usually
in stages, using a lever arm loaded with weights. Each stage is maintained until the axial
strain of the sample has stabilised. The test can either be drained (upstream and downstream
pore pressures controlled) or undrained (sample hydraulically isolated, pore circuits closed).
At regular intervals defined by the operator the following quantities are recorded:
– axial stress (force sensor),
– radial stress (oedometric cells equipped with strain gauges),
– upstream and downstream pore pressures (pressure transducers),
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 19

– piston displacement (Linear Variable Displacement Transducer), used to determine

the axial strain of the sample,
– temperature (it is recommended to place the oedometers in an air-conditioned room
since the pressures imposed are highly sensitive to temperature variations).

Displacement sa Downstream
sensor
pressure pump

Cell

Sample Porous discs

https://telegram.me/Geologybooks

Upstream
pressure pump

Figure 1.2.3
Diagram of an oedometric cell.

These tests are particularly well-suited to the study of elastoplastic behaviour of very
poorly permeable rocks. Since the sample is thinner than for a triaxial test, pressures reach
equilibrium more quickly. Note that, using oedometers, it is not generally possible to apply
loads large enough to cause sample failure.
Oedometric tests can be used to determine:
– the drained uniaxial compressibility,
– the creep properties (axial strain under constant load as a function of time),
– several poromechanical parameters [Bemer et al., 2001],
– the axial permeability.

1.2.4 Conventional triaxial test

The conventional triaxial test is carried out on cylindrical rock samples subjected to an axial
load, a uniform radial stress (via a pressurised confinement fluid) and an interstitial pressure
in the interconnected porous network. The sample is surrounded by an impermeable jacket
preventing any communication between the pore fluid and the confinement fluid. The pore
pressure must always be less than the confining pressure. The imposed stress field is there-
fore axisymmetrical and defined by the axial stress sa, the confining pressure pc and the
pore pressure pp (Figure 1.2.4). This load is associated with a strain state defined by the axial
strain ea and the radial strain er.
 20 Chapter 1 • Elements of Rock Mechanics. Fundamentals

sa

Downstream pore
pressure (pp)
Porous disc
Jacket

Sample pc

Upstream pore
pressure (pp)
https://telegram.me/Geologybooks

Figure 1.2.4
Stress and pressure states during a triaxial test.

The piston applying the axial stress can be loaded mechanically using a load frame or
hydraulically via an upper pressure chamber. The confining and pore pressures are imposed
by servo-controlled hydraulic pumps. The axial and radial stresses are controlled indepen-
dently. Consequently, any required stress path can be applied to the specimen mounted in
the cell.
The devices equipped with two pumps to control the pore pressure upstream and down-
stream from the sample can be used to generate a flow through the sample and therefore to
measure its vertical permeability. The porous discs placed at each end of the sample ensure
uniform distribution of the upstream and downstream pore pressures. Some laboratories
have cells that can be used to measure the horizontal permeability of the sample under triax-
ial load.
Various devices can be used to measure the axial and radial strains. A local measurement
of these two strains can be obtained using gauges stuck to the sample. Vertical displacement
sensors (LVDT or cantilever type) placed inside the confinement chamber will provide a
semi-local measurement of the axial strain (over much of the sample height). It is not recom-
mended to place these sensors outside the cell, since the global measurement carried out also
includes the cell strains, sensitive in particular to the confining pressure applied. The radial
strain can also be measured using horizontal displacement sensors (LVDT or cantilever
type) placed on the jacket half-way up the sample (local measurement) or a small chain
around the sample on a diameter (semi-local measurement). Note that the jacket strains must
be calibrated in order to use these two measuring instruments correctly.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 21

The results obtained are often studied in terms of mean stress p and deviatoric stress q,
associated with the volume strain ev and the deviatoric strain ed:

σa + 2 pc
p= εv = εa + 2εr (1.2.1)
3

q = σa − pc εd =
2
(ε −ε
3 a r
) (1.2.2)

The sample can be isolated by two valves located on the upstream and downstream pore
circuits. When these valves are open and the pore pressure is kept constant via a servo-con-
trolled pump, the test is said to be “drained”. When they are closed, the volume of fluid
inside the sample remains unchanged and the test is said to be “undrained”. Note that, in this
case, the pore pressure is controlled by the stress state imposed and the rock behaviour.
The first step of a triaxial test consists in saturating the sample with the test fluid, either
https://telegram.me/Geologybooks

mineral oil (which offers the advantage of avoiding any chemical interaction with the rock
constituents), water, brine, etc. The test generally starts by applying a high vacuum in the
sample previously placed in the triaxial cell. Then, under low initial axial stress and confin-
ing pressure, an upstream pore pressure (below the confining pressure) is applied, the down-
stream valve being kept closed. At this point, we simply monitor the volume of fluid
injected; the sample is considered to be saturated when the volume reaches a constant value.
Note that with a reservoir rock, this procedure assumes that the sample has first been cleaned
by removing all the in situ fluids. To date, no thorough studies have been conducted to deter-
mine the possible impact of this cleaning on the hydromechanical properties of the rock.
Note, however, that the rock wettability properties may be modified by the chemical treat-
ment applied.
The standard loading path is a drained path with constant pore pressure. The stresses are
increased first isotropically (hydrostatic loading) up to the required confining pressure, then
only the axial stress is further increased up to sample failure.
The axial stress is increased at constant axial strain rate (velocity control) or at constant
axial load rate dσa dt (force control). Unloading/reloading cycles can be carried out while
loading to estimate the elastic properties.
More complex loading paths can be performed, in particular to simulate the effects of
depletion on the behaviour of a reservoir (see § 3.1.3). An uniaxial strain test can be carried
out if the radial strain rate dεr dt is maintained at zero via a servo-control system. A varia-
tion of the confining pressure proportional to that of the axial pressure can also be applied.
When a constant axial load is imposed, a creep test is carried out; when a constant axial
strain is imposed, a relaxation test is carried out.

1.2.5 Effective stresses and drained triaxial test

Remember that the effective stress is given by σ ' = σ − α pp 1.

 22 Chapter 1 • Elements of Rock Mechanics. Fundamentals

By definition, the deviatoric stress is “effective”, it does not depend on the pore pressure:

( )(
q ' = σa '− pc ' = σa − α pp − pc − α pp = σa − pc = q )
We will now consider the results of a standard drained triaxial test carried out at constant
pore pressure. The end of hydrostatic loading is characterised by a mean stress pco , zero
deviatoric stress and a pore pressure pop . This point is generally chosen as reference state
when calculating strains. The stress-strain relations then concern the stress increments:

σa + 2 pco 3 pco q
q = σa − pco and Δp = p − pco = − = (1.2.3)
3 3 3
Since the pore pressure is constant, irrespective of the expression of the effective stress
considered, we obtain:
Δp ' = Δp
https://telegram.me/Geologybooks

In this precise case, therefore, characterisation of the stress-strain relations is indepen-

dent of the expression of the effective stress considered.
To determine the failure criterion which defines a domain in the stress space outside
which the rock cannot withstand the loads, various experimental studies have demonstrated
that, provided that the permeability is sufficient to ensure a homogeneous pore pressure dis-
tribution, rock failure is governed by Terzaghi’s effective stress:
σ ' = σ − pp 1 (1.2.4)

If we plot the values at failure obtained at various pore pressures on a p – q plane

(p = mean stress, q = deviatoric stress), we obtain a failure curve for each pore pressure level
(Figure 1.2.5). If we plot the same values in a p' – q plane (p' = Terzaghi’s mean effective
240 240

220 220

200 200

180 180
q (MPa)

q (MPa)

160 160

140 140
pp = 1 MPa
pp = 10 MPa
120 pp = 20 MPa 120 pp = 1 MPa
Failure line (pp = 1 MPa) pp = 10 MPa
100 Failure line (pp = 10 MPa) 100 pp = 20 MPa
Failure line (pp = 20 MPa) Failure line

80 80
20 40 60 80 100 120 140 20 40 60 80 100 120 140
p (MPa) p' = p–pp (MPa)

Figure 1.2.5
Deviatoric stress at failure as a function of mean stress and mean effective
stress on a limestone with 10% porosity [Vincké et al., 1998].
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 23

stress, q = deviatoric stress), the various curves lie on top of each other. In other words,
increasing the pore pressure by Δpp will produce the same reduction in axial stress at failure
as reducing the confining pressure by Δpp [Vincké et al., 1998].
The Mohr-Coulomb failure criterion traditionally used to describe brittle failure of rocks
can therefore be written in the form:
q = A + B p ' where p ' = p − pp (1.2.5)
parameters A and B being related to the (drained) cohesion c ' and the (drained) internal
angle of friction j ' of the rock by the following relations:

6 cos ϕ ' 6 sin ϕ '

A= c' B=
3 − sin ϕ ' 3 − sin ϕ '
Since we must reason in terms of global stress level applied to the rock in order to study
the failure, Terzaghi’s effective stress is generally adopted to represent all the results
https://telegram.me/Geologybooks

obtained during a triaxial test. For a loading path with pore pressure variations, different
effective stresses may need to be considered, including Biot’s effective stress which is
related to the rock elastic strains:
σ ' = σ − b pp 1 (1.2.6)
where b is Biot’s coefficient.

1.2.6 Rock behaviour during compression

The behaviour of a rock sample subjected to compression is highly dependent on:
– the rock type (sandstone, carbonate, stiff clays, shale, etc.),
– the initial rock porosity level,
– the loading path followed;
and various microstructural parameters such as, for the sandstones: grain size, cementation,
clay content, etc. [Wong et al., 1997]. The saturation state may also have a significant influ-
ence, in particular for chalks [Homand and Shao, 2000] [Matà, 2001].
This paragraph is largely based on the experimental studies conducted by Wong, Zhu
and David [Wong et al., 1997] [Zhu and Wong, 1997] for sandstones and by Yale and Craw-
ford (1998) for carbonates. Additional data can also be found in [Schutjens and de Ruig,
1996] [Schutjens et al., 1998] [Brawn and Jahns, 1998], etc.

1.2.6.1 Elastic behaviour and measurement of elastic parameters

Elastic behaviour is characterised by perfect reversibility of the strains associated with a
load increment. For a large number of rocks, this type of behaviour is encountered for mod-
erate stress ranges.
For a triaxial test, elastic behaviour is observed more especially when loading starts. For
small stresses, there may nevertheless be an initial tightening phase corresponding to closure
 24 Chapter 1 • Elements of Rock Mechanics. Fundamentals

of existing microcracks (generated in particular by sample coring) and associated with a

reduction in permeability (Figures 1.2.6 and 1.2.9). Unloading carried out during this phase
would result in a slight residual strain. Note that for some rocks, several stress-strain cycles
(seasoning) may be required to obtain reproducible behaviour.

p’ Linear elasticity p’ Non linear

with an initial elasticity
tightening phase

Tangent
modulus
Secant
modulus
https://telegram.me/Geologybooks

0 ev 0 ev

Figure 1.2.6
Stress-strain curves representative of elastic behaviour.

Above a certain level of loading some rocks, including crystalline rocks, sandstones of
low to medium porosity, compact limestones, chalks, etc. exhibit linear behaviour. The iso-
tropic elastic parameters of the rock are then easily determined using the stress-strain curves
obtained at constant pore pressure (drained test).
The drained bulk modulus Kd is defined by the slope of the mean effective stress-volume
strain curve, and the shear modulus G by one third of the slope of the deviatoric stress-devi-
atoric strain curve:
Kd = dp ' dεv (1.2.7)

3G = dq dεd (1.2.8)

The drained Young’s modulus Ed and the drained Poisson’s ratio nd are deduced from
the following relations:

9 Kd G 3 Kd − 2G
Ed = and νd = (1.2.9)
3 Kd + G 6 Kd + 2G

In the special case of hydrostatic loading (i.e. isotropic), only the bulk modulus can be
determined.
For purely axial loading, the drained Young’s modulus is defined by the slope of the
effective axial stress-axial strain curve: Ed = dσa ' dεv , and the drained Poisson’s ratio by
the opposite of the slope of the radial strain-axial strain curve: νd = − dεr dεa .
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 25

Biot’s coefficient can be determined by varying the pore pressure, keeping the axial
stress and the confining pressure constant. The ratio Kd b , defined as the opposite of the
slope of the pore pressure-volume strain curve, is then measured: Kd b = − dpp dεv .
Knowing Kd, we can deduce b.
Assuming undrained loading, we can calculate the undrained bulk modulus
Ku = dp dεv , the shear modulus 3G = dq dεd and Biot’s modulus, provided that Biot’s
coefficient Mb = dpp dεv has been determined.
Numerous rocks have nonlinear stress-strain curves in the elastic range, reflecting an
increase in bulk modulus as the material is compressed (Figure 1.2.7). This behaviour is
observed in particular with rocks whose structure can be schematised by an assembly of
cemented “grains”. The grains themselves exhibit linear elastic behaviour. The increase in
the contact surfaces introduces nonlinearities, however.
https://telegram.me/Geologybooks

60

50

40
p’ (MPa)

30

20

10

0
–0.010 –0.008 –0.006 –0.004 –0.002 0
Dev

Figure 1.2.7
Drained hydrostatic compression test on a reservoir sandstone with 19.8%
porosity.

In practice, nonlinear elasticity curves are interpreted in two ways. Depending on the extent of
the stress range associated with the phenomenon studied, the true behaviour may be represented
by a linear approximation by determining tangent or secant moduli of elasticity (Figure 1.2.6).
The nonlinear nature of the behaviour can also be taken into account via different models: Biot’s
semilinear model [Bemer et al., 2001] or Hertzian contact model [Vincké, 1994].

1.2.6.2 Sandstones (damage, failure)

Depending on the loading path followed, a sample of sandstone subjected to compression
may fail in two different ways.
 26 Chapter 1 • Elements of Rock Mechanics. Fundamentals

– At low mean effective stresses, brittle failure through the formation of a shear band
crossing the sample (shear localization) is observed. The result is a rapid drop in the
ability of the rock to withstand a load.
– At high mean effective stresses, failure occurs under the effect of homogeneous
microcracking (cataclastic flow). In this case we speak of ductile failure, with the rock
retaining an ability to withstand a load as the strain increases. It is important to note
here that this strain is due to slipping along a multitude of intersecting shear planes
and is therefore not true plastic strain.
Figure 1.2.8 schematises the corresponding failure surface in the q-p ' plane (deviatoric
stress-Terzaghi’s mean effective stress).

Brittle failure
"Shear localisation"
q (Pr)
Ductile failure
https://telegram.me/Geologybooks

"Cataclastic flow"

"Damage"
(C’)

pt p* p’

Figure 1.2.8
Failure surface of a sandstone.

Brittle and ductile failure modes are characterised by distinct critical stresses corre-
sponding to different steps in the development of the damage, which is associated with
acoustic activity of the rocks: microcracking, slipping with friction and pore collapse, etc.
These critical stresses can be identified by measuring the acoustic emissions during the test.
Figure 1.2.9 schematises the change in porosity (ΔΦ) as the mean effective stress (p ')
increases depending on the loading path followed. Note that in this case we are considering
the variation in Lagrangian porosity, i.e. calculated with respect to the initial volume of the
sample. The various curves show a tightening phase followed by an elastic phase (assumed
linear on the figure), these two phases being associated with a reduction in the permeability.
The third phase is nonlinear and occurs after crossing a critical effective pressure which
depends on the loading path (C ', C* or p*). The phenomena involved in this phase, known
as the post-elastic phase, are complex and may lead to a decrease or an increase in the per-
meability reflecting the competition under the effect of progressive microcracking of the
rock between firstly the greater connectivity and secondly the greater tortuosity and/or
reduction in the size of the pore throats.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 27

Start of post-elastic Hydrostatic

p’
phase compaction

p*
Dilatancy –
Shear Pc
localisation Ps C* Shear-enhanced
compaction –
Start of linear C’ Cataclastic flow
phase

0 DF

Dilatancy Contractancy

Figure 1.2.9
Behaviour of a sandstone subjected to compression.
https://telegram.me/Geologybooks

Hydrostatic loading
The curve Δφ − p ' presents a point of inflection at the critical effective pressure p*, associ-
ated with the onset of grain crushing and pore collapse induced by the growth of transgranu-
lar cracks and is accompanied by an increase in acoustic activity indicating a change of
physical mechanism. As long as p* is not reached, the rock behaviour corresponds to elastic
reduction of the porosity. Above p*, the post-elastic behaviour phase starts, systematically
inducing a decrease in the permeability in case of hydrostatic compression. The notion of
low or high mean stresses is defined with respect to the position of the ratio p '/p* on the
scale [0,1].

Ductile failure (failure by homogeneous microcracking or cataclastic flow)

Ductile failure is observed for deviatoric loads at high mean effective stress. Compared with
hydrostatic loading, the curve ΔΦ − p ' indicates a higher reduction in the porosity above an
effective compaction pressure C*, below p*. This higher reduction in the porosity, together
with a reduction in permeability more pronounced than in the hydrostatic case, is induced by
the deviatoric part of the stress field. We then speak of shear-enhanced compaction. The
curve ΔΦ − p ' presents a point of inflection at the effective collapse pressure Pc.
As long as C* is not reached, the rock behaviour corresponds to an elastic reduction of
the porosity independent of the deviatoric stress. Above C*, the acoustic emission rate
increases more rapidly, indicating the development of intragranular cracks initiated at the
points of contact between grains. When they become transgranular, these cracks increase the
stresses on the load-bearing grains, finally crushing the grains and causing the pores to
collapse. Collapse of the sample occurs at pressure Pc and is associated with a peak in the
acoustic emission rate. The non-hydrostatic nature of the loading leads to a preferential
orientation of the cracks approximately parallel to the axial stress.
 28 Chapter 1 • Elements of Rock Mechanics. Fundamentals

With consolidated sandstones, the stress levels associated with C* and Pc are very close
and only the effective compaction pressure C* is used to characterise ductile failure.

Brittle failure (failure by creation of a shear band or shear localisation)

Brittle failure is observed for deviatoric loads at low mean effective stress. Compared with
hydrostatic loading, the curve ΔΦ − p ' indicates dilating behaviour above an effective dila-
tancy pressure C ' marking the start of damage and well below p*. The stress curve then
shows a peak Ps, which corresponds to the formation of a shear band crossing the sample,
and beyond which we observe “softening” behaviour, indicating the fact that less force is
required to strain the sample at constant strain rate. This behaviour phase is sometimes
represented by elastoplastic strain softening models. It is important to bear in mind,
however, that it does not correspond to a homogeneous strain in the rock, but to slipping of
rigid blocks at the fractures.
As long as C ' is not reached, the rock behaviour corresponds to an elastic reduction of
https://telegram.me/Geologybooks

the porosity independent of the deviatoric stress. Intergranular cracks develop between C '
and Ps (probably due to cementation failure), inducing a faster increase in the acoustic emis-
sion rate. These cracks, whose preferential orientation is approximately parallel to the axial
stress, allow relative movement of the grains thereby causing an overall dilatation of the
porous space. Performing several unloading/reloading cycles between C ' and Ps would
demonstrate a reduction in the rock stiffness due to the effect of the damage. The change in
permeability is then related to the initial rock permeability.
– Rocks of low initial permeability, i.e. poorly connected rocks, will be highly sensitive
to the increased connectivity, due to the development of microcracks, and their per-
meability will therefore increase.
– For rocks of high initial permeability, i.e. well-connected rocks, the microcracks do not
modify the overall connectivity but allow grain movements, which increase the
tortuosity and/or reduce the size of the pore throats, thereby reducing the permeability.
Intragranular cracks develop on approaching the stress peak; subsequent coalescence of
these cracks finally results in failure through formation of a shear band crossing the sample.
The bulk permeability is then controlled by the properties of the localisation zone, which
may in particular obstruct the flow of fluids. The stress peak is marked by a peak in the
acoustic emission rate. The maximum acoustic emission rate associated with a brittle failure
(dilating strain) is less than the maximum rate associated with a ductile failure (compacting
strain).

1.2.6.3 Carbonates (plasticity, failure)

Unlike sandstones, carbonates may develop a significant “plastic” behaviour before failure.
The term plastic refers here to a behaviour showing irreversible strains without a decrease of
the elastic properties. These strains are induced by the coupling of the microfracturing (asso-
ciated with crushing and frictional slipping) and the plasticity of the crystals. Plasticity is
essentially encountered for loading paths at high effective mean stress and for rocks of high
initial porosity, in particular chalks.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 29

In addition, since very few events generate acoustic activity in carbonates, the effective
critical pressures must be determined using the stress-strain curves only, creating a larger
dispersion than for the sandstones [Vajdova et al., 2004].
This paragraph is based on the experimental results obtained by Yale and Crawford
(1998) on various oil-saturated carbonates of porosities ranging from 14% to 42%. The tests
followed various stress paths to approximate the in situ conditions corresponding to extrac-
tion of oil from a reservoir.

q
Hardening cap
Shear
localisation
https://telegram.me/Geologybooks

Initial yielding

p’

Figure 1.2.10
Carbonate failure/plastification model.

The stress paths with low mean effective stress resulted in brittle failure of the samples.
The stress paths with higher mean effective stress were used to determine two yield surfaces
(Figure 1.2.10). Engstrøm (1991) also observed the existence of progressive strain-
hardening of chalks for this type of loading.
According to Yale and Crawford (1998), the first yield surface (initial yielding) corre-
sponds to the stress state from which the curve ΔΦ − p ' deviates from its initial linear slope
by more than 10%. This surface is to be compared with the effective compaction pressures
C* associated with sandstones.
The second yield surface (hardening cap) is reached at the point of inflection of the curve
ΔΦ − p ' and is therefore to be compared with the collapse pressures Pc. The fundamental
difference between sandstones and carbonates lies in the significant gap between the two
yield surfaces for carbonates: the ratio between the critical hydrostatic pressures of
hardening cap and initial yielding reaches 1.8 for 20% porosity carbonates.
Strain hardening of the samples is positive (irreversible strain can only increase by increasing
the load) and mainly controlled by the porosity of the carbonates tested: for low-porosity
samples, the initial hardening threshold stresses are higher and the gap between the initial
yielding and the hardening is larger. In addition, dispersion of experimental results is more
pronounced for higher-porosity samples due to the greater sensitivity to small-scale heterogeneities.
In case of brittle failure, the stress levels associated with damage onset and the failure
peak are very close for carbonates. Failure therefore occurs quite suddenly after the elastic
phase. As for sandstones and with the same reserves, the post-peak behaviour is sometimes
represented by elastoplastic strain-softening models.
 30 Chapter 1 • Elements of Rock Mechanics. Fundamentals

1.2.6.4 Shales
The term shale designates rocks whose clay mineral content (clay volume divided by total
volume) is greater than 35% [Vernik et al., 1993] [Plumb, 1994]. It therefore includes rocks
of highly variable composition and structure. Clay minerals consist of stacks of elementary
sheets. Depending on the type of clay, the bonds between sheets are more or less strong.
Unlike kaolinite and illite, smectite has weak bonds, making it highly sensitive to water con-
tent; as a result it has the ability to fix water molecules between two neighbouring sheets and
exhibits a high swelling potential.
The study of swelling clays is highly complex and will not discussed here.
Very broadly speaking, we can identify two main categories of shales according to their
water content (weight of water divided by weight of solid particles) [Rousset, 1988]:
– “plastic” shales with high water content (> 15 %),
– “stiff” shales with low water content (< 10%).
https://telegram.me/Geologybooks

Since the elastic limit of plastic shales is virtually zero, they can withstand high strains
with no apparent macroscopic failure. After a strain hardening phase, loadings at low mean
effective stress lead to a “rounded” stress peak, followed by softening behaviour. We then
obtain a stress plateau demonstrating the residual strength retained by the shale.
This type of behaviour is generally represented by a Cam-Clay type yield surface elasto-
plastic model (Figure 1.2.11).

Strain softening

Strain hardening

p’

Figure 1.2.11
Cam-Clay model yield surface.

The behaviour of stiff shales is more like that of an elastobrittle rock. They may even
exhibit very high strengths [Lin, 1981]. At low mean stresses, a sharp peak is observed, fol-
lowed by a rapid drop in the stress. The brittle failure data of stiff shales can be represented by
a linear Mohr-Coulomb criterion or a nonlinear criterion [Green et al., 1973] [Lin, 1981] [Sar-
gand and Hazen, 1987] [Cook et al., 1991] [Steiger and Leung, 1992] [Gavito, 1996]. Few
experimental data are available, however, to close the yield surface for high mean stresses.
In addition, plastic shales exhibit pronounced delayed response. At a given stress state,
the rock may therefore undergo strains over a long period of time, which may stabilise or
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 31

not. Stiff shales may also develop more or less large delayed strains. Viscous behaviour
models may be required to take into account these phenomena [Boidy, 2002].

1.2.7 Simulating depletion in the laboratory

Depletion tests in the laboratory are generally conducted on cylindrical samples placed in a
triaxial cell, with test procedures leading to a stress state more or less close to the in situ con-
ditions. Depletion can be simulated by either reducing the pore pressure at constant axial
stress to simulate a constant weight of earth (“true depletion” tests), or by increasing the
effective stresses at constant pore pressure. Obviously, simulating depletion by reducing
pore pressure is more representative of reservoir production mode (see § 3.1.3).
One disadvantage with replacing the pressure reduction by an increase in effective
stresses is that it introduces Biot’s coefficient. Depletion tests are in fact mainly carried out
to characterise the strain behaviour of the rock and, in this case, the relevant effective stress
https://telegram.me/Geologybooks

is Biot’s.
The loading path followed in terms of the evolution in axial stress and confining pressure
may be:
– isotropic ( Δpc ' = Δσa ');
– oedometric (i.e. uniaxial, the confining pressure evolution is then controlled automat-
ically to maintain the condition Δεr = 0 );
– proportional ( Δpc ' = K Δσa ' ), in order to follow the reservoir stress path when known
(see § 3.1.1).
Numerous experimental studies demonstrate that the behaviour of reservoir rocks is
highly dependent on the stress path [Holt, 1990] [Warpinski and Teufel, 1990] [Rhett and
Teufel, 1992a] [Ruisten et al., 1996] [Schutjens and de Ruig, 1996] [Keaney et al., 1998]
[Schutjens et al., 1998]. Consequently, if the stress path is known, it should be followed in
the tests conducted to determine the reservoir rock characteristics.

1.3 MODELLING ROCK BEHAVIOUR

Warning! Throughout Chapter 1.3, compressive stresses are taken to be negative, according
to the solid mechanics convention.
We will examine the modelling of various aspects of the behaviour of a rock saturated
with a single fluid: a liquid or a perfect gas. Firstly, we will consider the case of a solid skel-
eton exhibiting isotropic linear elastic behaviour. We will then introduce nonlinear elastic
behaviour of the solid skeleton. Our intention here is not to provide an exhaustive descrip-
tion of the theoretical framework of poromechanical modelling, which interested readers can
find in [Coussy, 1991, 1995, 2004], but to show precisely where the nonlinearities occur in
the various cases considered. Lastly, we will present an example of a poroelastoplastic
behaviour model: the modified Cam-Clay model.
 32 Chapter 1 • Elements of Rock Mechanics. Fundamentals

1.3.1 Representation of the porous medium

We adopt the formalism developed by Coussy (1991, 1994, 2004). The porous medium is
represented as the superimposition of two continuous interacting media: the solid skeleton
and the interstitial fluid (Figure 1.3.1). It is considered as an open continuous medium and
its strains are described by those of its skeleton.

Occluded Connected
porosity porosity

= +
Solid matrix

Porous medium Skeleton Fluid

https://telegram.me/Geologybooks

Figure 1.3.1
Porous medium seen as the superimposition of two continuous media.

1.3.2 Fundamental assumptions

We make the following assumptions:
– quasistatic evolutions (to neglect inertia forces);
– infinitesimal transformations ( grad ξ << 1, where ξ is the displacement vector of
the skeleton particles, to ensure first order agreement of the Lagrangian and Eulerian
approaches);
– small displacements of the skeleton particles with respect to a characteristic length of
the porous medium Lo ( ξ Lo << 1, to merge the initial and current configurations for
the space arguments of the various fields);
and possibly:
(
– small variations in the fluid density ( ρ fl − ρofl ) ρofl << 1);

– small variations in the Lagrangian porosity ( Φ − Φo( ) Φo << 1).

1.3.3 Equilibrium equations

1.3.3.1 Conservation of mass
The continuity equation expresses conservation of the fluid mass:

dm
+ div w = 0 (1.3.1)
dt
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 33

The fluid mass content per initial volume unit m is simply related to the Lagrangian
porosity Φ by the following relation which expresses complete saturation of the medium:
m = ρ fl Φ (1.3.2)
where rfl designates the fluid density.
The relative flow vector of fluid mass w is directly proportional to the relative flow rate
of the fluid with respect to the skeleton:
⎛
w = ρ fl V V =φ ⎜ v fl − vs ⎞⎟ (1.3.3)
⎝ ⎠
where f is the Eulerian porosity, which is the ratio of the current porous volume to the cur-
rent total volume, and V is the filtration vector.
Assuming infinitesimal transformations, the Lagrangian and Eulerian porosities are sim-
ply related by:
( )
https://telegram.me/Geologybooks

Φ = 1+ ε φ

where ε = tr ε .

1.3.3.2 Equation of motion

Momentum balance leads to the equation of mechanical equilibrium:
div σ + ρ F = 0 (1.3.4)
( )
where ρ = ρs 1 − φ + ρ fl φ is the density of the porous medium and F represents the exter-
nal volume forces often reduced to the gravitational forces g .
Assuming small variations in fluid density and porosity, the equation of mechanical
equilibrium can be simplified to:
div σ + ρo F = 0

( )
where ρo = ρso 1 − Φo + ρofl Φo .

1.3.4 Fluid constitutive equations

The fluid constitutive equations can be defined by applying the first and second laws of ther-
modynamics to the interstitial fluid. In particular, they can be expressed using the fluid-spe-
cific free enthalpy gfl. In this case, they relate the density rfl and the fluid-specific entropy sfl to
the thermodynamic pressure pp (associated with the pore pressure) and to the temperature T.
dpp
dg fl = − s fl dT (1.3.5a)
ρ fl

or (
g fl = g fl pp , T )
 34 Chapter 1 • Elements of Rock Mechanics. Fundamentals

1 ∂g fl ∂g fl
= − s fl = (1.3.5b)
ρ fl ∂pp ∂T

Note that the fluid-specific enthalpy is related to its specific free energy yfl:
pp
g fl = ψ fl + (1.3.6)
ρ fl
By differentiating the constitutive equations (1.3.5b), we obtain:
d ρ fl dpp
= − 3α fl dT (1.3.7a)
ρ fl K fl

dpp dT
ds fl = −3α fl + cp (1.3.7b)
https://telegram.me/Geologybooks

ρ fl T

where Kfl designates the fluid bulk modulus, 3α fl the fluid volumetric thermal expansion
coefficient and cp the fluid volumetric specific heat capacity at constant pressure.
For a low compressibility liquid and small pressure and temperature variations, we can
obtain a linearised expression of the fluid constitutive equations by integrating equations
(1.3.7) with Kfl, 3α fl and cp constant:

⎡ pp − pop ⎤
1
=
1 ⎢
ρ fl ρo ⎢
1−
K fl
+ 3α fl T − To ( )⎥⎥
fl ⎣ ⎦

pp − pop T − To
s fl = sofl − 3α fl + cp
ρofl To

where the index o refers to the initial state.

This linearisation is no longer relevant, however, for a high compressibility gas. If this
gas is represented by a perfect gas model, we obtain:
pp T
o
ρ fl = ρofl (1.3.8)
pop T

Differentiation of equation (1.3.8) and comparison with equation (1.3.7a) demonstrate

that the bulk modulus and the volumetric thermal expansion coefficient of a perfect gas are
equal respectively to: K fl = pp and 3α fl = 1 T .
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 35

1.3.5 Clausius-Duhem inequality and thermal equation

By applying the first law of thermodynamics to the porous medium seen as the superimposi-
tion of the solid skeleton and the interacting interstitial fluid, we obtain the energy equation,
which expresses the conservation of energy:

dE
dt
=σ :
dε
dt
(
− div hfl w + q + F .w ) (1.3.9)

where E designates the internal energy density of the porous medium per initial volume unit,
hfl = g fl − T s fl the specific enthalpy of the fluid and q the outgoing heat flow vector.
The second law of thermodynamics leads to the entropy balance:

( )
dS ⎛ q⎞
+ div s fl w ≥ − div ⎜ ⎟ (1.3.10)
dt ⎝T⎠
https://telegram.me/Geologybooks

where S designates the entropy density of the porous medium per initial volume unit. The ≥
sign shows that there is a spontaneous production of entropy: ϕ T , where ϕ is the overall
dissipation.
By introducing the porous medium density of free energy ψ = E − T S , using the above
two relations, we derive the Clausius-Duhem inequality which shows that the dissipation ϕ
is non-negative:
ϕ = ϕs + ϕ fl + ϕth ≥ 0 (1.3.11)

dε dm dT dψ
ϕs = σ : + g fl −S − (1.3.12)
dt dt dt dt
represents the intrinsic dissipation, which turns out to be related to the solid skeleton alone.
If we introduce the entropy and free energy of the skeleton per initial volume unit,
Ss = S − m s fl and ψ s = ψ − m ψ fl , the fluid constitutive equations (1.3.5a) and the relation
m = ρ fl Φ lead to the following expression of the intrinsic dissipation:

dε dΦ dT dψ s
ϕs = σ : + pp − Ss − (1.3.13)
dt dt dt dt

ϕ fl = ⎡ − grad pp + ρ fl F ⎤ ⋅V (1.3.14)
⎣ ⎦
represents the dissipation associated with the fluid and more particularly its movement with
respect to the skeleton.
q
ϕth = − ⋅ gradT (1.3.15)
T
represents the thermal dissipation.
 36 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Due to the highly different nature of these dissipations, we adopt the decoupling assump-
tion which consists in replacing inequality (1.3.11) by three separate inequalities:

dε dΦ dT dψ s
ϕs = σ : + pp − Ss − ≥0 (1.3.16)
dt dt dt dt

ϕ fl = ⎡ − grad pp + ρ fl F ⎤ ⋅V ≥ 0 (1.3.17)
⎣ ⎦

q
ϕth = − ⋅ grad T ≥ 0 (1.3.18)
T
After identifying the spontaneous production of entropy ϕ T , inequality (1.3.10) finally
leads to the thermal equilibrium equation:

⎡ dS
( )
⎤
T ⎢ + div s fl w ⎥ = − div q + ϕM
https://telegram.me/Geologybooks

⎣ dt ⎦
where ϕM = ϕs + ϕ fl represents the mechanical dissipation density.

1.3.6 Conduction laws

1.3.6.1 Fluid conduction law

Flow dissipation expression (1.3.17) demonstrates that the law of fluid conduction must
relate the filtration vector to the flow force − grad pp + ρ fl F . In this case, its simplest form
is the linear relation given by Darcy’s law. For an isotropic medium, it is written:

w
ρ fl (
=V = Λ − grad pp + ρ fl F ) (1.3.19)

Note that the volume forces often prove to be negligible compared with the hydraulic
gradient.
For a low compressibility interstitial fluid, the hydraulic conductivity L is expressed as a
function of the intrinsic permeability k, which depends solely on the geometry of the porous
network, and of the dynamic viscosity of the interstitial fluid mfl:

k
Λ= (1.3.20)
μ fl

An additional interstitial pressure dependence must nevertheless be taken into account

when the interstitial fluid is a gas [Patron, 1999]:

k kg
Λ= + (1.3.21)
μ fl pp
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 37

When these models are included in finite elements simulation software, we may intro-
duce variation laws of the hydraulic parameters k and kg as a function of the porosity, which
represents a simplified way of taking into account the evolution of the porous network and
thereby of including this aspect of the hydro-mechanical coupling in the numerical model.

1.3.6.2 Heat conduction law

Thermal dissipation expression (1.3.18) shows that the heat conduction law must relate the
heat flow vector to the force responsible for heat production − grad T . In this case, its sim-
plest form is the linear relation given by Fourier’s law. For an isotropic medium, it is writ-
ten:
q = − λT grad T (1.3.22)
where lT is the thermal conductivity.
https://telegram.me/Geologybooks

1.3.7 Constitutive equations of the skeleton and the porous medium

Intrinsic dissipation expression (1.3.13) shows that, due to its nature, the skeleton density of
free energy depends on the external state variables: ε , Φ and T (observable variables whose
variation can be controlled externally), their associated variables being σ , pp and Ss.
According to the postulate of local state [Coussy, 2004], the skeleton density of free energy
is independent of the evolution rates and can finally be expressed in the form:

(
ψ s = ψ s ε , Φ, T , χ1 , χ2 ,..., χN ) (1.3.23)
where the variables ci are “internal” state variables (their variation cannot be controlled
externally).
Since the intrinsic dissipation cannot be negative, we have:

⎛ ∂ψ s ⎞ ⎛ ∂ψ s ⎞ ⎛ ∂ψ s ⎞ ∂ψ s
⎜σ − ⎟ :d ε + ⎜ pp − ⎟ dΦ − ⎜ S + ⎟ dT − ⋅dχ ≥ 0 (1.3.24)
⎝ ∂ε ⎠ ⎝ ∂Φ ⎠ ⎝ ∂T ⎠ ∂χ

We will now consider evolutions where the internal variables are constant. Since ine-
quality (1.3.24) must be satisfied for any evolution in the external variables independently of
each other, we obtain the following constitutive equations:

∂ψs ∂ψs ∂ψs

σ= pp = S=− (1.3.25)
∂ε ∂Φ ∂T

If we now introduce the potential Gs = ψ s − pp Φ , the constitutive equations can be written:

∂Gs ∂Gs ∂Gs

σ= Φ=− Ss = − (1.3.26)
∂ε ∂pp ∂T
 38 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Note that equations (1.3.25) and (1.3.26) remain valid for evolutions including a varia-
tion of internal variables. Complementary evolution laws governing the evolution of internal
state variables need to be introduced however.
Intrinsic dissipation expression (1.3.12) leads to the expression of the density of free
energy and the constitutive equations of the porous medium relating the external state vari-
ables ε , m and T to their associated variables σ , gfl and Ss:

(
ψ = ψ ε , m, T , χ1 , χ2 ,..., χN ) (1.3.27)

∂ψ ∂ψ ∂ψ
σ= g fl = S=− (1.3.28a)
∂ε ∂m ∂T

For porous media with no chemical activity, however, it is simpler to introduce the fluid
mass content directly into the constitutive equations (1.3.26):
https://telegram.me/Geologybooks

∂Gs m ∂Gs ∂Gs

σ= =− S=− + m s fl (1.3.28b)
∂ε ρ fl ∂pp ∂T

1.3.8 Linear poroelastic behaviour

1.3.8.1 Skeleton constitutive equations

We will now consider a porous medium whose skeleton exhibits isotropic linear elastic
behaviour. We can therefore adopt for ys a quadratic expression ψ sel which depends solely
on the first and second invariants of the strain tensor.
By differentiating the constitutive equations (1.3.26) with Gs = Gsel = ψ sel − pp Φ , we
obtain the porous skeleton thermoporoelastic constitutive equations which involve only
coefficients which are independent of the interstitial fluid:

⎛ 2G ⎞
d σ = ⎜ Kd − d ε 1 + 2G d ε − b dpp 1 − 3αd Kd dT 1
⎝ 3 ⎟⎠
dpp
d Φ = b dε + − 3αΦ dT (1.3.29)
N
dT
dSs = 3αd Kd d ε − 3αΦ dpp + Cε
T
where Kd is the skeleton bulk modulus, G the skeleton shear modulus, b Biot’s coefficient,
1/N the pore compressibility, 3ad the skeleton volumetric thermal expansion coefficient,
3aΦ the pore volumetric thermal expansion coefficient and Cε the skeleton volumetric heat
capacity at constant strain.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 39

Assuming that the solid matrix is homogeneous and isotropic, we can determine expres-
sions relating coefficients b, N and aΦ, to the characteristics of the solid matrix [Coussy,
2004]:
Kd 1 b − Φo
b = 1− = (1.3.30)
Ks N Ks

αd = αs (
αΦ = b − Φo αs) (1.3.31)

where Ks is the bulk modulus of the solid matrix and as the thermal expansion coefficient of
the solid constituent.

1.3.8.2 Porous medium constitutive equations

The thermoporoelastic constitutive equations of the porous medium considered as an open
https://telegram.me/Geologybooks

thermodynamic system involve the fluid mass content m. They are obtained from constitu-
tive equations (1.3.28b) and the fluid constitutive equations (1.3.7).
Differentiating relation (1.3.2), we obtain:

dm d ρ fl
= dΦ + Φ
ρ fl ρ fl

The fluid constitutive equations then lead to the relations:

dm dpp
= dΦ + Φ − 3Φα fl dT
ρ fl K fl

( )
d m s fl = s fl dm − 3Φα fl dpp + m cp
dT
T
By introducing the above relations in equations (1.3.29), we can eliminate the
Lagrangian porosity variation, replacing it by the variation in fluid mass content. Finally, we
obtain the constitutive equations:

⎛ 2G ⎞
d σ = ⎜ Kd − d ε 1 + 2G d ε − b dpp 1 − 3αd Kd dT 1
⎝ 3 ⎟⎠
dm dpp
= b dε + − 3αm dT (1.3.32)
ρ fl M
dT
dS − s fl dm = 3αd Kd d ε − 3αm dpp + Cεd
T

1 1 Φ
with = + (1.3.33)
M N K fl
 40 Chapter 1 • Elements of Rock Mechanics. Fundamentals

αm = αΦ + Φα fl and Cεd = Cε + m cp (1.3.34)

The skeleton characteristics Kd, G, b and ad, prove to be drained properties of the porous
medium, i.e. they can be measured from tests where the interstitial pressure is maintained
constant ( dpp = 0). The drained volumetric heat capacity at constant strain Cεd takes into
account the fluid mass content via the term m cp .
Biot’s modulus M and the differential thermal expansion coefficient am depend on the
interstitial fluid. For a porous medium saturated by a low compressibility fluid, Biot’s
modulus and the differential thermal expansion coefficient can be considered as constant to
the first order of approximation. When the interstitial fluid is highly compressible, however,
they depend on the interstitial pressure and the temperature.
The drained constitutive equations (1.3.32) can be combined to obtain an undrained
expression of these equations:
https://telegram.me/Geologybooks

⎛ 2G ⎞ dm
d σ = ⎜ Ku − ⎟ d ε1 + 2Gd ε − bM 1 − 3αm Ku dT 1
⎝ 3 ⎠ ρ fl
dpp dm
= − bd ε + + 3αm dT (1.3.35)
M ρ fl
dm dT
dS − s fl dm = 3α Ku d ε − 3α m M + Cεu
ρ fl T

with Ku = Kd + M b2, 3αu Ku = 3αd Kd + 3αm M b and Cεu = Cεd − 9 T αm

2 M . K is the und-
u
rained bulk modulus and 3au the undrained volumetric thermal expansion coefficient, in
other words these parameters can be measured from tests where the fluid mass content is
maintained constant.

1.3.8.3 Equation governing the variation in elementary Eulerian porosity

The Lagrangian and Eulerian porosities are related by the Jacobian of the transformation:
Φ = J φ . Assuming infinitesimal transformations, we can easily obtain the following relation:

(
Φ = 1+ ε φ )
A first order approximation yields:
d Φ = dφ + Φo d ε

The constitutive equation (1.3.29) then leads to the equation governing the variation in
elementary Eulerian porosity f:
dpp
(
dφ = b − Φo d ε + ) N
− 3αΦ dT (1.3.36)
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 41

1.3.9 Nonlinear poroelastic behaviour

We will now adopt Biot’s semilinear model (1973) which assumes that the nonlinear
behaviour of some porous media is due to local geometry changes (variation in contact
surfaces, closure of cracks, etc.) and that the behaviour of the solid matrix remains linear.
We will also consider an isotropic porous medium and isothermal transformations. In
addition, the reference state is assumed to be free of strains and pressure.
The semilinear constitutive equations are deduced by decomposing the stress state into
an effective part acting on the “dry” medium (no fluid) and an isotropic part acting on the
solid matrix [Bemer et al., 2001]:
σ = σ − pp 1 (1.3.37)

pp = 0 + pp
Note that the effective stress σ defined in this way corresponds to Terzaghi’s effective stress
https://telegram.me/Geologybooks

(Figure 1.3.2).
s s — PP1

pp 0 pp
pp 0 pp
= +
pp pp
0

Figure 1.3.2
Load decomposition.

Since the matrix behaviour is assumed to be linear and isotropic with a bulk modulus Ks,
decomposition of the stresses (1.3.37) is associated with the following decomposition of the strains:
pp
ε =ε − 1 (1.3.38)
3 Ks
where ε represents the strains produced by σ .
Figure 1.3.3 illustrates the validity of Biot’s semilinearity assumption for a 19.8%
porosity reservoir sandstone. This experimental curve is obtained from a hydrostatic test
where the confining pressure and the pore pressure are increased simultaneously, thereby
avoiding any variation in the effective stress and the associated nonlinear strain.
Since only the strains ε include nonlinear variations, potential Gs is taken to be the sum
of the quadratic potential of the linear poroelasticity Gsel leading to the constitutive equa-
tions (1.3.29) and a third-order nonlinear potential H which is purely a function of ε :

( ) ( )
Gs ε , pp = Gsel ε , pp + H ε ( ) (1.3.39)
 42 Chapter 1 • Elements of Rock Mechanics. Fundamentals

120

100

80
PP (MPa)

60

40

Experimental curve
20
Model

0
–0.004 –0.003 –0.002 –0.001 0

Figure 1.3.3
https://telegram.me/Geologybooks

Illustration of the linear behaviour of the matrix.

The constitutive equations (1.3.26) then lead to the relations:

∂Gsel ∂H
σ= +
∂ε ∂ε

∂Gsel 1 ∂H
Φ=− − tr
∂pp 3 Ks ∂ε

Hence,
⎛ 2G ⎞ ∂H
σ = ⎜ Kd − ⎟ ε 1 + 2G ε − b pp 1 +
⎝ 3 ⎠ ∂ε
(1.3.40)
pp1 ∂H
Φ − Φo = b ε + − tr
N 3Ks ∂ε

Equations (1.3.40) can be decomposed into a hydrostatic part and a deviatoric part:

1 ∂H
σ = Kd ε − b pp + tr
3 ∂ε
∂H 1 ∂H
s = 2Ge + − tr (1.3.41)
∂ε 3 ∂ε
pp 1 ∂H
Φ − Φo = b ε + − tr
N 3 Ks ∂ε
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 43

tr σ ε
where σ = , s = σ − σ 1 and e = ε − 1. Note that s = s and e = e .
3 3

Biot (1963) adopts for H the following third-order expression:

H=D
ε3
3
)(
+ ( F − D ε I2 − 3 I3 ) (1.3.42)

where I2 = ε11 ε22 + ε22 ε33 + ε33 ε11 − ε12 ε21 − ε23 ε32 − ε31 ε13 and I3 = det (e). F and D are
constants which have the same dimensions as stress.

Taking the principal stress coordinate system, we obtain the constitutive equations:

D � 2F 2
s = Kd e - b pp � e
3
https://telegram.me/Geologybooks

s11 � 2G e11 �
F-DÈ 2
� �
e � 2e11e22 � e332 � 2e11e33 - 2 e112 � 2e22 e33 ˘˙ (1.3.43a, b, c)
3 ÍÎ 22 ˚
pp D � 2 F
F - Fo � b e � - e2
N 3 Ks

The expressions associated with the other principal stress directions are obtained simply
by circular permutation.
After differentiating equations (1.3.43a) and (1.3.43c) we can explicitly define a tangent
drained bulk modulus, a tangent Biot’s coefficient and a tangent pore compressibility func-
tion of e . We obtain:

� � � �
ds = Kdt e d e - bt e dpp
t (1.3.44a, b)
Ê 1ˆ
� �
d F � bt e d e � Á ˜
ËN¯
� e � dpp
D � 2F
� �
Kdt e � Kd � 2
3
e

D � 2F
with � �
bt e � b - 2
3 Ks
e (1.3.45a, b, c)

t
Ê 1ˆ 1 D � 2F
� �
ÁË N ˜¯ e � N -
3 Ks2
e

t
Kdt Ê 1ˆ bt - Fo
Note that the following relations are satisfied: bt � 1 - and Á ˜ � .
Ks ËN¯ Ks
 44 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Equation (1.3.43a) can be used to express e as a function of the mean effective stress
s � s � pp and therefore relate the tangent drained bulk modulus, the tangent Biot’s coeffi-
cient and the tangent pore compressibility to the effective mean stress alone:

D � 2F
� �
Kdt s � Kd2 � 4
3
s

� �
bt s � 1 - �1 - b �2 - 4 D3�K2 F Ks (1.3.46a, b, c)
s s

t 2
Ê 1 ˆ 1 - Fo Ê 1- b ˆ D � 2F s
ÁË N ˜¯ � K - Á K ˜ - 4
s Ë s ¯ 3 Ks2 Ks2

Figure 1.3.4 shows that the tangent drained bulk modulus does in fact vary as a function
of the effective means stress s � s � pp .
https://telegram.me/Geologybooks

10000

8000

6000
Kot (MPa)

4000

2000 pp = 1 MPa
pp = 51 MPa

0
0 20 40 60
s + pp (MPa)

Figure 1.3.4
Variation law of the tangent drained bulk modulus on a sandstone with 19%
porosity [Bemer et al., 2001].

It is more difficult to define a tangent shear modulus using Biot’s semilinear model. From
the deviatoric part of the constitutive equations (1.3.43b), we must obtain a relation of type

��
d s � 2Gt e de (1.3.47)
i.e., according to the principle direction 1,

��
ds11 � 2Gt e de11 (1.3.48)
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 45

However, by differentiating equation (1.3.43b), we obtain:

ds11 � 2G de11

�2
F-DÈ
3 Î
� � � � � �
- 2e11 � e22 � e33 d e11 � e11 � e22 - 2e33 d e22 � e11 - 2e22 � e33 d e33 ˘
˚
To obtain an expression of type (1.3.48), relation e33 � e22 must necessarily be satisfied,
which amounts to having a condition of axisymmetry around principal axis 1. In this special
case, equation (1.3.43b) leads to relation:

q � 3G ed �
9
4
� �
F - D ed2 (1.3.49)
2
� �
where q � s 22 - s11 is the deviatoric stress and ed � e22 - e11 � ed the associated devia-
3
toric strain.

A tangent shear modulus function of ed can be defined explicitly by differentiating equa-

https://telegram.me/Geologybooks

tion (1.3.49). We obtain:

� �
dq � 3Gt ed d ed (1.3.50)

with � �
Gt ed � G � 3
F-D
2
ed (1.3.51)

Equation (1.3.49) can be used to express ed as a function of the deviatoric stress q and
thereby relate the tangent shear modulus to the deviatoric stress alone:

��
Gt q � G2 � F - D q� � (1.3.52)

To date, the validity of this variation law has not been checked experimentally.

1.3.10 Poroelastoplastic behaviour

The poroelastic model described here is the modified Cam-Clay model [Roscoe and Bur-
land, 1968]. This model was developed to represent the isothermal behaviour of normally
consolidated or slightly overconsolidated clays. It is an isotropic poroelastoplastic behaviour
model with isotropic strain hardening, defined for a water-saturated porous medium and
therefore initially associated with a model describing the viscous flow of a low compress-
ibility fluid. Extending poroelastoplasticity to unsaturated porous media results in relatively
complex behaviour models. Some examples can be found in [Alonso et al., 1990] and [Dan-
gla et al., 1997].

1.3.10.1 Plastic strains, plastic porosity and trapped energy

Apart from interstitial pressure and strain-hardening parameter(s), the yield function of an
isotropic material depends on the first three invariants of the stress tensor. Models often
 46 Chapter 1 • Elements of Rock Mechanics. Fundamentals

implement yield functions which depend only on the first two invariants. The mean stress p
and the deviatoric stress q are generally used in rock mechanics [Charlez, 1994]:
tr s 3 tr s
p� and q � s : s with s � s - 1
3 2 3
The volume strain ev and the deviatoric strain ed correspond respectively to the mean and
deviatoric stresses:
2 tr e
ev � tr e and ed � e : e with e � e - 1
3 3
Stresses p and q are the variables associated with strains ev and ed by the density of free
energy ys. We obtain:
s :d e � p d ev � q d ed
https://telegram.me/Geologybooks

Since the intrinsic dissipation is positive (1.3.16), we obtain the inequality:

d ev d ed d F dy s
js � p �q � pp - �0 (1.3.53)
dt dt dt dt
When considering evolutions where the internal variables are constant, the following
constitutive equations can be deduced from (1.3.53):
�y s �y s �y s
p= q= pp � (1.3.54)
�ev �ed �F

Coussy defines poroelastoplasticity as being “the ability of porous materials to undergo

permanent strains and permanent changes in porosity and, as a consequence, permanent
changes in fluid mass content” [Coussy, 2004]. The strain increments can thus be decom-
posed into an elastic part and a plastic part:

d ev � d eve � d evp and d ed � d ede � d edp (1.3.55)

Rather than a plastic variation in fluid mass content, it is generally preferred to introduce
a plastic Lagrangian porosity variation [Dangla et al., 1997]. We then write:

d F � d Fe � d F p (1.3.56)
A thermodynamic state is characterised by the external state variables: ev, ed and Φ, the
internal state variables: evp , edp , Φp and variables characterising the strain hardening c :

�
y s � y s ev , ed , F, evp , edp , F p , c �
Using the energy separation assumption, we can express the free energy of the skeleton
in the following form [Coussy, 2004]:

�
y s � y sel ev - evp , ed - edp , F - F p � U c� � � (1.3.57)
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 47

where y sel represents the proportion of the free energy of the skeleton recoverable as
mechanical energy and therefore depends only on reversible strains and porosities. U repre-
sents the quantity of energy trapped through the strain-hardening phenomena and is there-
fore related to the strain-hardening variables.
By introducing equations (1.3.54) and (1.3.57) in inequality (1.3.53), we obtain the fol-
lowing intrinsic dissipation expression:

d evp d edp d F p �U d c
js � p �q � pp - � �0 (1.3.58)
dt dt dt � c dt

which demonstrates that the thermodynamic forces associated with evp , edp , Φp and c , are
respectively p, q, pp, and the thermodynamic force V defined by the relation:

�U
V�- (1.3.59)
�c
https://telegram.me/Geologybooks

1.3.10.2 Cam-Clay model constitutive equations

Using the same quadratic expression adopted in the context of linear poroelasticity for y sel ,
we can deduce from equations (1.3.54) and (1.3.57) the following expression of the constitu-
tive equations:
dp ' � d ( p � dpp ) � K0 ( d ev - d evp )

dq � 3G ( d ed - d evp ) (1.3.60a, b, c)
dpp
d F - d F p � b( d ev - d evp ) �
N
where p ' � p � b pp represents the elastic mean effective stress.
For a low compressibility interstitial fluid, equation (1.3.60c) can also be expressed as a
function of the fluid mass content m:
dm
rofl
�
- d F p � - b d ev - d ev p � � dp
M
(1.3.61)

The constitutive equations (1.3.60) define a model whose elastic part is linear. The
model defined by Roscoe, however, assumes that the constitutive equations are in fact non-
linear. This special formulation is based on Terzaghi’s hypothesis, which assumes elastic
and plastic incompressibility of the solid matrix. The elastic and plastic effective stresses are
then identical to Terzaghi’s effective stresses, defined by: s ' � s � p1 [Charlez, 1994].
The choice of a nonlinear constitutive equation is based on experimental observations
which show that, under hydrostatic loading, the void ratio decreases in direct proportion to
the logarithm of the absolute value of the mean effective stress. Hence,

d p'
dee � -k (1.3.62)
p'
 48 Chapter 1 • Elements of Rock Mechanics. Fundamentals

where k is the swelling coefficient. For small strains and an incompressible matrix, the void
ratio is related to the volume strain by d ev � de 1 � eo . The nonlinear volumetric constitu-
tive equation of the modified Cam-Clay model is then written:

k d p'
d ev - d evp � - (1.3.63)
1 � eo p '

The deviatoric constitutive equation is assumed to be linear, thus:

dq
d ed - d edp � (1.3.64)
3G
The model formulated in this way is thermodynamically admissible [Coussy, 2004].
In view of the various constitutive equations defined above, two formulations of the
Cam-Clay model can be considered:
https://telegram.me/Geologybooks

– a linear formulation defined by equations (1.3.60a, b),

– a semilinear formulation defined by equations (1.3.63) and (1.3.64).
Unlike the linear model, the semilinear model is based on Terzaghi’s hypothesis. Note
that constitutive equation (1.3.60a) is obtained by linearising relation (1.3.63) around a
mean effective stress po'.

1.3.10.3 Plasticity criterion

The plasticity criterion of the Cam-Clay model is defined by the following yield function:

� � 12 ÈÍÍÎ� p � b p � � q ˘
2 2
f p, q, pp , pcr � p
� pcr - pcr2 ˙
2
˙˚

where b and M are constants and pcr is a positive strain-hardening variable called the (cur-
rent plastic) critical effective pressure which is associated with the scalar thermodynamic
force V .
pb ' � p � b pp is the plastic mean effective stress, not to be confused with the elastic
mean effective stress p ' � p � bpp . The yield surface f is then written:

� � 12 ÈÍÍÎ� p '� p � � q ˘
2 2
f pb ', q, pcr � b cr
- pcr2 ˙ (1.3.65)
2
˙˚

� � � �
In plane pb ', q , the plasticity criterion f pb ', q, pcr � 0 defines a family of ellipses

passing through the origin, having a maximum value at pb ' � - pcr and cutting the x-axis at

pb ' � - 2 pcr (Figure 1.3.5). The latter point, which corresponds to the current elastic limit of
the material under hydrostatic loading, is called the effective consolidation pressure. Double
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 49

Zero dilatancy
q
H=0
Plastic
Plastic dilatancy H < 0
contractancy
H>0

M
Compression Tension

–2pcr –pcr Pb'

Figure 1.3.5
Cam-Clay model yield surface.
https://telegram.me/Geologybooks

the value of the critical pressure represents in fact the absolute value of the maximum plastic
mean effective pressure previously undergone by the material. A material is therefore said to
be normally consolidated (or overconsolidated) if the current plastic mean effective stress is
equal to (or less than in absolute value) the consolidation effective pressure.

1.3.10.4 Flow rule

The flow rule defines the evolution of plastic strains when the plasticity criterion has been
reached. The Cam-Clay model considers a normal flow rule associated with the plasticity
criterion:
�f �f �f
d evp � d l d edp � d l dFp � dl
�p �q � pp

Hence, �
d evp � d l pb '� pcr �
q
d edp � d l (1.3.66)
2

�
d F p � d l b pb '� pcr �
with d l �
1 È
HÎÍ � �
pb '� pcr dpb '�
q dq ˘
2 ˚
˙ � 0, where H is the strain-hardening modulus.

From the above expressions, we deduce that d F p � b d evp .

We will now consider the straight line of equation pb '� pcr � 0 . This straight line is a
line of symmetry for the yield surface which it separates into two equal semi-ellipses.
 50 Chapter 1 • Elements of Rock Mechanics. Fundamentals

For points on the yield surface belonging to the semi-ellipse located in the half-plane
defined by pb '� pcr � 0 , flow rule (1.3.66) shows that d evp � 0 and d F p � 0 . The material
is then plastically compacting and the plastic porosity decreases.
Inversely, for points on the yield surface belonging to the semi-ellipse located in the
half-plane defined by pb '� pcr � 0, we have d evp � 0 and d F p � 0. The material is then
plastically dilating and the plastic porosity increases.
For the point located at the intersection between the straight line of equation pb '� pcr � 0
and the ellipse, flow rule (1.3.66) shows that d evp � 0 and d F p � 0 . Plastic strain then takes
place with no variation in volume and the plastic porosity remains unchanged.

1.3.10.5 Strain-hardening equations

� �
The Cam-Clay model introduces a potential h p, pp , pcr not associated with the criterion to
specify the evolution equations governing the scalar hardening variable c:
https://telegram.me/Geologybooks

� � 1 b - Fo
� �
2
h p, pp , pcr � p � b pp � pcr (1.3.67)
� �
2 2
1 - Fo

c is then related to the critical pressure pcr by:

d c � dl
�h
� dl
b - Fo
p � b pp � pcr � � (1.3.68)
� pcr
� �
2
1 - Fo

Relations (1.3.66) and (1.3.68) show that variable c can be identified with the plastic
void ratio ep, which represents the irreversible variation of void ratio with respect to a given
initial reference state:
F p - Fo evp
ep � (1.3.69)
� �
2
1 - Fo

The expression of pcr as a function of ep remains to be determined. Experience suggests

using an equation of type [Charlez, 1994] (Figure 3.1.6):

d pcr
de p � - l - k � � pcr
with l � k (1.3.70)

where l is the (plastic) compressibility coefficient.

Relation (1.3.70) can equally be expressed as a function of the plastic volume strain:

d pcr 1 b - Fo
�- d evp (1.3.71)
l -k
� �
pcr 2
1 - Fo
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 51

0.0

Oedometric curve
–0.2

–0.4
Volumetric strain (%)

k
1 + eo l
–0.6 1 + eo
1
–0.8

–1.0

–1.2 1

–1.4
1 10 100
https://telegram.me/Geologybooks

Effective mean stress (MPa)

Figure 1.3.6
Poroplastic behaviour of an argillite in an oedometric test
(under Terzaghi’s hypothesis).

Relations (1.3.59) and (1.3.70) lead to the following expression of trapped energy:

È p ˘
� � � �
o exp - e
U e p � l - k pcr Í ˙
ÍÎ l - k ˙˚
(1.3.72)

The strain-hardening modulus H is then:

H�
1 b - Fo
�
p p � b pp � pcr �� p � b p � (1.3.73)
� �
l -k 2 cr p
1 - Fo
Consequently, for yield points such that the material exhibits plastically compacting
behaviour, the hardening modulus is positive and the critical pressure increases, which
corresponds to a strain-hardening phase. Inversely, for yield points such that the material has
a plastically dilating behaviour, the hardening modulus is negative and the critical pressure
decreases, which corresponds to a strain-softening phase. For yield points such that the
plastic strain takes place with no variation in volume or plastic porosity, the hardening
modulus is zero and the material behaves as a plastically perfect material. For these points,
the strains which are now only plastic occur at constant stresses and interstitial pressure. The
state is said to be critical.
The locus of critical states is defined by the following relations:

È p ˘ È p ˘
o exp - e
p � b pp � pcr o exp - e
Í ˙�0 q-  pcr Í ˙�0
ÍÎ l - k ˙˚ ÍÎ l - k ˙˚
 52 Chapter 1 • Elements of Rock Mechanics. Fundamentals

� �
In plane pb ', q , it is the straight line of equation q � -  pb (Figure 1.3.5).

1.3.10.6 Incremental expression of the flow rule

Introducing expression (1.3.73) of the hardening modulus in flow rule (1.3.66) leads to the
following incremental relations:

�1 - Fo � � l - k � ÈÍÊÁ 1
2
˘
1 ˆ q
d evp � � ˜ dpb '� dq ˙
b-F ÍÁ p pb ' ˜¯ 2 pcr pb ' ˙
o ÎË cr ˚

�1 - Fo � � È ˘
2
Í q2 ˙
d edp � l -k Í �q
dpb '�
� �
dq ˙ (1.3.74)
b-F 2
Í  pcr pb '
4
 pcr pb ' pcr � pb '
https://telegram.me/Geologybooks

o ˙
Î ˚
df p � b d evp

Eliminating pcr by using relation

h � q pb ' , we finally obtain:
� �
f pb ', q, pcr � 0 and introducing parameter

�1 - Fo � � l - k � ÈÍ- dpb ' - 2 h dh ˘˙

2

d evp �
b-F Í p ' 2 � h2 ˙
o Î b ˚

�1 - Fo � � l - k � 2 h ÈÍ- dpb ' - 2 h dh ˘˙

2

d edp � (1.3.75)
b-F 2 - h2 Í p ' 2 � h2 ˙
o Î b ˚
df p � b d evp

1.4 DETERMINATION OF IN SITU STRESSES

It is impossible to measure the stress tensor directly. However, some components can be
estimated by observing the strain processes associated with the action of the stresses on the
material considered.
In the absence of tectonic effects, the major principal stress is assumed to be the vertical
stress sv due to the weight of the overburden. The two other principal stresses are horizontal
and equal to Ksv. According to the theory of elasticity, if we assume that there is no hori-
zontal strain during burial (oedometric condition) and that the medium is isotropic and
homogeneous, then we obtain:

νd
σH
' = σv' , where s ' are effective stresses and nd the drained Poisson’s ratio.
1 − νd
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 53

In most cases, the in situ stress state does not correspond to this ideal scheme and the
stress distribution is more complex, depending on geological history. We identify:
– regional stresses which are related to major geological phenomena (plate tectonics);
indications of their orientation and sometimes their amplitudes are now available
[Reinecker et al., 2005],
– local stresses which are affected by local accidents such as folds, faults, salt domes;
stress rotations may then lead to orientations and amplitudes differing considerably
from the original scheme.
In foothill regions, for example, tectonic forces dominate and the maximum horizontal
stress may be the highest of the principal stresses [Last et al., 1995].

1.4.1 Determination from in situ measurements

https://telegram.me/Geologybooks

We assume that the lithostatic vertical stress is one of the principal stresses.

1.4.1.1 Horizontal stress directions

Two methods can be used to determine the in situ horizontal stress directions in vertical
boreholes:
– The first is based on analysis of borehole breakouts. Breakouts form at different
depths when the concentration of compressive stresses exceeds the rock strength. The
axis defined by breakout corresponds to the minimum horizontal stress azimuth.
– The second method uses the traces of drilling-induced hydraulic fractures (tensile
failures). Their azimuth is a reliable indicator of the maximum horizontal stress direc-
tions (Figure 1.4.1).

Maximum horizontal
stress (sHmax)
Drilling induced
fracture

Minimum horizontal Breakout

stress (shmin)
Borehole

Figure 1.4.1
Breakouts and fractures induced in the stress field.
 54 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Breakouts and drilling-induced fractures can be detected by systems recording the hole
geometry, multi-arm callipers and well imaging probes.

1.4.1.1.1 Oriented calliper data (HDT, SHDT, HEXDIP)

HDT (High resolution Dipmeter Tool) and SHDT (Stratigraphic High resolution Dipmeter
Tool) are Schlumberger tools with four hydraulically operated orthogonal pads. Each pad is
equipped with a microresistivity electrode (1 for the HDT, 2 for the SHDT).
The HEXDIP (Hexagonal Dipmeter) manufactured by Baker has six pads.
The 4-pad calliper data are composed of:
– calliper C1-3: diameter between pad 1 and pad 3,
– calliper C2-4: diameter between pad 2 and pad 4,
– P1 azimuth: bearing of pad 1 from the north,
– relative bearing: bearing of the top side of the tool from the north,
– hole azimuth: direction of the borehole in a horizontal plane from the north,
https://telegram.me/Geologybooks

– deviation: hole deviation from the vertical.

As it is brought up to the surface, the calliper logging tool is rotated constantly. In zones with
breakouts, the pads tend to remain blocked and the tool bearing remains frozen (Figure 1.4.2).
Figure 1.4.3 shows an example of a calliper and the common interpretation in terms of
hole geometry which is made in terms of in-gauge hole, with breakout, eroded with formation
of washout, deformed on one side only (keyseat) by the drillstring [Plumb & Hickman, 1985].
The criteria for accurate identification of breakout are:
– the borehole deviation must be less than 10°,
– the log must show tool rotation above and below the breakout,
– the tool rotation in the zone concerned must be less than 5°,
– the difference between the two callipers must be more than 1 cm,
– the zone must be more than 3 m long.

1.4.1.1.2 Imaging data

Acoustic imaging is extremely useful for identification of breakouts that appear as dark
bands on opposite sides of a wellbore wall (Figure 1.4.4):
– UBI (Ultrasonic Borehole Imager) manufactured by Schlumberger,
– CBIL (Circumference Borehole Imaging Log) manufactured by Baker Atlas,
– CAST (Circumferential Acoustic Scanning Tool) manufactured by Halliburton.
Resistive imaging also produces good results for identification of breakouts and drilling-
induced fractures (Figure 1.4.5):
– FMS (Formation MicroScanner), FMI (Formation MicroImager), OBMI (Oil-Base
MicroImager) manufactured by Schlumberger. The FMI has a network of 24 electro-
des on each of its 4 orthogonal pads [Ma et al., 1993].
– STAR (Simultaneous Acoustic and Resistivity) manufactured by Baker Atlas includes
acoustic and resistive imaging sensors in a single instrument. The resistive probe is
equipped with a network of 24 electrodes on each of its six pads.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 55

P1 azimuth

C1-3

C2-4
https://telegram.me/Geologybooks

Figure 1.4.2
Example of breakout detection.

Calliper Calliper
Depth

Depth

In gauge hole Breakout

Calliper Calliper
Depth

Depth

Breakout
Washout
and washout

Calliper
Depth

KeySeat

Figure 1.4.3
Interpretation of CALLIPER data [Plumb & Hickman, 1985].
 56 Chapter 1 • Elements of Rock Mechanics. Fundamentals

N E S W N N E S W N

270 1 2 3 4 5 6 90

180
https://telegram.me/Geologybooks

Figure 1.4.4
Interpretation of UBI (a) and FMI (b) images in terms of breakouts
[Zoback et al., 2003].

Figure 1.4.5
FMI images interpreted in terms of drilling-induced fractures. These fractures
are characterised by low vertical extension and irregular shape and can only
appear on one side of the hole (HEF petrophysical consulting site).
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 57

1.4.1.1.3 Shear wave anisotropy

Anisotropy of shear wave propagation velocities may indicate stress anisotropy, which must
not be confused with structural anisotropy [Plona et al., 2000]. Fast shear azimuth is
assumed to coincide with the direction of maximum horizontal stresses. The latest genera-
tion tools (MSIP, sonic scanner) are even very accurate [Pistre et al., 2005] (Figure 1.4.6).

MSIP fast shear azimuth OBMI induced fracture strike

0 0
100 20%
330 30 330 30
80
15%

60
300 60 300 10%
60
40

5%
20 40° 45°
270 90 270 5% 10% 15% 20% 90
https://telegram.me/Geologybooks

240 120 240 120

210 150 210 150

180 Rosette strike 180 Induced fracture

Cumulated interval 1500 - 1700m Cumulated interval 1500 - 1700m

Figure 1.4.6
Comparison between MSIP fast shear azimuth and OBMI induced fractures
[Franco et al., 2005].

1.4.1.2 Stress amplitude

The amplitude of in situ stresses is generally determined by assuming that the principal
stresses are vertical and horizontal. Starting from this assumption, the analysis is carried out
to obtain the vertical stress on the basis of a density log, assuming that the vertical stress is
equal to the weight of overburden. The horizontal stresses are obtained from well tests
(Mini-Frac tests, LOT, ELOT, SRT).
The stress tensor measurement obtained is local information; the distribution at field
scale can be disturbed by faults, fractures and heterogeneities of mechanical properties.

1.4.1.2.1 Determining the amplitude of vertical stresses

The vertical stress can only be determined by integrating the density log. In this case, several
problems may occur:
– availability of density logs over a reduced thickness,
– error when interpreting the values,
– possible arching effect in an anticline, increasing horizontal stresses and decreasing
vertical stresses.
 58 Chapter 1 • Elements of Rock Mechanics. Fundamentals

In offshore, the density log can be calibrated on measurements taken in geotechnical sur-
veys performed before installation of platforms or subsea pipelines.

1.4.1.2.2 Determining the amplitude of horizontal stresses

Amplitude of the minimum horizontal stress

The minimum stress amplitude can only be measured directly by using hydraulic fracturing
tests. Fractures propagate perpendicular to the minimum stress.
In the hydraulic fracturing test, a section of the borehole, less than one metre long, is
sealed off by two packers. The sealed-off section is then slowly pressurised with a fluid,
water or mud. Figure 1.4.7 illustrates the results obtained.

Breakdown pressure
https://telegram.me/Geologybooks

ISIP
Pressure

Leakoff pressure

Fluid flow rate

Time

Figure 1.4.7
Hydraulic fracturing test.

The pressure is increased by injecting a fluid at constant flow rate. The pressure curve
can be subdivided into five phases: a linear increase phase, a nonlinear phase during which
the rate of increase slows down, the maximum pressure (breakdown pressure), a more or
less steady decrease indicating the opening of a fracture, lastly the relaxation curve after the
injection pump stops. The Leak-Off Pressure (LOP) corresponds to the end of the linear sec-
tion with change of slope. During the relaxation phase, the Instantaneous Shut In Pressure
(ISIP) or closure pressure is assimilated to the minimum stress in situ. The shut-in pressure
corresponds to the level of pressure exerted by the fluid inside the fracture, just sufficient to
keep the edges of the fracture open. Consequently, the minimum stress amplitude can be
determined by analysing the pressure during fracturing and closure of the fracture. Several
methods are available for determining closure pressure from the fall off curve. One of these
uses a change in linearity of the pressure decay on a graph of pressure plotted against the
square root of time [Desroches and Kurkjian, 1998, Zoback, 2008]. If the amplitude of the
closure pressure is less than the overburden at this depth, then the smallest principal stress
will correspond to the horizontal minimum stress (Shmin).
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 59

The Leak-Off Test (LOT) is conducted during the drilling operations after cementing a
casing. It consists in increasing the pressure at the casing shoe to a value greater than the
LOP but less than the rock breakdown pressure (Figure 1.4.8). This operation is mainly used
to evaluate the maximum permissible mud pressure. In a variant of the LOT, the Extended
Leak-Off Test (ELOT), the pressurisation phase continues beyond the formation breakdown
pressure and the fracture propagates. The test consists of two or more pressurisation cycles
(Figure 1.4.9). Extended leak-off tests use procedures from leak-off tests, therefore open
bottomhole or “barefoot” well configuration is pressurised without using packers.
In practice, the quality of 50% of the LOTs is poor. There are several reasons for this:
open section too long, mud compressibility artefact, pumping stopped too soon or relaxation
phase badly recorded [Chardac et al., 2005]. LOT and ISIP generally overestimate the mini-
mum stress.
A good estimate of the minimum stress amplitude can generally be obtained from a
series of fracturing tests. ELOTs are reliable tests compared with LOTs, but they are not as
https://telegram.me/Geologybooks

precise as mini-frac tests with several pressurisation cycles.

After several pressurisation cycles, when injection is stopped, fluid is quickly withdrawn
from the fracture to close it in the vicinity of the wellbore only. The remainder of the frac-
ture stays pressurised above the closure pressure. A pressure rebound can then be observed
when the fracture produces back until pressure decreases again. If the maximum pressure
reached during flow back is greater than the hydrostatic pressure, this is a good indication
that the fracture has been created [Desroches and Kurkjian, 1999] and, for an impermeable
formation, this leads to an estimation of the minimum horizontal stress.
SRTs (Step Rate Tests) are carried out to determine the maximum water injection rate
possible without fracturing a permeable formation. The test is conducted by recording the
injection pressure and injection rate at several increasing injection rates, allowing the

5.0
4.5
ISIP
4.0 LOP
3.5
Pressure (MPa)

Volume (litres)

3.0
2.5
2.0
1.5
900
1.0
0.5
0.0 0
11700 12000 12300 12600 12900 13200 13500 13800 14100

Time (sec)

Figure 1.4.8
Leak-off test (credit ITC).
 60 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Breakdown Residual tension

pressure Strenght stress Fracture
Pressure

concentration Fracture propagation

propagation Final
Initial shut in
shut in

2nd shut in
Leak-off

Fracture initiation stage 1st re-opening stage 2nd re-opening stage Time
Pump rate
https://telegram.me/Geologybooks

Time = 1 hour

Figure 1.4.9
Typical ELOT result [Enever et al., 1996].

pressure to stabilise at each step. All steps must last the same time, for example 30 min. The
fracture pressure can be identified from a graph of injection pressures against flow rates. The
SRTs obviously give an upper limit of the minimum horizontal stress.
In all cases an approach with a “reconciliation” diagram between all the tests performed
is necessary [Desroches and Kurkjian, 1999].

1.4.1.2.3 Maximum horizontal stress

Currently, no method is available to measure the maximum horizontal stress. In some spe-
cific cases an approximation can be found, especially if the vertical stress is the major stress.
We have the relation:
( 3σ H min − σ H max + Tc − 2 Pp )
Pb − Pp =
K
with
Pb: breakdown pressure,
Pp: pore pressure,
Tc: tensile strength,
K: poro-elastic constant which depends on the permeability and compressibility of the
formation,
K = 2 is applicable in permeable formations in which the injected fluid penetrates the
porous space during the test.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 61

K = 1 is applicable in impermeable formations where the pore pressure remains

unchanged in the porous space around the borehole [Lowrey & Steiner, 1993].
When the stress Sh min is known, the fracture reopening pressure can be used to deter-
mine the maximum stress. This pressure is expressed as: Preopening = 3Sh min − SH max − Pp
(K = 1 and Tc = 0).
When the rock strength is known, the breakout angular width [Zobak et al., 2003] mea-
sured on acoustic image log can be used.

(C + 2 Pp + ΔP + σ ΔT ) − Sh min (1 + 2 cos 2θb )

SH max =
1 − 2 cos 2θb
with
2θb = π − wbo with wbo angular breakout opening,
C:rock cohesion,
https://telegram.me/Geologybooks

Pp:pore pressure,
σ ΔT :thermal stress,
ΔP:difference between the well pressure during drilling and the pore pressure.
The measurements of shear wave velocities made with crossed dipole sonic logging tools
such as DSI, MSIP, Sonic scanner, Wavesonic (fast and slow Dt shear) supply anisotropy
indicators. These indicators are expressed as percentages. For example, DT-based anisot-
ropy is:
Dtslow * − Dt fast
* 100
( Dtslow + Dt fast ) / 2

Calibration on mini-frac values or poroelastic models is necessary to convert these indi-

cators in terms of minimum and maximum stress amplitudes [Klimentos et al., 1999, Plona
et al., 2000, Sinha et al., 2000]. Anisotropies as low as 1% can be measured [Franco et al.,
2005].
If LOT data are available in several inclined boreholes (with different inclinations and
azimuths) distributed in a homogeneous field in a specific zone, an inversion method can be
used to estimate the orientation and amplitude of the horizontal stresses [Aadnoy, 1988].

1.4.2 Determination from core measurements

After extraction, a core expands more in the direction of the maximum stresses. This expan-
sion generates microcracks perpendicular to the three principal stresses. Two methods are
based on this relaxation principle:
– ASR (Anelastic Strain Recovery),
– DSCA (Differential Strain Curve Analysis).
With these methods, the core must be oriented, which is often costly, and highly accurate
strain measurements are required. With ASR, the measurements must be taken immediately
 62 Chapter 1 • Elements of Rock Mechanics. Fundamentals

after recovery since anelastic strains develop in less than 50 hours. A detailed interpretation
of the tests is provided in the book by Fjaer et al. (1991). DSCA requires isotropic recon-
finement. Although inaccurate, these two methods provide an idea of the orientation, and
even the amplitude, of the stresses [Cui et al., 2006]. The measurements become very diffi-
cult to interpret in anisotropic and/or fractured rocks.

1.4.3 How to constrain the stress tensor

Geological strata contain faults, fractures and discontinuities at several scales. Friction in
these existing discontinuities limits the stresses.
At equilibrium, Jaeger and Cook demonstrated that the ratio of the maximum and mini-
mum effective stresses had to remain less than a certain value which depends on the coeffi-
cient of friction of the discontinuity m:
https://telegram.me/Geologybooks

2
σ '1 ⎡ 1 ⎤
σ '3 ⎢
(
≤ ⎢ μ2 + 1 ) 2 + μ⎥
⎥
⎣ ⎦
The values of s '1 and s '3 depend on the tectonic system:
– normal fault stress regime (sv > sH max > sh min): s '1 =�s 'v and s '3 =�s 'h min,
– strike-slip fault stress regime (sH max > sv > sh min): s '1 =�s 'H max and s '3 =�s 'h min,
– reverse stress regime (sH max > sh min >�sv): s '1 =�s 'H max, s '3 =�s 'v.
According to Byerlee [1978], the coefficient of friction of the discontinuities m lies
between 0.6 and 1.
Geomechanical modelling tools can be used to constrain the stress tensor, checking the
stability or instability of the boreholes where clearly identified breakouts and induced frac-
tures appear.
These tools require the following input data:
– the assumed directions and amplitudes of the minimum and maximum horizontal stresses,
– the pore pressure,
– the geomechanical properties of the rocks where the breakouts appear,
– a failure criterion (e.g. Mohr Coulomb).
Indications such as mud losses during drilling may also be used.

1.5 GEOMECHANICAL MODELLING ELEMENTS

The discipline of oil reservoir geomechanical modelling is in full expansion. The first
studies were conducted to explain the significant subsidence observed in some mature fields
with highly porous (North Sea chalk) or poorly consolidated (Zuata field, Venezuela)
reservoirs [Charlez, 1997]. The scope of geomechanical modelling is much broader and
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 63

concerns all stress-dependent problems: modification of the mud window, instability of the
borehole walls, modification of flow directions, reactivation of fractures or faults,
interpretation of modifications in 4D seismic data, microseismicity. The stress modifications
are induced by reservoir depletion or, on the contrary, injection of fluids during hydrocarbon
production or gas storage (natural or CO2).

1.5.1 Geology and griding

The zone considered for geomechanical modelling must include a volume larger than the
reservoir since the mechanical properties of surrounding formations have an impact on the
mechanical behaviour of the reservoir. The width of the sideburden will be ideally 3 times
the horizontal extension of the reservoir region. The height of the underburden will be of the
same order of value as the depth of the reservoir region.
Geomechanical modelling may be more or less complex, 2D or 3D depending on the
https://telegram.me/Geologybooks

problem to be solved. The geometry of the various layers is derived from seismic interpreta-
tion, constrained by the available borehole data. The number of stratigraphic units to be
taken into account depends on the borehole data available, especially the seismic impedance
contrast (product of the formation density by the compression wave velocity). The dynamic
elastic properties from the sonic logs and the static elastic properties used in the geomechan-
ical models are in fact closely related.
The mesh will be oriented according to the orientation of the principal stresses.
The geological strata exhibit discontinuities which may be more or less detectable and
more or less cemented or open. They share the following characteristics: low shear strength,
negligible tensile strength and hydraulic conductivity often much greater than the wall rock.
Several categories can be identified:
– diaclases: rock fractures, due to compressive, tensile or shear failures related to the
tectonic stresses; the two parts of the rock have not moved, however;
– faults: fractures identical to the diaclases but which caused a relative movement of
two parts of the wall rock. A slip has therefore occurred along this fault and the throw
varies from a few tens of centimetres to a few hundred metres;
– sedimentary joints: these are the joints separating two strata deposited at different
times under different conditions.
Discontinuities are generally areas of high deformability, where failure is easier and
where the fluids can flow.
All realistic models of a structure and its environment must take into account the possi-
ble discontinuities. Geometric discontinuities are generally correctly taken into account.

1.5.2 Geomechanical properties

Geomechanical properties must be attributed at least to each rock layer identified in the
model. Seismic, log analysis and laboratory data can be used.
 64 Chapter 1 • Elements of Rock Mechanics. Fundamentals

Resistivity, nuclear and gamma ray logs give a good estimation of the mineralogy and, in
particular, the location of the shale layers [Chardac et al., 2005].
The compression and shear wave velocities (Vp and Vs if available) and the density logs
provide information on the dynamic elastic properties, but the plasticity or viscosity proper-
ties can only be obtained by conducting laboratory tests.
A few calibration points can be obtained from the laboratory data, for the conversion of
sonic logs into static elastic properties.
If no S-wave records are available, the Vs velocity is estimated from Vp using empirical
formulae that depend on the lithology [Castagna et al., 1985].
The Vp and Vs velocities are used to calculate the undrained dynamic moduli:

⎛ 4 ⎞
Kudyn = ρ ⎜Vp2 − Vs2 ⎟ and Gdyn = ρVs2, with
⎝ 3 ⎠

Kudyn: rock undrained dynamic bulk modulus

https://telegram.me/Geologybooks

Gdyn: rock dynamic shear modulus

r: overall rock density (deduced for example from the density log)

Eudyn = ρVS2
( 3VP2 − 4VS2 ) and νdyn = (VP2 − 2VS2 )
(VP2 − VS2 ) u
2 (VP2 − VS2 )

The static properties can be deduced from the dynamic properties [Vidal-Gilbert et al.,
2005]. The procedure consists of two phases. Firstly, “fluid saturation” correction is carried
out to change from undrained dynamic elastic moduli to dynamic drained moduli. Secondly,
dynamic drained moduli are changed to static drained moduli.
One of the problems frequently studied in rock physics, especially for analysis of logs,
samples and seismic data, is the prediction of seismic velocities in rocks saturated with a
given fluid from another saturation state. This fluid substitution problem can be predicted
using the Biot-Gassmann theory (1951). Gassmann calculated, for a quasi static load, the
difference in elastic modulus between a drained porous medium and the same medium in
undrained state, in which local variations in fluid content are nil during mechanical loading.
The Gassmann equation can be expressed as:

Ku Kd K fl
= +
Ks − Ku Ks − Kd (
φ Ks − K fl )
with
Ku: undrained bulk modulus
Kd: drained bulk modulus
Ks: bulk modulus of the matrix (mineral bulk material forming the rock)
Kfl: fluid bulk modulus
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 65

The following assumptions are used in the Biot-Gassmann equation:

1) the porous medium is macroscopically homogeneous and isotropic,
2) all pores are connected,
3) the pores are filled with a low-viscosity fluid,
4) the solid-fluid system is closed (undrained),
5) as the wave passes, there is no coupling (relative movement) between the solid phase
and the liquid phase,
6) the fluid does not interact with the solid, so that the shear modulus remains constant.
A practical expression of the Biot-Gassmann equation involves the Biot coefficient b and
the Biot modulus M:
Ku = Kd + b2 M

⎛K ⎞
with: b = 1− ⎜ d ⎟
https://telegram.me/Geologybooks

⎝ Ks ⎠
1 b−φ φ
and = +
M Ks K fl

In this expression, we can see that the first term in the expression of 1/M is often much
smaller that the second since Ks is always much bigger than Kfl. Neglecting this first term,
we obtain a very simple expression:
b2
Ku ≈ Kd + K , [Rasolofosaon and Zinszner, 2003].
φ fl
The error thus committed has been estimated to be less than 5%.
In order to apply the Gassmann equation to wave propagation (dynamic case), the fre-
quency must be low enough for the variation in saturating fluid induced by passage of the
seismic wave to be uniform throughout the porous space (no gradient). This low frequency
limitation explains why the Biot-Gassmann equation is well suited to seismic frequencies
(< 100 Hz). While this relation can also be applied for the log analysis frequency band, it
may prove less relevant for ultrasound measurements on samples. The latter point is open to
discussion since Rasolofosaon and Zinszner (2004) demonstrated that the experimental
results on samples seemed to be in agreement with the Biot-Gassmann theory, for various
fluids saturating the porous medium.
 66 Chapter 1 • Elements of Rock Mechanics. Fundamentals

The Biot-Gassman equation is then frequently used to estimate the dynamic drained
moduli from undrained dynamic moduli:
2
⎛ K dyn ⎞
K fl ⎜ 1 − u ⎟
⎜⎝ Ks ⎟⎠ Eudyn
Kddyn = Kudyn − with Kudyn = where:
⎛
⎜1 −
K fl ⎞
⎟φ−
K fl ⎛
⎜1 −
Kudyn ⎞
⎟
(
3 1 − 2νu )
⎜⎝ Ks ⎟⎠ Ks ⎜⎝ Ks ⎟⎠

Ks: bulk modulus of the solid matrix (of the material forming the grains),
K fl : bulk modulus of the fluid saturating the pores,
f: porosity (deduced from the density measurements, for example).
The bulk modulus of the rock matrix Ks depends on the mineral composition, but also to
https://telegram.me/Geologybooks

the second order on pressure and temperature. To a first approximation, the orders of magni-
tude indicated in Table 1.5.2 (dynamic values) will be used.

Table 1.5.2 Mean values of dynamic bulk and shear moduli in the main minerals forming sedimentary rocks
[Zinszner and Pellerin, 2007].

Mineral class r(kg/m3) Ks(GPa) Gs(GPa)

Clay estimation 2,750 25 9
Quartz 2,650 37 45
Calcite 2,710 70 30
Dolomite 2,870 80 50
Siderite 3,960 120 50

If the mineral composition is known, homogenisation methods can be applied to estimate

the mean effective modulus of the matrix, or at least determine its upper and lower limits
[Berryman et al., 1995, Hu et al., 2006]. This is rarely the case outside the reservoir.
The fluid bulk modulus Kfl depends on the nature of the saturating fluid. For a water-oil
mixture for example, it can be calculated using the following formula:

1
K fl =
Sw (1 − Sw )
+
Kw Koil

with:
Kw : water bulk modulus,
Koil : oil bulk modulus,
Sw : water saturation, determined using log analyses, for example.
 Chapter 1 • Elements of Rock Mechanics. Fundamentals 67

The water bulk modulus varies with salinity, pressure and temperature (Figure 1.5.1).
For brine at a concentration of 15 g/L, Kfl = 2.2667 + 0.0063 p in GPa with p = pressure
expressed in MPa [Zinszner and Rasolofosaon, 2008].

100 MPa PPM = 0

5.0
50 PPM = 150000
4.5 PPM = 300000
0.1
4.0
Bulk modulus (GPa)

3.5
3.0
2.5
2.0
1.5
https://telegram.me/Geologybooks

1.0
0.5
0 50 100 150 200 250 300 350
Temperature (°C)

Figure 1.5.1
Bulk modulus of a brine against pressure, temperature and salinity
[Batzle and Wang, 1992].

The bulk modulus of oil depends on its density, as well as the pressure and temperature
(Figure 1.5.2).

3000 ro = 1.00 (10 deg. API)

ro = 0.88 (30 deg. API)
ro = 0.78 (50 deg. API)
2500 0.
1
M
Oil bulk modulus (MPa)

Pa 25
M
Pa 50
2000 M
Pa

1500

1000

500

0
0 50 100 150 200 250 300 350
Temperature (°C)

Figure 1.5.2
Bulk modulus of oil against pressure, temperature and composition
[Batzle and Wang, 1992].
 68 Chapter 1 • Elements of Rock Mechanics. Fundamentals

The estimated dynamic drained moduli can then be changed to static drained moduli
using empirical correlations [Wang, 2000].
For reservoir rocks, we can use the correlations determined by Wang, who identifies two types:
– if Edstat < 15 GPa then Edstat = 0.4145 Eddyn − 1.0595 (GPa ) and
– if Edstat > 15 GPa, then Edstat = 1.1530 Eddyn − 15.1970 (GPa ) .

For shales, we can use correlations determined for all lithologies [Lacy, 1997, Cooper
and Hatherly, 2003] or for shales only [Horsrud, 2001], according to the following
sequences:
– Vp Vs: P and S wave velocities,
– Ed: dynamic modulus,
– Es: static modulus,
– UCS: unconfined compressive strength.
For example:
https://telegram.me/Geologybooks

– Ed = 0.265(Vp )2.04 [Lacy, 1997] with Vp in km/s and Ed in millions of psi (all lithologies),
– UCS = A.Vp3 [Cooper & Hatherly, 2003],
– ES = 0.0428( Ed )2 + 0.2334 Ed (for clays),

– UCS = 0.2787( ES )2 + 2.458 ES (all lithologies),

Es = 0.076(Vp )3,23 G = 0.03(Vp )3.3 and UCS = 0.77(Vp )2.93 (for shales) [Horsrud, 2001].

The static Young’s moduli are lower than the dynamic moduli [Wang, 2000; Yale et al.,
1995]. The static Poisson’s ratios are greater than or equal to the dynamic Poisson’s ratios.
For both static and dynamic properties, laboratory measurements remain preferable
when determining local or regional relations of the zone of interest.
Obviously no matter which approach is chosen, the measurements cannot be considered
as completely reliable. Current geomechanical models are deterministic. The values
assigned to the geomechanical properties must take into account the risk factor associated
with the problem studied.

1.5.3 Boundary conditions

The traditional displacement boundary conditions consist in blocking the horizontal dis-
placements of the vertical edges and the vertical displacements of the horizontal edges,
unless the upper edge coincides with that of the natural terrain.
The vertical displacements are fixed at the bottom and the horizontal displacements on
the lateral sides.
The contact surfaces between layers and the faults are generally fixed in a first simula-
tion. Friction can then be introduced, or not, depending on the resulting stress state.
 2

Geomechanics, Drilling
and Production
https://telegram.me/Geologybooks

2.1 DRILLING PERFORMANCE

Drilling bits can be divided into two broad categories: roller-cone bits and PDC (Polycrys-
talline Diamond Compact) bits (Figure 2.1.1).
Roller-cone bits have several toothed cones which are free to turn as the bit is rotating.
Tri-cone is the dominant type. The teeth can either be ground in the matrix or consist of
tungsten carbide buttons inserted into holes in the cones. The teeth are designed to crush the
rock as the bit rotates (Figure 2.1.2).
PDC bits consist of several discs containing numerous cutting edges made of tungsten
and diamond. They cut the rock by shearing, layer after layer. These bits, more modern than
roller-cone bits, can reach high drilling rates in relatively soft formations. They are also
more efficient in hard rocks, provided that certain recommendations are respected (speed
and weight on bit), especially for new bits which are more fragile.
Drillers have two main concerns:
– reaching maximum efficiency under ordinary conditions,
– solving “problems” under exceptional conditions which could eventually damage the
bit, in particular.
The “problems” most frequently encountered are excessive vibrations, whose effects at
the surface arouse the driller’s attention, although lateral vibrations and precession some-
times remain unnoticed.
These excessive vibrations are generally observed when drilling through hard rocks at
low rate of penetration (regular bounce or precession), but the TRAFOR recordings taken by
IFP Energies nouvelles have also demonstrated bounce activity (more chaotic than the pre-
vious activity), while drilling rapidly through soft rocks; expulsion of the cuttings in “plugs”
seems to be the most likely cause for the excitation [Putot, 1995].
Until recent years in the drilling industry, “rock drillability” and “drillstring dynamics”
were quite independent disciplines [Pessier and Fear, 1992].
 70 Chapter 2 • Geomechanics, Drilling and Production

a b
Pilot pin
Locking ring
Journal

Special Shirtail
metal inlays Gage
Bit leg
Nozzle
O-ring seal boss

Stretch type Shank

pressure
compensator Breaker
Nozzle slot
Compensator retaining ring
protector
Nozzle
«O» ring
Lubricant
reservoir cap Jet nozzle

Marking on top of shank

Bit type/ Bit size
https://telegram.me/Geologybooks

Trademark

Figure 2.1.1
Examples of bits: roller-cone (a) and PDC (b) [Nguyen, J-P., 1993].

Steel
Steel
F

(a) (b)

Figure 2.1.2
Rock destruction modes: punching for the roller-cone bit, shearing for the new
PDC bit [Hélène Geoffroy, 1996].

“Rock drillability” concerns the ability of the drill bit to penetrate and remove the rock,
in the context of the bottomhole, controlled from the surface via the drillstring. This special-
ity generally deals with ordinary drilling situations and aims at optimising bit concepts with
respect to cut and hydraulics.
“Drillstring dynamics” studies examine the types of vibration which would damage the
drilling system equipment (drillstring and bit), using in particular surface indicators more or
less able to detect the various types of vibration at the bottom. This speciality also analyses
the conditions leading to a vibratory malfunction.
 Chapter 2 • Geomechanics, Drilling and Production 71

With traditional rock drillability, it is impossible to correctly study the reduced drilling
rate regimes with high dissipation at the cutting face and therefore validly assess the drilling
efficiency [Warren and Smith, 1985]; the advance speed is reduced considerably by the
vibrations; in addition, the dynamics of the bottomhole drilling system, in particular, is
strongly related to the aggressivity of the cutters and the conditions under which the cuttings
are expelled. The distinction between the two disciplines therefore becomes quite arbitrary.
Drillability in dynamic regime is the end result of combining the two disciplines.
As regards drillability, one direction consists in analysing drillings in relatively soft forma-
tions. Tool performance, mainly that of monobloc “PDC” tools, is generally not a major problem,
provided that the cuttings are suitably evacuated, which nevertheless demands well-designed tool
hydraulics, good control and, in particular, sufficient hydraulic power [Wardlaw, 1971].
It seemed useful to describe the phenomenon associated with a tool cleaning fault,
through balance equations both as regards the divided solid (rock) and the drilling mud,
relying on simplified behaviour laws [Putot, 1995]. The evolution of the mixture of fluid and
rock particles is governed by a “source” term including the conditions under which the rock
https://telegram.me/Geologybooks

is broken by the cutters, together with terms of retention on the particles crushed but not
necessarily expelled, function of the entrainment capacity of the drilling mud.
Note that the hydraulics is a determining factor in soft rocks for evacuation of the cuttings
but also essential in “hard” rocks for good cooling of the cutters, guarantee of their integrity.

2.1.1 Empirical drilling efficiency: weight on bit, rate of penetration

and torque

2.1.1.1 Axial kinematics of the bit: normalised rate of penetration

The first means of analysing the drilling performance are modelled on the concerns of the
driller who, as a priority, pays special attention to the axial kinematics of the bit.
Main notations (with variants):
W Weight On Bit (WOB)
T Torque On Bit (TOB)
R Rate Of Penetration (ROP)
N bit speed of rotation (RPM)
d = R/N penetration per revolution
Penetration per revolution divided by weight on bit (d/W) is an intuitive concept and a
key indicator: a high value is a sign of good performance, while a low value often indicates
an accumulation of cuttings and clogging.
Historically, the first refinement made to the driller’s good judgement – do not drill too
quickly or the hole will clog up – consisted in recording the d-exponent indicator:
log R N
d= (2.1.1)
log W D
where D is the bit diameter.
 72 Chapter 2 • Geomechanics, Drilling and Production

Under normal conditions, the d-exponent indicator can be used to “subtract” the effects
of modifications of the “loading” W and N. Its “logarithmic” definition offers the advantage
of slightly reducing the experimental dispersion, since the signal is generally very noisy due
to fluctuations related to the mineralogical characteristics of the rocks at grain scale and to
the pore pressure.
Plotted against the depth for a given lithology, the d-exponent provides an image of the
compaction. In a normal situation, the greater the depth, the greater the compaction. It can
be used to detect any compaction anomaly, for example areas of pore overpressure
[Mouchet and Mitchell, 1989].
Real-time drilling interpretation is commonly based on observing the normalised rate of
penetration RD alone, which provides information on the lithology changes and would also
give an indication on bit operation. The following expression was proposed by Warren (1984):
n
⎛ W ⎞
RD = k ⎜ (2.1.2)
⎝ σ D2 ⎟⎠
https://telegram.me/Geologybooks

where:
s is the unconfined compressive strength of the rock, frequently compared with a hardness.
k is an empirical proportionality factor related to the bit efficiency (ability to break the rock).
n is an exponent characterising, in principle, the local failure mechanism but which, in
practice, is more representative of a degree of uncertainty related to the position on the curve
d = f(W) (Figure 2.1.3).
n = 2 is the value generally used to characterise field data; it corresponds to a normal
drilling regime, for which increasing the weight on bit W significantly increases the perfor-
mance, i.e. the penetration per revolution, d = R/N. We simply take the logarithm of relation
(2.1.2) to obtain the d-exponent (2.1.1).
A decrease in n indicates a priori a deterioration of drilling conditions, generally due to
difficult expulsion of rock cuttings and sometimes more or less sudden damage to the drill
bit. The weight then has less effect on the normalised penetration d. These approximations
reflect the existence of an “S-shaped curve” whose upper plateau corresponds to saturation
of d with W (Figure 2.1.3).
For both roller-cone bit and PDC, Falconer et al. (1988) demonstrated that using the
assumptions adopted in his model, RD (ability of the bit to penetrate the rock) is inversely
proportional to the shear strength t of the rock; paradoxically, for the roller-cone bit,
although the kinematics consists mainly of punching, the efficiency of the shear mechanism
is the determining factor when assessing penetration. The paradox can be explained by the
fact that the implicit assumption t/s = constant is often made.
Falconer et al. (1988) proposed a quantity FORS proportional to the drilling resistance
defined later by Detournay and Defourny (1992) for the PDCs:

W σ
FORS = = (2.1.3)
RD D2 k
 Chapter 2 • Geomechanics, Drilling and Production 73

d = R/N

n<1

n=1

n=2
W
https://telegram.me/Geologybooks

Figure 2.1.3
Sigmoid shape of the graph of penetration per revolution d against weight W:
the exponent n reflects the local behaviour.

FORS is the ratio of rock penetration resistance s divided by bit efficiency factor k.
Cunningham (1978) proposed a numerical frame very similar to expression (2.1.2),
where the exponents assigned to the hardness s and the weight W in the expression of their
dependency with respect to the normalised penetration differ – which is not the case in
(2.1.2) – and are fully defined by the rock hardness; he improves his description by explic-
itly introducing the drop in performance resulting from the pressure difference between mud
pressure and pore pressure, using the following empirical relation where he assumes that the
deviations from linearity R(N) are due to effects of pressure difference Δp:

NW a
R= (2.1.4)
k1σ 1.5 a
d + NW Δp ( )0.75
Numerous laboratory experiments have corroborated these expressions.
The previous literal expression, essentially empirical, showing the effect of interstitial pres-
sure, explains the role of this pressure in the shape of the curves R(N) at fixed W; it provides
information on the decrease in penetration d = R/N with speed N, useful to characterise the
dependency of torque on bit with angular speed, especially to assess the stick-slip mechanism.
As an illustration, Cunningham observes:
– linearity R(N) for Δp = 0,
– R(N) in N0.65 for Δp = 20 MPa (soft rock),
– R(N) in N0.85 for Δp = 20 MPa (hard rock).
For more recent studies concerning the effect of pore pressure on drilling performance,
interested readers can refer in particular to Peltier and Atkinson (1986), Lesso and Burgess
(1986), Detournay and Atkinson (1991), Hanson (1991).
 74 Chapter 2 • Geomechanics, Drilling and Production

2.1.1.2 Circumferential response and representation in the plane RD – TD

The second generation of performance indicators is no longer based on a single scalar quan-
tity but on the definition of an operation line in the plane RD – TD, the energy component of
the torque TD forming a second indicator just as valuable as the axial component. This cir-
cumferential component indicates the shear strength, generally moderate for roller-cone bits
and much higher for PDC type monobloc bits. It is the energy response of the torque, since
the weight does not act directly in the drilling process (Figure 2.1.4).
The adimensional variable TD introduced by par Falconer et al. (1988) can be written
simply:
1 T
TD = (2.1.5)
DW
It suggests a friction law (Amonton’s 2nd law postulates that friction is proportional to
the normal load). Using the Pessier and Fear (1992) approximations for a simplified bit
https://telegram.me/Geologybooks

model, we then obtain:

μ
TD = (2.1.6)
3
where m is the “coefficient of friction”.
TD is in fact the transposition, at bit scale, of the ratio t/s shear strength divided by
punching strength. This quantity indicates the drilling efficiency.
A reasonable value of m for a rock-mechanical system interaction is about 0.6 [Scholz,
1990], so we can expect TD to have values of about 0.2.
In practice, Falconer records values in the region of 0.1 for a roller-cone bit operating
normally, in marls, and between 0.20 and 0.35 for a new PDC bit; equivalent friction, so the
performances of the two types of bit are therefore a priori quite different, at least when new.
These differences can be explained by the fact that the rock destruction modes are radically
different: punching for the roller-cone bit (little friction), shear for the PDC (high interac-
tion).
Values of TD significantly higher than the previous reference values generally indicate
porous formations (e.g. some sandstones) if, at the same time, RD is relatively high.
Inversely, significantly smaller values of TD indicate stiff rock layers: indurated sandstones,
compact limestones, dolomites, etc. provided that they are correlated with relatively low val-
ues of RD.
The mechanical characterisations of the behaviour of roller-cone bits have been formu-
lated by Falconer et al. (1988) and have been represented analytically in the plane of vari-
ables RD and TD.
The authors’ analysis is based mainly on the mechanics of a circular cone interacting, on
a contact surface close to a generator, with a flat block. The model had been presented previ-
ously by Warren (1984, 1985), with a proposal for extension to PDC bits made by Falconer
without making any special adaptations to the theory.
 Chapter 2 • Geomechanics, Drilling and Production 75

0.5

0.4

0.3
TD

0.2

0.1

0
0 0.005 0.010 0.015 0.020 0.025 0.030 0.035
RD
https://telegram.me/Geologybooks

Figure 2.1.4
Example of TD – RD values in a shale layer at a depth of approximately 1 800 m.

Detournay and Defourny (1992) completely revised the formulation for PDC bits by
introducing assumptions on the local mechanics of the cutters and by including the forces on
the entire tool. To characterise the results, variables S and E quite similar to the previous one
are used, called respectively drilling resistance and specific energy. Detournay and
Defourny also characterise the operation of roller-cone bits in the same system of variables,
if only to clearly identify their specificity.
Note that, for PDC bits, the analysis strategy used by Kuru and Wojtanowicz (1988) is
very similar although the experimental verification does not seem to have been as thorough
as that conducted by Detournay with Glowka’s data (1989).
Falconer and Detournay use a similar approach to characterise normal system operation,
by aligning points representative of bit performance in the plane of variables. This straight
line is the analogue of a rock failure criterion under global load. In the remainder of the
document, we will refer only to Falconer’s “roller-cone bit” terminology. Warren’s
mechanical model, expressed in response variables RD and TD is then written:

T D = c1 + c 2 R D (2.1.7)

We see immediately the similarity of (2.1.7) with Coulomb’s law, provided that in relation
(2.1.2) n is given a value of 2, which corresponds to the “normal” regime. However, it is proba-
bly more correct to position (2.1.7) with respect to a simple Amonton type interface friction law.
These normal conditions are represented on Figure 2.1.5 by a straight line.
We assume that neither the rock parameters determining the characteristics of the
straight line (ratio t/s of the shear strength to the compressive strength) nor the bit condition
 76 Chapter 2 • Geomechanics, Drilling and Production

parameters are affected during the observation interval; only the two response components
RD and TD (or S and E) vary according to the disturbances on the input: weight on bit W and
speed of rotation N, which are nevertheless already mean values. Variations in rock “hard-
ness” s may also affect the response.

TD

0.1
https://telegram.me/Geologybooks

0
0 0.1 RD

Figure 2.1.5
Normal conditions in the TD – RD plane.

The slope representative of normal operating conditions lies between 0.07 and 0.15 for “ref-
erence shales”. These rocks are considered as favourable to analyse the wear factor, for exam-
ple, since the ratio t/s changes very little within this class, unlike the other rock categories.
The deviations on TD and on the slope values are attributable to a “non-shale” facies or to
transfers within the weight/torque torsor which are generally symptomatic of dysfunctions.
Refer to the publication by Falconer and Defourny for the quantitative effects of the wear
on the operating points: the drilling resistance is increased, at constant weight on bit W,
according to relation (2.1.3), which decreases RD. The complementary effect is a substantial
drop in TD indicating a decrease in shearability (Figure 2.1.6).
The criterion indicates the increasing energy contribution with axial thrust, which can be
interpreted as increasing “friction” at the borehole wall with confinement.
Another type of improvement was obtained by taking into account bit wear, since this
has an important effect on the torque. This correction was therefore introduced in the estima-
tions as an indicator TD for the torque, sensitive to cutter wear, like the more traditional indi-
cator RD which corresponds to the axial dimension. Combined use of these two indicators
yields a more reliable diagnosis than with a single indicator, especially when studying tran-
sitions to degraded or dangerous regimes.
 Chapter 2 • Geomechanics, Drilling and Production 77

New bit
TD (sharp punch)

Worn bit
(blunt punch)
https://telegram.me/Geologybooks

RD1/2

Figure 2.1.6
Effects of wear on the relation of bit behaviour in the plane of reduced
variables, especially on the slope and the Y-intercept [Falconer and Defourny,
1992].

2.1.1.3 Cutter wear, a major concern

Studies have only recently turned to an additional parameter, that of cutter wear. It repre-
sents a determining factor in drilling performance, governing dynamics and rate of penetra-
tion. Presence of a wear flat (Figure 2.1.7) cannot be avoided, and the general trend in the
past which consisted in exclusively studying new tools is insufficient. Moreover, optimised
drilling must take into account bit wear which can be identified. Monobloc bits (PDC type)
are much more sensitive to wear. It is quite well-known that new PDC bits are highly sensi-
tive to weight and must treated carefully, especially during the “running-in” phase, to avoid
premature damage [Kuru and Wojtanowicz, 1988] [Fear et al., 1995]. This is not the case
with roller-cone bits, which are much sturdier in this respect.
Developing PDC bits capable of drilling a highly abrasive rock efficiently, while offer-
ing sufficient lifetime, has therefore been a true challenge for bit manufacturers specialised
in diamond impregnated cutters, even though this technology is thoroughly controlled. New
bits are considered as highly efficient for such conditions but they are so aggressive that they
suffer rapid damage if incorrectly controlled when first put into service (gentle control is
required during the “running in” phase). The density of the cutters and their arrangement can
be designed so that the bit is less vulnerable initially, but this strategy fails as wear
progresses and the tool generally becomes ineffective.
For further reading on this subject, refer to [Geoffroy et al., 1997, 1998a, 1998b].
Studies in the field of elastoplastic rock behaviour laws have focused more specifically on
the distribution of forces exerted on the cutter, in particular on the wear flat.
 78 Chapter 2 • Geomechanics, Drilling and Production

Figure 2.1.7
Example of PDC type bit, showing a wear flat (DBS document).
https://telegram.me/Geologybooks

This action is in line with the studies also conducted by SANDIA [Glowka, 1987, 1989]
and the University of Minnesota [Detournay, 1991, 1996]. The distribution of elementary
forces on the diamond impregnated cutters and therefore the ability of the bit to cut down the
rock have been predicted more accurately. It has led to a strategy for positioning the cutters
which is better than that associated with the new bit.

2.1.2 Drillability and detailed analysis of failure mechanisms

at the working face

2.1.2.1 Rock cutting mechanisms

Numerous experimental and theoretical studies have been conducted on the fundamental
problems associated with the drilling process. They are focused on the mechanical action of a
bit: in the sense used by Miners, not Drillers! Miners are concerned directly by the cutter
which, in fact, they call “a pick” and not by the rock drilling machine, whereas Drillers see the
tool more globally and are less concerned by the cutting “studs”. This tool punches the surface
of a rocky bed, like the tool used in hardness tests. However, the mechanism involved in the
hardness test is more one of breaking the rock into numerous fragments of various sizes, rather
than a cutting action as such (Figure 2.1.8). The destruction process is therefore more like a rock
crushing process. In processes such as these involving a notion of volume of crushed material,
the energy – volume relations, resulting in the definition of specific energy, are extremely useful
and considered as the significant index of the mechanical efficiency of the drilling process.
In the petroleum industry, “ductile” failure processes are favoured by the high confine-
ment of drilling situations.
Despite the widespread and well-accepted nature of the standard hardness test, the basic
understanding of the hardness test in terms of the mechanisms actually involved has only
improved slowly. B. Lawn and R. Wilshaw (1975) studied the appearance of the cracks
caused by punching with a hard sphere.
 Chapter 2 • Geomechanics, Drilling and Production 79

Crushed rock

Flat
Grain and
Damaged area powder

Crushed rock

Shear band
Intact rock

Crack Plastic deformed Damaged area

rock
Compact material
https://telegram.me/Geologybooks

Damaged area

Figure 2.1.8
On the left: Cutting mechanisms observed in sandstone with new PDC blades.
On the right: Representation of the damaged region below the flat [Geoffroy,
1996].

Since spherical symmetry (Hertz’s theory) is only an approximation which he considers

as sufficient, Detournay assumes that there is a core under hydrostatic pressure immediately
underneath the punch, extending with the applied load. This core is surrounded by a region
in plasticity state obeying a well-defined criterion. Beyond this region we find the elastic
matrix. We can demonstrate that this set of states generates a mechanism through which the
compressive stress field under the punch is then converted away from the punch into a stress
field with a tensile component.
Via a certain degree of plasticity microcracking equivalence, this model introduces the
dissipation related to the rock crushing process in the hydrostatic compressive region.
Drillers are obviously interested in propagating a crack by simple extension, which
generates a well-defined chip. The two dissipation modes present in punching: diffuse
mode, associated with plastic deformation of the rock, and localised mode related to
propagation of a tensile crack, are both clearly present in Detournay’s model (1996).
Various authors have studied the rock cutting mechanism in a rather brittle perspective; for
example, Nguyen Minh (1974) for oil applications and Déliac (1986) for mining applications.

2.1.2.2 Shear cutting regimes

Detournay drew up a list of various geometrical cutting situations with associated elemen-
tary models [Detournay, 1996].
Detournay considers three mechanisms depending on the penetration per revolution d
(regimes I, II and III below) in increasing order of depth; r designates a characteristic size
related to the grain dimension (Figures 2.1.9 and 2.1.10).
 80 Chapter 2 • Geomechanics, Drilling and Production

I Grain yielding= II Grain boundary

abrasive regime sliding and failure=
inefficient mechanism efficient mechanism
d<r r<d<l

d r
d

III Brittle fracture

d>l
efficient mechanism
but poor cleaning conditions

d KIc
2

l
https://telegram.me/Geologybooks

Figure 2.1.9
Cutting mechanisms depending on grain size r and penetration per revolution d
[Detournay, 1996].

Bit signature:
wear and blunting decreasing Rock signature:
performance = abrasive regime specific energy with threshold l

d
III Insufficient hydraulics
II Shear cutting poor cleaning conditions
efficient regime inefficient extrusion of
crushed material

New bit
III
sharp cutters
II Blunt PDC
cutters

Improving control by
l
choosing specific design

W
I Abrasive action,
inefficient rubbing

Figure 2.1.10
Bit penetration d against weight W exerted [Detournay, 1996].
 Chapter 2 • Geomechanics, Drilling and Production 81

Regime I: d  r
The first “cutting” regime is that of surface spalling and inefficient abrasion, exhibiting high
specific energies. Grain plastification occurs. In the penetration per revolution – weight on
bit axis system, this regime marks out a weight efficiency threshold which increases with the
size of the wear flat [see also Kuru and Wojtanowicz, 1988]. The order of magnitude of the
penetration per revolution is less than that of the grain size.

Regime II: r  d� l

The second regime corresponds to the cutting mode as such, ductile. This is the preferred
mode in practice, which acts efficiently on cohesion, cementation between grains by separa-
tion: “grain joint plastification” occurs, consuming less energy than grain plastification.
For the two above mechanisms (regimes I and II), plasticity is not confined.

Regime III: d  l
The third regime is considered as brittle. It appears from a depth l proportional to Irwin’s
https://telegram.me/Geologybooks

length [Detournay, 1996]:

2
⎛K ⎞
λ ∝ ⎜ Ic ⎟
⎝ q ⎠
where KIc is the rock toughness (ability of the material to withstand sudden propagation of
a crack) and q its unconfined compressive strength. l is high for ductile rocks. The propor-
tionality factor, mainly related to the bit geometry, not indicated in the previous relation,
also governs the ductility. For a sandstone of unconfined compressive strength about
50 MPa, the typical depth l is about 1 to 2 millimetres.
Regime III can still be considered as efficient, since the hydraulics are sufficient.
The behaviour of the punching model [Johnson, 1985], Figure 2.1.11, follows the well-
known progression of “strain stages” when the force applied increases: elasticity, contained
plasticity, unconfined flow.
Damjanac and Detournay recently (1995) resumed studies on stage II with contained
plasticity (Figure 2.1.12).
This analysis can also be compared with that conducted by Kendall (1978) who,
considering “the complexity of compression failure modes” according to the area over
which the compressing force is applied (w plate width) compared with the cross-sectional
area of the test piece (d), also identifies three failure mechanisms. Spalling, which prevails at
low values of w/d, axial splitting with elastic failure criterion involving KIc for intermediate
values, plastic buckling or barrel failure at high values.

2.1.2.3 Consequences and applications

Using the almost straight line relation between force and penetration per revolution in duc-
tile regime II (Figure 2.1.10), the mechanical properties of the rock can be identified from a
simple core. This property led to the design of the RSD (Rock Strength Device) whose oper-
ation is similar to that of a cutting tool.
 82 Chapter 2 • Geomechanics, Drilling and Production

P P
R

a a b a a

3 E
Sphere D
(FE)

C F Cone Cavity
Sphere B G Sphere model
Mean pressure Pm/Y

Cone A
https://telegram.me/Geologybooks

Elastic Elastic-plastic Fully-plastic

0
1.0 10 100
E* tan E*a
Non-dimensional strain or
Y YR

Figure 2.1.11
Johnson model (1985).

2a

Elastic Plastic

Crack l

Figure 2.1.12
Normal indentation of a wedge into rock
[Damjanac and Detournay model, 1995].
 Chapter 2 • Geomechanics, Drilling and Production 83

In practice, the characteristic of the global rock drilling bit system in the required opera-
tion is also a straight line, in the plane with the weight on the x-axis and the penetration per
revolution on the y-axis. The slope of the line is representative of the intrinsic specific
energy of the rock. It is very high for a sandstone, the specific energy being very low. If the
weight on bit is not precisely controlled, we rapidly enter abrasion mode (regime I) or mode
III which, due to accumulation of chips, soon causes cleaning problems.
In contrast, the specific energy is high in shales, increasing with confining pressure. The
bit is controlled better with the weight.
The weight efficiency threshold is related to the presence of a more or less pronounced
wear flat. It is worthwhile converting this information in a plane containing the reduced
variables weight and torque and interpreting using Detournay’s formalism.
The link between the specific energy and the cutter angle of attack may also be
considered; if very large, energy is lost while cutting, but there is better tool control.
The cutting test indicated previously can be used to obtain the rock properties, provided
https://telegram.me/Geologybooks

that we are in regime II.

2.1.2.4 Punching models for bit behaviour

Determination of a model of the punching process and especially of the plastic zone
development phase is critical in order to obtain a good understanding of the tensile crack
initiation and propagation stage. Using the cavity expansion model as a basis, Detournay
demonstrated (1996) that the process is controlled by a single scalar g measuring the “degree
of elastoplasticity” and governing the size of the plastic zone:
2( K p + 1)Gtg β
γ =
πq
b is the punch angle,
q the unconfined compressive strength,
G the shear modulus,
KP a coefficient related to the angle of friction.
This factor g is quite similar to the typical strain e (load-dependent) described by Johnson
(1985) for the spherical indentation (Figure 2.1.11):

E* a
ε=
YR
where E* is the equivalent modulus of elasticity, R the indentation radius, a the contact cir-
cle radius, Y the yield stress.
For the conical indentation, this strain, independent of the load, is:
E * tg β
Y
The punch is conical, with half-angle at the top p/2-b. The radius of the contact circle is
a. Sharp cones (b close to p/2) favour plasticity, as do small radii R.
 84 Chapter 2 • Geomechanics, Drilling and Production

E*/Y is the ratio of the modulus of elasticity to a yield criterion, generally the shear
failure modulus. For a highly cohesive rock this ratio will be moderate (several hundred), for
a poorly cemented rock it will be higher (several thousand). Note that E*/Y is the reciprocal
of the strain at failure.
The formalism on the transition of mechanisms provides a mechanical explanation of the
Johnson or Detournay indices. Including the specific fracture work, it introduces competi-
tion between failure and plasticity. This is in fact the meaning of the cutting mechanism
boundary l indicated by Detournay.

2.1.3 Drilling performance including evacuation of cuttings and bit dynamics

Drillability is traditionally studied in terms of steady state regimes, with limited fluctuation
of the characteristic quantities of the system.
By definition, the study of dysfunctions concerns transitions, whether due to a poorly
https://telegram.me/Geologybooks

managed control parameter – generally the weight, but possibly the speed of rotation or the
hydraulics – whose untimely variation triggers the phenomenon, or through the sudden
appearance of a rock formation transition for which the current control is unsuitable, thereby
inducing an unfavourable effect.
This description is also valid for dysfunctions such as excessive vibration in hard rocks
or for an incident related to clogging in soft rocks.
As we have seen in 2.1.1.2, normal behaviour is characterised by an operation line in a
suitable plane of variables (Figure 2.1.5) relating mean quantities (RD – TD) and which sug-
gests a Mohr-Coulomb type failure criterion with, in particular, a steady increase in the nor-
malised torque TD when the confinement effect RD increases (Falconer representations for
the roller-cone bit, Detournay for the PDC).
A significant departure of the operation points away from the “normal” line may come
from:
(i) A radical modification of the rock behaviour at failure (sudden lithological transition).
(ii) A vibratory dysfunction during the interaction of the bit with a very hard rock or on
the contrary degenerating into clogging/jamming with a soft rock for which the con-
trol was not adapted on time. Clogging designates conditions under which the
balance of divided solid material (cuttings) is in excess, jamming refers to conditions
which are “even more irreversible” for which physicochemistry makes the bonds
between aggregates and bit structure even stronger.
(ii) is in fact an exaggerated form of (i).
(i) In this case, after facies transition, the bit is no longer under optimised conditions
allowing efficient drilling but we are nevertheless outside the field of normal opera-
tion.
(ii) In this case, operation is disturbed on leaving the transition and drilling cannot
continue in the long term; an inappropriate modification of the control or lack of
reactivity by the driller during a facies transition may cause these untimely and often
irreversible excursions.
 Chapter 2 • Geomechanics, Drilling and Production 85

2.1.3.1 Modelling of jamming during drilling

Jamming or clogging is a major problem encountered by drillers [Garnier and van Lingen,
1958] [Hebert and Young, 1972]. The problem occurs when, due to defective hydraulics or
unfavourable lithology, rock chips build up at the working face; various descriptive mecha-
nisms have been proposed, some privileging the bit and some the formation as “catalytic”
site of the aggregation process. Nevertheless, existing models [Wardlaw, 1971] [Warren,
1984] cannot be used as a predictive base and a fortiori as a means of control to avoid it.
One original study consisted in using conservation laws and simplified constitutive rela-
tionships [Eronini et al., 1982] to determine the evolution equations giving the global accu-
mulation rate at the working face as a function of time. The critical regimes, as a function of
the weight, speed of rotation and flow rate control, were identified; excessive weights
quickly lead to clogging, with large cuttings (Figure 2.1.13); insufficient weights result in
slow pollution of the working face by fine particles and also induce an inefficient regime
that can be qualified as jamming, thereby completing, from a hydraulic point of view, the
https://telegram.me/Geologybooks

description of the well-known effect of mechanical efficiency threshold [Putot, 1995, 1997].
Putot’s model represents a significant improvement over Warren’s semi-empirical
representations in case of defective hydraulics since it considers, in a simplified way, the
phenomena actually involved. Obviously, the disadvantage is that a large number of parameters
must be identified. However they all have a physical significance. The advantage of the analysis
concerns a realistic description of the conditions leading to a bit clogging mechanism.

Figure 2.1.13
Examples of damage to roller-cone bits. Top left: broken cone. Top right: worn
tooth. Bottom left: broken tooth. Bottom right: jamming.
 86 Chapter 2 • Geomechanics, Drilling and Production

A dynamic version of the bit coupled with the cuttings concentration has been developed
[Putot and Constantinescu, 1997]: the state of the working face has two dynamic variables (in the
longitudinal and circumferencial directions) and a concentration variable; the solution of the
differential system combining these state variables is as much one of a problem of dynamics as
one of a problem of concentration evolution. The influence of the initial conditions and the
severity of the ramps forming the loading paths (in particular weight and torque control) have been
studied. Compared with the quasi-static version, this model offers highly significant advantages:
influence of bit dynamics on concentration evolution – and improved understanding of the drilling
system dynamics in case of clogging “problems”; but also, introduction of the rotation degree of
freedom and consequently torque taken into account in the response observation criteria.
Validation has been carried out semiquantitatively on a GdF Suez site [Putot et al., 2000].

2.1.3.2 Drilling efficiency and vibrations of the drilling system

Symptoms of drillstring vibrations vary widely. Sometimes, the drillstring bounces or
https://telegram.me/Geologybooks

rotates so irregularly that the drill team is clearly aware that something is wrong. In other
cases, for example in case of transverse vibrations, the anomaly is undetected at the surface
and can lead to the failure of the equipment, observed too late!
The problems due to vibrations are varied and often extremely serious. Frequently, the
damaged equipment must be brought to the surface, the borehole is irregular and, during
coring operations, the samples are damaged (e.g. core discing). Vibrations are correlated
with a reduced rate of penetration. Limiting the vibrations generally involves higher costs,
due to the need for more sophisticated equipment [Jardine et al., 1994].
It is important to dispel the notion that vibrations may benefit the rock drilling process
since, while the rate of penetration may increase in the short term, this comes at the expense
of shorter lifetime of the bit and drillstring. Although vibrations such as bit noise are used as
a seismic signal while drilling, these are usually of a different frequency and amplitude from
the destructive vibrations.
Drillstring vibrations can be divided into three types: axial, torsional and transverse:
– Axial vibrations cause bit bounce and rough drilling, behaviour that destroys bits,
damages Bottom Hole Assemblies (BHA), increases total drilling time and may be
detected at surface.
– Torsional vibrations cause irregular downhole rotation that fatigues drill collar con-
nections, damages the bit and slows drilling. The vibrations are recognised at the drill-
floor by fluctuations in the power needed to maintain a relatively constant rate of sur-
face rotation.
– The more destructive transverse vibrations may be unleashed with no sign at surface.
Deep in the hole, the BHA interacts with the borehole wall, generating very violent
transverse shocks; this leads to an out-of-gauge and rough hole and the shocks can
damage components of the BHA. The bottomhole equipment (optional) used to assess
the severity of this type of vibration significantly increases the drilling costs.
Vibrations of all three types may occur simultaneously or one at a time, depending on the
case, during rotary drilling and are more violent in vertical or low-angle wells, where the
drillstring may move more freely than in high-angle wells with a long offset.
 Chapter 2 • Geomechanics, Drilling and Production 87

Many studies of drillstring vibrations employ one-dimensional, linear harmonic analysis

that largely ignores the dynamics of wellbore wall contact. This approach has proved rela-
tively successful for axial and torsional vibration. Studies have concluded that there are two
types of torsional vibrations: transient and stationary. Transient vibrations correlate with
variations in drilling conditions such as heterogeneity in the rocks. Stationary vibrations are
caused by the natural resonance of the drillstring and are most likely to cause problems.
During rotary drilling, the most recognisable manifestation of stationary torsional vibration
is stick-slip. Because of friction between the bit and BHA (Bottom Hole Assembly) and the
wellbore and the spring-like nature of the drillstring, the bit actually stops rotating, even though
the drillpipe is still being turned at a more or less constant rate on surface. This is known as the
stick phase. Sufficient torque is progressively stored up in the system, until it finally overcomes
friction and releases the bit which accelerates suddenly, reaching angular speeds far greater than
the nominal speed. This is known as the slip phase. This phase may last for several seconds,
depending in particular on the length of the drillstring and, to a lesser extent, the mean speed of
rotation. The speed then decreases to start another stick phase. These more or less periodic oscil-
https://telegram.me/Geologybooks

lations exert high cyclic stresses on the drillpipe and slow drilling [Pavone and Desplans, 1994].
The high speeds achieved during the slip phase are also one of the causes of transverse
vibrations due to centrifugal force [Théron et al., 2001].
Axial vibrations are also excited, since the axial drilling rate depends on the angular
speed. Lateral accelerations are frequently 10 times greater than axial accelerations.
However, despite the relation with rotation, analysis of transverse vibrations cannot rely
on a harmonic model. This is important since, in the past, it was believed that by “fine-
tuning” the drillstring – varying its rate of rotation and weight on bit – to avoid a resonant
frequency, transverse vibrations that have been initiated downhole may be stopped. Studies
by ANADRILL reveal, on the contrary, that this does not always work.
Field observations have led to the conclusion that transverse vibration problems appear
in hard or abrasive formations. Shocks are often seen as the MWD (Measurement While
Drilling) or adjacent stabiliser passes from a soft to a hard formation, without changing the
speed of rotation.
At the heart of the generation of transverse vibrations is a self-sustaining interaction
between the rotating drillstring and the wellbore [Théron et al., 2001]. Once initiated, this
interaction is hard to stop. Once shocks are observed downhole, varying the speed of rota-
tion will often affect only the number and severity of the shocks. Shocks are not eliminated
until the speed of rotation is reduced to a much lower value than the rate at which they com-
menced. Thus, significant shocks may be sustained at all but the lowest speeds of rotation.
Variation in transverse vibration with lithology is due to changes in the rock’s coeffi-
cients of friction and restitution. As friction between the drillstring and rock increases, more
energy per impact is transformed from rotary to transverse motion. In some cases, friction
may be exaggerated through the digging action of stabilisers clashing with the borehole.
The coefficient of restitution determines what proportion of the kinetic energy at impact
is absorbed by the formation. Low restitution results in significant energy absorption.
Because limestones and sandstones have high coefficients of friction and restitution, they
are more likely to generate high shocks than soft shales – an intuitive result.
 88 Chapter 2 • Geomechanics, Drilling and Production

2.2 BOREHOLE STABILITY DURING DRILLING

2.2.1 Instabilities and drilling

2.2.1.1 Description of drilling modes

Two drilling modes can be used: OBD (Over Balanced Drilling) and UBD (Under Balanced
Drilling). OBD is a procedure used to drill wells where the mud pressure is greater than the
pore pressure. Most wells are drilled using this method. In UBD, the mud pressure is less
than the pore pressure and production takes place while drilling. In the remainder of this
document, we will implicitly assume that OBD is used.

2.2.1.2 Borehole instability mechanisms while drilling

When a well is drilled, the rock surrounding the borehole must take up the load previously
https://telegram.me/Geologybooks

supported by the rock that has been removed. This results in the development of a stress con-
centration at the borehole wall. If the rock is not strong enough, the wall will fail.
In OBD, the mud density is determined to prevent a fluid influx from the formation
(Pmud > Ppore). If the mud density is too low compared with the mechanical stability crite-
rion, compressive failures could occur, possibly causing the well to collapse. The mud den-
sity must therefore be adjusted according to the most restricting criterion (pore pressure or
compressive stability).
The mud density cannot be increased infinitely, however, since if it is too high, tensile
failures caused by hydraulic fracturing will develop in the rock (Figure 2.2.1).

Formation Lost Wellbore

breakdown circulation pressure

Friable
sandstone
Hole
collapse
Brittle
Compressive shale
failure

Hole
closure Salt

Figure 2.2.1
Failure by stress concentration around the borehole [Last & McLean, 1995].
 Chapter 2 • Geomechanics, Drilling and Production 89

The mud pressure must therefore lie within a window (Figure 2.2.2) whose:
– lower limit is the larger of two values, which are the pore pressure Ppore and the mini-
mum mud pressure necessary to ensure stability of the walls Pcollapse,
– the upper limit is given by the fracture propagation pressure Pfrac in the terrain, i.e.:
max (Ppore, Pcollapse) < Pmud < Pfrac

Wellbore stability chart

mud weight window

Fracture

Stable
https://telegram.me/Geologybooks

region
Depth

Collapse

Drilling fluid density

Figure 2.2.2
Mud weight window.

The mud weight window may be very narrow when the – abnormal – pore pressure is close
to the hydraulic fracturing pressure, a situation observed in numerous parts of the world, for
example the North Sea [Ward et al., 1994] and the Potwar Basin in Pakistan [Kadri, 1991].
The true mud pressure in the borehole depends on the static weight of the mud column
increased by the dynamic effect of the flow (known as ECD – Effective Circulating
Density), together with occasional fluctuations as the drillstring moves (pistoning or
suction). In view of these fluctuations, the borehole stability conditions are often borderline.
In OBD, the aim is to work at a pressure slightly above the lower limit for several reasons:
– to detect impregnated reservoirs,
– to limit the risks of fracturing and mud losses,
– to increase drilling performance.
Compressive failures are highly dependent on the in situ stress field. Failures occur pro-
gressively. Spalling appears towards the minor stresses in the first phases followed by stabi-
lisation with borehole breakout or continued enlargement of the borehole if the spalls come
loose from the wall and fall off (Figure 2.2.3).
 90 Chapter 2 • Geomechanics, Drilling and Production

Diameter (mm)
200 240 280 320 360
s 2385
mm NS
3°W

Lon
ga 2395
xis

mm
=3
40

24
mm

Depth (m)
=2 2405
xis
rt a
o
Sh

Drilled diameter 2415

222 mm

2425
https://telegram.me/Geologybooks

a) b) Long axis Short axis

Figure 2.2.3
Schematic illustration of spalling in the damaged area around a borehole.
Detection of borehole breakout using a Calliper [McLellan, 1994].

Maury and Sauzay (1987) identify more precisely three destabilisation mechanisms cor-
responding to three different stress models at the wall (sa = axial stress, sq = tangential
stress, radial stress = mud pressure):
– mode A, with “intermediate” axial stress (Pmud < sa < sq), indicated by the presence
of narrow, elongated spalls,
– mode B, with “intermediate” tangential stress sq (Pmud < sq < sa), producing curved
spalls concentric with drilling,
– mode C, shear failure when the mud pressure is high enough to open the natural frac-
tures (sq < Pmud < sa), generating small wedge-shaped spalls (Figure 2.2.4),
– mode D, extension failure.
The temperature variations induced by the circulation of mud may have considerable
influence on borehole stability [Maury, 1994]. Local cooling (at well bottom) releases the
compressive stresses around the borehole, possibly leading to temporary stability which will
disappear if the temperature rises (e.g. if the mud circulation stops). Inversely, local heating
(at the top of the well) may cause instabilities. Maury mentions that, in some cases, mud
cooling devices have proved an efficient way of reducing instability. It is important to know
the temperature distribution, or at least have an estimation (Figure 2.2.5). A number of soft-
ware programs capable of predicting the temperature variations in the borehole are now
available for operators.
 Chapter 2 • Geomechanics, Drilling and Production 91

Isotropic Lateral in-situ stresses Anisotropic

Internal pressure
Intermediate stress (or mud weight)
Mode A

Nil or low High

"A" shear
Borehole elongation
https://telegram.me/Geologybooks

Intermediate

Medium Medium

"B" shear

Intermediate Toroïdal spalls

C
Low

High

"C" shear

Very high D

"D" extension

Hydraulic fracturing

Figure 2.2.4
Failure modes and spalling process [Maury & Sauzay, 1987].
 92 Chapter 2 • Geomechanics, Drilling and Production

Temperature (°C)
20 40 60 80 100 120 140 160 180

100
17"1/2 Drilling phase (Q —~ 3300 l/min)
12"1/4 Drilling phase (Q —~ 2400 l/min)

1000 8"1/2 Drilling phase (Q —~ 1000 l/min)

2000
Static
17"1/2
Depth (m)

temperature

13"3/8 3000
70

12"1/4
4000
9"5/8
85
https://telegram.me/Geologybooks

8"1/2 5000
173°C/
7" liner 135 5000m

Figure 2.2.5
Temperature profiles during mud circulation [Maury, 1994].

2.2.2 Instabilities by formation type

2.2.2.1 Shale formations

Shales are highly compressible sedimentary rocks, composed of interstratified clay minerals
(kaolinite, illite, chlorite, montmorillonite) which can absorb water through their structural
network. The cationic exchange capacity is a good estimation of the swelling smectite content.
Shales represent 75% of the drilled formations and are responsible for over 90% of well
stability problems [Steiger & Leung, 1992].
If water-based drilling muds are used, migration of water into the shales is likely to
induce high swelling stresses. In situ stresses, however, limit water penetration and hydra-
tion generally remains restricted to the surface [Onaisi et al., 1994]. Note that diffusivity
increases if the rock is exposed to a mud hotter than the formation.
In case of imbalance between the formation and mud water compositions, ionic migra-
tion occurs as well as water migration. We may therefore consider the propagation of three
fronts – pressure, water and ion – which may locally modify the effective stresses and the
mechanical strength of shales [Fam and Dusseault, 1998; Chen et al., 2003a]. These fronts
propagate slowly and instability phenomena may only appear after several days [Mouton
et al., 1998]. They worsen over time, however (Figure 2.2.6). Reduction in local mechanical
properties and increase in pore pressure favour failure and produce spalling, enlarging the
borehole and creating numerous associated problems.
 Chapter 2 • Geomechanics, Drilling and Production 93

The choice of drilling fluid is important for the swelling shales; a high electrolyte con-
centration may reduce crystalline swelling but it is recommended to use polymers and seal
narrow fractures with gilsonite.

1.5
Normalised enlarged hole size

1.4

1.3

1.2

1.1
Model-predicted extent of yielded zone
https://telegram.me/Geologybooks

Calliper-measured enlarged hole size

1.0
0 10 20 30 40 50
Time (day)

Figure 2.2.6
Hole enlargement against time in a shale formation
[Hawkes and McLellan, 1997].

2.2.2.2 Unconsolidated formations (erosion)

These formations are less unstable than shales. Numerous incidents may occur while drill-
ing, however:
– caving caused by hydraulic erosion,
– deposit of a thick cake, risk of differential pressure sticking,
– mud losses due to excess pressure in the annulus,
– splitting, sliding for shales.
Problems are also encountered in arctic regions when drilling in the permafrost. The
heating caused by mud circulation considerably reduces the strength of the layers.

2.2.2.3 Fractured/cracked formations

Limestones, dolomites and other brittle formations may be naturally fractured or become
fractured while drilling. Invasion of the fracture network may cause loosening of the blocks
and their erosion (Figure 2.2.7). Blocks of all sizes may then break away from the walls and
fall into the annulus. Damage increases with time and may cause the well to collapse.
These fractured formations may have Unconfined Compressive Strengths (UCS) greater
than 50 MPa and the dominant failure mode may be due to friction of the fractures
[McLellan and Cormier, 1996, Chen et al., 2003b]. The walls deteriorate over time, as can
be seen on acoustic or electric images.
 94 Chapter 2 • Geomechanics, Drilling and Production

a) 2303.6 m b) 2309.9 m c) 2707.3 m

Natural Spalled rock Breakout
fracture wedge directions

Bedding
plane

Spalling Bedding
breakout plane

1m

Figure 2.2.7
https://telegram.me/Geologybooks

Borehole ultrasonic images in a cracked material [McLellan and Cormier, 1996].

In this type of formation, changing from a water-base mud to an oil-base mud has no
effect and increasing the mud density favours mud invasion of the fracture network
[Santarelli et al., 1992]. This problem is solved by adapting the mud rheology.
In some extreme cases, fault invasion will reduce the local effective stresses and the
shear stresses may cause lateral displacement of the borehole axis [Maury, 1994] with the
consequences we can well imagine: drillstring sticking, or even breaking (Figure 2.2.8).

u
sn
a) t b)

a) Ahead of the tool b) Behind the tool or stabilisers

Consequences: Consequences:
Tight hole Stuck in holes
Difficulties to run-in-hole Difficulties to pull out of hole
Abnormal torque Drillstring failures (?)
Need for reaming Need for back reaming

Figure 2.2.8
Lateral displacement of the borehole axis induced by shear relaxation along
pre-existing faults or fractures [Maury, 1994].
 Chapter 2 • Geomechanics, Drilling and Production 95

2.2.2.4 Creeping formations (salt or plastic shales)

The properties of evaporitic rocks (salt) vary depending on the component minerals (halite,
carnallite, sylvinite, kieserite and bischofite). Firstly, they are highly water-soluble (bischofite
is 10 times more soluble in hot water than halite (NaCl – table salt). Washouts beyond the
range of calliper logs of have been observed with water-base mud [Williamson et al., 1997].
The behaviour of salt under stress is similar to that of a highly viscous fluid. The strain
rate depends on stress (depth), temperature, mineralogical composition, water content and
the presence of impurities such as clays. Rates of 2 cm/hour have been recorded.
Some formations may be highly fractured, resulting in mud losses.
The operational problems have been identified: fluid kicks (brine, oil, gas, salt), losses
(filtration, fractures), drillstring sticking, casing collapse (Figure 2.2.9).
Large plastic strains may be observed in soft shales.
https://telegram.me/Geologybooks

Salt
washout

Salt

Dolomite

Salt Kicks
movement

Brine/oil/
gas/H2S

Fracture Filtration
losses losses

Mud /
brine

Mud /
brine

Figure 2.2.9
Instabilities in salt [Williamson et al., 1997].
 96 Chapter 2 • Geomechanics, Drilling and Production

2.2.2.5 Formations in tectonic zones

Stress distribution in tectonic zones differs from that observed in calm zones. The vertical
stress may be less than the horizontal stresses, which may be highly anisotropic. Variability
is generally high in the presence of faults.
In case of instability, spalls and breakouts spread in the direction of the lowest stresses,
for example parallel to the mountain ranges in the Columbia and Alberta foothills [Last and
McLean., 1995, Plumb et al., 1998].

2.2.3 Stability calculation

2.2.3.1 Analytical solution to the elasticity problem

The analytical solution to the elasticity problem for an inclined wellbore in a system exhibit-
ing anisotropic horizontal stress is described in numerous books and publications [Fjaer
https://telegram.me/Geologybooks

et al., 1992; Charlez, 1997]. The main equations given by Aadnoy and Chenevert (1987)
will be mentioned below.
The following assumptions are made: linear elasticity and isotropy of the in situ rock mechani-
cal properties. We assume that the principal stresses in the formation are: sv vertical stress, sH
maximum horizontal stress, and sh minimum horizontal stress. Stresses near the wellbore are
described in a coordinate system where the z-axis is parallel to the wellbore and the y-axis is hori-
zontal (Figure 2.2.10). Transformation of coordinates (sv, sH, sh) to (x, y, z) is carried out by two
rotations, a rotation “a” around the sv axis and a rotation “i” around the y-axis. Angle i represents
the wellbore deviation and angle a the azimuth with respect to the maximum horizontal stress.
Pw represents the mud pressure in the wellbore of radius R.

sv z

sh i
y

sH a
sz
x

Pw

sr
sq

Figure 2.2.10
Transformation system and notation.
 Chapter 2 • Geomechanics, Drilling and Production 97

The stresses in cylindrical coordinates are given by the following expressions as a func-
tion of the distance r to the centre of the wellbore.
⎛ σx + σ y ⎞ ⎛ R2 ⎞ ⎛ σ x − σ y ⎞ ⎛ R4 R2 ⎞
σr = ⎜ ⎟ ⎜1 − 2 ⎟ + ⎜ ⎟ ⎜ 1 + 3 4 − 4 2 ⎟ cos 2θ
⎜⎝ 2 ⎟⎠ ⎝ r ⎠ ⎜⎝ 2 ⎠⎟ ⎝ r r ⎠

⎛ R4 R2 ⎞ R2
+ τ xy ⎜ 1 + 3 − 4 ⎟ sin 2θ + p
⎝ r 4 2
r ⎠ r2 w

⎛ σx + σ y ⎞ ⎛ R2 ⎞ ⎛ σ x − σ y ⎞ ⎛ R4 ⎞ ⎛ R4 ⎞ R2
σθ = ⎜ ⎟ ⎜1 + 2 ⎟ − ⎜ ⎟ ⎜ 1 + 3 4 ⎟ cos 2θ − τ xy ⎜ 1 + 3 4 ⎟ sin 2θ − 2 pw
⎜⎝ 2 ⎟⎠ ⎝ r ⎠ ⎜⎝ 2 ⎠⎟ ⎝ r ⎠ ⎝ r ⎠ r

( ) Rr cos 2θ − 4ντ
2 R2
σ z = σ zz − 2ν σ x − σ y sin 2θ
https://telegram.me/Geologybooks

2 xy
r2

⎡⎛ σ − σ ⎞ ⎤⎛ R4 R2 ⎞
x y
τrθ = ⎢⎜ ⎟ sin 2θ + τ xy cos 2θ ⎥ ⎜ 1 − 3 4 + 2 2 ⎟
⎢⎜⎝ 2 ⎟⎠ ⎥⎝ r r ⎠
⎣ ⎦

⎛ R2 ⎞
τrz = ⎡τ xz cos θ + τ yz sin θ ⎤ ⎜ 1 − ⎟
⎣ ⎦⎝ r2 ⎠

⎛ R2 ⎞
τθ z = ⎡ −τ xz sin θ + τ yz cos θ ⎤ ⎜ 1 + ⎟
⎣ ⎦⎝ r2 ⎠

With

( )
σ x = σ H cos2 a + σ h sin2 a cos2 i + σv sin2 i

(
σ y = σ H sin2 a + σ h cos2 a )
( )
σ zz = σ H cos2 a + σ h sin2 a sin2 i + σv cos2 i

( ) )
τ yz = 0.5 σ H − σh sin ( 2 a sin i

( )
τ xz = 0.5 σ H cos2 a + σ h sin2 a − σv sin 2i

( ) )
τ xy = 0.5 σ H − σh sin ( 2 a cos i
 98 Chapter 2 • Geomechanics, Drilling and Production

At wellbore wall r = R, so the equations become:

σr = pw

σθ = σ x + σ y − 2(σ x − σ y )cos 2θ − 4τ xy sin 2θ − pw

( )
σ z = σ zz − 2ν σ x − σ y cos 2θ − 4ντ xy sin 2θ

τrθ = 0

τrz = 0

τθ z = 2 ⎡ −τ xz sin θ + τ yz cos θ ⎤
⎣ ⎦
In an anisotropic horizontal stress field (sH max > sh min), concentrations of maximum
circumferential stress sq are observed on the wellbore walls in the direction of the minimum
https://telegram.me/Geologybooks

horizontal stress (Figure 2.2.11). We can therefore see that breakouts will occur in these
zones, depending on the hardness of the rock.
160
sqq
140
Stress at wellbore wall (MPa)

120

100

80

60
szz
40

20
srr
0
0 90 180 270 360
Angle around the hole (from south)

Figure 2.2.11
Stresses sq around a vertical wellbore under an east-west compressive stress
field [Zoback et al., 2003].

sz and sq are not principal stresses, which are calculated as follows:

σ + σz 1
σ1 = θ + (σθ − σ z )2 + 4τθ2zz
2 2
σ + σz 1
σ2 = θ − (σθ − σ z )2 + 4τθ2zz
2 2
σ3 = pw
 Chapter 2 • Geomechanics, Drilling and Production 99

Any compressive failure criterion can be applied to estimate the collapse failure: Mohr
Coulomb, which does not take into account the intermediate stress, Von Mises, Drucker
Prager, etc. [McLean and Addis, 1990].

2.2.3.2 Wall instability model

Various software programs are now available to calculate wall stability for any wellbore
angle and any stress field.
These tools run rapidly on PC with a simple poroelastic model, sometimes taking into
account fluid diffusion against time as well as thermal and chemical effects.
They are often conservative [Van Oort et al., 2001]. Rocks are not purely elastic and
continue to have a load bearing capacity after failure. Operators therefore use these linear
elastic models for fast analysis and optimisation of the wellbore trajectory, but still require
more sophisticated models involving finite element codes, with poroelastoplasticity models
(Cam-Clay or Laderock type).
https://telegram.me/Geologybooks

These models have demonstrated in particular that horizontal wellbores are mechanically
more stable if they are drilled parallel to the maximum horizontal stress [Zoback et al., 2003].

2.3 SAND PRODUCTION

Sand production designates transport of grains by the production fluid, from the reservoir
rock into a wellbore.
In gas and light oil reservoirs, or in offshore production, above a certain proportion, sand
production causes a number of undesirable problems such as damage to wellbore pumps and
wellhead erosion (a leak may be created by sandblasting in less than an hour in a gas well),
plugging of perforations or even total invasion of the production column. Several methods
are available to address these problems, but they are expensive to implement and often
reduce the production rate.
In heavy oil reservoirs, if sand production prevention systems are used, the oil flow rates
are considerably reduced (from 7/15 m3 a day to 0.5 m3 a day, for example) and the well
becomes uneconomical. In this case, it is acceptable to produce oil and sand simultaneously,
as with the Cold Heavy Oil Production with Sand (CHOPS) method which may prove
profitable under certain conditions. Due to the high viscosity and low production rates,
pump and valve erosion problems are minor and the sand can be transported up to the
surface without difficulty using progressive cavity pumps. If thermal stimulation is used, for
example Steam Assisted Gravity Drainage (SAGD), the oil viscosity is lowered and systems
must be installed to prevent sand production.
In all cases, therefore, anticipating the onset of sand production is a major challenge.
The physical problem is extremely complex and involves numerous parameters: rock
and reservoir characteristics, wellbore completion type and production conditions. Some
triggering mechanisms are well-known, for example damage to soft formations by drilling
or depletion, and damage under the effect of hydrodynamic forces; other, such as water cut,
erosion at the surface of cavities, etc., are much less well-known.
 Next Page

100 Chapter 2 • Geomechanics, Drilling and Production

2.3.1 Types of sand production

We could speak of “solid” production since this term is less restricting than “sand” which
refers to a geological and grain size classification. According to Dusseault and Santarelli
(1989), chalk and coal may also produce solids, ranging from sands to silts and clays. The
fact remains that, in the large majority of cases, it is sand production which occurs.
The term sand production includes a wide range of phenomena. On the basis of field
observations, Veeken et al. (1991) suggested three types of solid production: transient, con-
tinuous and catastrophic.

2.3.1.1 Transient solid production

Transient solid production refers to a solid concentration declining with time under constant
well production conditions. This phenomenon is frequently observed during clean-up after per-
forating or acidising, after pullout or after water breakthrough. The rate of sand production may
https://telegram.me/Geologybooks

vary considerably, for example a sand volume of 1 to 200 L over a period of 1 to 500 hours.

2.3.1.2 Continuous production

In a large number of fields, continuous levels of sand production are observed. What constitutes an
acceptable level of sand production depends on operational constraints (erosion, separator capacity,
sand disposal, lift, etc.). For example, the following levels can be considered as acceptable:
– 6-600 g/m3 in oil wells,
– 16 kg/Mm3 in gas wells.
The well may nevertheless fill up with sand and require regular cleaning. Sand production
increases the volume of perforations and may even create large cavities around the casing.

2.3.1.3 Catastrophic solid production

Catastrophic solid production refers to events where a high influx of solids, sometimes sev-
eral dozen m3, causes the well to choke and/or die.
The following correlations are observed:
– The grain size distribution of the sand produced is identical to that of the formation.
– Sand production is more likely to occur with poorly consolidated, low compressive
strength (Rc < 2 MPa) rocks at shallow depth.
– Sand production is triggered or increases with reservoir depletion, when the effective
stresses inside the reservoir increase.
– Sand production occurs when the pressure drawdown (difference between pore pres-
sure far from the wellbore and pressure in the flowing well) reaches or exceeds a cer-
tain critical threshold; the critical pressure drawdown is associated with the rock
mechanical strength. Moricca et al. (1994) speak of total pressure drawdown, the dif-
ference between the initial pore pressure far from the wellbore and the pressure in the
flowing well. Field observations show that sand production continues while the stress
conditions which triggered it remain and stops when these conditions disappear.
– Watercut is often correlated with sand production.
 3

Geomechanics
and Reservoir
https://telegram.me/Geologybooks

It has become standard practice in reservoir engineering to use a geocellular 3D model for
the reservoir and the surrounding rocks in order to study the possible combined effects
between fluid production and the mechanical strains of the structure. It is in fact very impor-
tant to predict possible subsidence of the surface as oil and gas production proceeds.
Even a slight variation in pore pressure in a reservoir may cause significant displace-
ments in the subsurface layers, especially if the reservoir is very poorly or weakly cemented,
which increases the risk of collapse and the development of shear zones. Combined prob-
lems such as these are encountered when studying the stability of CO2 sequestration sites,
where it is important to check that the injection of large quantities of gas into the structure
will not destroy its long-term mechanical stability.
Exploitation of an oil reservoir is very expensive and entails a high degree of uncer-
tainty. These costs and uncertainties can be reduced by reservoir studies conducted by the oil
companies before or during exploitation. These reservoir studies are generally based on
numerical simulations of flows in the reservoir in order to predict the quantities of hydrocar-
bons present and their production over time, optimise reservoir production and assess the
risks associated with production. Reservoir simulators give a highly accurate description of
the fluid and heat transfers in the reservoir but do not model (or do so in a very simplified
manner only) the mechanical behaviour of the reservoir during its production. Production in
certain types of reservoir (poorly consolidated reservoirs, fractured or faulted reser-
voirs, etc.) is accompanied by major mechanical effects likely to threaten the integrity of the
reservoir and its wells, generate subsidence of the surface layers and modify the flows of the
reservoir fluids, and therefore any production predictions made by the reservoir engineers.
Modelling these mechanical effects therefore requires a geomechanical model which must
be coupled with the reservoir simulation.
We will first describe the hydraulic diffusivity equation, a classical model of the single-
phase and isothermal flows in porous media. This diffusivity equation involves two
important parameters: the compressibility and the permeability reservoir rock. We will then
demonstrate that these two parameters may be highly dependent on the stress state in the
reservoir. In this case, it becomes essential to model the mechanical behaviour of the oil
 136 Chapter 3 • Geomechanics and Reservoir

reservoir during production. We then describe how the coupling between the reservoir and
geomechanical models can be achieved. Two examples of coupling between reservoir and
geomechanical simulators are described for a high porosity reservoir and a faulted reservoir.
Lastly, we will study the evolution of stresses in the reservoir during depletion and the
measurements available to validate it.

3.1 MODELLING THE GEOMECHANICAL AND RESERVOIR COUPLING

3.1.1 Flow and permeability

3.1.1.1 Hydraulic diffusivity equation
The phenomena involved during production of oil or gas from a reservoir and when
injecting fluids (water, gas, etc.) to enhance production are mainly related to the flow of
https://telegram.me/Geologybooks

fluids in porous media (see Bear and Buchlin, 1991, for example). To simplify the
description, we will restrict ourselves to single-phase isothermal flows in porous media. We
write f the porosity of the porous medium (Eulerian porosity, which is the ratio of the pores
to the current total volume) and r the density of the saturating fluid. The conservation of
fluid mass can then be expressed as:
∂
( ρφ ) + div ( ρφ v ) = 0 (3.1.1)
∂t
with v the seepage velocity of the fluid in the porous medium. The latter is generally related
to the fluid pressure p via Darcy’s law which, in the absence of gravity, can be written:

k
φv = − grad p (3.1.2)
μ
with k the rock intrinsic permeability and m the fluid dynamic viscosity.
The fluid is assumed to be compressible, of compressibility Cf defined by:

∂ρ = ρ0C f ∂p (3.1.3)
with r0 the density of the fluid under fluid pressure p0.
A reservoir rock “compressibility” Cr is introduced at the same time, to take into account
a variation of the porous medium porosity as a function of the saturating fluid pressure:

∂φ = φ0Cr ∂p (3.1.4)
with f0 the initial porosity of the porous medium given for the initial fluid pressure p0. Intro-
ducing equations (3.1.2) to (3.1.4) in the conservation of mass equation (3.1.1), we obtain
after linearisation:
∂p k
φ0 (C f + Cr ) − Δp = 0 (3.1.5)
∂t μ
 Chapter 3 • Geomechanics and Reservoir 137

Associated with representative boundary and initial conditions, equation (3.1.5) gives the
evolution of saturating fluid pressure in the porous medium. It is important to note, however,
that the compressibility Cr is not an intrinsic property of the reservoir rock but depends on
the stress path followed by the reservoir during its depletion. In addition, the permeability of
the porous medium may vary significantly during depletion of a reservoir due to changes in
effective stress in the reservoir.

3.1.1.2 Reservoir rock compressibility

Equation (3.1.5) shows that the evaluation of recovery is based mainly on the estimation of
reservoir rock compressibility, which represents the strain of the porous volume induced by
the variation in reservoir pressure:
1 ΔVp
Cr =
https://telegram.me/Geologybooks

(3.1.6)
Vp Δpp

So defined, the reservoir rock compressibility depends on the stress path followed by the
reservoir during production. This stress path may be quite different from the isotropic or
oedometric loading paths traditionally adopted in the laboratory to evaluate Cr. In particular,
it may be heterogeneous and vary during depletion (see Chapter 3.1.3).
Concerning this point, note that the compressibilities defined by Zimmerman describe
exclusively the volume response of a rock subjected to hydrostatic loading σ = − pc 1 and to
a pore pressure pp [Zimmerman 1991]:

1 ∂Vb 1 ∂Vp
Cbc = − Cpc = −
Vb ∂p pp Vp ∂p pp
c c

1 ∂Vb 1 ∂Vb
Cbp = − Cpp = − (3.1.7)
Vb ∂p pc Vb ∂p pc
p p

where Vp designates the porous volume and Vb the bulk volume, the compressibility closest
to definition (3.1.6) being obviously Cpp.
Assuming linear poroelastic behaviour of the reservoir rock, the compressibilities so
defined can be expressed as a function of the drained bulk modulus Ko and Biot’s
coefficient b [Boutéca, 1992]:

1 b Cbp
Cbc = Cpc = =
K0 φ0 K0 φ0

b b − φ0 (1 − b )
Cbp = Cpp = (3.1.8)
K0 φ0 K0
 138 Chapter 3 • Geomechanics and Reservoir

For an incompressible solid matrix (i.e. b = 1), we obtain simply:

1 1
Cbc = Cbp = Cpc = Cpp = (3.1.9)
K0 φ0 K0

Still assuming linear poroelastic behaviour of the reservoir rock, we can determine
expressions for Cr as a function of the loading path considered. The three types of loading
path generally followed to simulate depletion in the laboratory are the hydrostatic,
oedometric and proportional paths (see Chapter 3.1.3):

1 ⎛ b2 b − φo ⎞
Crhydrostatic = ⎜ + ⎟
φo ⎝ Ko Ks ⎠

⎛ ⎞
1 ⎜ b2 b − φo ⎟
https://telegram.me/Geologybooks

Croedometric = ⎜ + ⎟ (3.1.10)
φo ⎜ 4G Ks ⎟
Ko +
⎝ 3 ⎠

1 ⎛ 1 + 2 K b2 b − φ ⎞
o
Crproportional = ⎜ + ⎟
φo ⎝ 3 Ko
Ks ⎠

For a hydrostatic loading path, we obtain the expression for Cpp expressed in this case as
a function of the bulk modulus of the solid matrix.
Table 3.1.1 compares the compressibilities estimated from expressions (3.1.8) and
(3.1.10). The poroelastic parameters used are extracted from the thesis of D. Bévillon (2000)
and are representative of a highly compactable reservoir chalk. Note that the values obtained
largely depend on the assumptions made concerning the loading path.

Table 3.1.1 Estimation of chalk compressibility as a function of the loading path followed.

Assumed stress path Calculated compressibility (GPa–1)

Hydrostatic path (Cbc) 0.71
Hydrostatic path (Cpp) 1.76
Oedometric path 0.88
Proportional path (K = 0.5) 1.18

In addition, for some rocks, the compressibility may depend, like the drained bulk modu-
lus, on the pore pressure and stress state [Bemer et al., 2001].
 Chapter 3 • Geomechanics and Reservoir 139

3.1.1.3 Intrinsic permeability variation

The experimental data obtained on various reservoir rocks (sandstones and carbonates) show
that the permeability of a rock subjected to compression decreases during the initial
nonlinear phase when the existing microcracks close and the linear phase associated with
pore elastic deformation. The change in permeability during the post-elastic phase is more
complex: examples where the permeability decreases [Zhu and Wong 1997] [Brawn and
Jahns 1998] [Yale and Crawford 1998] and examples when the permeability increases
[Rhett and Teufel 1992a] [Keaney et al. 1998] [Brawn and Jahns 1998] [Yale and Crawford
1998] have been encountered in the laboratory. The results obtained demonstrate that the
change in post-elastic permeability depends on the loading path followed and the initial rock
porosity and permeability levels.
It is important to note that the various studies mentioned are based on traditional triaxial
tests where the stress state is axisymmetric and where the permeability is measured along
the axial direction. Microcracks develop preferentially along the axial direction, which is
https://telegram.me/Geologybooks

also that of the flow. Consequently, when the authors observe an increase in permeability
with the dilatancy of the sample, the permeability is measured parallel to the main axis of the
cracks. The petroleum industry, however, is more concerned by the change in horizontal
permeability.
Schutjens and de Ruig (1996) used a special cell to measure the change in radial perme-
ability (flow perpendicular to the preferential orientation of the cracks). Sandstone samples
with high initial porosity and permeability levels (porosity between 21% and 32%, radial
permeability between 0.1 mD and 500 mD) have been tested under hydrostatic and oedo-
metric loading paths. The true depletion tests showed a decrease in radial permeability.
However, the test conditions (loading paths and initial permeability levels of the samples)
correspond a priori to cases where the axial permeability also decreases. By comparing tests
with measurement of the axial permeability and tests with measurement of the radial perme-
ability, conducted on the same rock of low initial porosity and along a deviatoric loading
path with low mean effective stress, it would be possible to check whether the radial perme-
ability increases in the same way as the axial permeability.
Only very few data are available in the literature on the behaviour of a rock subjected to
a real triaxial loading path and on the anisotropy of the permeability tensor. Al-Harthy et al.
(1998) conducted depletion tests on various sandstones for three different stress states: true
triaxial (po; 0.8*po; 0.6*po), triaxial (po; 0.8*po; 0.8*po) and hydrostatic (po; po; po). The
results have shown a decrease in the vertical permeability (direction of the major principal
stress) with the drop in pressure, this loss of permeability being lower for true triaxial load-
ing than for traditional triaxial loading or hydrostatic loading.
Al-Harthy et al. (1999) completed these results by studying, still for sandstones, the
anisotropy of the permeability tensor and the influence of the loading path on the change in
permeabilities in the three main directions of the true triaxial test. For samples exhibiting
horizontal stratification, they measured vertical anisotropy factors (ratios between the verti-
cal permeability and a horizontal permeability) of up to 40; this factor is obviously greater
for true triaxial loading (1.25*po; po; 0.75*po) than for hydrostatic loading (po; po; po). The
horizontal anisotropy factors (ratios between the two horizontal permeabilities) measured
 140 Chapter 3 • Geomechanics and Reservoir

are more limited and lie between 1.1 and 1.7, but this difference in horizontal permeabilities
may prove extremely important when installing a horizontal well. The results obtained do
not indicate a clear trend as to the influence of depletion on the anisotropy of permeabilities;
a slight decrease in anisotropy with increase in effective stresses has nevertheless been
observed on a reservoir sample exhibiting horizontal stratification.
Some authors have tried to define variation laws (mainly empirical) expressing perme-
ability as a function of the stress path followed. The small permeability variations in the
elastic range can be estimated using the experimental law proposed by David et al. (1994):

( )
⎡− a p '− p ' ⎤
k = ko e⎣ o ⎦

where ko is the permeability measured at po'.

A variation law was also proposed by Boutéca et al. (2000) to describe the change in per-
meability between the damage and brittle failure criteria in the case where permeability
decreases, i.e. for high initial permeabilities. This law involves the stress path followed
https://telegram.me/Geologybooks

defined by the derivative of the function p ' = f ( q) (effective mean stress – deviatoric stress):
1 ∂k ∂p '
− =a +b
k0 ∂q ∂q
where ko is the permeability before damage.
For low porosity rocks, permeability is likely to increase as the rock is subjected to
shear-induced dilation. This increase may be related to the strains in the rock, the dilating
strain in a given direction being associated with the width of the crack created.
In case of ductile failure of the reservoir rock (very unlikely in situ), no permeability
reduction law is available.
Currently therefore, there is no generic law describing the change in intrinsic permeabil-
ity of a reservoir rock as a function of the stress path.

3.1.2 Coupling between geomechanical and reservoir models

3.1.2.1 Conventional reservoir approach

Reservoir simulations are based on a reservoir model which describes the reservoir
geometry, heterogeneities of petrophysical properties in the reservoir, thermodynamics of
the reservoir fluids, flows in porous media, etc. The models used to predict the production of
reservoir fluids are more or less complex: single-phase model, Black-Oil model,
compositional model, treatment of thermal effects, case of flows in fractured reservoirs.
Traditionally in reservoir engineering, the mechanical behaviour of the reservoir is ignored
and modelling focuses on the fluid flows in the reservoir layers. In the simplest case of a
single-phase isothermal flow, the reservoir simulator solves an equation similar to the
hydraulic diffusivity equation (3.1.5). In this approach, the rock permeability and
compressibility are generally constant, which amounts to neglecting the influence of the
rock strains and setting a priori the stress path followed (generally with uniaxial or isotropic
 Chapter 3 • Geomechanics and Reservoir 141

strain) for each mesh of the reservoir model [Boutéca, 1992]. Some simulators nevertheless
allow these parameters to vary with the reservoir fluid pressure.

3.1.2.2 Limitations of the conventional reservoir approach

The pressure, temperature and saturation variations generated by exploitation modify the
distribution of stresses in the reservoir. The stress variations associated with reservoir
production may be extremely high in poorly consolidated reservoirs and fractured and faulted
reservoirs. Consequently, the porosity, compressibility and permeability of the reservoir rock
are likely to vary while the reservoir is producing. While conventional reservoir simulators
assume that these parameters are constant or depend only on the fluid, in actual fact they
depend on the stress path followed by the reservoir rock. The pressure dependence of porosity
imposed via the rock compressibility introduced in conventional reservoir simulators cannot
alone describe the behaviour of the reservoir rock during production [Gutierrez and Lewis,
1998]. To obtain a precise description of the change in reservoir porosity and permeability, the
geomechanical behaviour of the reservoir during production must be modelled.
https://telegram.me/Geologybooks

The main geomechanical effects affecting production are the compaction and associated
increase in the recoverable quantities of fluid. Obviously, these effects are more noticeable
for highly compactable reservoir rocks (poorly cemented sandstones generally qualified as
poorly consolidated, highly porous chalks, etc.). Compaction is due to the decrease in reser-
voir pressure and the associated increase in effective compressive stresses in the reservoir.
The mechanical behaviour of poorly consolidated reservoir rocks is complex, being highly
nonlinear and depending on the stress path and the temperature. With highly porous chalks,
the mechanical behaviour of the reservoir rock may also be considerably affected by varia-
tions in water saturation (“water weakening”) [Hermansen et al., 2000; Homand, 2000;
Matà, 2001]. Compaction phenomena play a significant role in the production of reservoir
fluids, as demonstrated by the Bachaquero fields [Merle et al., 1976] and Zuata [Charlez,
1997] in Venezuela. For highly compactable reservoirs, therefore, it is essential to predict
the change in reservoir porosity during production using strain or stress data provided by a
geomechanical model associated with the reservoir model.
The stress variations resulting from reservoir production modify the porous structure of
the rock and therefore its permeability. The permeability of highly consolidated and natu-
rally fractured or faulted reservoirs is extremely sensitive to stress changes during produc-
tion. The thermoporoelastic effects in these reservoirs may modify the conductivities of the
fractures and make waterflooding much less effective [Guttierez & Makurat, 1997]. For
these reservoirs, the thermoelastic effects are generally concentrated around injector wells,
where the temperature variations are higher, whereas hydromechanical effects affect the
entire reservoir. The thermohydromechanical behaviour of the fracture system is complex.
Macroscopic behaviour depends on the density and orientation of the fracture system, the
initial stress state and the stress path in the reservoir during production. Due to the high elas-
tic rigidity of the matrix (rock intact), the strains are mainly located on the fracture planes,
thereby generating a change in their hydraulic conductivity [Koutsabeloulis et al., 1994;
Heffer et al., 1994]. Geomechanical modelling of the reservoir is therefore required to esti-
mate the permeability changes in the reservoir, whether they are related to changes in the
porous structure or to activation of fractures or faults.
 142 Chapter 3 • Geomechanics and Reservoir

3.1.2.3 Geomechanical approach

Hydrocarbon production modifies the stress and strain fields around the reservoir, so the
geomechanical model must therefore take into account the behaviour of the surrounding for-
mations. The reservoir overburden is generally modelled up to ground level, where a free
surface condition is imposed. The sideburden and underburden taken into account in the
geomechanical model are assumed to be quite extensive so that the zero normal displace-
ment conditions can be imposed at the boundaries of these surfaces. Osorio et al. (1998)
demonstrated that the conditions at the mechanical limits imposed on the boundaries of the
surrounding formations influenced the permeability changes in the reservoir.
To take into account the above-mentioned mechanical effects, the equations of equilibrium
in the reservoir and its surrounding formations must be solved, in addition to the hydraulic
problem. The equations of this problem can be obtained from the poroelastic constitutive
laws proposed by Coussy (1995) for an isotropic material. Using the tensorial notation and
the convention of mechanics of continuous media (negative compression), these equations
https://telegram.me/Geologybooks

can be written as:

2
∂σ = (K0 − G)∂ε1 + 2G ∂ε − b∂p1 (3.1.11)
3
1 1
∂p = − b∂ε + ∂m (3.1.12)
M ρ0

with M Biot’s modulus, b Biot’s coefficient, G the rock shear modulus and K0 the drained
bulk modulus of the rock. σ and ε designate respectively the stress and strain tensors, 1
the Kronecker symbol, p the pore pressure and m the fluid mass intake with respect to the
initial total volume. The first poroelastic constitutive law is introduced in the mechanical
equilibrium equation which, in the absence of body force, is written:
div σ = 0 (3.1.13)
The strain tensor is therefore expressed as a function of the displacement vector u . The
previous equation therefore becomes:
G
G Δu + ( K0 + )grad u = b grad p (3.1.14)
3
In addition, the fluid mass intake m with respect to the fluid initial volume is expressed
from the mass balance and Darcy’s law as:
∂m ⎛ k ⎞
∂t
( )
= − div ρφ v = − div ⎜ ρ grad
⎝ μ
p⎟
⎠
(3.1.15)

The latter equation is introduced in the second poroelastic constitutive law, to give after
linearisation:
1 ∂p k ∂ε
− Δp = − b v (3.1.16)
M ∂t μ ∂t
with εv = tr ε = div u , the volume strain.
 Chapter 3 • Geomechanics and Reservoir 143

The geomechanical problem therefore consists in solving the following system of equa-
tions system which describes the flows and elastic strain of the porous medium [see also
Boutéca, 1992; Coussy, 1995; Lewis and Schrefler, 1998]:

G
G Δu + ( K0 + )grad u = b grad p (3.1.17)
3

1 ∂p k ∂ε
− Δp = − b v (3.1.18)
M ∂t μ ∂t

The pore pressure p is the main unknown of equation (3.1.18) and the displacement u is
the main unknown of equation (3.1.17). The displacement u appears in the right member of
the “hydraulic” equation (3.1.18) via the volume strain. Similarly, the fluid pressure modi-
fies the mechanical equilibrium via the right member of equation (3.1.17). Equations
(3.1.17) and (3.1.18) are therefore coupled.
https://telegram.me/Geologybooks

3.1.2.4 Coupling between fluid flow and geomechanics

To identify the various levels of coupling between the hydraulic and mechanical aspects, we
examine the discretised form of the system of equations (3.1.17) and (3.1.18). After discreti-
sation in space and time, the system of equations (3.1.17) and (3.1.18) is written in the fol-
lowing matrix form [Settari and Walters, 1999]:

⎡K L ⎤ ⎡ Δt u ⎤ ⎡ F ⎤
⎢ T ⎥⎢ ⎥=⎢ ⎥ (3.1.19)
⎢⎣ L E ⎥⎦ ⎢⎣ Δt p ⎥⎦ ⎣ R ⎦

with:
– u the discretised mechanical unknown,
– p the discretised pressure unknown,
– Δt the time difference operator,
– K the mechanical rigidity matrix or stiffness matrix,
– L the coupling matrix between mechanical and hydraulic unknowns,
– F the force vector,
– E the hydraulic flow matrix in porous medium,
– R the source term of the flow problem.
The hydraulic flow matrix E is generally decomposed in the form E = T – D, where T is a
transmissivity matrix and D the accumulation diagonal matrix. It is important to note that in
the case described, matrices K and E are considered as linear operators whereas for a more
realistic problem (elastoplastic behaviour, compaction, multiphase flow, etc.), these opera-
tors are nonlinear.
Various coupling levels can be considered when solving the matrix system (3.1.19):
– Conventional reservoir approach: in this approach, the coupling term in the second
equation of matrix equation (3.1.19) is reformulated as a function of the pressure
variation so that matrix D uses the rock and fluid compressibilities. The limitations of
 144 Chapter 3 • Geomechanics and Reservoir

this approach have already been discussed. It can nevertheless be used to obtain an
estimation of the stress variations when the pressure variations calculated by the reser-
voir simulator are transmitted to the second member of the first equation of matrix
equation (3.1.19) via a geomechanical simulator. In this case, we speak of one-way
coupling or decoupling.
– Partially coupled: in this approach, the equations of matrix system (3.1.19) are
solved separately on different simulators. A reservoir simulator solves the hydraulic
problem and a mechanical simulator the geomechanical problem. Coupling is carried
out by exchanging results between the two simulators at different exchange periods.
In practice, coupling consists in calculating the pressure field p from the displacement
field u calculated during the previous geomechanical simulation with the equation:
E Δt p = R − LT Δt u (3.1.20)

This first calculation is carried out with a conventional reservoir simulator using a
https://telegram.me/Geologybooks

porosity correction calculated from a mechanical variable [Settari and Mourits 1994,
Settari and Mourits 1998, Bévillon 2000, Mainguy and Longuemare 2002]. Then, using
the pressure field found using (3.1.15), the displacement vector is calculated from:
K Δt u = F − L Δt p (3.1.21)

This second calculation is carried out using a geomechanical simulator in order to

apply the reservoir model pressure variations as external load. This type of coupling is
said to be explicit if it is not iterated up to convergence of mechanical and hydraulic
unknowns over the same period [Fung et al. 1992, Tortike and Farouq Ali, 1993,
Settari and Walters, 1999] In contrast, we speak of iterative coupling if the process is
iterated up to convergence of unknowns [Settari and Mourits 1994, Chin and Thomas,
1999]. The main advantage of loose coupling is that reservoir and geomechanical
simulators, already highly developed in their respective domains, can be reused. This
method also allows the use of different time and space discretisations for the
hydraulic and geomechanical problems.
– Fully coupled: in this approach, equations (3.1.19) are solved simultaneously in the
same simulator. The space discretisation method is then the same for both equations
and may be either finite difference type [Stone et al., 2000] or finite element type
[Lewis and Schrefler 1998, Koutsabeloulis and Hope 1998]. This approach provides a
consistent solution to the coupled problem but involves solving all equations at each
time step.

3.1.2.5 Example of external coupling in porosity on a highly compactable reservoir

This section illustrates the iterative loose coupling methodology where the porosity injected
in the reservoir model is calculated using strains calculated by a geomechanical model. The
results presented are extracted from the thesis of D. Bévillon (2000) and obtained from the
SARIPCH reservoir simulator and ABAQUS finite element software on a synthetic reservoir
case. The example illustrates the depletion of a highly compactable homogeneous reservoir,
of simplified geometry with a producing well located in its centre. The geometry of the
 Chapter 3 • Geomechanics and Reservoir 145

reservoir and its finite volume mesh are shown on Figure 3.1.1. The finite element mesh
used in ABAQUS includes the reservoir and the side, over and underburdens (Figure 3.1.2).
The simplified reservoir geometry and the choice of finite volume and finite element
discretisations simplify data exchange between the two models.

1000 m

Borehole
https://telegram.me/Geologybooks

100 m

3000 m

Figure 3.1.1
Reservoir mesh composed of 55 finite volume cells.

7000 m

3000 m

100 m
900 m

9000 m

Figure 3.1.2
Geomechanical mesh composed of 1296 Q20 finite elements.
 146 Chapter 3 • Geomechanics and Reservoir

The reservoir pressure is initially 48 MPa and the reservoir is produced from the central
producing well whose bottomhole pressure is fixed at 10 MPa. The reservoir fluid is
assumed to be poorly compressible. The reservoir porosity is initially high (40%) and the
mechanical behaviour is assumed to be nonlinear and represented by a Cam-Clay model.
The behaviour of the surrounding formations is represented by a linear poroelastic model
which provides the changes in pressure and stress field. Porosity and permeability are low in
the surrounding formations.
https://telegram.me/Geologybooks

Figure 3.1.3
Deforming finite element mesh after 33 years production.

U3 Value
-7.93E-01
-7.32E-01
-6.71E-01
-6.10E-01
-5.49E-01
-4.88E-01
-4.27E-01
-3.66E-01
-3.05E-01
-2.44E-01
-1.83E-01
-1.22E-01
-6.10E-02
+2.73E-23

Figure 3.1.4
Vertical displacement isovalues on the surface after 33 years production.
 Chapter 3 • Geomechanics and Reservoir 147

The results of the coupled simulations indicate compaction of the reservoir rock resulting
as surface subsidence (Figures 3.1.3 and 3.1.4). After 33 years production, the vertical
displacement reaches a maximum value of 75 cm on the surface and in the centre of the
reservoir. The hydromechanical calculations provide the change in pore pressure in the
surrounding formations. These results indicate overpressure in the area of the surrounding
formations located at the periphery of the reservoir, reaching 7 MPa afters 12 years
production (Figure 3.1.5). This overpressure is due to stress transfer from the reservoir to the
area of the surrounding formations in contact with the reservoir, since their rigidity is higher
than that of the reservoir. This phenomenon is generally called “arch effect” or “stress
arching”, [Osorio et al., 1998]. This stress transfer increases the pore pressure near the
reservoir, which cannot be dissipated due to the low permeability of the surrounding
formations.
https://telegram.me/Geologybooks

Figure 3.1.5
Pore pressure isovalues in the reservoir and sideburdens after 12 years
production.

It is worthwhile comparing coupled geomechanical-reservoir simulations with conven-

tional reservoir simulations to determine the impact of the reservoir mechanical behaviour
on production. The previous results were therefore compared with conventional reservoir
simulations for which various reservoir rock compressibility values were used. The varia-
tions laws of compressibility as a function of pore pressure which have been tested are
shown on Figure 3.1.6. The “Cp elastic” and “Cp oedometric” compressibilities are obtained
for an oedometric loading path assuming that the reservoir rock behaviour is respectively
elastic and elastoplastic (modified Cam-Clay model), the “Cp mean compressibility” corre-
sponding to the mean value between the maximum elastoplastic compressibility and the
elastic compressibility. Figure 3.1.7 compares the recovery rates obtained for the various
simulations. We observe widely differing changes between the conventional reservoir simu-
lations and the coupled geomechanical-reservoir model. The latter calculation predicts far
 148 Chapter 3 • Geomechanics and Reservoir

higher recovery than the reservoir calculations at start of simulation (which indicates that the
reservoir calculation underestimated the effects of pore compression), then lower recovery
after a long period of time (which indicates that the reservoir calculation overestimated the
effects of pore compression).

16
Cp oedometric local
14
Cp mean
Cp elastic
12
Cp (1.E-9Pa-1)

10

6
https://telegram.me/Geologybooks

0
0 5 10 15 20 25 30 35 40 45 50
Pore pressure (MPa)

Figure 3.1.6
Various pore compressibilities used in conventional reservoir calculations.

24
22 Partial coupling
Cp elastic
20
Cp mean
18 Cp elastoplastic
Recuperation (%)

16
14
12
10
8
6
4
2
0
0 2000 4000 6000 8000 10000 12000
Time (day)

Figure 3.1.7
Comparison of the coupled calculation with various conventional reservoir cal-
culations using various compressibilities [Bévillon, 2000].
 Chapter 3 • Geomechanics and Reservoir 149

3.1.2.6 Example of coupling in permeability

This section illustrates the explicit loose coupling methodology with the reservoir model per-
meability updated using the geomechanical model of a faulted reservoir. The results pre-
sented were obtained using the ATHOSTM reservoir simulator and the VISAGETM geome-
chanical simulator [Longuemare et al., 2002]. The reservoir studied is a highly heterogeneous
and compartmentalised limestone reservoir located in the Middle East. The geomechanical
study was conducted to find out whether the changes in effective stresses and temperature
resulting from production and injection of cold water could explain premature watercuts.
https://telegram.me/Geologybooks

Figure 3.1.8
Reservoir geometry [Longuemare et al., 2002].

Figure 3.1.8 shows the reservoir geometry taken into account in the reservoir simulator. The
reservoir model is a simple porosity model including faults whose permeability is described by
transmissivity multipliers. The faults taken into account in the geomechanical model associated
with the reservoir model are shown on Figure 3.1.9. Figure 3.1.10 shows the distribution
between the two types of rock identified in the reservoir according to their dolomite content.
The first type, with low dolomite content, has smaller drained Young’s modulus and effective
cohesion (Eo = 12 GPa, C' = 4 MPa) than the second (Eo = 24 GPa, C' = 11.3 MPa).
The reservoir simulations model the thermohydraulic behaviour of the reservoir over a
period of 5 years. The geomechanical model calculates the associated stress variations and the
strains normal and tangential to the fault planes (Figure 3.1.11). The latter are related to the
permeability changes in the faults via a phenomenological law [Koutsabeloulis and Hope,
1998]. The permeability changes in the faults are then used to update the fault transmissivities
included in the reservoir model. Figures 3.1.12 and 3.1.13 show the fault transmissivity
multipliers of the reservoir model after 5 years production. The activated faults (whose
permeability is increasing) correspond to the faults on Figure 3.1.11 stressed due to the pressure
and temperature variations in the reservoir and their orientation with respect to the direction of
maximum stress.
 150 Chapter 3 • Geomechanics and Reservoir

Figure 3.1.9
Faults of the geomechanical model [Longuemare et al., 2002].
https://telegram.me/Geologybooks

Figure 3.1.10
Distribution of the two types of rock in the reservoir (dark blue E = 12 GPa,
light blue E = 24 GPa) [Longuemare et al., 2002].

Model: TFEA00
L005: Time/Months 59,90
Nodal 1-STRN Normal
Max = .43E-4 Min = 0

.391E-4
.352E-4
.313E-4
.274E-4
.234E-4
.199E-4
.156E-4
.117E-4
.781E-5
.391E-5

Figure 3.1.11
Strain isovalues normal to the fault planes after 5 years production
[Longuemare et al., 2002].
 Chapter 3 • Geomechanics and Reservoir 151

Figure 3.1.12
Transmissivity multiplier in the x direction [Longuemare et al., 2002].
https://telegram.me/Geologybooks

Figure 3.1.13
Transmissivity multiplier in the y direction [Longuemare et al., 2002].

3.1.3 Stress path followed by the reservoir during depletion

Primary production of hydrocarbons leads to a decrease in reservoir pressure. Depletion
increases the effective stresses, which finally generates strains inducing modifications of the
mechanical parameters and petrophysical characteristics of the rock (note: in this chapter,
the stresses are taken to be positive in the direction of compression).
The in situ stress changes obviously take place in all directions. Since the minimum
horizontal stress can be measured (see Chapter 1.4), the response of a reservoir during
depletion is generally described by the ratio between the variations in minimum horizontal
stress sh and reservoir pressure pp, called the stress path [Addis et al., 1997, 1998] [Kenter
et al., 1998] [Hettema et al., 1998] [Santarelli et al., 1998]:
 152 Chapter 3 • Geomechanics and Reservoir

Δσ h
γ = (3.1.20)
Δpp
with Δpp < 0 for depletion. Note that expression (3.1.20) is not based on any behaviour or
boundary condition assumption.
Some authors adopt a different definition of the stress path, expressed as a function of
the variations in minimum effective horizontal stress sh' and effective vertical stress sv'
[Rhett and Teufel 1992a] [Schutjens and de Ruig 1996] [Hettema et al., 1998] [Schutjens
et al., 1998] [Yale and Crawford, 1998]. Giving, assuming elastic behaviour,
Δσ 'h Δ(σ h − bpp )
K= = (3.1.21)
Δσ 'v Δ(σv − bpp )
where b is Biot’s coefficient.
If we assume that the axial stress does not vary during depletion (no stress arching), we
https://telegram.me/Geologybooks

obtain the following relation:

γ = b(1 − K ) (3.1.22)

3.1.3.1 In situ measurements and assumption of uniaxial compaction

The assumption most frequently accepted to estimate the stress path of a reservoir during
depletion based on the reservoir rock characteristics is uniaxial reservoir compaction, i.e.
with no horizontal strain. If in addition we adopt an isotropic poroelastic behaviour model
for the reservoir rock, we obtain:
1 − 2 vo
γ =b (3.1.23a)
1 − vo

vo
K= (3.1.23b)
1 − vo
where no is the drained Poisson’s ratio.
In situ measurements taken on different fields (sandstones and chalks) show that the
Δσ h
stress path γ = followed by a reservoir during depletion varies between 0.2 and 1
Δpp
[Addis et al., 1998] and may be quite different from the path predicted assuming uniaxial
compaction. Thus, for the Groningen field (consolidated sandstone with 15%-20%
porosity), the stress path estimated using in situ measurements is γ = 0.4 , to be compared
with the value of 0.8 obtained using equation (3.1.23a) [Hettema et al., 1998]. The compila-
tion of 20 years in situ data gives a stress path K = 0.2 for the Ekofisk field, whereas uniax-
ial compaction tests in laboratory lead to values ranging from K = 0.4 for chalks in the Tor
formation, to K = 0.5 − 0.6 for the Lower Ekofisk formation (Rhett and Teufel, 1992a).
 Chapter 3 • Geomechanics and Reservoir 153

It is important to note that use of expressions (3.1.23) to predict the stress path during
depletion from laboratory measurements is based on strong assumptions:
– uniaxial reservoir compaction,
– elastic behaviour of the rock during depletion,
– homogeneous drained Poisson’s ratio at reservoir scale, identical to that measured at
laboratory scale.

Various factors influencing reservoir behaviour during depletion question the above
assumptions; they are described in the next paragraph. From the point of view of scale
effects between the laboratory samples and the reservoir, note for example that in order to
calibrate the predicted subsidence values with in situ data, the rock stiffness obtained in the
laboratory must be divided by 2 for Norwegian chalk reservoirs and, inversely, multiplied
by 2 for the Groningen field sandstone [Santarelli et al., 1998]. Upscaling from laboratory
data to reservoir scale is a central problem, and one of increasing interest to rock mechanics
engineers in the petroleum community.
https://telegram.me/Geologybooks

3.1.3.2 Factors influencing reservoir behaviour during depletion

Influence of surrounding formations and reservoir geometry

Analytical and numerical models taking into account the reservoir and its surrounding
formations (upper, lower and side layers) have demonstrated that the rocks surrounding the
reservoir could have a significant influence on its behaviour during depletion. Refer to
[Bévillon, 2000] for a numerical approach combining a reservoir model, describing the flow
of the various fluid phases, and a geomechanical model, providing a rigorous mechanical
framework to describe compaction.
For a reservoir represented by an ellipsoidal inclusion, Addis et al. (1998) calculated that
the stress path g decreases when the ratio between the drained Poisson’s ratios of the reser-
voir rock and the host rock increases (all other data remaining constant). For a disc-shaped
reservoir of radius Rr, included in a host rock with a lateral condition of zero horizontal
strains at radius Rh in the host rock, they obtain:

Rr
– for = 0.1, a stress path γ = 0.84 independent of the ratio between the drained
Rh
Young’s moduli of the reservoir rock and the host rock;
Rr
– for = 0.01, a stress path varying between γ = 0.77 for E0r = E0h and γ = 0.93
Rh

E0h
for E0r = .
20
Kenter et al. (1998) generated a 3D mesh of the Shearwater field using seismic data,
producing a better representation of the true geometry of the reservoir, the surrounding
 154 Chapter 3 • Geomechanics and Reservoir

layers, and the fractures. They demonstrated that, due to the special geometry of the
reservoir studied (inclined and elongated), depletion induces stress arching which reduces
the stresses at the top of the reservoir and increases them at its boundaries.

Plastic behaviour of the reservoir rock

In case of reservoir repressurisation, the stress path may be quite different from the path fol-
lowed during depletion, thereby indicating stress path irreversibility. This is the case in par-
ticular of the North Sea poorly consolidated sand reservoir studied by Santarelli et al. (1998):
they obtain gdepletion = 0.4 or 0.7, depending on the zone considered and grepresurisation ≈ 0 (i.e.
Δσ h ≈ 0). The Venice bay follows a stress path during repressurisation one order of magni-
tude lower than the stress path during depletion.
Kenter et al. (1998) represented the behaviour of the Shearwater field using a Cam-Clay
model. The simulations carried out demonstrate that the curve showing the change in mini-
mum horizontal stress as a function of the pressure decrease changes slope when the reser-
https://telegram.me/Geologybooks

voir rock elastic limit is reached. The stress path (calculated according to the stress and pres-
sure state before depletion) changes from g = 0.7 – 0.8 when the plasticity criterion is
reached to g = 0.3 – 0.4 at end of depletion.

Influence of fractures
The presence of fractures may result in a non-homogeneous stress path in the reservoir. San-
tarelli et al. (1998) report, in particular, measurements taken on a North Sea poorly consoli-
dated sand reservoir which reveal the existence of two distinct zones: the first zone follows a
stress path g ≈ 0.7, while the second follows a lower stress path g ≈ 0.4 due to the presence of
a greater number of fractures (faults observable on the seismic data).

3.2 ENVIRONMENTAL ASPECTS

3.2.1 Subsidence

3.2.1.1 Introduction and famous examples

Subsidence is the lowering of the ground surface due to natural causes or induced by human
activity.
Natural subsidence may be caused by lowering of the lithosphere. In this case, there are
three types of associated subsidence:
– tectonic subsidence, which may be related to crustal thinning (mainly tensile stress
regime) and synclinal folding (mainly compressive stress regime) [Perrodon, A.,
1989]. This type of subsidence is encountered in rifts and active margins;
– thermal subsidence, resulting from thermal contraction of the lithosphere;
– flexural subsidence, related to a topographic load acting on the lithosphere (e.g. pre-
sence of relief) and generating local bending of the lithosphere.
 Chapter 3 • Geomechanics and Reservoir 155

In addition to lowering of the lithosphere, other natural phenomena may contribute to

natural subsidence. This is the case in particular of natural compaction of sediments
resulting from their consolidation with expulsion of the interstitial water. This compaction,
which maybe particularly pronounced in case of coals and clays, may generate subsidence of
the ground surface. Dissolution of underground formations may also contribute to the
appearance of subsidence. For example, dissolution or carbonate rocks in contact with
rainwater loaded with carbonic acid or dissolution of a salt layer by an unsaturated brine.
Natural subsidence nevertheless remains a slow phenomenon, the annual rate of
subsidence generally varying from a fraction of a millimetre to a few millimetres per annum.
Subsidence related to human activity is generally faster. It may be caused by mining
(extraction of coal, salt or other minerals) or generally by the production of in situ fluids. In
the latter case, we can distinguish between exploitation of aquifers and production of hydro-
carbons from a reservoir. Generally speaking, measurement of anthropogenic subsidence
includes natural subsidence. This chapter focuses on subsidence caused by the production of
https://telegram.me/Geologybooks

underground fluids. The articles by Christensen et al. (1988) and Boutéca et al. (1996)
report the main cases of subsidence related to pumping in aquifers and production of hydro-
carbons. Some of the most famous cases of subsidence will be mentioned below.

A. Subsidence in the city of Venice, Italy

The city of Venice is slowly sinking with respect to sea level, a problem due to three phe-
nomena: the rise in the average level of the Adriatic sea (eustatism), natural subsidence
related to consolidation of deposits in the lagoon substrate, and subsidence induced for
exploitation of the aquifer for industrial purposes. The amplitudes of these phenomena differ
considerably [Gatto and Carbognin, 1981]. From 1908 until 1980, the average sea level rose
by 9 cm while during the same period, natural subsidence was about 3 cm. Exploitation of
the aquifer started in 1925, resulting in maximum anthropogenic subsidence of about 14 cm
in 1969. Pumping ceased in 1969, arresting subsidence and causing elastic rebound of the
ground surface of about 2 cm in 1975. Due to these three phenomena, the sea level rose by
about 23 cm, a highly critical situation for Venice whose average altitude is about 1 metre
above sea level [Lewis and Schrefler, 1998].

B. Subsidence of San Joaquin valley, California, United States

Subsidence of the land surface in San Joaquin valley, California (USA), is mainly due to inten-
sive pumping of the aquifer to irrigate farmland. This subsidence is worsened by subsidence
related to the compaction of surface sediments which had little exposure to water after deposi-
tion, and which compact in contact with irrigation water. This subsidence varies between 1.5 to
3 metres and may locally exceed a value of 4.5 metres [Bull, 1964; Lofgren, 1975]. Subsidence
due to the production of oil and gas reservoirs (less than 1 metre) and tectonic subsidence con-
tribute to a lesser extent to lowering the ground surface. Subsidence due to extraction of ground-
water started in the 1920’s, reaching a maximum level of 9.1 metres in 1981 (Figure 3.2.1). In
1970, 13,500 km2 of the San Joaquin valley were affected by subsidence, exceeding 0.3 m over
an area of approximately 11,000 km2 [Poland and Lofgren, 1984]. This subsidence is mainly
due to compaction of highly compressible clay beds, resulting from the increase in effective
 156 Chapter 3 • Geomechanics and Reservoir

stress associated with the pressure drop in the aquifer. Although the ground surface subsided
slowly, there are numerous negative effects: fractured well casings, damage to roads and canals
and irrigation networks due to differential subsidence associated with high horizontal strains.
https://telegram.me/Geologybooks

Figure 3.2.1
Joseph Poland beside a telephone pole in San Joaquin valley approximately
marking the position of maximum subsidence in 1981
(Photo: Galloway et al., 1999, p. 23).

C. Subsidence of the Bachaquero, Lagunillas and Tia Juna fields in Venezuela

These three heavy oil fields are located on the east coast of Lake Maracaibo, still known as the
Bolivar coast. Production of the Lagunillas field started in 1926 and the first signs of
subsidence, flooding due a breech in the dyke, were observed in 1929. Subsidence above the
three fields affects an area of about 450 km2; maximum subsidence of 4 to 5 metres in 1986
was measured above each field. Subsidence above the Lagunillas field is the most problematic
since there is significant subsidence near the coastal dyke [Finol and Z. A. Sancevic, 1995].
The dyke has had to be raised on several occasions and a drainage and pumping network has
been installed to evacuate the rainwater which may build up in the subsidence bowl.
 Chapter 3 • Geomechanics and Reservoir 157

D. Subsidence of the Wilmington field, California, USA

The Wilmington reservoir is an oil field consisting of seven producing zones divided by
5 main faults. Intensive production of the field started in 1941 and 1942, leading to maxi-
mum subsidence of 9 metres in 1968 and affecting a total land area of 75 km2. Water injec-
tion started in 1958, reducing the rate of subsidence from 75 cm/year in 1952 to 0 cm/year in
1968. Subsidence generates large horizontal displacements. On the Wilmington field, a hor-
izontal displacement of 3.66 metres was measured between 1937 and 1966. The conse-
quences of ground subsidence were extremely expensive for the town of Long Beach, esti-
mated at 100 million dollars in 1962 [Marsden and Davis, 1967]. In particular, subsidence
caused major damage to the roads and buildings in the town, to the production wells, and
flooding. Dykes and retention walls had to be built and structures raised. The subsidence
also created earth tremors of medium amplitude (2.4 to 3.3) due to the sudden release of
shear stresses in the soft shaly levels.
https://telegram.me/Geologybooks

E. Subsidence of the Groningen field, The Netherlands

Groningen, a vast 900 km2 gas field in The Netherlands, went into production in 1963. The
subsidence area is an ellipsoid of 56 ¥ 70 km2; subsidence measured about 8 cm in 1982 for
a depletion of about 14 MPa [Hettema et al., 2002]. Subsidence of the Groningen field is
difficult to manage due to the low ground level and the proximity of the sea, the maximum
possible subsidence being 25 cm to 30 cm. Greater subsidence would involve digging a
very expensive draining channel. The Groningen subsidence is a striking example of how
important it is to monitor subsidence measurements and predict long-term subsidence. An
extensive measuring system has been set up to monitor the compaction and subsidence. The
subsidence monitoring system is based in particular on 8 annual levelling measurements
over a large area and 14 wells wired down to a depth of 400 metres measuring the compac-
tion of Tertiary and Quaternary sediments. Prediction of the subsidence above the field
proved difficult. The first estimation which predicted 90 cm subsidence in 2000 soon turned
out to be incorrect, since the measurements were three times less than the predictions
[Geertsma, 1989]. The prediction models were later improved to take into account the
effects of loading rate on the compaction of sandstones and their degree of cementation
[De Waal and Smith 1988, Van Hasselt 1992]. The difference between prediction and mea-
surement was resorbed by these improvements and subsidence at end of operation was rees-
timated at 60-70 cm.

F. Subsidence of the Ekofisk field, North Sea, Norway.

The Ekofisk reservoir includes the Ekofisk (Danian) and Tor (Cretaceous) formations, com-
posed of highly porous chalk (frequently above 40%) and separated by a low-porosity layer
(tight zone). The high porosity of the chalk at the extreme depths of the formations can be
explained by the abnormally high pressure in the reservoir. The Ekofisk and Tor formations
exhibit low matrix permeability of the order of the millidarcy, and the productivity of the
formations is due to a network of natural fractures. The reservoir overburden is heteroge-
neous, but consists mainly of clay and shales. Production from the oil field, which started in
1971, increased significantly from 1974.
 158 Chapter 3 • Geomechanics and Reservoir

The first subsidence problems appeared end 1984, maximum subsidence was about 3
metres at end 1985, with a subsidence rate of 40 to 45 cm/year. Numerous subsidence-
induced casing failures were observed in the reservoir and the overburden [Yudovich and
Chin, 1989]. In addition, due to the marine environment the subsidence consequences were
more expensive. During the summer of 1987, seven platforms had to be raised 6 metres due
to the significant decrease in the air gap between the platform and the sea [Sulak and
Danielsen, 1989]. Seabottom subsidence results from reservoir compaction associated with
elastic then plastic strains (pore collapse). This strain generates a high irreversible reduction
of the porosity when the pore pressure decreases [Johnson et al. 1989]. Reservoir subsidence
may be increased by consolidation of clays in the overburden. Several reinjection operations
were carried out to limit the subsidence. Some of the gas produced was reinjected from 1985
to 1987 [Sulak and Danielsen, 1989]. While this limited the subsidence rate to about 26 cm/
year, it had to be stopped for financial considerations.
Water injection was then carried out to stabilise the pressure in the reservoir and enhance
hydrocarbon recovery. Water injection was first started in the Tor formation from end 1987
https://telegram.me/Geologybooks

then extended to the Ekofisk formation from 1994. In 1996, the reservoir pressure was
perfectly stabilised (even very slightly increased) although reservoir subsidence continued,
reaching a maximum value of about 7 metres (Figure 3.2.2). This permanent strain of the
chalk at constant pressure was put down to water weakening (weakening caused by the
increase in water saturation of the chalk) [Maury et al., 1996; Sylte et al., 1999;
Matà, 2001].

Reservoir depletion (MPa)

0 5 10 15 20 25
0
1971 1979
100
1981
200
Subsidence (cm)

300
1986
400

500

600
1996
700

800

Figure 3.2.2
Subsidence as a function of depletion for the Ekofisk field (Hetema et al.
2002).
 Chapter 3 • Geomechanics and Reservoir 159

3.2.1.2 Analysis of subsidence related to the production of underground fluids

Production of underground fluids is generally accompanied by a significant decrease in the
pressure of the fluids saturating the reservoir rock pores. This decrease in pore pressure
increases the effective stress imposed on the porous medium. This leads to settling or com-
paction of the reservoir rock, which may be quite considerable in case of highly porous, brit-
tle or poorly cemented, highly compressible rocks and plastically deformable rocks. The
parameters influencing reservoir compaction are the pressure drop in the reservoir, its thick-
ness and mechanical behaviour.
While reservoir compaction is a determining factor in the subsidence process, it is not
the only one. If, for example, the reservoir is covered by an overburden which is rigid and
therefore difficult to deform, the displacements of the reservoir layers resulting from reser-
voir compaction are totally absorbed by the overburden and no subsidence is generated at
the surface. Subsidence only appears if the geometric characteristics (especially thickness)
and mechanical characteristics of the overburden allow transmission of the reservoir dis-
https://telegram.me/Geologybooks

placements up to the land surface [Christensen et al., 1988].

Large, shallow reservoirs covered by an overburden of very low rigidity are therefore
most prone to subsidence. Since compaction is generally irreversible, this is also true of sub-
sidence. If no other phenomena are involved, maintaining pressure in the reservoir will stop
compaction and subsidence. Generally, however, repressurising the reservoir rock will only
reduce the subsidence very slightly (elastic rebound) since an increase in reservoir pressure
is accompanied purely by elastic strains (Settari, 2002, for example).
Figure 3.2.3 is a simplified representation of the horizontal (Uh) and vertical (Uv) dis-
placements associated with subsidence. The generally dish-shaped subsidence area is called a
subsidence bowl. The vertical displacement is zero at large distances from the subsidence
zone, progressively increasing up to is maximum value at the centre of the subsidence zone.
The horizontal displacement is directed towards the point of maximum subsidence. The max-
imum horizontal displacement is located about halfway up the flanks of the subsidence bowl,
from where it decreases progressively to zero at both the centre and periphery of the bowl.

Uh

Uv

Figure 3.2.3
Idealised representation of the horizontal and vertical displacements on the
surface of a subsidence zone.
 160 Chapter 3 • Geomechanics and Reservoir

A delay of a few months to a few years is frequently observed between the pressure drop
in the reservoir and its consequences on the surface in terms of subsidence. As shown on
Figure 3.2.4, the linear poroelastic model cannot account for this delay, which must there-
fore be attributed to nonlinearity. This delayed effect can be explained by the usual vis-
coelastic behaviour of the rocks or by compaction of possible clay levels in the reservoir.
Due to the delay between reservoir depletion and the start of subsidence, long-term
monitoring of the displacements will be required after production stoppage and a subsidence
measuring system will therefore have to be maintained for many years.

Linear extrapolation

Field measurements
Subsidence
https://telegram.me/Geologybooks

Subsidence-depletion
delay

Prognosis based on
linear elasticity

Pressure depletion

Figure 3.2.4
Diagram of the subsidence-depletion delay effect [Hetema et al., 2002].

The main risks associated with subsidence are illustrated in the above-mentioned
examples. Subsidence is a serious problem for offshore platforms (reduced air gap) and near
seas and lakes (risk of flooding). In addition, Boutéca et al. (1996) point out that the risks
are highly dependent on the environment: ten centimetres subsidence is critical at Gronin-
gen, for example, but quite acceptable at other fields. Vertical subsidence is associated with
horizontal displacements which may damage buildings and surface facilities. The UK
National Coal Board classifies damage according to the size of the structure and the ampli-
tude of the displacements on the ground [Geertsma and Van Opstal, 1973]. Horizontal and
vertical strains between the surface and the reservoir may also fracture casings, destroy
numerous boreholes and reactivate faults.

3.2.1.3 Subsidence measurement and prediction

A. Subsidence measurement
Boutéca et al. (1996) propose a review of subsidence and compaction measurement
methods. On the ground, subsidence is traditionally measured by levelling survey or GPS
positioning. More recently, Brink et al. (2002) used an InSAR (Interferometric Synthetic
 Chapter 3 • Geomechanics and Reservoir 161

Aperture Radar) method to determine subsidence rates at the Lost Hills field in California.
This technique, based on processing images obtained using SAR radar on board satellite,
represents a powerful ground movement estimation tool. At sea, measurement can be carried
out less precisely by measuring the height variations between the platform and sea level.
Measurements of underground displacements in the reservoir and the overburden can also be
used. These measurements can be taken using conventional log analyses (measurement of
natural gamma radiation) or specialised log analyses (variation in casing lengths,
measurement using radioactive marker techniques, etc). For shallow fields, reservoir
compaction can be measured using a taut wire at the bottom of the well.

B. Subsidence prediction
Geertsma (1973) proposed an analytical model to provide an estimation of the subsidence
resulting from compaction of a cylindrical reservoir, of thickness h and radius R located at
depth z. Reservoir compaction is given by:
https://telegram.me/Geologybooks

h
C = Δh = ∫ cm ( z ) Δp( z )dz (3.2.1)
0

with cm the reservoir rock compressibility in oedometric condition (uniaxial condition and con-
stant vertical stress) which can be measured in laboratory and Δp the pressure variation in the
reservoir. If cm and Δp are homogeneous in the reservoir, the reservoir compaction is given by:
C = cm Δph (3.2.2)

Geertsma (1973) then gives the analytical expressions of the horizontal and vertical
elastic displacements on the surface in case of infinitely rigid underburden, overburden with
the same mechanical characteristics as the reservoir and reservoir with uniaxial behaviour.
The vertical displacement at the centre of the reservoir can be expressed as follows:

⎛ η ⎞ z
S = −2(1 − ν ) ⎜ 1 − ⎟ C with η = (3.2.3)
⎜⎝ 1 + η2 ⎟⎠ R

According to the latter equation, the ratio between maximum subsidence and reservoir
compaction (S/C) is determined by the ratio between depth of burial and the lateral extent of
the reservoir. Expression (3.2.3) shows in particular that small deeply buried reservoirs (large
h) will present no significant subsidences, even if their reservoir compaction is high. On the
contrary, large reservoirs at low depths (small h) may be potential candidates for subsidence.
Morita et al. (1989) later proposed a semi-analytical solution based on finite element cal-
culations, which does not assume that the reservoir and its confining formations have the
same mechanical properties.
The previous solutions are based on simplifying assumptions which do not allow accu-
rate calculation of the subsidence for a given field. In particular, the complex reservoir
geometry, heterogeneities in the reservoir and the surrounding formations and elastoplastic
constitutive laws must be taken into account to predict subsidence accurately. More accurate
 162 Chapter 3 • Geomechanics and Reservoir

subsidence predictions therefore require coupled geomechanical-reservoir simulations (see

Chapter 3.1) to precisely describe reservoir compaction during production and the strains in
the overburden [Geertsma and Van Opstal, 1973].
Note that this type of modelling requires numerous data not only on the reservoir but also
on the surrounding formations (porosity distribution, mechanical properties, etc.) and on the
reservoir production. This type of study is therefore reserved for the most serious cases of
subsidence (e.g. the Groningen field).

3.2.2 CO2 sequestration

For the last century or so, human activity, in particular combustion of fossil hydrocarbons, has
increased the levels of greenhouse gases in the atmosphere. The CO2 concentration has increased
and could reach 1,000 ppm by the end of the century if no steps are taken to reduce emissions.
Experts agree on the fact that this increase is responsible for noticeable climate changes and vari-
https://telegram.me/Geologybooks

ability, with potentially disastrous impacts on the ecosystems and human activities.
As pointed out in the 2005 IPCC (Intergovernmental Panel on Climate Change) report,
Carbon Capture and Storage (CCS) is an efficient way of reducing CO2 emissions. If imple-
mented, the rate of greenhouse gas emission into the atmosphere could start to stabilise sig-
nificantly in the near future.
Storage in end-of-life oil or natural gas reservoirs offers several advantages, the most
important being that it is a known technique and virtually leaktight. These natural reservoirs
have proven their ability to store hydrocarbons for several millions of years. CO2 has already
been reinjected for many years in depleted reservoirs to reduce the oil viscosity, improve its
mobility and enhance recovery. While this is an EOR (Enhanced Oil Recovery) technique,
for sequestration of large quantities of CO2, the pressures considered exceed the pressures
that existed in situ before any human intervention, which is never the case with EOR.
CO2 storage in deep aquifers represents another possibility. Numerous deep aquifers
exist, onshore and offshore, some with areas of up to several thousand km2. Formed from
porous, permeable rocks, filled with brine unsuitable for human consumption, they could
store large quantities of CO2. Some are already used for this purpose (for example the Utsira
formation in the North Sea).

3.2.2.1 Short-term geomechanical risks

CO2 injection will therefore be carried out on a massive scale, under pressure. A distinction
must be made between the local pressure increase at the injectors and the average pressure
increase in the reservoir. These pressure increases modify the in situ stresses locally and glo-
bally, thereby deforming the reservoir and the surrounding layers with the following poten-
tial consequences:
– hydraulic fracturing,
– reactivation of faults or fractures,
– shear failure,
– possible leaks to the surface.
 Chapter 3 • Geomechanics and Reservoir 163

Hydraulic fracturing may be increased by the fact that CO2 injection will generate a
priori thermal perturbation with a cold front around the injectors. In some cases, fracturing
may propagate in the cap rock (see Chapters 2.4 and 2.5).
Pressure injection reduces the effective stresses normal to the faults and fractures and
therefore reduces their shear strength. The injector wells must be located far away from faults.
Shear stresses may be created at the top of the reservoir at the interface with the cap rock
[Hawkes et al., 2004].

3.2.2.2 Modelling
To quantify the risks, the most reliable approach is to conduct precise coupled geomechani-
cal-reservoir simulations. As when calculating subsidence, it will be essential to take into
account the geometry and mechanical properties of the reservoir and the surrounding forma-
tions.
Knowledge of the initial in situ stresses is also fundamental [Hawkes et al., 2005].
https://telegram.me/Geologybooks

The few studies already conducted indicate possible surface uplift in case of shallow
reservoir or aquifer (Figure 3.2.5) [Rutqvist et al., 2007, Vidal-Gilbert et al., 2009].

SMB15 SMB15

SMB17 SMB17

a) b)
U3 (m) U3 (m)
+5 = 810e-03 +2 = 054e-03 +4 = 409e-03 +2 = 709e-03
+5 = 814e-03 +1 = 428e-03 +4 = 126e-03 +2 = 425e-03
+4 = 558e-03 +8 = 019e-04 +3 = 842e-03 +2 = 142e-03
+3 = 932e-03 +1 = 759e-04 +3 = 559e-03 +1 = 859e-03
+3 = 306e-03 -4 = 500e-04 +3 = 276e-03 +1 = 575e-03
+2 = 680e-03 -1 = 076e-03 +2 = 992e-03 +1 = 292e-03
+2 = 054e-03 -1 = 702e-03 +2 = 709e-03 +1 = 008e-03

Figure 3.2.5
Variations of vertical displacement (m) a) at the reservoir top and b) at surface
during CO2 injection [From Vidal-Gilbert et al., 2009].
 164 Chapter 3 • Geomechanics and Reservoir

3.2.2.3 Medium-term geomechanical risks

Injection of supercritical CO2 in a geological reservoir will necessarily modify the physico-
chemical equilibrium in the host rock. Dissolution of CO2 in the formation water causes
acidification. This phenomenon accelerates the dissolution-precipitation reactions and may
modify the mechanical and hydraulic properties of the rock. The fluid characteristics change
on moving away from the injector well and will also change over time, passage of the acid
reactive front, then a drying or dehydrating front. We can expect to find dissolution
phenomena with more or less fingered profiles as we move further away. The dissolved
materials may also precipitate out again at a variable distance from the initial dissolution
zone. These phenomena could change the geomechanical properties of the host rock.
In carbonate rocks, with the high injection rates expected around the well, the develop-
ment of networks of wormholes could be observed [Egerman et al., 2006]; farther away,
lower rates would lead to more compact dissolution figures.
In sandstone reservoirs with carbonate cement, CO2 injection may generate a number of
https://telegram.me/Geologybooks

damaging phenomena. In particular, progressive dissolution of a carbonate cement could

lead to considerable mechanical weakening of the rock.
Generally, alteration of the reservoir matrix by the CO2-brine mixture could result in
compaction of the reservoir rock and development of cracks in the reservoir and the cap
rock. Data obtained from recent experiments show clear trends of mechanical weakening
[Bemer, 2008].

3.2.3 Well abandonment

Wells are drilled for various reasons: exploration, oil and gas production, water injection,
storage, etc. At some stage, all these wells will be permanently abandoned. Producer wells
have a lifetime of about twenty years, corresponding to the production period of an oil field.
After this phase, the well must be abandoned and plugged with suitable material, typically
cement-based. Even though the well has been plugged, there is a serious risk of leakage in
the long term which represents a threat to the environment. For example, migration of
hydrocarbons along the borehole may pollute the groundwater, the marine ecosystem or the
atmosphere in case of toxic gases. These leaks can be sealed, but repair costs are extremely
high for offshore installations. The main objective of oil well abandonment is therefore to
prevent the migration of any hydrocarbon reserves remaining at end of production by
limiting the flow of fluids along the borehole between the various geological layers.
Well closure and abandonment is therefore a critical issue regarding protection of the
environment and in particular protection of water resources. Since the techniques involved
in closure require a detailed analysis of the geological context, the well, variations in bot-
tomhole conditions expected over the long term, reservoir fluids, etc., each abandonment
procedure is unique.
 Chapter 3 • Geomechanics and Reservoir 165

3.2.3.1 Abandonment procedure

The operator responsible for well closure and abandonment must comply with local laws
(aimed at protecting the environment) while also respecting financial constraints. It is not
financially viable, for example, for the operator to pour a cement plug down the entire depth
of the reservoir and also seal all the annuli from well bottom up to the surface. The abandon-
ment procedure varies depending on whether or not some of the steel tubes (production tub-
ing and some of the casings) are abandoned with the well. Offshore wells are frequently
abandoned with some of their steel tubes if no drilling equipment is available.
Kelm and Faul (1999) recommend installing several plugs to optimise well isolation (see
Figure 3.2.6). In particular, these authors recommend installing cement plugs at the casing
perforations, above the oil zones located above the perforations, at abnormal pressure zones
and above the noncemented annuli. Plug length varies from 15 to 60 metres, the longest
plugs being recommended for the abnormal pressure zones and the surface plugs (i.e. the
last plug preventing surface pollution).
https://telegram.me/Geologybooks

Typical
Typical «Best practices»
production well plug and abandonment

Drive
Bridge Plug No. 5
pipe
plug Surface isolation
Freshwater Plug No. 4
Isolate freshwater

Bridge
Saltwater
plug
Cut and pull
Surface
intermediate casing
casing
below freshwater base

Cement
retainer Plug No. 3
Intermediate Squeeze/isolate
casing noncemented annulus
Cut and pull production
casing near intermediate
Top of cement
casing shoe
Production
tubing Bridge
Plug No. 2
plug
Safety plug
Production
packer

Perforations Plug No. 1

Set-through squeeze
of perforations
Production casing

Figure 3.2.6
Oil well abandonment procedure using cement plugs [Kelm and Faul, 1999].
 166 Chapter 3 • Geomechanics and Reservoir

Installation of the well plug is a determining factor in closure quality. Wells are gener-
ally plugged with Portland cement to which sand and bentonite (for expanding cements) can
be added. Cement set accelerators can also be added, as well as dispersing agents to improve
the durability of the materials. For very large diameter wells, operators may also use con-
crete. With water wells, use of bentonite-based clay materials is sometimes recommended
[Calvert and Smith, 1994]. Presence of fluid in the well makes installation of the cement
plug difficult since it favours dilution and the creation of preferential flow paths of the
cement slurry [Smith et al. 1984]. The density difference between the cement slurry and the
fluid in the well may cause the cement slurry to flow to the bottom of the well, the result
being that no plug is really formed [Calvert and Smith, 1994, Crawshaw and Frigaard,
1999]. This type of flow can be avoided by installing a mechanical plug under the cement
plug, to ensure the cement slurry remains in place while setting. The installation procedure
must also ensure perfect adherence between plug and wall to avoid risks of leakage at the
interfaces. To improve adhesion between the casing and the cement plug, Evans and Carter
(1963) and Kelm and Faul (1999) recommend cleaning the tubing before installing the
https://telegram.me/Geologybooks

cement plug.

3.2.3.2 Behaviour of the well plug after abandonment

After abandonment, the well plug is subject to variations in the bottomhole conditions which
may prejudice well closure due to chemical, physical and mechanical mechanisms. In the
medium and long terms, this may result in a loss of integrity of the closure and consequently
risks of flow along the borehole. We will now describe the main risks of loss of integrity of
the closure.

A. Behaviour associated with installation of the plug

– Well plugs are generally made of cement, a material chosen for its high durability and
low cost. Hydration of the cement slurry injected into the well solidifies the material
and forms a plug. With ordinary cements, the hydration reactions are accompanied by
a decrease in plug volume (i.e. hydration shrinkage) which occurs during the first few
hours of hydration. Hydration shrinkage may generate high tensile stresses in the
material causing the cement to crack. Plug shrinkage may also result in separation at
the plug/casing interface. These two phenomena must be limited or avoided to pre-
serve well closure integrity.
– The amplitude of the shrinkage is affected by the temperature and pressure to which
the cementation material is subjected [Parcevaux and Sault, 1984]. To avoid the nega-
tive effects of cement shrinkage during setting, service companies have developed
expanding cements which prevent the formation of microannular spaces between the
casing and the cemented annulus [Seidel and Greene, 1985; Baumgarte et al., 1999].
To avoid the formation of microannular spaces between the plug and the casing,
Bosma et al. (2000) propose installing an elastomer plug of thermal expansion coeffi-
cient greater than that of the surrounding rock. Thermal expansion of the plug occurs
due to progressive heating of the rock and the elastomer after a cooling phase carried
out before injecting the liquid elastomer. Bosma et al. (2000) studied in particular the
 Chapter 3 • Geomechanics and Reservoir 167

influence of this cooling phase to avoid excessive thermal expansion of the plug,
which could damage the primary cementation of the annulus (Figure 3.2.7).

0
-.3E7
-.6E7
-.9E7
-.12E8
-.15E8
-.18E8
-.21E8
-.24E8
-.27E8
-.30E8
-.33E8
-.36E8
-.39E8
-.42E8
-.45E8
-.48E8
-.51E8
-.54E8
-.57E8
-.60E8

Figure 3.2.7
Evolution of the radial stress around the well a) before cooling, b) after cooling
and c) after installing the plug and heating [Bosma et al., 2000].
https://telegram.me/Geologybooks

B. Behaviour associated with reservoir production and abandonment

– Reservoir production modifies the pressure, temperature and stress in the reservoir
and more especially around the boreholes. If a well is abandoned and reservoir pro-
duction maintained, these variations will generate thermal, hydraulic and mechanical
loads on the well plug. If production is stopped completely and all wells in the reser-
voir abandoned, the return to thermal, hydraulic and mechanical equilibrium will be
accompanied by temperature, pressure and stress variations which will also act as
loads on the plugs. These loads are difficult to estimate since they depend on each
individual case. In particular, presence of an active aquifer under the reservoir may
cause the pressure to rise in the reservoir; in wells operating with injection of cold
water during production, the temperature around the injector well will rise when the
well is abandoned and, inversely, in wells implementing enhanced recovery processes
based on steam injection, the area around the well will cool down when the well is
abandoned.
– The mechanical properties of the well plug must therefore be adapted to variations in
bottomhole conditions after abandonment. These variations can be estimated using
coupled geomechanical-reservoir simulations which supply the pressure, temperature
and stress variations after abandonment of the well and/or the reservoir.
– Using this technique, Mainguy et al. (2004) were able to estimate the stress variations
in the cap rock of a reservoir where it is recommended to install a plug [UKOOA,
2001]. This study shows in particular that the cap rock behaviour is a determining fac-
tor for durability of the well plug, since a reduction in the effective compression stress
during the abandonment phase in this cap rock (Figure 3.2.8) induces tensile stress in
the plug if it was not initially prestressed. This type of load is particularly detrimental
for cement based plugs which have low tensile strength and may also favour separa-
tion of the plug from the casing or surrounding rock.
– The behaviour of cement plugs is generally modelled using a constitutive law which
combines a smeared cracking model [De Borst and Nauta, 1985] with an elastoplastic
 168 Chapter 3 • Geomechanics and Reservoir

constitutive law in compression. In particular, this is the model used by Bosma et al.
(1999) and Ravi et al. (2002) to study the behaviour of the annuli cemented during the
well completion and production phases. In case of deep wells exposed to high
temperatures, the temperatures will modify the cement microstructure and therefore
affect the mechanical behaviour of the plug. It is therefore important to characterise
the mechanical behaviour of well plugs at high temperatures so that the plug
behaviour can be modelled during abandonment using suitable models.

4
400 years

3 150 years

125 years
q (MPa)

2
https://telegram.me/Geologybooks

100 years

1
75 years
50 years
0
-6 -4 -2 0 2
p' (MPa)

Figure 3.2.8
Example of tensile load applied in the shale caprock above an injector well
[Mainguy et al., 2007]. Evolution of the deviatoric and effective mean stresses
from end of production to 400 years after abandonment.

C. Plug ageing under high pressure and temperature conditions

– Well plugs installed in the reservoir are exposed to high pressure and temperature
conditions which could damage the cement plug in the long term. Ageing of the
cement slurry under severe pressure and temperature conditions irreversibly damages
its mechanical properties (Young’s modulus, breaking strength, etc.). Studies on
ageing of cement materials under high pressure and temperature conditions are
generally conducted over periods of 1 to 2 years [Eilers and Root, 1974, Noik and
Rivereau, 1999]. These data must nevertheless be extrapolated over longer time scales
for use in studies on abandonment over longer periods.

D. Leaching of cement plugs

– In contact with a neutral or slightly basic solution (as compared with the highly basic
interstitial solution of a cement slurry), the hydrated phases of the cement, in
particular portlandite (Ca(OH)2), dissolve. This cement degradation process (called
leaching) combines dissolution of the chemical compounds in the cement and transfer
of chemical species in liquid phase by diffusion [Adenot, 1992]. Chemical
 Chapter 3 • Geomechanics and Reservoir 169

degradation of the cement is accompanied by degradation of its elastic mechanical

[Gérard, 1996] and breaking [Carde, 1996] properties. Nevertheless, the aggressive
water near the plug must be renewed if degradation of the cement plugs by leaching is
to take place. This scenario is only possible with plugs installed near an aquifer or in
contact with runoff water.

E. Carbonation of cement plugs

– Cement carbonation designates all chemical reactions between carbon dioxide and the
compounds in the hardened cement produced by cement hydration reactions. Carbon
dioxide penetrates the cement in gaseous form or in solution. Carbonation occurs in
aqueous medium, by reaction of the carbon dioxide dissolved in the interstitial solu-
tion of the cement with portlandite that has gone into solution. This reaction first pro-
duces calcium carbonate, then finally calcite. C-S-H compounds (silicates) and alumi-
nates may also react with carbon dioxide. Progression of carbonation towards the
https://telegram.me/Geologybooks

centre of a cement paste is generally assumed to be proportional to the square root of

the carbon dioxide exposure time.
– The mechanical properties of the cement paste deteriorate with carbonation. Chi et al.
(2002) studied the influence of the depth of carbonate cement on the loss of unconfi-
ned compressive strength. Beaudoin (2000) studied the reduction in Young’s modulus
of a carbonate cement slurry as a function of the increase in its porosity resulting from
the dissolution of hydrates. Carbonation of cement plugs installed in reservoirs seems
unlikely, however. This type of phenomenon could be observed in cement plugs used
to close CO2 storage wells.

3.3 GEOMECHANICAL MONITORING OF RESERVOIRS

Reservoir monitoring is a production technology designed to measure, check and predict the
reservoir performance. Monitoring generally includes measurements of temperature, pres-
sure, flow rate and constituents, as well as surface deformation, 4D seismics and microseis-
micity which no longer concern the well alone but consider the reservoir as a whole with the
surrounding formations. These records can be used to map the progression of a heat front in
3D, identify reactivation of a fracture or fault and check the integrity of the cap rock.

3.3.1 Measurement of surface deformations

Volume variations in the reservoir may cause deformations in all the surrounding geological
layers as well as surface deformations. This occurred with subsidence, but it may also be the
case when the reservoir is shallow and operated using thermal methods.
Tiltmeters can be installed in a fixed network on the surface or moulded in the wells.
Modern tiltmeters can measure angles as small as few nanoradians. They are so sensitive
that they record the Earth tides generated by the moon. They have been used to record sur-
face deformations during SAGD processes (Figure 3.3.1).
 170 Chapter 3 • Geomechanics and Reservoir

TOTAL DCEL 16 Jul-2007 to 01-Oct-2007

50
Tiltmeter site
Reference point
Tilt vector (0 m/uR) 40
Well trajectories
Contour level (2 mm)
Max movement (mm) 30
Min movement (mm)

Vertical displacement (mm)

20

10

-10

-20
https://telegram.me/Geologybooks

-30

-40

-50
200 m

Figure 3.3.1
Surface deformation over a period of 3.5 months as computed from tiltmeter
measurements on a SAGD field [Dubucq et al., 2008].

“Precision” GPS instruments and interferometric aperture radar techniques can also be
used to measure surface displacements, but with less accuracy.
The InSAR (Interferometric Synthetic Aperture Radar) technique uses two synthetic
aperture radars simultaneously, or the same radar used at different times. The point to point
phase differences in the images generated are then studied to find the altitude of the land.
Centimetric displacements can be measured in areas where the signal remains coherent [Mei
and Frosses, 2007].

3.3.2 Seismic monitoring

The seismic technique is the geophysical method most commonly used to define sub-surface
structures. The most common implementation of this technique consists in reflection seismic
with multiple coverage. This method provides imaging of the sub-surface in 2 or 3
dimensions. Lithology and fluid information can be extracted from seismic data (wave
propagation velocities, amplitudes, etc.).
 Chapter 3 • Geomechanics and Reservoir 171

Time-lapse seismic reservoir monitoring is the process of acquiring and analysing multi-
ple seismic surveys, repeated at the same site over calendar time, in order to image fluid-
flow effects in a producing reservoir. If each survey is “3-D seismic,” then the resulting set
of time-lapse data is termed “4-D seismic,” where the extra fourth dimension is calendar
time [Lumley, 1995].
In addition to 4-D seismic, there are other viable methods of time-lapse seismic monitor-
ing including repeated 2-D surface seismic, surface-to-borehole VSP (vertical seismic pro-
filing), and crosswell tomography.
Figure 3.3.2 shows two seismic cross-sections acquired at different times during the pro-
duction of the Duri field, Indonesia: before steam injection and after 5 months of steam
injection. The last seismic cross-section is the difference between the two previous ones.
The well outline visible in the third cross-section is due to a shallow borehole steam
leak.The time lapse seismic images show some changes in amplitudes along baseline reflec-
tors, due to a variation in fluid content.
https://telegram.me/Geologybooks

No steam 5 months steam Difference

Pre-steam: migrated inline Steam: migrated inline Migration difference

0 0 0

B
Time (s)

Time (s)

Time (s)

0.1 0.1 0.1

0.2 0.2 0.2

30

35

40

45

50

55

60

30

35

40

45

50

55

60

30

5
60
3

5
0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

X distance (km) X distance (km) X distance (km)

Figure 3.3.2
Seismic cross-sections: (left) before steam injection, (centre) during injection,
(right) difference between the two cross-sections. S: steam zone, B: borehole
heating, P: polarity reversal [Lumley, 1995].
 172 Chapter 3 • Geomechanics and Reservoir

Similarly, Figure 3.3.3 shows the maps of amplitude differences between the seismic
surveys carried out in 2001, 2004 and 2005 on the Christina Lake field operated by SAGD.
The amplitude differences indicate the changes in the steam chamber around the wells.

4D:2001-04 4D:2001-05

a) 2004 b) 2005

Figure 3.3.3
https://telegram.me/Geologybooks

Maps of amplitude difference a) between 2001 and 2004 and b) between 2001
and 2005 [Zhang et al., 2005].

The images provided by reservoir seismic monitoring may help identify by-passed oil
volumes, which will then be drilled. The reserves extracted can therefore be increased,
thereby extending the economic life of an oil field.
4D seismic can be used to track the progress of the injected fluid front (water, gas, steam,
CO2, etc.) and may play a role in optimising the injection programs.
More recently, time-lapse seismic has been applied for objectives other than monitoring
saturation changes, such as pressure monitoring and compaction detection. Over the past
few years, seismic monitoring has been widely used to track fluid movements (saturation
effect). This approach is no longer sufficient and pressure effects must be taken into account
to improve the interpretation of 4D seismic data [Landro, 1999]. More precisely, the pres-
sure effect does not concern the pore pressure alone, but also the induced geomechanical
effects.
Decreasing the pore pressure in the reservoir will increase the mean effective stress; this
will lead to rock compaction and increase the P-wave velocity Vp in the reservoir. The
overburden and underburden will deform to fill the space created by reservoir compaction.
Vp may be reduced in these zones due to the arching effect, a phenomenon observed by
Hatchell et al. (2004) on a North Sea field. The velocity variations in the cap rock were even
greater than those observed in the reservoir.
Coupled geomechanical-reservoir simulations can be used to determine the porosity, in
situ stresses and saturations which will then be substituted in the Gassmann equation to cal-
culate the P- and S-wave velocity field not only in the reservoir but also in the entire block
considered. The studies conducted in this field are highly promising [Vidal et al., 2002,
Minkoff et al., 2004].
 Chapter 3 • Geomechanics and Reservoir 173

3.3.3 Passive seismic monitoring

The various terms used, seismic or microseismic monitoring, Passive Seismic Monitoring
(PSM) and microseismicity refer to this technique, based on the recording of induced seismic
signals. The term induced implies that the seismic activity recorded results from a man-
made action or operation on the medium, as opposed to natural seismicity related to
tectonics.
Microseismic events correspond to sudden readjustments of mechanical stresses
(Figure 3.3.4). Stress rebalancing may be caused by a variation in pore pressure (oil produc-
tion, storage of fluid in underground aquifer, extraction of material mining, washout), or
variations in the temperature of the medium (injection of fluid which is hot or cold with
respect to the medium: geothermy, enhanced recovery by water flooding, thermal methods).
In some cases, these phenomena may be superimposed.

Confining stress
https://telegram.me/Geologybooks

Energy release
Pore
pressure

Shear stress

Figure 3.3.4
Microseismic failure mechanism.

Initially observed in Colorado after injection of fluid waste for underground disposal in
the 1960’s, induced microseismicity soon became a subject of interest to oil producers in
order to map hydraulic fractures, the first publications dating back to the 1970’s.
In the following years, microearthquakes were observed in the highly depleted oil zones,
especially the Groningen (the Netherlands), Gazli (Uzbekistan) and Lacq (France) fields.
Since these were naturally seismic risk areas, implementation of monitoring was in some
respects expected to distinguish between natural seismicity and induced seismicity. Surface
monitoring was carried out, using seismology equipment.
In the 1980’s, Passive Seismic Monitoring (PSM) became widely used in the mining
industry to prevent the risk of cave-in, especially in South Africa where extraction of various
ores (coal as well as coal and diamonds) is carried out at greater and greater depths. In the
field of underground storage, this technique has been used to monitor storage in salt cavities.
In these fields, the measurement means are traditional and inexpensive, consisting of
geophones connected to recording systems. The frequencies used ranges from a few Hz (less
in seismology) to several hundred Hz. The most recent developments take into account
three-component sensors.
 174 Chapter 3 • Geomechanics and Reservoir

In the oil industry, the depth of the reservoir may limit the use of this technique due to
signal attenuation particularly in the near surface.
Ideally, information derived from interpretation of microseismic data can be used to:
– locate possible sliding over a natural fault plane,
– locate the compacting zones,
– locate the hydraulic fracturing zones and determine the characteristics (width, length,
height, orientation, growth, complexity, etc.) of the fractures created,
– identify the natural fractures or faults contributing to fluid displacements,
– map the fluid displacements,
– locate the zones badly drained or not drained by the fluids injected,
– identify the mechanically active levels likely to play a role in well failure (casing),
– map the thermal fronts, etc.

3.3.3.1 Temporary or permanent passive seismic monitoring

https://telegram.me/Geologybooks

“Short-term” applications mainly concern injection operations that are highly localised in
time and space. The operator expects interpretation of the measurements immediately after
the test, for hydraulic fracturing or injection control.
Monitoring of hydraulic fracture propagation (see 2.4.5) depends on the instruments
implemented (number of sensors) and the knowledge available on wave propagation veloci-
ties in the different layers.
Injection control consists in checking that injection of a given fluid (production water,
wastewater, industrial fluids, drilling mud, steam, CO2, etc.) in a reservoir or an
underground aquifer carried out in limited proportions (limited injection duration) proceeds
correctly. Under these conditions, the operator does not want to fracture the geological units
surrounding the reservoir layer or the aquifer, to avoid polluting the other layers.
Carried out near the injection zone, PSM can track possible fracturing of the rocks or
reactivation of natural fractures or faults. If this type of phenomenon is observed, injection
can be contained in the initial layer by simply adapting the injection parameters. On the con-
trary, if there is no microseismic activity and no brutal injection pressure change, there is a
strong likelihood that the rock has not been fractured.
To date, “long-term” or permanent applications are operational in mining and storage of
hydrocarbons in saline cavities. Their main role is to ensure the safety of people and
property. These applications are also implemented for safety considerations in two other
cases: oil production where permanent surface measures have already been imposed on
some operators (especially NAM in the Netherlands) and storage of nuclear waste (NIREX).
In traditional oil and gas applications, long-term PSM proves interesting for gas storage
in underground aquifers [Vidal et al., 2000; Deflandre et al., 2002], monitoring of oil pro-
duction, monitoring of steam chamber propagation in thermal production [Snell et al.,
1999], reinjection of production water and more recently CO2 sequestration.
Long-term applications can be expected to play an increasing role in meeting restrictions
on protection of the environment [Snell et al., 1999]. It is in fact likely that PSM will even-
tually be implemented in oil well abandonment procedures.
 Chapter 3 • Geomechanics and Reservoir 175

3.3.3.2 Surface PSM

A ground surface network of sensors can be used for detecting micro-seismic events. Surface
PSM applications include fields for which measurements are taken on the surface and measure-
ments that are easy to access (e.g. mine galleries). In practice, the same instruments are used.
In mine galleries, the main objective is to predict the risk of collapse during operation. In
particular, the aim is to precisely locate the dislocation zones.
The oil industry may also be concerned by PSM surface applications. Further to the
earthquakes of magnitude 5 on the Gutenberg-Richter scale, recorded around the Groningen
field in the Netherlands, the government obliged the oil companies to implement surface
monitoring to predict the seismic risk induced by massive extraction of hydrocarbons from
the subsoil [Mobach and Gussinklo, 1994]. In this case, the major risk is failure of a dyke,
which could have disastrous consequences. Studies have therefore been conducted to predict
the maximum seismicity level that exploitation of the neighbouring fields could induce.
https://telegram.me/Geologybooks

3.3.3.3 Downhole PSM

Given the low amplitude of microseismic signals and the uncertainties over wave propaga-
tion velocities in the surface layers, it became clear that buried sensor networks are recom-
mended. More events can be recorded and located with greater accuracy. Furthermore, since
the signals are less attenuated due to shorter trajectories, the frequency bandwidth is richer
and, as a result, mechanical phenomena can be described with higher resolution.
Several types of instrument are used in wells: VSP, acoustic probes or permanent sensors
(Figure 3.3.5).

Wellhead sensor

Microseismic
Permanent downhole source
geophones

Ds = (D P, Dq)

Figure 3.3.5
Principle of passive seismic monitoring with downhole instruments.
 176 Chapter 3 • Geomechanics and Reservoir

A. VSP (Vertical Seismic Profile) tools

The advantage with VSP tools is that an array of geophones is placed vertically in a well-
bore. However, the instruments must be installed in a well dedicated to measurement (pro-
duction stopped and production column removed) and the acquisition systems are not
always suitable for permanent monitoring. On the Ekofisk and Valhall fields in the North
Sea, records are taken by systematic triggering every 30 seconds (CGG-BSD seismic
recording system) and a 2-second reset time is required. The microseismic events are
detected when rereading the records. This method does not allow real-time monitoring.

B. Acoustic probes equipped with geophones

Several tools have been developed to monitor the recording of acoustic signals during hydrau-
lic fracturing. These tools are mono- or multi-components. The tool developed by IFP Ener-
gies nouvelles (Figure 3.3.6) was designed for use in injection well and in observation well.
The logging probe includes an acoustic module equipped with a triaxial accelerometer and a
https://telegram.me/Geologybooks

clock and an electronic module equipped with a pressure sensor and a temperature probe.
This system allows bottomhole fracturing temperature and pressure logging during and
after injection and real-time logging of a large number of events due to the proximity of the
emission zone. In addition, proximity of the emission zone reduces the uncertainty on the

SIMFRAC® Tool SIMFRAC® Tool

2" (50mm) max OD 2" (50mm) max OD
Through tubing and packer cased hole only

Pumping unit Logging truck Tubing 2"7/8 to CS 9"5/8

Mono-conductor
Logging cable
BOP
Tubing Motor

Electronic
module Two arms latch assembly
Downhole pressure
Digital transmission
Mono conductor
logging cable Downhole temperature

Packer Decoupling
Casing device Decoupling control
Microseismic sensor
source
Permanent magnet

Seismic sensors X Y Z Accelerometers

Reservoir module
layer (OD 2")
Pendulum Rb rotation
SIMFRAC tool
Weight Permanent magnet
XYZ accelerometers
Pressure - temperature
Pendulum

Fractured zone

Figure 3.3.6
Diagram of the SIMFRAC logging probe during a fracturing experiment.
 Chapter 3 • Geomechanics and Reservoir 177

location of microseismic events and produces a more accurate map of the structure. Lastly,
system installation costs are low since the tubing does not have to be pulled up to take the
measurement.
In the frac-treated well, however, the microseismicity activity emitted during injection
cannot be analysed due to the noise level in the well. The probe can only be used during the
minifracturing phase preceding injection of the proppant.
Acoustic probes equipped with three geophones can be lowered down on a cable in a
cased well not equipped with a production column. Coupling is provided by the opening of
two motor-driven arms.

C. Well permanent seismic sensors

Passive seismic monitoring currently benefits from recent instrument developments with, in
particular, well permanent seismic sensors and bottomhole numerical acquisition and trans-
mission systems (Figure 3.3.7).
https://telegram.me/Geologybooks

The big advantage with well permanent seismic sensors is that measurements can be
taken without affecting production. This advantage will be increasingly important when
operating deep sea wells due to their specificities (oil type, low pressure and temperature).
Comprehensive systems cover the requirements of conventional well seismics and pas-
sive seismic monitoring.

Cable for digital

transmission
to surface
Casing
Electronic
module

Tubing

Analogic links
Decoupling
device Bow-spring
for coupling

Frame 3-component
geophone
module

Digital
transmission to
deeper levels

Figure 3.3.7
Well permanent seismic sensors [Laurent et al., 1999].
 Introduction

At all times during the exploitation of an oil field, the engineers strive to optimise drilling
times, reduce uncertainties and production costs, and optimise the use of available data.
Geomechanics, which includes all thermohydromechanical phenomena, plays an impor-
tant role in every operation involved in the exploitation of hydrocarbons, from drilling to
production and right up to the time the wells are abandoned. Pressure changes in the reser-
voir modify the in situ stresses and cause strains not only to the reservoir but also to the
https://telegram.me/Geologybooks

entire sedimentary column. In turn, these stress modifications and strains change the stabil-
ity parameters of the walls of the wells to be drilled and the flow properties of the fluids.
The main purpose of this book is to provide engineers involved in oil field development
and petroleum engineering students, whether drillers or producers, the information they need
about the techniques and approaches currently used in geomechanics.
The book is divided into three broad sections.
The first section provides information on rock mechanics, describes the characterisation
of rocks in the laboratory and the modelling of their mechanical behaviour. This chapter also
describes how geomechanics determine the stresses applied to the material in place and why
they need to model the reservoirs and the surrounding material.
The second section deals with the role of geomechanics in the optimisation of drilling
(drillability and wellbore stability) and production. Sand production, which brings numerous
undesirable problems, can be predicted and in some cases controlled. Hydraulic fracturing
can be encouraged to stimulate production or on the contrary avoided so as not to reactivate
a fault, in case of CO2 sequestration.
The third section deals with the coupling between the production of fluids in the reser-
voir and the mechanical strains of the complete oil structure. The environmental aspects of
this interaction are discussed with respect to subsidence of the surface layers, CO2 seques-
tration and well abandonment. Geomechanical monitoring of reservoirs may prove essential
in some cases to optimise production or check cap rock integrity.
Warning! In rock mechanics the compressive stresses are positive, whereas in continuous
media mechanics, stresses are negative. The latter convention is adopted in the numerical models.
Each chapter will indicate whether the positive or negative compressive stresses are being used.
 References
https://telegram.me/Geologybooks

Aadnoy, B.S. (1988) Inversion technique to determine the in situ stress field from fracturing data,
Paper SPE 18023 presented at the 63rd Annual Technical Conference and Exhibition, Houston,
Texas, October 2-5.
Aadnoy, B.S. and Chenevert, M.E. (1987) Stability of highly inclined boreholes, Paper SPE 16052
presented at the SPE/IADC Drilling Conference held in New Orleans, March 15-18.
Abaqus Analysis User’s Manual, Abaqus Version 6.4, Simulia, Dassault Systèmes, 2005.
Abass, H.H., Hedayati, S. and Meadows, D.L. (1996) Nonplanar fracture propagation from a horizontal
wellbore: experimental study, Paper SPE 24823, SPE Production and Facilities, pp. 133-137.
Addis, M. A. (1997) The stress-depletion response of reservoirs, Paper SPE 38720 presented at the
SPE Annual Technical Conference and Exhibition held in San Antonio, Texas, October 5-8.
Addis, M.A., Choi, X. and Cumming, J. (1998) The influence of the reservoir stress-depletion res-
ponse on the lifetime considerations of well completion design, Paper SPE 47289 presented at
the SPE/ISRM Eurock’98 held in Trondheim, Norway, July 8-10.
Adenot, F. (1992) Durabilité du béton: Caractérisation des processus physiques et chimiques de dégra-
dation du ciment, PhD. Thesis, Université d’Orléans.
Alfenore, J. (2001) How to control high productivity wells in deep water reservoirs, SPE Distinguished
Lecturer Program.
Al-Harthy, S.S., Jing X.D., Marsden, J.R. and Dennis, J.W. (1999) Petrophysical properties of
sandstones under true triaxial stresses I: Directional transport characteristics and pore volume
change. Paper SPE 57287 presented at the SPE Asia Pacific Improved Oil Recovery Conference
held in Kuala Lumpur, Malaysia, October 25-26.
Alonso, E. E., Gens, A. and Josa, A. (1990) A constitutive model for partially saturated soils, Géotech-
nique, Vol. 40, n°3, pp. 405-430.
Batzle, M. and Wang, Z. (1992) Seismic properties of pore fluids, Geophysics, Vol. 57, N° 11,
pp. 1396-1408.
Baumgarte, C., Thiercelin, M. and Klaus, D. (1999) Case studies of expanding cement to prevent
microannular formation, Paper SPE 56535 presented at the Annual Technical Conference and
Exhibition, Houston, Texas, October 3-6.
Bear, J. and Buchlin, J.M. (1991) Modelling and applications of transport phenomena in porous media,
Kluwer, Academic Publishers, Dordrecht.
Beaudoin, J.J. (2000) Calcium hydroxide in cement matrices: physico-mechanical and physico-
chemical contributions, In Materials Science of Concrete: Calcium Hydroxide in Concrete,
Special Volume, pp. 131-142.
 180 References

Bemer, E., Boutéca, M., Vincké, O., Hoteit, N. and Ozanam, O. (2001) Poromechanics: from linear to
nonlinear poroelasticity and poroviscoelasticity, Oil & Gas Science and Technology – Rev. IFP,
Vol. 56, N° 6, pp. 531-544.
Bemer, E. and Lombard, J.-M. (2008) An experimental investigation of the evolution of rock porome-
chanical properties associated to chemical alteration processes, Proceedings of Third Internatio-
nal Conference on Coupled T-H-M-C Processes in Geo-systems (GEOProc 08), Lille, France.
Berryman, J. G. (1995) Mixture theories for rock properties, in Rock physics and phase relations – A
handbook of physical constants, edited by T. J. Ahrens, American Geophysical Union,
Washington, pp. 205-228.
Bévillon, D. (2000) Couplage d’un modèle de gisement et d’un modèle mécanique: Application à
l’étude de la compaction des réservoirs pétroliers et de la subsidence associée. PhD Thesis, Uni-
versité des Sciences et Technologies – Lille.
Biot, M. A. (1941) General theory of three-dimensional consolidation, Journal of Applied Physics,
Vol. 12, pp. 155-164.
Birch, F. (1961) The velocity of compressional waves in rocks to 10 kilobars, Part 2. J. Geophys. Res.,
66, pp. 2199-2224.
Boidy, E. (2002) Modélisation numérique du comportement différé des cavités souterraines, PhD The-
https://telegram.me/Geologybooks

sis, Université Joseph Fourier – Grenoble 1.

Bosma, M.G.R., Cornelissen, E.K. and Schwing, A. (2000) Improved experimental characterisation of
cement/rubber zonal isolation materials, Paper SPE 64762 presented at the International Oil and
Gas Conference and Exhibition in China held in Beijing, China, November 7-10.
Bourbié, T., Coussy, O. and Zinszner, B. (1986) Acoustique des milieux poreux, Editions Technip,
Paris.
Boutéca, M., Lessi, J. and Sarda, J-P. (1983) Stress changes induced by fluid injection in a porous
layer around a wellbore, Proceedings 24th US Symposium on Rock Mechanics, College Station,
Texas.
Boutéca, M. and Sarda, J-P. (1987) Etat de l’art en fracturation hydraulique, Oil & Gas Science and
Technology – Rev. IFP, Vol. 42, N° 1, pp. 39-75.
Boutéca, M. (1988) Hydraulic fracturing model based on a closed form : tests , analysis of fracture
geometry and containment, SPE Production Engineering, Vol. 3, N° 4, pp. 445-454.
Boutéca, M. (1989) Three-dimensional hydraulic fracture propagation: derivation of closed forms
from potential theory, Proceedings ISRM/SPE International Symposium on Rock at Great
Depth, Vol. 2, pp. 971-979, Ed. V. Maury & D. Fourmaintraux, A.A. Balkema.
Boutéca, M. (1992) Elements of poro-elasticity for reservoir engineering, Oil & Gas Science and
Technology – Rev. IFP, Vol. 47, N° 4, pp. 479-490.
Boutéca, M., Sarda, J.-P. and Schneider, F. (1996), Les subsidences provoquées par la production des
fluides souterrains, Oil & Gas Science and Technology – Rev. IFP, Vol. 51, N° 3, pp. 349-364.
Boutéca, M.J., Sarda, J-P. and Vincké, O. (2000) Constitutive law for permeability evolution of
sandstones during depletion, Paper SPE 58717 presented at the SPE International Symposium
on Formation Damage Control held in Lafayette, Louisiana, February 23-24.
Bratli, R.K. and Risnes, R. (1981) Stability and failure of sand arches, Paper SPE 8427-PA, SPE Jour-
nal, Vol. 21, N° 2, pp. 236-248, April 1981.
Braun, R. and Jahns, E. (1998) Cracking and compaction behavior of sandstone, Paper SPE 47377 pre-
sented at the 1998 SPE/ISRM Conference on Rock Mechanics in Petroleum Engineering,
Trondheim, Norway, July 8-10.
Brink, J.L., Patzek, T.W., Silin, D.B. and Fielding, E.J. (2002) Lost Hills field trial – Incorporating
new technology for reservoir management, Paper SPE 77646 presented at the Annual Technical
Conference and Exhibition held in San Antonio, Texas, 29 September-2 October.
Bull, W. B. (1964) Alluvial fans and near-surface subsidence in western Fresno County, California,
USGS Professional Paper. 437-A, pp. 1-71.
 References 181

Burger J., Gadelle, C.P. and Marrast, J.R. (1986) Development of a chemical process for sand control,
SPE 15410 presented at the SPE Annual Technical Conference and Exhibition, New Orleans,
Louisiana, October 5-8.
Byerlee, J.D. (1978) Friction of rocks, Pure and Applied Geophysics, Vol. 116, pp. 615-626.
Calvert, D.G. and Smith, D.K. (1994) Issues and techniques of plugging and abandonment of oil and
gas wells, Paper SPE 28349 presented at the Annual Technical Conference and Exhibition, New
Orleans, September 25-28.
Carde, C. (1996) Caractérisation et modélisation de l’altération des propriétés mécaniques due à la
lixiviation des matériaux cimentaires, PhD INSA Toulouse.
Carter, B.J., Desroches, J., Ingraffea, A.R. and Wawrzynek, P.A. (2000) Simulating fully 3D hydraulic
fracturing, In Modeling in Geomechanics, Ed Zaman, Booker and Gioda, Wiley Publishers.
Castagna, J.P., Batzle, M.L. and Eastwood, R.L. (1985) Relationships between compressional-wave
and shear-wave velocities in clastic silicate rocks, Geophysics, Vol. 50, N° 4, pp. 571-581.
Cerasi, P., Papamichos, E. and Stenebraten, J. F. (2005) Quantitative sand-production prediction: Fric-
tion-dominated flow model, Paper 94791 presented at the SPE Latin American and Caribbean
Petroleum Engineering Conference held in Rio de Janeiro, Brazil, June 20-23.
Chardac, O., Murray, D., Carnegie, A. and Marsden, J.R. (2005) A proposed data acquisition program
https://telegram.me/Geologybooks

for successful geomechanics projects, Paper SPE 93182 presented at the 14th SPE Middle East
Oil and Gas Show and Conference, Bahrain, March 12-15.
Charlez, P.A. (1991) Rock mechanics, Volume 1, Theoretical fundamentals, Editions Technip, Paris.
Charlez, P.A. (1994) Exemple de modèle poroplastique: le modèle de Cam-Clay, Ecole de Mécanique
des Milieux Poreux, Aussois, France, 30 Mai-3 Juin 1994.
Charlez, P.A., Lemonnier, P., Ruffet, C., Boutéca, M. and Tan, C. (1996) Thermally induced fractu-
ring: analysis of a field case in North Sea, Paper SPE 36916 presented the SPE European Petro-
leum Conference held in MiIan, Italy, October 22-24.
Charlez, P.A. (1997) Rock mechanics, Petroleum Applications. Editions Technip, Paris.
Chavez, J-C., Carruthers, J. and McMurdy, P. (2005) Water flooding efficiency in a scenario of
multiple induced fractures, an applied geomechanical study, Paper SPE 97526 presented at the
SPE International Improved Oil Recovery Conference in Asia Pacific held in Kuala Lumpur,
Malaysia, December 5-6.
Chen, G., Chenevert, M., Sharma, M. and Yu, M. (2003) A study of wellbore stability in shales inclu-
ding poroelastic, chemical, and thermal effects, Journal of Petroleum Science and Engineering,
Vol. 38, N° 3-4, June 2003, pp. 167-176.
Chen, X., Tan, C.P. and Detournay, C. (2003) A study on wellbore stability in fractured rock masses
with impact of mud infiltration, Journal of Petroleum Science and Engineering, Vol. 38, N° 3,
pp. 145-154.
Chi, J.M., Huang, R. and Yang, C.C. (2002) Effects of carbonation on the mechanical properties and
durability of concrete using accelerated testing method, Journal of Marine Science and Techno-
logy, Vol. 10, No. 1, pp. 14-20.
Chin, L.Y. and Thomas, L.K. (1999) Fully coupled analysis of improved oil recovery by reservoir
compaction, Paper SPE 56753 presented at the SPE Annual Technical Conference and Exhibi-
tion, Houston, Texas, October 3-6.
Christensen, O., Janbu, N. and Jones, M.E. (1988) Subsidence due to oil-gas production, Proceedings
of the International Conference on Behaviour of Offshore Structures (BOSS’ 88), Trondheim,
Norway, V1, pp. 143-157.
Cipolla, C.L. and Wrigth, C.A. (2002) Diagnostic techniques to understand hydraulic fracturing: what?
Why? and How?, Paper SPE 75359, SPE Production & Facilities, Vol. 17, N° 1, pp. 23-35.
Cook, J. M., Sheppard, M. C. and Houwen, O. H. (1991) Effects of strain rate and confining pressure
on the deformation and failure of shale. Paper SPE 19944 presented at the 1990 IADC/SPE Dril-
ling Conference held in Houston, Texas, February 27-March 2.
 182 References

Cooper, G. A. and Hatherly, P. (2003) Prediction of rock mechanical properties from wireline data and
their use in drilling simulation, Paper SPE 83509 presented at the SPE Western Regional/AAPG
Pacific Section Joint Meeting held in Long Beach, California, May19-24.
Cossé, R. (1993) Basic of reservoir engineering, Editions Technip, Paris.
Coussy, O. (1991) Mécanique des milieux poreux, Editions Technip, Paris.
Coussy, O. (1995) Mechanics of Porous Continua, John Wiley & Sons Ltd., Chichester.
Coussy, O. (2004) Poromechanics, John Wiley & Sons Ltd., Chichester.
Crawshaw, J. P. and Frigaard, I. (1999) Cement plugs: stability and failure by buoyancy-driven mecha-
nism, Paper SPE 56959 presented at the Offshore Europe Conference held in Aberdeen, Scot-
land, September 7-9.
Cunningham, R.A. (1978) An empirical approach for relating drilling parameters, Paper SPE 6715,
Journal of Petroleum Technology, Vol. 30, N° 7, pp. 987-991, July 1978
Damjanac, B. and Detournay, E. (1995) Numerical modeling of normal wedge indentation in rocks, in
Rock Mechanics, Proceedings of the 35th U.S. Symposium, University of Nevada, Reno,
pp. 349-354. J.J.K. Daemen and R. A. Schultz, Balkema Editions, Rotterdam.
https://telegram.me/Geologybooks

Dangla, P., Malinski, L. and Coussy, O. (1997) Plasticity and imbibition-drainage curves for unsatura-
ted soils: A unified approach, Proc. of the 6th International Symposium on Numerical Models in
Geomechanics (NUMOGVI), Montréal, Quebec, July 2-4, Edited by S. Pietruszczak and G.
Pande. A.A. Balkema, Rotterdam, the Netherlands. pp. 141-147.
De Borst, R. and Nauta, P. (1985) Non-orthogonal cracks in a smeared finite element model. Enginee-
ring Computations, Vol. 2, N° 1, pp. 35-46.
De Sa, A. N., Tavares, A. F. C. and do Carmo Marques (1989) Gravel Pack in offshore wells, Proc. of
the 21th Annual Offshore Technology Conference, Houston, paper 6041.
De Waal, J.A. and Smits, R.M.M. (1988) Prediction of reservoir compaction and surface, Paper SPE
14214, SPE Formation Evaluation, Vol. 3, N° 2, pp. 340-346.
Deflandre, J-P. and Huguet, F. (2002) Microseismic monitoring on gas storage reservoirs: a ten-year
experience, Proc. 17th World Petroleum Congress, Rio de Janeiro, Brazil.
Deliac, E (1986) Optimisation des machines d’abattage à pics, PhD Thesis, Université Pierre et Marie
Curie.
Desroches, J. and Kurkjian, A.L. (1998) Applications of wireline stress measurements, SPE 48960
presented at the SPE Annual Technical Conference and Exhibition, New Orleans, September
27-30.
Desroches, J., Marsden, J.R. and Colley, N., M. (1998) Wireline open-hole stress tests in a tight gas
sandstone, Proceedings of the International Gas Research Conference, Cannes, France, Novem-
ber 6-9.
Detienne, J-L., Danquigny, J., Lacourie, Y. and Espy, M. (2002) Produced water re-injection on a low
permeability carbonaceous reservoir, Paper SPE 78482 presented at the 10th Abu Dhabi Interna-
tional Petroleum Exhibition and Conference, October 13-16.
Detournay, E. and Atkinson, C. (1991) Influence of pore pressure on the drilling response of PDC bits,
Rock Mechanics as a Multidisciplinary Science, Ed. by J.C. Roegiers, Balkema, Rotterdam,
pp. 539-547.
Detournay, E. and Defourny, P. (1992) A phenomenological model for the drilling action of drag bits,
International Journal of Rock Mechanics and Mining Sciences & Geomechanics, Vol. 29, N° 1,
pp. 13-23.
Detournay, E. (1996) Models for rock cutting, Proceedings of the Tool-Rock Interaction Workshop
held during the 2nd North American Rock Mechanics Symposium (NARMS-2), Montreal, June
22, Editions Balkema.
 References 183

Dubucq, D., Chalier, G. and Xavier, J-P. (2008) Steam chamber monitoring using surface deformation
from radar (PS-InSAR) and tiltmeter measurements, Proceedings of the Canadian CSPG CSEG
CWLS Convention.
Dusseault, M. B. and Santarelli, F. J. (1989) A conceptual model for massive solids production in
poorly consolidated sandstones, Proceedings ISRM/SPE International Symposium on Rock at
Great Depth, Vol. 2, Balkema, pp. 789-798.
Dusseault, M, Geilickman, M. and Spanos, T. (1998) Mechanisms of sand production in heavy oils,
Proceedings 7th UNITAR Conference on Heavy Oil and Tar Sands, Beijing, N° 1998.123.
Economides, M.J., Watters, L.T. and Dunn-Norman, S. (1998) Petroleum well construction, John
Wiley & Sons.
Egermann, P., Bemer, E. and Zinszner, B. (2006) An experimental investigation of the rock properties
evolution associated to different levels of CO2 injection like alteration processes, Paper SCA
2006-34, Proceedings of the International Symposium of the Society of Core Analysts, Trond-
heim, Norway, September 12-16.
Eilers, L.H. and Root, R.L. (1974) Long-term effects of high temperature on strength retrogression of
cements, SPE, Paper 5028 presented at the 49th Annual Fall Meeting of the SPE of AIME,
Houston, Texas, USA, October 6-9.
https://telegram.me/Geologybooks

Elkins, L., Morton, D. and Blackwell, W. (1972) Experimental fireflood in a very viscous oil-consoli-
dated sand reservoir, Paper SPE 4086 presented at the Fall Meeting of the Society of Petroleum
Engineers of AIME, 8-11 October 1972, San Antonio, Texas.
Ellis, R.C. (1998) An overview of frac packs: a technical revolution (evolution) process, Journal of
Petroleum Technology, Vol. 50, N° 1, pp 66-68.
Enever, J.R., Yassir, N., Willoughby, D.R. and Addis, M.A. (1996) Recent experience with extended
leak-off tests for in situ stress measurements in Australia, Australian Petroleum Production and
Exploration Association Journal, pp. 528-535.
Engstrøm, F. (1991) Rock mechanical properties of danish North Sea Chalk, Proceedings of the Fourth
North Sea Chalk Symposium. Deauville, France, September 21-23.
Eronini, I. E., Somerton, W. H. and Auslander, D. M. (1982) A Dynamic model for rotary rock dril-
ling, Transactions of ASME, Journal of Energy Resources Technology, Vol. 104, N° 2, pp. 108-
120.
Evans, G.W. and Carter, L.G. (1963) Bonding studies of cementing compositions to pipe and forma-
tions, drilling and production practice 1962, American Petroleum Institute, New York.
Falconer, I.G., Burgess, T.M. and Sheppard, M.C. (1988) Separating bit and lithology effects from
drilling mechanics data, Paper SPE 17191 presented at the SPE/IADC Drilling Conference, Dal-
las, Texas, February 28 -March 2.
Fam, W.A. and Dusseault, M.B. (1998) Borehole stability in shales: a physico-chemical perspective,
Paper SPE 47301 presented at the SPE/ISRM Conference on Rock Mechanics in Petroleum
Engineering, Eurock’98, Trondheim, Norway, Balkema, Rotterdam, pp. 285-290.
Fear, M.J., Knowlton, R.H., Meany, N. and Warren, T.M. (1995) Experts discuss drill-bit design, field
performance, Journal of Petroleum Technology, 100-104, pp. 142-143.
Finol, A.S. and Sancevic, Z. A., (1995) Subsidence in Venezuela, in subsidence due to fluid
withdrawal, developments in petroleum science, edited by G. V. Chilingarian, E. C. Donaldson
and T.F. Yen, Elsevier Science, Amsterdam.
Fjær, E., Holt, R. M., Horsrud, P., Raaen, A. M. and Risnes, R. (1992) Petroleum related rock mecha-
nics, Elsevier, Amsterdam.
Franco, J.L.A., De La Torre, H.G., Ortiz, M.A.M., Wielemaker, E., Plona, T.J., Saldungaray, P. and
Mikhaltseva, I. (2005) Using shear-wave anisotropy to optimize reservoir drainage and improve
production in low-permeability formations in the North of Mexico, Paper SPE 96808 presented
at the SPE Annual Technical Conference and Exhibition held in Dallas, Texas, October 9-12.
 184 References

Fung L. S. K., Buchanan, L., and Wan, R. G. (1992) Coupled geomechanical-thermal simulation for
deforming heavy-oil reservoirs. Proceedings of the CIM Annual Technical Conference, Calgary,
June 7-10.
Galloway, D.L., Hudnut, K.W., Ingebritsen, S.E., Phillips, S.P., Peltzer, G., Rogez, F. and Rosen, P.A.
(1998) InSAR detection of aquifer-system compaction and land subsidence, Antelope Valley,
Mojave Desert, California, Water Resources Research., Vol. 34, 2573-2585.
Galloway, D.L., Jones, D.R. and Ingebritsen, S.E. (1999) Land subsidence in the United States: U.S.
Geological Survey Circular 1182.
Garnier, A.J. and van Lingen, N.H. (1958) Phenomena affecting drilling rates at depth, Paper SPE
1097 presented at 33rd SPE Annual Meeting, Houston, in Petroleum Transactions, AIME,
Volume 216, 1959, pages 232-239.
Gassmann, F. (1951) Über die Elastizität Poröser Medien, Veirteljahrsschrift der Naturforschenden
Gesellschaft, in Zurich, 96, 1-23.
Gatto, P. and Carbognin, L. (1981) The Lagoon of Venice: natural environmental trend and man-
induced modification, Hydrological Sciences Bulletin, 26/4/12, pp. 379-391.
Gavito, D. G. (1996) A new rock strength model and its practical applications, Paper SPE 35322 pre-
sented at the International Petroleum Conference and Exhibition of Mexico City, March 5-7.
https://telegram.me/Geologybooks

Geertsma, J. (1973) A basic theory of subsidence due to reservoir compaction: the homogeneous case,
Verhandelingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig Genootschap,
Vol. 28, pp. 43-62.
Geertsma, J., and Van Opstal, G. (1973) A numerical technique for predicting subsidence above com-
pacting reservoirs, based on the nucleus of strain concept, Verhandlingen van het Koniklijk
Nederlands Geologisch Mijnbouwkundig Genootschap, Vol. 28, pp. 63-78.
Geertsma, J. (1989) On the alert for subsidence, Agip Review, Vol. 6, pp. 39-43.
Geilickman, M. and Dusseault, M. (1997) Dynamics of wormholes and enhancement of fluid produc-
tion, Paper 97-09 presented at the 48th Annual Technical Meeting of the Petroleum Society in
Calgary, Canada.
Geilickman, M. Dusseault, M. and Dullien, F.A. (1994) Fluid saturated solid flow with propagation of
a yielding front, Proceedings of the SPE/ISRM Conference on Rock Mechanics in Petroleum
Engineering, Eurock’94, Delft, Balkema, Rotterdam.
Geoffroy, H. (1996) Etude de l’interaction roche-outil de forage: influence de l’usure sur les
paramètres de coupe, PhD thesis, Ecole Polytechnique, Palaiseau, France.
Geoffroy, H., Nguyen Minh, D. and Putot, C. (1997) Study on interaction between rocks and worn
PDC’s cutter, Proceedings of the 36th Symposium US on Rock Mechanics ISRM, New York.
Geoffroy, H., Nguyen Minh, D., Bergues, J. and Putot, C. (1998a) Frictional contact on cutters wear flat
and evaluation of drilling parameters of a PDC bit, Paper SPE 47323 presented at the SPE/ISRM
Conference on Rock Mechanics in Petroleum Engineering, Trondheim, Norway, July 8-10.
Geoffroy, H., Nguyen Minh, D., Maitournam, H., Bergues, J. and Putot, C. (1998b) Evaluation of dril-
ling parameters of a PDC bit, in Advances in Rock Mechanics, World Scientific Publishing Co
Ed., pp. 301-314.
Gérard, B. (1996) Contribution des couplages mécanique-chimie-transfert dans la tenue à long terme
des ouvrages de stockage des déchets radioactifs, PhD Thesis, Ecole Normale Supérieure,
Cachan, France.
Gidley, J.L., Holditch, S.A., Nierode, D.E. and Veatch, R.W. (1989) Recent advances in hydraulic
fracturing, Society of Petroleum Engineers, Richardson, TX, USA., Monograph SPE Vol. 12.
Glowka, D. A. (1989) Use of single-cutter data in the analysis of PDC bit designs: Part 1 – Develop-
ment of a PDC cutting force model. Part 2 – Development and use of the PDCWEAR computer
code, Journal of Petroleum Technology, August. Vol. 41, N° 8.
Glowka, D.A. (1987) Development of a method for predicting the performance and wear of PDC drill
bits, Sandia Report 86-1745, Sandia National Laboratories.
 References 185

Green, S. J., Griffin, R. M. and Pratt, H. R. (1973) Stress-strain and failure properties of a porous
shale. Paper SPE 4242 presented at the SPE Drilling and Rock Mechanics Conference, Austin,
Texas, January 22-23.
Guéguen, Y. et Palciauskas, V. (1992) Introduction à la physique des roches, Editions Hermann, Paris.
Guéguen, Y. and Boutéca, M. (1999), Mechanical properties of rocks: Pore pressure and scale effects.
Oil & Gas Science and Technology – Rev. IFP, Vol. 54, N° 6, pp. 703-714.
Guillemot, J. (1986) Elements de géologie, Editions Technip, Paris.
Gutierrez, M. and Lewis, M. (1998) The role of geomechanics in reservoir simulation, Proceedings of
the SPE/ISRM Conference on Rock Mechanics in Petroleum Engineering, Trondheim, Norway,
Eurock’98, Trondheim, Norway, July 8-10, pp. 439-448.
Hanson, J.M. (1991) Pore pressure ahead of the bit, Paper SPE 21916 presented at SPE/IADC Drilling
Conference, Amsterdam, Netherlands, March 11-14.
Hatchell, P.J., van den Beukel, A., Molenaar, M.M., Maron, K.P., Kenter, C., Stammeijer, J.G.F., van
der Velde, J.J. and Syers, C.M. (2003) Whole earth 4D: reservoir monitoring geomechanics,
Expanded Abstracts of the 73rd Annual International Meeting of the Society of Exploration
Geophysicists, Dallas, pp. 1330-1333.
https://telegram.me/Geologybooks

Hawkes, C.D. and McLellan, P.J. (1997) A new model for predicting time-dependent failure of shales:
theory and application, Paper presented at the 48th Technical Meeting of the Petroleum Society
of CIM, Calgary, June, also published in the Journal of Canadian Petroleum Technology,
Vol. 30, N° 13, December 1999.
Hawkes, C.D., McLellan, P.J. and Bachu, S. (2005) Geomechanical factors affecting geological sto-
rage of CO2 in depleted oil and gas reservoirs, Journal of Canadian Petroleum Technology
vol. 44, N° 10, pp. 52-61.
Hebert, W.E. and Young, F.S. (1972) Estimation of formation pressure with regression models of dril-
ling data, Journal of Petroleum Technology, Volume 24, Number 1, pp. 9-15.
Heffer, K.J., Last, N.C., Koutsabeloulis, N.C., Chan, H.C.M., Gutierrez, M. and Makurat, A. (1994)
The influence of natural fractures, faults and earth stresses on reservoir performance – Geome-
chanical analysis by numerical modelling. North Sea oil and gas reservoirs – Norwegian Geo-
technical Institute (Revue), Vol. 192, pp. 201-211.
Hermansen, H., Landa, G.H., Sylte, J.E. and Thomas L.K. (2000) Experiences after 10 years of water-
flooding the Ekofisk Field, Norway. Journal of Petroleum Science and Engineering, Vol. 26,
N° 1-4, pp. 11-18.
Hettema, M.H.H., Schutjens P.M.T.M., Verboom B.J.M. and Gussinklo H.J. (1998) Production-indu-
ced compaction of sandstone reservoirs: The strong influence of field stress, Paper SPE 50630
presented at the European Petroleum Conference, 20-22 October 1998, The Hague, Netherlands.
Hettema, M., Papamichos, E. and Schutjens, P. (2002) Subsidence delay: field observations and analy-
sis, Oil & Gas Science and Technology – Rev. IFP, Vol. 57, N° 5, pp. 443-458.
Homand, S. (2000) Comportement mécanique d’une craie très poreuse avec prise en compte de l’effet
de l’eau: de l’expérience à la simulation, PhD Thesis, University of Lille.
Horsrud, P. (2001) Estimating mechanical properties of shale from empirical correlations, Paper SPE
56017, Journal SPE Drilling & Completion, Vol. 16, N° 2, pp. 68-73.
Horsrud, P., Holt, R., Sonstebo, E.F., Svano, G. and Bostrom, B. (1994) Time dependent borehole sta-
bility: laboratory studies and numerical simulation of different mechanisms in shale, Paper SPE
28060 presented at the Conference on Rock Mechanics in Petroleum Engineering, Eurock 94,
29-31 August, Delft, Netherlands.
Hu, T., Fournier, F., Royer, J.J. and Joseph, C. (2006) Impact of uncertainties in rock mechanical pro-
perties on subsidence evaluation, Proc. of the XIth International Congress of the International
Association for Mathematical Geology, University of Liège – Belgium.
 186 References

Huang, W.S., Marcum, B.E., Chase, M.R. and Yu, C.L. (1998) Cold production of heavy oil from
horizontal wells in the Frog Lake field, Paper SPE 52636, Journal SPE Reservoir Evaluation &
Engineering, Vol. 1, N° 6, pp. 551-555.
Jaeger, J. C. and Cook, N. G. W. (1979) Fundamental of rock mechanics – Third Edition, Chapman
and Hall, London.
Jardine, S., Malone, D. and Sheppard, M. (1994) Putting a damper on drilling’s bad vibrations, Oilfield
Review, January 1994.
Johnson, K.L (1985) Contact mechanics, Cambridge University Press.
Johnson, J.P., Rhett, D.W. and Siemers, W.T. (1989) Rock mechanics of the Ekofisk reservoir in the
evaluation of subsidence, Paper SPE 17854, Journal of Petroleum Technology, Vol. 41, N° 7,
pp. 717-722.
Kadri, I. B. (1991) Abnormal formation pressures in post-eocene formation, Potwar Basin, Pakistan,
Paper SPE 21920 presented at the SPE/IADC Drilling Conference, Amsterdam, Netherlands,
11-14 March.
Keaney, G.M.J., Meredith, P.G. and Murell, S.A.F. (1998) Laboratory study of permeability evolution
in a ‘tight’ sandstone under non-hydrostatic stress conditions. Paper SPE 47265 presented at the
https://telegram.me/Geologybooks

SPE/ISRM Conference on Rock Mechanics in Petroleum Engineering, Trondheim, Norway, 8-

10 July.
Kelm, C. H. and Faul, R. R. (1999) Well abandonment – A “best practices” approach can reduce
environmental risk, Paper SPE 54344 presented at the SPE Asia Pacific Oil and Gas Conference
and Exhibition, Jakarta, Malaysia, April 20-22.
Kendall, K. (1978) Complexities of compression failure, Proceedings of the Royal Society of London.
Series A, 361, pp. 245-263.
Kenter, C.J., Blanton, T.L., Schreppers, G.M.A., Baaijens, M.N. and Ramos, G.G. (1998) Compaction
study for Shearwater Field, Paper SPE 47280 presented at the SPE/ISRM Conference on Rock
Mechanics in Petroleum Engineering, Trondheim, Norway, July 8-10.
Klimentos, T., Badri, M., Hassan, S. and Zeid, A. (1999) Determination of earth stresses, azimuthal
rock-mechanical properties and fracture permeability using dipole shear sonic logging in the
Gulf of Suez, Egypt, Proceedings of the 40th Annual SPWLA Logging Symposium, Oslo,
Norway, May 30-June 3.
Koutsabeloulis, N.C., Heffer, K.J. and Wong, S. (1994) Numerical geomechanics in reservoir enginee-
ring, Computer methods and advances in geomechanics, Siriwardane and Zaman Editions, Balk-
ema, Rotterdam.
Koutsabeloulis, N.C. and Hope S. A. (1998) Coupled stress/fluid/thermal multi-phase reservoir simu-
lation studies incorporating rock mechanics, Paper SPE 47393 presented at the SPE/ISRM Con-
ference on Rock Mechanics in Petroleum Engineering, Trondheim, Norway, July 8-10, pp. 449-
454.
Kovary, K., Tisa, A., Einstein, H. H. and Franklin, J. A. (1983) Suggested methods for determining the
strength of rock materials in triaxial compression: revised version. Int. J. Rock Mech. Min. Sci.
& Geomech. Abstr., Vol. 20, No.6, pp. 283-290.
Kuru, E. and Wojtanowicz, A.K. (1988) A method for detecting in-situ PDC bit dull and lithology
change, Paper SPE 17192 presented at the SPE/IADC Drilling Conference, Dallas, Texas,
28 February-2 March 1988.
Lacy, L.L. (1997) Dynamic rock mechanics testing for optimized fracture designs, Paper SPE 38716
presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas,
October 5-8.
Lambe, T.W. and Whitman, R.W. (1969) Soil Mechanics, John Wiley &Sons, Inc, New York.
Landro, M. (1999) Discrimination between pressure and fluid saturation changes from time-lapse seis-
mic data, 69th Annual International Meeting, SEG, Expanded Abstracts, pp. 1651-1654.
 References 187

Last, N., Plumb, R., Harkness, R., Charlez, P., Alsen, J. and McLean, M. (1995) An integrated
approach to evaluating and managing wellbore instability in the Cuisiana field, Colombia, South
America, Paper SPE 30464 presented at SPE Annual Technical Conference and Exhibition held
in Dallas, Texas, October 22-25.
Laurent, J., Deflandre, J.P, Dubois, J.C. and Huguet, F. (1999) Permanent Downhole Geophones: an
efficient approach for reservoir monitoring, Proceedings of the European Applied Research
Conference on Natural Gas 1999 (“Eurogas 99”) Bochum, Germany.
Lavergne, M. (1986) Méthodes sismiques, Editions Technip, Paris.
Lawn, B.R. and Wilshaw, T.R. (1975) Fracture of brittle solids, Cambridge University Press.
Lesso, J.R. and Burgess, T.M. (1986) Pore pressure and porosity from MWD measurements, Paper
SPE 14801 presented at the SPE/IADC Drilling Conference, Dallas, Texas, February 9-12.
Le Tirant, P. and Gay, L. (1972) Manuel de fracturation hydraulique, Editions Technip, Paris.
Lewis, R. W. and Schrefler, B.A. (1998) The Finite element method in the static and dynamic defor-
mation and consolidation of porous media, Second Edition, Wiley & Sons, New York.
Lin, W. (1981) Mechanical behavior of Mesaverde shale and sandstone at high pressure, Paper SPE
9835 presented at the SPE/DOE Low Permeability Gas Reservoirs Symposium, Denver, Colo-
https://telegram.me/Geologybooks

rado, May 27-29.

Lofgren, B. E. (1975) Land subsidence due to ground-water withdrawal, Arvin-Maricopa area, Cali-
fornia, USGS Professional Paper 437-D, Washington, D.C.
Longuemare, P., Détienne, J-L., Lemonnier, P., Boutéca, M. and Onaisi, A. (2001) Numerical mode-
ling of fracture propagation induced by water injection/re-injection, Paper SPE 68974 presented
at the SPE European Formation Damage Conference held in The Hague, The Netherlands, May
21-22.
Longuemare, P., Mainguy, M., Lemonnier, P., Onaisi, A., Gérard, Ch. and Koutsabeloulis, N. (2002)
Geomechanics in reservoir simulation: overview of coupling methods and field case study, Oil
& Gas Science and Technology – Rev. IFP, Vol. 57, No. 5, pp. 471-483.
Lumley, D. E. (1995) Seismic time-lapse monitoring of subsurface fluid flow, PhD. Thesis, Stanford
University.
Mainguy, M. and Longuemare, P. (2002) Coupling fluid flow and rock mechanics: formulations of the
partial coupling between reservoir and geomechanical simulators, Oil & Gas Science and Tech-
nology – Rev. IFP, Vol. 57, N° 4, pp. 355-367.
Mainguy, M., Longuemare, P., Audibert, A. and Lécolier, E. (2007) Analysing the risk of well plug
failure after abandonment, Oil & Gas Science and Technology – Rev. IFP, Vol. 62, N° 3,
pp. 311-324.
Marsden, S.S., and Davis, S.N. (1967) Geological subsidence, Scientific American, 216(6), pp. 93-108
Matà, C. (2001) Etude expérimentale et modélisation mécanique des effets du balayage à l’eau dans
une craie saturée d’huile – Application à l’industrie pétrolière. PhD Thesis, Ecole Nationale des
Ponts et Chaussées, Paris.
Maury, V. and Sauzay, J.M. (1987) Borehole instability: case histories, rock mechanics approach and
results, Paper SPE 16051 presented at the SPE/IADC Drilling Conference, New Orleans, Loui-
siana, March 15-18.
Maury, V. (1994) Rock mechanics over the last ten years in oil and gas production: safety and eco-
nomy, 14th World Petroleum Congress, Stavanger, Norway, Paper N° 26135.
Maury, V., Piau, J.-M. and Halle, G. (1996) Subsidence induced by water injection in water sensitive
reservoir rocks: the example of Ekofisk, Paper SPE 36890 presented at the SPE European Petro-
leum Conference held in Milan, Italy, October 22-24.
McLean, M.R. and Addis, M.A. (1990) Wellbore stability: the effect of strength criteria on mud
weight recommandations, Paper SPE 20405 presented at the 65th SPE Annual Technical
Conference and Exhibition, New Orleans, September 23-26.
 188 References

McLellan, P. (1994) Assessing the risk of wellbore instability in horizontal and inclined wells, Paper
HWC94-14 presented at the Canadian SPE/CIM/CANMET International Conference on Recent
Advances in Horizontal Well Applications, Calgary, Canada, March 20-23.
McLellan, P.J. and Cormier, K. (1996) Borehole instability in fissile, dipping shales, northeastern
British Columbia, Paper SPE 35634 presented at the SPE Gas Technology Symposium, Calgary,
Alberta, Canada, 28 April-1 May.
Mendelsohn, D.A. (1984) A review of hydraulic fracture modeling, part 1: general concepts, 2D
models, motivation for 3D modeling, Journal of Energy Resources Technology, Vol. 106, N° 3,
pp. 369-376.
Mendelsohn, D.A. (1984) A review of hydraulic fracture modeling, part II: 3D modeling and vertical
growth in layered rock, Journal of Energy Resources Technology, Dec 1984, Vol. 106, N° 4,
pp. 543-553.
Merle, H.A., Kentie, C.J.P., van Opstal, G.H.C. and Schneider, G.M.G. (1976) The Bachaquero study
– A composite analysis of the behavior of a compaction drive/solution gas drive reservoir,
Journal of Petroleum Technology, September, Vol. 28, N° 9, pp. 1107-1115.
Mindlin, R. D. (1949) Compliance of elastic bodies in contact, Journal of Applied Mechanics, Vol. 16,
https://telegram.me/Geologybooks

pp. 259-262.
Minkoff, S. E., Stone, C. M., Bryant, S. and Peszynska M. (2004) Coupled geomechanics and flow
simulation for time-lapse seismic modelling, Geophysics, Vol. 69, N° 1, pp. 200-211.
Mobach, E. and Gussinklo, H. J. (1994) In situ reservoir compaction monitoring in the Groningen
field, Proceedings of EUROCK’94, Rock Mechanics in Petroleum Engineering, A. A. Balkema
Publ., Rotterdam, The Netherlands, pp. 535-547.
Moricca, G. Ripa, G., Sanfilippo, F. and Santarelli, F.J. (1994) Basin scale rock mechanics: field
observations of sand production, Paper SPE 28066 presented at the SPE/ISRM Rock Mechanics
in Petroleum Engineering Conference Delft, The Netherlands, August 29-31.
Morita, N., Whitfill, D.L., Nygaard, O. and Bale, A. (1989) A quick method to determine subsidence,
compaction and in-situ stresses induced by reservoir depletion, Journal of Petroleum Techno-
logy, January 1989, SPE 17150, Vol. 41, N° 1, pp. 71-83.
Mouchet, J. P. and Mitchell, A. (1989). Abnormal pressure while drilling. Manuals techniques 2,
Boussens, France, Elf Aquitaine Editions.
Mouton, D.E., Wilton, B.S. and Ramos, G.G. (1998) Optimizing drilling strategies in a tectonic belt,
Pagerungan field, North of Bali, Paper SPE 39357 presented at the IADC/SPE Drilling Confer-
ence, Dallas, Texas, March 3-6.
Nguyen, J.P. (1993) Le forage, Editions Technip, Paris.
Nguyen Minh, D. (1974) Contribution à l’étude de la taille des roches, PhD thesis, University of Paris VI.
Noik, C. and Rivereau, A. (1999) Oilwell cement durability, Paper SPE 56538 presented at the SPE
Annual Technical Conference and Exhibition, Houston, Texas, October 3-6.
Nolte, K.G. (1979) Determination of fracture parameters from fracturing pressure decline, Paper SPE
8341 presented at the SPE Annual Technical Conference and Exhibition, Las Vegas, Nevada,
September 23-26.
Nolte, K.G. and Smith, M.B. (1981) Interpretation of fracturing pressures, Journal of Petroleum Tech-
nology, Vol. 33, N° 9, Sept 1981, pp. 1767-1775.
Nouri, A., Vaziri, H., Belhaj H. and Islam, R. (2004) Sand production control: a new set of criteria for
modeling based on large scale transient experiments and numerical investigation, SPE Jounal,
Vol. 11, N° 2, pp. 227-237.
Nouri, A., Vaziri, H., Kuru, E. and Islam, R. (2006) A comparison of two sanding criteria in physical
and numerical modeling of sand production, Journal of Petroleum Science and Engineering,
Vol. 50, N° 1, pp. 55-70.
 References 189

Nur, A. and Murphy, W. (1981) Wave velocities and attenuation in porous media with fluids, Procee-
dings of the 4th International Conference on Continuum Models of Discrete Systems, Stoc-
kholm, pp. 311-327.
Onaisi, A., Durand, C. and Audibert, A. (1994) Role of hydration state of shales in borehole stability
studies, Paper SPE 28062 presented at the SPE/ISRM Conference on Rock Mechanics in Petro-
leum Engineering, Eurock’94, Delft, Netherlands.
Osorio, J. G., Chen, H.Y., Teufel, L. and Schaffer, S. (1998) A two-domain, fully coupled fluid-flow/
Geomechanical simulation model for reservoirs with stress-sensitive mechanical and fluid-flow
properties. Paper SPE 28062 presented at the SPE/ISRM Conference on Rock Mechanics in
Petroleum Engineering, Eurock’98, Trondheim, Norway, July 8-10.
Papamichos, E., Stenebraten, J., Cerasi, P., Lavrov, A., Vardoulakis, I., Fuh, G.H., Brignoli, M.,
Goncalves de Castro, C.J. and Havmoller, O. (2008) Rock type and hole failure pattern effects
on sand production, Proceedings of the 42nd US Rock Mechanics Symposium, San Francisco,
June 29-July 2, Paper ARMA 08-217.
Patron, E. (1999) Poromécanique des joints en graphite expansé matricé, PhD Thesis, Ecole Nationale
des Ponts et Chaussées, Paris.
https://telegram.me/Geologybooks

Parcevaux, P. A. and Sault, P. H. (1984) Cement shrinkage and elasticity: a new approach for a good
zonal isolation, Paper SPE 13176 presented at the 59th SPE Annual Technical Conference and
Exhibition, Houston, September 16-19.
Pavone, D. R. and Desplans, J. P. (1994) Application of sampling rate downhole measurements for
analysis and cure of stick-slip in drilling, Paper SPE presented at the SPE Annual Technical
Conference and Exhibition, New Orleans, Louisiana, September 25-28.
Peltier, B. and Atkinson, C. (1986) Dynamic pore pressure ahead of the bit, Paper SPE 14787, SPE
Drilling Engineering, Vol. 2, N° 4, pp. 351-358.
Perkins, T. K. and Weingarten (1988) Stability and failure of spherical cavities in unconsolitated sand
and weakly consolidated rock, Paper SPE 18244 presented at SPE Annual Technical Conference
and Exhibition, Houston, Texas, October 2-5.
Perkins, T.K. and Gonzalez, J.A. (1985) The effect of thermoelastic stressses on injection well fractu-
ring, Paper SPE 11332, SPE Journal, Vol. 25, N° 1, pp. 78-88.
Perrodon, A. (1989) Dynamique et méthode d’étude des bassins sédimentaires, Editions Technip,
Paris.
Pessier, R.C (1994) Different shales dictate fundamentally different strategies in hydraulics, bit selec-
tion and operating practices, Paper SPE 28322 presented at the 69th Annual Technical Confer-
ence and Exhibition, New Orleans, Louisiana, September 25-28.
Pessier, R. C. and Fear, M. J. (1992) Quantifying common drilling problems with mechanical specific
energy and a bit-specific coefficient of sliding friction, Paper SPE 24584 presented at the SPE
Annual Technical Conference and Exhibition, Washington, D.C., October 4-7.
Pistre, V., Kinoshita, T., Endo, T., Schilling, K., Pabon, J., Sinha, B., Plona, T., Ikegami, T. and John-
son, D. (2005) A modular wireline sonic tool for measurements of 3D (azimuthal, radial and
axial) formation acoustic properties, Proceedings of the 46th SPWLA Annual Logging Sympo-
sium held in New Orleans, Louisiana, United States, June 26-29.
Plona, T.J., Kane, M.R., Sinha, B.K., Walsh, J. and Viloria, O. (2000) Using acoustic anisotropy,
Proceedings 41th Annual SPWLA Logging Symposium.
Plumb, R.A. (1994) Influence of composition and texture on the failure properties of clastic rocks.
Proceedings of the Eurock 1994. Rock Mechanics in Petroleum Engineering, Balkema,
Rotterdam, pp. 13-20.
Plumb, R., Papanastasiou, P. and Last, N. (1998) Constraining the state of stress in tectonically active
settings, Proceedings of the SPE/ISRM Conference on Rock Mechanics in Petroleum Enginee-
ring: Eurok 98, Trondheim, July 8-10.
 190 References

Plumb, R.A. and Hickman, S.J. (1985) Stress induced borehole elongation: a comparaison between the
four arm dipmeter and yhe borehole televiewer in the Auburn geothermal well, Journal of Geo-
physical Research, Vol. 90, N° B7, pp. 5513-5522.
Poland, J. F. and Lofgren, B. E. (1984) Case History No 9.13 San Joaquin Valley, California, USA, in
Poland, J.F., ed., Guidebook to studies of land subsidence due to ground-water withdrawal,
UNESCO Studies and Reports in Hydrology, N°. 40, pp. 263-277.
Putot, C.J.M., Guesnon J., Perreau P.J. and Constantinescu A. (2000) Quantifying drilling efficiency
and disruption: field data vs. theoretical model., Journal SPE Drilling & Completion, Vol. 15,
N° 2, pp. 118-125.
Putot, C.J.M. and Constantinescu, A. (1997) Modèle de foration fondé sur le couplage des effets
d’évacuation des déblais et de coupe de la roche, Symposium Saint-Venant, Editions ENPC,
Paris.
Putot, C. (1995) Modèle de foration fondé sur le couplage de la destruction de la roche et d’évacuation
des débris, Oil & Gas Science and Technology – Rev. IFP, Vol. 50, N° 4, pp. 491-516.
Putot, C.J.M., Perreau P.J. and Constantinescu A. (1998) Field data vs. Theoretical model to quantify
drilling efficiency and disruption, Paper SPE 50579 presented at the SPE European Petroleum
Conference, The Hague, Netherlands, October 20-22.
https://telegram.me/Geologybooks

Ramos, G.G., Katahara, K.W., Gray, J.D. and Knox, D.J.W. (1994) Sand production in vertical and
horizontal wells in a friable sanstone formation, North Sea, Paper SPE 28065 presented at the
SPE/ISRM Conference on Rock Mechanics in Petroleum Engineering, Eurock’94, Delft,
Netherlands.
Rasolofosaon, P. and Zinszner, B. (2002) Vérification expérimentale de la formule de Gassmann dans
les calcaires poreux, Oil & Gas Science and Technology, Rev. IFP., Vol. 57, N° 2, pp. 129-138
Rasolofosaon, P., and Zinszner, B. (2004) Laboratory petroacoustics for seismic monitoring feasibility
study, The Leading Edge, Vol. 23, N° 3, pp. 252-258.
Ravi, K., Bosma, M. and Gastebled, O. (2002) Safe and economic gas wells through cement design for
life of the well, Paper 75700 presented at the SPE Gas Technology Symposium, Calgary,
Alberta, Canada, April 30-May 2.
Reinecker, J., Heidbach, O., Tingay, M., Sperner, B. and Müller, B. (2005) The release 2005 of the
world stress map (available online at www.world-stress-map.org)
Renard, G., Nauroy, J.-F., Deruyter, Ch., Moulu, J.-C., Sarda, J.-P. and Le Romancer, J.-F (2000)
Production froide des huiles visqueuses, Oil & Gas Science and Technology – Rev. IFP, Vol. 55,
No. 1, pp. 35-66.
Rhett, D.W. and Teufel, L.W. (1992a) Effect of reservoir stress path on compressibility and permeabi-
lity of sandstones, Paper SPE 24756 presented at the SPE Annual Technical Conference and
Exhibition, Washington, D.C., October 4-7.
Rhett, D.W. and Teufel, L.W. (1992b) Failure criteria for high porosity North Sea Chalk. Proceedings
of the Fourth North Sea Chalk Symposium, Deauville, France, September 21-23.
Roscoe, K.H. and Burland, J.B. (1968) On the generalised stress-strain behaviour of ‘wet clay’, Engi-
neering Plasticity, Eds Heyman J. and Leckie, F.A., Cambridge University Press, pp. 535-609.
Rousset, G. (1988) Comportement mécanique des argiles profondes – Application au stockage de
déchets radioactifs, PhD Thesis, Ecole Nationale des Ponts et Chaussées.
Ruisten, H., Teufel, L.W. and Rhett, D. (1996) Influence of reservoir stress path on deformation and
permeability of weakly cemented sandstone reservoirs, Paper SPE 36535 presented at the SPE
Annual Technical Conference and Exhibition, Denver, Colorado, October 6-9.
Rutqvist, J., Birkholzer, J.T. and Chin-Fu Tsang, (2007) Coupled reservoir – geomechanical analysis
of the potential for tensile and shear failure associated with CO2 injection in multilayered
reservoir – caprock systems, International Journal of Rock Mechanics & Mining Sciences,
Vol. 45, pp. 132-143.
 References 191

Santarelli, F.J., Dardeau, C. and Zurdo, C. (1992) Drilling through highly fractured formations: a pro-
blem, a model, and a cure, Paper SPE 24592 presented at the SPE Annual Technical Conference
and Exhibition, Washington, D.C., October 4-7.
Santarelli, F.J., Tronvoll J.T., M. Svennekjaer, Skeie H., Henriksen H. and Bratli R.K. (1998) Reser-
voir stress-path: The depletion and the rebound, Paper SPE 47350 presented at the SPE/ISRM
Conference on Rock Mechanics in Petroleum Engineering, Trondheim, Norway, July 8-10.
Sarda, J-P. (1984) Fracturation hydraulique, Rapport IFP 31911.
Sargand, S. M. and Hazen, G. A. (1987) Deformation behaviour of shales, International Journal Rock
Mechanics Mining Sciences, Vol. 24, No.6, pp. 365-370.
Schofield and Wroth (1968) Critical state soil mechanics, McGraw Hill, London.
Scholz, C. H. (1990) The mechanics of earthquakes and faulting, Cambridge University Press.
Schutjens, P.M.T.M. and de Ruig, H. (1996) The influence of stress path and permeability of an over-
pressurised reservoir sandstone: some experimental data, Physics and Chemistry of the Earth,
Vol. 22, N° 1-2, pp. 97-103.
Schutjens, P.M.T.M., Blanton T.L., Martin J.W., Lehr, B.C. and Baaijens, M.N. (1998) Depletion-
https://telegram.me/Geologybooks

induced compaction of an overpressurised reservoir sandstone: an experimental approach. Paper

SPE 47277 presented at the SPE/ISRM Rock Mechanics in Petroleum Engineering, Trondheim,
Norway, July 8-10.
Seidel, F. A. and Greene, T. G. (1985) Use of expanding cement improves bonding and aids in elimi-
nating annular gas migration in Hobbs Grayburg-San Andres wells, Paper SPE 14434 presented
at the 60th Annual Technical Conference and Exhibition held in Las Vegas, NV, Sept. 22-25.
Servant, G., Nauroy, J-F. and Bemer, E. (2007) Coupled geomechanical transport modelling of sand
production in petroleum reservoirs, Proceedings of the 2nd International Conference on Porous
Media and its Applications in Science and Engineering, ICPM2, Kauai, Hawaii, USA.
Settari, A. and Mourits, F.M. (1994) Coupling of geomechanics and reservoir simulation models, In
Computer Methods and Advances in Geomechanics, Siriwardane and Zaman Editors, Balkema,
Rotterdam, pp. 2151-2158.
Settari, A. and Mourits, F.M. (1998) A coupled reservoir and geomechanical simulation system, Paper
SPE 50939, SPE Journal, September, pp. 219-226.
Settari, A. and Walters, D.A. (2001) Advances in coupled geomechanical and reservoir modelling with
applications to reservoir compaction, Paper SPE 74142, SPE Journal, Vol. 6, N° 3, pp. 334-342.
Settari, A. (2002) Reservoir compaction, Paper SPE 76805, Journal of Petroleum Technology,
Vol. 54, N° 8, pp. 62-69.
Shao, J.F. and Marchina, P. (2002) A damage mechanics approach for the modelling of sand
production in heavy oil reservoirs, Paper SPE 78167 presented at the SPE/ISRM Rock
Mechanics Conference, Irving, Texas, October 20-23.
Sinha, B.K., Kane, M.R. and Frignet, B. (2000) Dipole dispersion crossover and sonic logs in a limes-
tone reservoir, Geophysics, Vol. 65, N° 2, pp. 390-407.
Smith, R.C., Beirute, R. M. and Holman, G. B. (1984) Improved method of setting successful cement
plugs, paper SPE 11415, Journal of Petroleum Technology, Vol. 36, N° 11, pp. 1897-1904.
Steiger, P. S. and Leung, P. K. (1992) Quantitative determination of the mechanical properties of sha-
les. Paper SPE 18024, SPE Drilling Engineering, Vol. 7, N° 3, pp. 181-185.
Stone, T., Bowen, G., Papanastasiou, P. and Fuller, J. (2000) Fully coupled geomechanics in a com-
mercial reservoir simulator, Paper SPE 65107 presented at the SPE European Petroleum Confe-
rence, Paris, October 24-25.
Sulak, R.M. and Danielsen, J. (1989) Reservoir aspects of Ekofisk subsidence, Paper SPE 17852,
Journal of Petroleum Technology, Vol. 41, N° 7, pp. 709-716.
 192 References

Sylte, J.E., Thomas, L.K., Rhett, D.W., Bruning, D.D. and Nagel, N.B. (1999) Water induced compac-
tion in the Ekofisk Field, Paper SPE 56426 presented at the SPE Annual Technical Conference
and Exhibition held in Houston, Texas, October 3-6.
Theron, A., De Langre, E. and Putot, C. (2001) The effect of dynamical parameters on precession in
rotary drilling, Journal of Energy Resources Technology, Vol. 123, N° 3, pp. 181-186.
Tixier, M.P., Loveless, G.W. and Anderson, R.A. (1975) Estimation of formation strength from the
mechanical properties log, Paper SPE 4532-PA., Journal of Petroleum Technology, Vol. 27,
N° 3, pp. 283-293.
Tortike, W.S. and Farouq, A. (1993) Reservoir simulation integrated with geomechanics. Journal of
Canadian Petroleum Technology, Vol. 32, N° 5, pp. 28-37.
Tremblay, B., Sedgwick, G. and Forshner, K. (1996) Imaging of sand production in a horizontal pack
by X-ray computed tomography, Paper SPE 30248-PA, SPE Formation Evaluation, Vol. 11,
N° 2, pp. 94-98.
Tremblay, B., Sedgwick, G. and Forshner, K. (1996) Simulation of cold production in heavy oil reser-
voirs: wormhole dynamics, Paper SPE 35387 presented at the SPE/DOE 10th Symposium on
Improved Oil Recovery held in Tulsa, Oklahoma, April 21-24.
Tremblay, B., Oldakowski, K. and Settari, A. (1997) Geomechanical properties of oil sands at low
https://telegram.me/Geologybooks

effective stress, Paper 97-07 presented at the 48th Annual Technical Meeting of the Petroleum
Society in Calgary, Alberta, Canada.
Tremblay, B., Sedgwick, G. and Forshner, K. (1998) Modelling of sand production from wells on pri-
mary recovery, Journal of Canadian Petroleum Technology, Vol. 37, N° 3.
Tremblay, B., Sedgwick, G., Vu, D., Lillico, D., Jossy, E., Yuan, J-Y, Babchin, A. and Sawatzky, R.
(1998) Cold production in heavy oil reservoirs, Poster presented at the 15th World Petroleum
Congress, Beijing, China.
Vajdova, V., Baud, P. and Wong, T-F. (2004) Compaction, dilatancy and failure in porous carbonate
rocks, Journal of Geophysical Research, Vol. 109, N° B5, B05204.
Van Hasselt, J.P. (1992) Reservoir compaction and surface subsidence resulting from oil and gas-
production: a review of theoretical and experimental research approaches, Geologie En
Mijnbouw, Vol. 71, N° 2, pp. 107-118.
Van Oort, E., Nicholson, J. and D’Agostino, J. (2001) Integrated borehole stability studies: key to
drilling at the technical limit and trouble cost reduction, Paper SPE 67763 presented at the SPE/
IADC Drilling Conference, Amsterdam, Netherlands, February 27 -March 1.
Vardoulakis, I., Stavropoulou, M. and Papanastasiou, P. (1996) Hydro-mechanical aspects of the sand
production problem, Transport in porous media, Vol. 22, N° 2, pp 225-244.
Vaziri, H., Barree, B., Xiao, Y., Palmer, I. and Kutas, M. (2002) What is the magic of water in produ-
cing sand?, Paper SPE 77683 presented at the SPE Annual Technical Conference and Exhibi-
tion, San Antonio, Texas, September 29 -October 2.
Vaziri, H. (2004) Integrated soil and rock mechanics: a major step towards consistent and improved
sand production prediction and management, Paper 04-461 presented at the 6th North America
Rock Mechanics Symposium (NARMS), Houston, Texas, June 5-9.
Veeken, C.A. M., Davies, D. R., Kenter, C. J. and Kooijman, A. P. (1991) Sand production prediction
review: developing an integrated approach, Proceedings 66th Annual Technical Conference of
the SPE, Dallas, Paper SPE 22792 presented at the SPE Annual Technical Conference and Exhi-
bition, Dallas, Texas, October 6-9.
Vernik L., Bruno M. and Bovberg, C. (1993) Empiricial relations between compressive strength and
porosity of siliciclastic rocks. International. Journal of Rock Mechanics and Mining Science,
Vol. 30, No.7, pp. 677-780.
Vidal, S., Longuemare, P. and Huguet, F. (2000) Integrating geomechanics and geophysics for reser-
voir seismic monitoring, feasibility studies, Paper SPE 65157 presented at the SPE European
Petroleum Conference, Paris, October 24-25.
 References 193

Vidal, S., Huguet, F. and Mechler, P. (2002) Characterising reservoir parameters by integrating seis-
mic monitoring and geomechanics, The Leading Edge, Vol. 21, pp. 295-301.
Vidal-Gilbert, S., Huguet, F., Assouline, L. and Longuemare, P. (2005) Hydromechanical modelling
of reservoir behaviour during underground gas storage exploitation, Proceeding of the 68th
European Association of Geoscientists and Engineers Conference & Exhibition, Madrid, Spain,
June 13-16.
Vidal-Gilbert, S., Nauroy, J-F. and Brosse, E., (2009) 3D geomechanical modelling for CO2 geologic
storage in the Dogger carbonates of the Paris Basin, International Journal of Greenhouse Gas
Control, Vol. 3, N° 3, pp. 288-299.
Vincké, O. (1994) An estimation of bulk moduli of sandstones as a function of confining pressure using
their petrographic and petrophysic description, Paper SPE/ISRM 28038 presented at the SPE
Conference on Rock Mechanics in Petroleum Engineering, Delft, Netherlands, August 29-31.
Vincké, O., Boutéca, M. J., Piau, J. M. and Fourmaintraux, D. (1998) Study of the effective stress at
failure, Proceedings of the Biot Conference on Poromechanics, Louvain-La-Neuve, Belgique,
September 14-16, Thimus et al Editions, Balkema, Rotterdam, pp. 635-639.
Wang, Z. (2000) Dynamic versus static elastic properties of reservoir rocks, in seismic and acoustic
velocities in reservoir rocks, SEG Geophysics Reprint Series, No 19.
https://telegram.me/Geologybooks

Ward, C.D., Coghil, K. and Broussard, M.D. (1994) The application of petrophysical data to improve
pore and fracture pressure determination in North Sea Central Graben HPHT wells, Paper SPE
28297 presented at the SPE Annual Technical Conference and Exhibition, New Orleans, Loui-
siana, September 25-28.
Wardlaw, H. W. R (1971) Optimization of rotary drilling parameters, PhD thesis, Texas University, Austin.
Warpinski, N.R., Moschovidis, Z.A., Parker, C.D. and Abou-Sayed, I.S. (1994) Comparison study of
hydraulic fracturing models test case: GRI staged field experiment N° 3, Paper SPE 275890,
SPE Production & Facilities, Vol. 9, N° 1, pp. 7-16.
Warren, T.M. (1984) Factors affecting torque for a roller cone bit, Paper SPE 11991-P.A., Journal of
Petroleum Technology, Vol. 136, N° 9, pp. 1500-1508.
Warren, T.M. and Smith, M.B. (1985) Bottomhole stress factors affecting drilling rate at depth, Paper
SPE 13381, Journal of Petroleum Technology, August, pp. 1523-1535.
Williamson, M.D., Murray, S. J., Atkins, W.S., Hamilton, T.A.P. and Copland, M.A. (1997) A review
of Zechstein drilling issues, Paper SPE 51182, SPE Drilling & Completion, Vol. 13, N° 3,
pp. 174-181.
Wong T-F., David C. and Zhu, W. (1997) The transition from brittle to cataclastic flow: mechanical
deformation, Journal of Geophysical Research, Vol. 102, N° B2, pp. 3009-3025.
Wright, C.A., Weijers, L., Davis, E.J. and Mayerhofer, M. (1999) Understanding hydraulic fracture
growth: tricky but not hopeless, Paper SPE 56724 presented at the SPE Annual Technical Con-
ference and Exhibition, Houston, Texas, October 3-6.
Wriput, D. and Zoback, M.D. (2000) Fault reactivation and fluid flow along a previously dormant
normal fault in the northern North Sea, Geology, Vol. 28, pp. 595-598.
Xu, W., Le Calvez, J. and Thiercelin, M. (2009) Characterisation of hydraulically-induced fracture
network using treatment and microseismic data in a tight-gas formation: a geomechanical
approach, Paper SPE 125237 presented at the SPE Tight Gas Completions Conference held in
San Antonio, Texas, June 15-17.
Yalamas, T., Nauroy, J.-F., Bemer, E., Dormieux, L. and Garnier, D. (2004) Sand erosion in cold
heavy oil production, Paper SPE 86949 presented at SPE International Thermal Operations and
Heavy Oil Symposium and Western Regional Meeting, Bakersfield, California, March 16-18.
Yale, D. P., Nieto, J.A. and Austin, S. P. (1995) The effect of cementation on the static and dynamic
mechanical properties of the Rotliegendes sandstone, In: Daemen, J.J.K. & Schulz, R. A. (ed.)
Proceedings of the 35th U.S. Symposium on Rock Mechanics, Balkema, Rotterdam,
pp. 169-175.
 194 References

Yale, D.P. and Crawford, B. (1998) Plasticity and Permeability in Carbonates: Dependance on Stress
Path and Porosity, Paper SPE 47582 presented at the SPE/ISRM Conference on Rock Mecha-
nics in Petroleum Engineering, Trondheim, Norway, July 8-10.
Yamamoto, K., Gutierrez, M., T. Shimamoto and Maezumi (2000) Verification of a 3D hydraulic frac-
turing model against a field case, Paper SPE 59373 presented at the SPE/DOE Improved Oil
Recovery Symposium, Tulsa, Oklahoma, April 3-5.
Yudovich, A., Chin, L.Y. and Morgan, D.R. (1989) Casing deformation in Ekofisk, Paper SPE 17856,
Journal of Petroleum Technology, Vol. 41, N° 7, pp. 729-734.
Zhang, W., Youn, S. and Doan, Q. (2005) Understanding reservoir architectures and steam chamber
growth at Christina lake, Alberta, by using 4D seismic and crosswell seismic imaging, Paper
SPE 97808 presented at the SPE/PS-CIM/CHOA International Thermal Operations and Heavy
Oil Symposium, Calgary, Alberta, Canada, November 1-3.
Zhu, W. and Wong, T.F. (1997). The transition from brittle to cataclastic flow: permeability evolution,
Journal of Geophysical Research, Vol. 102, B2, Feb. 10, pp. 3027-3041.
Zimmerman, R. W. (1991) Compressibility of sandstones, Elsevier Science Publishers B.V.,
Amsterdam.
Zinszner, B. and Pellerin, F-M. (2007) A Geoscientist’s guide to petrophysics, Editions Technip, Paris
https://telegram.me/Geologybooks

Zoback, M.D., Barto, C.A., Brudy, M.O., Castillo, D.A., Finkbeiner, T., Grollimund, B.R., Moos,
D.B., Peska, P., Ward, C.D. and Wiprut, D.J. (2003) Determination of stress orientation and
magnitude in deep wells, International Journal of Rock Mechanics & Mining Sciences, Vol. 40,
pp. 1049-1076.
 Main Symbols
https://telegram.me/Geologybooks

ρ density Kd drained bulk modulus (bulk modulus

Φ Lagrangian porosity of the skeleton)
φ Eulerian porosity Ku undrained bulk modulus
ε void ratio Ks bulk modulus of the solid matrix
ρs solid matrix density Kfl bulk modulus of the fluid
ρfl fluid density αd skeleton thermal linear coefficient of
expansion
k permeability
αu thermal linear coefficient of expansion
μ fluid viscosity
αΦ pore thermal linear coefficient of
η dynamic viscosity expansion

e elastic strain tensor αfl pore fluid thermal linear coefficient of

expansion
s stress tensor Vp compression wave velocity
s ' effective stress tensor Vs shear wave velocity
Pp pore pressure
λT thermal conductivity
T temperature
λ plastic compressibility coefficient
νd drained Poisson’s ratio κ swelling coefficient
νu undrained Poisson’s ratio g free enthalpy
E Young’s modulus S, s entropy
G shear modulus H, h specific enthalpy
ψ free energy
b Biot’s coefficient Gs potential
 INDEX

Index Terms Links

Abandonment 164
Abnormal pressure 165
Acid fracturing 117
Acoustic emission 27
Anelastic Strain Recovery 61
https://telegram.me/Geologybooks

Anisotropy 14
Aquifer 167
Arching effect 57 172
Axial permeability 139
Axial splitting 81
Axial vibrations 87

Biot’s coefficient 23 25 31
Biot-Gassmann 64
Borehole breakouts 53
Borehole imaging 128
Boundary conditions 68
Brazilian test 18
Breakdown pressure 58 119 120
Breakout 89 96
Brittle failure 26 28
Byerlee 62

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Calliper 54
Cam-Clay 30 47
Carbonates 28
Carbonation 169
Carter’s model 123
Caving 93
Cement 165
Cement plugs 169
Cement slurry 166 168
CHOPS 99
https://telegram.me/Geologybooks

Cleaning problems 83
Clogging 84
Closure pressure 58
CO2 sequestration 135 162
Coal gasification 118
Compact growth 104
Compaction 141
Compression and shear wave velocities 64
Compressive failures 88
Conduction laws 36
Consolidation 155
Consolidation treatment 117
Core discing 86
Coupled geomechanical-reservoir simulations 147 172
Coupling 136
Coussy 32
Creeping formations 95
Cunningham 73

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Cutter wear 76 77
Cuttings 71

Damage 140
Darcy 3 142
Decompacted 102
Depletion 136 151
Detournay and Defourny 75
Deviatoric stress 6
d-exponent 71
https://telegram.me/Geologybooks

Differential Strain Curve Analysis 61

Discontinuities 63
Dissolution-precipitation reactions 164
Drained bulk modulus 24
Drillability 70
Drilling resistance 75
Drilling-induced hydraulic fractures 53
Ductile failure 26 27
Dynamic elastic properties 64
Dynamic moduli 13

Effective Circulating Density 89

Effective stress 10
Elastic behaviour 23
Erosion 109 110
Eulerian porosity 3
Expandable sand screens 114

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Extended Leak-Off Test 59

External plugging 132

Falconer 74
Fault transmissivity 149
Faults 63
Filter cake 132
Fluid loss coefficient 123
FORS 72
Frac-pack 115
https://telegram.me/Geologybooks

Fracture monitoring 121

Fracture propagation 122 134
Fracture propagation pressure 89 120
Fracture tenacity 120
Fractured/cracked formations 93
Fractures 154
Fracturing tests 59
Fully coupled 144

Gassmann equation 172

Geertsma and de Klerk 123
Geomechanical model 141
Geomechanical modelling 62
Geomechanical monitoring 169
Geothermal technologies 118
Grain crushing 27
Gravel pack 114

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Hardening cap 29
Heavy oil reservoirs 99
Hollow cylinder test 111
Hydraulic conductivity 141
Hydraulic diffusivity equation 135
Hydraulic fracture propagation 174
Hydraulic fracturing 88 115 117
163 176
Hydrostatic loading 27
https://telegram.me/Geologybooks

Imaging data 54
In situ stress 52 53
Inclinometry 121
Initiation pressure 119
Instantaneous Shut In Pressure 58 120
Interferometric Synthetic Aperture Radar 160
Internal plugging 132
Iterative coupling 144

Jamming 84

Lagrangian porosity 3 26
Leakoff 119 120 125
Leak-Off Pressure 58

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Leak-Off Test 59
Limestone 17
Loading path 21 138
Loss of integrity 166

Mechanical rigidity matrix 143

Mesh 63
Microcracks 139
Microseismic 177
Microseismicity 121 128 169
https://telegram.me/Geologybooks

Mini-frac tests 120

Minimum horizontal stress 58
Modelling 31
Mohr-Coulomb failure criterion 23
Mud density 88
Mud rheology 94
Mud weight window 89

Oedometric test 18
One-way coupling 144
Over Balanced Drilling 88
Overburden 52 142

Passive Seismic Monitoring 128 173 177

Perkins, Kern, Nordgren model 125
Permanent passive seismic monitoring 174

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Permeability 3
Plastic buckling 81
Plasticity 28
Plasticity criterion 48
Plug ageing 168
Poisson’s ratio 11
Pore overpressure 72
Poroelastic behaviour 38
Poroelasticity 41
Poroelastoplastic behaviour 45
Porosity 3
https://telegram.me/Geologybooks

Pre-pack sand screens 114

Proppant 116 118
Punching 79
Punching model 81 83

Radial permeability 139

Rate Of Penetration 71
Reactivation of faults 162
Reopening pressure 61
Repressurisation 154
Reservoir simulator 140
Rock compressibility 136 137
Rock crushing 78
Rock toughness 81

SAGD 99 172

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Sand flow test 106

Sandstone 17 25
Seismic monitoring 170 172
4D seismics 169 172
Shale 17 30 92
Shear band 26 28
Shear failure 162
Shear modulus 11 24
Shear wave anisotropy 57
Sideburden 142
Slotted liners 114
https://telegram.me/Geologybooks

Softening behaviour 28
Solid matrix 41
Spalling 81 89
Specific heat capacity 34
Static moduli 11
Steam chamber 172
Steam injection 167 171
Step Rate Tests 59
Stick-slip 87
Strain-hardening 29 50
Strains 8
Stress path 8 137 151
Stress tensor 5
Subsidence 135 153 154

Tensile failures 88
Tensile strength 120
Thermal effects 130

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Thermal expansion coefficient 34 166

Thermal Induced Fracturing 132
Thermodynamics 35
Thermoelastic effects 141
Thick Walled Cylinder 106 111
Tiltmeters 128
Torque On Bit 71
Transverse vibrations 87
Triaxial 15 19

U
https://telegram.me/Geologybooks

Unconfined Compressive Strength 16

Unconsolidated formations 93
Under Balanced Drilling 88
Underburden 142
Undrained 21
Uniaxial compression test 16

Vertical Seismic Profile 176

Vertical stress 53 57
Vibration 84
Vibratory dysfunction 84

Washouts 95
Water injection 130
Watercut 101 116
Waterflooding 141

This page has been reformatted by Knovel to provide easier navigation.

 Index Terms Links

Weight On Bit 71
Wire-wrapped sand screens 114
Wormholes 101 104 164

Yield surface 30
Young’s modulus 11
https://telegram.me/Geologybooks

This page has been reformatted by Knovel to provide easier navigation.