Advanced Process Control

Advanced Process Control
Final Project
1. What if Scenarios
2. Move Suppression Tuning
3. MPC Literature Review
Submitted by
RatulDas L20315495
ObakoreAgbroko L20336679
DijalaFeludu L20341969
Muralidher Reddy Yenugu L20344366

What If Scenarios
• What If Case #1
It was assumed that the plant is long in fuel gas so that the incremental costs of fuel gas is zero. The cost of fuel
gas was changed to zero and the resulting effect on the LP cost was observed and an attempt is made to explain
the observation. Carrying out the first what if scenario, the feed temp LP cost decreased from $0.96 to-$1.59.
No change was observed in the LP cost of FC2001 and FC2002. FC 2004 remained unchanged as well. The
excel sheet for this what if study is depicted below.
Figure 1 What If Case#1 LP Cost Calculation

Disturbance variable FI-2005 was changed from 5 MBbl/d
to 6 MBbl/d.
Figure 2 What If Case#1 Step 1 Figure 3 What If Case#1 Step 150

AI-2020 limits were set to 2.5
The move suppressions were changed to [1 0.2 0.2 1].

• What If Case #2
It is assumed that the plant that uses the energy from the middle reflux exchanger is down, so there for no credit
can be given for heat recovery. The value of recovered duty was set to zero and the response of the LP costs to this
change is observed.
Figure 6 What If Case#2 LP Cost Calculation
In this scenario the major change in LP cost is noticed only in the FC2002. It increases from a value of -
1.14 to 0.002 dollars

Disturbance variable FI-2005 was changed from 5 MBbl/d
to 6 MBbl/d.

AI-2020 limits were set to 2.5
The move suppressions were changed to [1 0.2 0.2 1].

Move Suppression Tuning
In this step move suppression were tuned for each MV. The suggested controller move size guidelines according
to Appendix C are given in Table 1.
Variables Move Size
FIC-2001 ≤±0.2 MBbl/d
FIC-2002 ≤±0.15 MBbl/d
TIC-2003 ≤±0.5 ˚F
FIC-2004 ≤±0.5 Mbbl/d
Table 1: Controller Move Size
These suggested move sizes were in response to:
 A change in feed flow rate (FI-2005) of ±1.0 MBbl/d
 A change in the set point of AI-2020 of ±0.5 mol%

Change in Disturbance (FI-2005)
Initially each move suppression value is set to 0.2
The disturbance FI-2005 increased from 5-6 MBbl/d and the subsequent current moves were observed
From the very 1st step we can see that the current move has exceeded the
allowable move, so there is no use in continuing with these move suppression
values, we set new move suppression values as (0.4,0.4,0.2,0.4).
The same disturbance was again introduced and the subsequent current moves
were observed, we kept on changing the Move suppression values until the current
didn’t exceed the allowable move.
The Final Move Suppression Values for this step was - (1, 1.2, 0.2, 0.8), after this no
violation was observed.
Figure 11 Step 1 after changing FI-2005 from 5 to 6 MBbl/d
Figure 12 Step 1 with new move suppression values (no violation)

Change in Disturbance (FI-2005)
The disturbance FI-2005 increased from 5-4 MBbl/d and the subsequent current moves were observed
We see that the current move value of FIC-2002 was exceeding its allowable limit,
so the move suppression values were set to (1, 1.6, 0.2, 0.8), and subsequent
current moves were observed after introducing the same disturbance., we kept on
changing the Move suppression values until the current didn’t exceed the
allowable move.
Figure 13 Step 1 after changing FI-2005 from 5 to 4 MBbl/d

Change in AI-2020 set point
The Feed was brought back up to its initial value of 5 MBbl/d. The lower and upper limits of AI-2020 was set to 2.5% from 2%.
We see that the current move value of FIC-2004 was exceeding its allowable limit,
so the move suppression values were set to (1, 1.6, 0.6, 0.8), and subsequent
current moves were observed after introducing the same disturbance, we kept on
changing the Move suppression values until the current didn’t exceed the
allowable move.
Figure 15 Step 1 after changing AI-2020 limits form 2% to 2.5%

Change in AI-2020 set point
The limits of AI-2020 is brought to its initial value of 2% and the simulation was initialized. The limits of AI-2020 is now
decreased to 1.5% from 2%,
No violations are observed in any current move values of any MV, so the
final Move Suppression values are for the MV's are given in the table
Figure 17 Step 1 after changing AI-2020 limits form 2% to 1.5%
Table 2: Final Move Suppression Values
Manipulated Variable Move Suppresion
FIC-2001 1
FIC-2002 1.6
TIC-2003 0.6
FIC-2004 0.8

Literature Survey - MPC Technology
Figure 18 Genealogy of MPC Algorithms
The following MPC technologies are examined in detail:
 MAX APC - Cutlertech
 UPID - Cutlertech
 CTC-SIM - Cutlertech
 RMPCT (Now known as Profit Controller) – Honeywell
 Connoisseur – Invensys
 DMCplus – AspenTech
 SMOC – Shell Global Solutions
 3dMPC (now known as Predict and Control) – ABB

Max APC
MAX APC utilizes the special features of UPID to have SPs,
PVs, and OPs as independent manipulated variables. When a
PV is used as an independent variable, the calculated change
in the PV is passed through a PV to OP transform to obtain
the change in the OP. MAX APC typically up dates the
transforms at a 10 second frequency to adapt the transforms
for the changes in the upstream and downstream pressures
on the control valves. The update is necessary due to
switching pumps, changing the lineup to storage, opening
bypasses, etc. The adaptation of the transform permits MAX
APC to handle the most common non-linearity in a
controller model.

UPID
UPID is an exciting new product that allows you to develop models for both
operator training simulators and advanced process control applications in
ways never before possible. UPID uses newly patented technology developed
by Dr. Charles Cutler, the creator of the Model Predictive Control algorithm
and one of the pioneers in advanced process control. By removing the
dynamics of the regulatory PID controllers, the models can be reconfigured in
a multitude of ways. With UPID, it is possible to adapt an MPC controller
model to a new regulatory control configuration without the time and
expense of retesting the unit!
Benefits
• Keep Your Controllers Up and Running
in Tip-Top Shape
• Avoid Costly Retests
• Build Operator Training Simulators at
Minimal Cost Figure 18 UPID Interface

CTC-Sim
CTC-Sim is a complete plant simulation package that includes: Plant Sim, an
off-line training simulator; Design Sim, a powerful database and tool set for
building simulations and an Operator Advisor [OA], which assist operators in
making operating decisions. [OA], using newly patented technology, offers
simulation speeds of 50-100 times faster than real time! With such fast
simulation speeds, the operator can check the consequences of any potential
changes in real-time. The software’s “What If” function allows for safe
experimentation and training on the process, using live data. [OA] can easily
be integrated into DCS systems. In the off-line version of the simulator and the
background mode for the on-line system, you can run scenarios and keep
score!
Benefit of OA™
• Improved Process Safety
• Improved Product Quality
• Reduced Unit Downtime
• Higher Stream Factor for Multivariable Controls
Improve Profitability
Figure 19 CTC Interface

Figure 20 Value Addition by MAX APC.
Examples of Implementation
 Advanced Technology Middle East (ATME) has replaced
eight multivariable controllers with MAX APC in a
Kuwait Refinery.
 KNPC and SABIC added CTC to their approved vendor
list for process Control and Optimization.
 Enterprise Business Solutions (EBS) has reported
successful completion of a MAX APC control project for
an Acetic Acid for Sipchem in Al Jubail Saudi Arabia.
 APEX Engineering has completed installation of a MAX
APC controller on a LNG Plant for Kleenheat in
Australia.

Honeywell Profit Controller (RMPCT)
Profit Controller is an integrated component of Honeywell’s Profit Suite for
Advanced Control and Optimization. It includes the tools necessary to design,
implement and maintain multiple input/multiple-output (MIMO) advanced
control applications. It has the unique ability to maintain superior process
control even with significant model mismatches that result from underlying
process changes. Profit Controller utilizes a dynamic process model to drive
maximum value by predicting future process behavior. It ensures optimal
control response by using the minimum manipulated variable movement
necessary to bring all variables within limits or to set points. With Profit
Controller, users not only benefit from project payback periods of less than a
year, but also from sustained benefits that exceed the industry norm.
What Problems Does It Solve?
Profit Controller utilizes a dynamic process model to drive maximum value
through the following steps:
 Predict future process behavior
 Control the process using the minimum manipulated variable movement
necessary to bring all process variables within limits or to set points
 Optimize the process with the remaining degrees of freedom to drive the
process to optimum operation

Benefits
Maximum Process Efficiency – The advanced multivariable control algorithm
balances performance and robustness objectives against process economics to
minimize costly process movement.
Flexibility to Meet Process Needs – A configurable control response path allows
tailoring of control performance to meet process objectives.
Optimum Control Performance - Independent feed-forward and feedback control
tuning provides optimum control performance for changes in both control targets
and process disturbances.
Enhanced Robustness – The configurable funnel-based approach to range control
delivers enhanced robustness versus target-only approaches, while providing
flexibility in control performance.
Best-in-Class Operator Interface – Profit Controller provides unmatched man-
machine interface capabilities by offering both Profit Suite™ Operator Station (a .net-
based environment compatible with all modern DCS) and the HMIWeb APC Shape
Library (for use with Experion® R31x and later). Both environments provide
maximum flexibility in the design of the user environment, workflow integration
with existing operator work processes, and diagnostic tools to promote increased
understanding of the APC applications controlling their plant. The end result is a net
increase in operator effectiveness, higher application uptimes and more appropriate
utilization of your plant’s APC investment.
Easy Maintenance - Range control design enables easier tuning and enhanced
performance. Robust control design reduces tuning needs.
Figure 21 Profit Controller Interface

Connoisseur Advanced Process Control
Connoisseur™ model-based predictive control is comprehensive, advanced
process control (APC) software that improves process profitability and
control by enhancing quality, increasing throughput, and reducing energy
usage. Applications of Connoisseur include mining, power, chemicals and
refining, among others.
Key Benefits
 Increase throughput between 1-5%
 Improve process yields by 2-10%
 Reduce specific energy consumption by 3-10%
 Enable quicker, more effective start-ups-10%
Key Capabilities
 Process Modeling - Quantifies cause and effect relationships by
accurately representing process behavior to provide better
understanding of problems and assist in controlling them
 Controller Generation - Allows the system to automatically generate a
robust and accurate multi-variable controller
 Real Time Adaptive Control - Enables the control system to be adapted
to prevailing process conditions on-line
 Constrained Optimization - Permits operation within the physical
constraints of the process, allowing Connoisseur to maximize process
potential
CONNOISSEUR FEATURES
• Easier, non-linear identification
• Process modeling
• Controller design and simulation
• Real-time adaptive control
• Constrained linear economic optimization
• Non-linear optimization control
• RBF Neural nets, can be mixed with step test derived models
• Director Executive to detect and supervise the controller mode changes
needed for changing process states
• Fully automated PRBS Testing
• ARx model option for superior load rejection performance
• Online performance monitoring and controller performance reporting
• Model ill conditioning assessment

DMCplus is the “new generation” multivariable control product
developed by Aspen Technology following its merger with Dynamic
Matrix Control Corporation and Setpoint, Inc. DMCplus continues the
tradition of technology leadership established by these companies
previously in over 1000 control applications. DMCplus is built from
proven parts:
• The [DMC]™ engine, which has demonstrated reliability and power in
hundreds of applications, maximizing client benefits.
• The SMCA™ graphical user interface and environment, which
pioneered the use of modern tools in multivariable controller design
and operation.
Product Description
DMCplus integrates an off-line system for analysis and design with an
on-line system for implementation. Off-Line System: Comprised of an
integrated suite of three programs:
• DMCplus Model — Graphical-based modeling tool with improved
data handling and analysis capabilities. Modeling is intuitive. Model
validation is powerful with prediction error simulations.
• DMCplus Build — Graphical controller configuration tool. When used
with Model, this component expedites the implementation,
management and maintenance of the controller. Context-sensitive help
screens are included to enhance productivity, especially for new users.
• DMCplus Simulate — Graphical tool for interactive evaluation and
testing of controller performance. During simulation, default plot
definition speeds development while custom plots can be quickly
configured. On-line tuning is accomplished simply by selecting the
desired window and entering new values into a spreadsheet-like form.
On-line snapshots of the controller can be uploaded to Simulate (e.g., for
initializing simulations).

Figure 22 DMCplus Architecture
Figure 23 DMCplus Graphic User Interface.

Shell Multivariable Optimising Controller
SMOC is Shell Global Solutions’ Multivariable Optimisation and Control suite
of software packages. SMOC provides the tools necessary to design,
implement and maintain multivariable Advanced Control strategies to
effectively improve plant stability and maximise plant profitability for the
Hydrocarbon Processing and Chemicals industries.
Key characteristics of SMOC:
• Highest Uptimes in industry (i.e. highest benefits)
• Use of unmeasured disturbance models and grey box models to include
apriori process know how resulting in high robustness
• Easy to use design and simulation kit (off line)
• Embedding in DCS (no special interface software or doubling of databases)
Shell Global Solutions Advanced Control engineers, with the support of in-
house instrumentation and process experts, offer the whole spectrum of
services for successfully
Figure 24 Benefit from SMOC
Figure 25 SMOC Design Interface

3dMPCTM - ABB
The 3-dimensional Multivariable Predictive Controller (3dMPC) software
suite is a process optimization package which enables production engineers
and management to achieve consistently optimum conditions throughout the
production line. The controller operates through existing instrumentation and
control equipment. No major investment or interruption of production is
required. At an attractive price, the 3dMPC package will quickly benefit any
process plant.
Figure 26 3dMPC Architecture

The 3dMPC controller provides constant predictions of process conditions
and appropriate corrections resulting in minimum deviation from optimum
conditions. Benefits include:
 Improved product quality – closer to specification targets and greater
consistency.
 Reduced raw material consumption.
 Reduced energy consumption.
 Increased production.
Figure 27 3dMPC Controller Cycle

Conclusion
Over the last decade a mathematically clean formulation of MPC emerged which allows researchers to address problems
like feasibility, stability and performance in a rigorous manner. In the non-linear area a variety of issues remain which
are technically complex but have potentially significant practical implications for stability and performance and the
computational complexity necessary to achieve them. The new software tools, however, which are becoming available for
developing first-principle models efficiently have led to a steady increase in the use of non-linear MPC in industry. There
have been several innovative proposals how to achieve robustness guarantees but no procedure suitable for an industrial
implementation has emerged. While a resolution of the aforementioned issues will undoubtedly change our
understanding of MPC and be of high scientific and educational value, it may never have more than a minor effect on the
practice of MPC. Seemingly peripheral issues like model identification and monitoring and diagnostics will continue to be
decisive factors if MPC will or will not be used for a certain application. By generalizing the on-line MPC problem to
include integer variables it will be possible to address a number of practical engineering problems directly which may
lead to a qualitative change in the type of problems for which MPC is used in industry.

Advanced Process Control

More Related Content

What's hot (20)

Viewers also liked (17)

Similar to Advanced Process Control (20)

Advanced Process Control