Cover
Title Page
Figure List
Table List
Preface
1 Introduction of Model Predictive Control
1. 1.1 Purpose of Process Control in Chemical Process Industries (CPI)
2. 1.2 Shortcomings of Simple Regulatory PID Control
3. 1.3 What Is Multivariable Model Predictive Control?
4. 1.4 Why Is a Multivariable Model Predictive Optimizing Controller Necessary?
5. 1.5 Relevance of Multivariable Predictive Control (MPC) in Chemical Process Industry in Today’s Business Environment
6. 1.6 Position of MPC in Control Hierarchy
7. 1.7 Advantage of Implementing MPC
8. 1.8 How Does MPC Extract Benefit?
9. 1.9 Application of MPC in Oil Refinery, Petrochemical, Fertilizer, and Chemical Plants, and Related Benefits
10. References
2 Theoretical Base of MPC
1. 2.1 Why MPC?
2. 2.2 Variables Used in MPC
3. 2.3 Features of MPC
4. 2.4 Brief Introduction to Model Predictive Control Techniques
5. References
3 Historical Development of Different MPC Technology
1. 3.1 History of MPC Technology
2. 3.2 Points to Consider While Selecting an MPC
3. References
4 MPC Implementation Steps
1. 4.1 Implementing a MPC Controller
2. 4.2 Summary of Steps Involved in MPC Projects with Vendor
3. References
5 Cost–Benefit Analysis of MPC before Implementation
1. 5.1 Purpose of Cost–Benefit Analysis of MPC before Implementation
2. 5.2 Overview of Cost–Benefit Analysis Procedure
3. 5.3 Detailed Benefit Estimation Procedures
4. 5.4 Case Studies
5. References
6 Assessment of Regulatory Base Control Layer in Plants
1. 6.1 Failure Mode of Control Loops and Their Remedies
2. 6.2 Control Valve Problems
3. 6.3 Sensor Problems
4. 6.4 Controller Problems
5. 6.5 Process‐Related Problems
6. 6.6 Human Factor
7. 6.7 Control Performance Assessment/Monitoring
8. 6.8 Commonly Used Control System Performance KPIs
9. 6.9 Tuning for PID Controllers
10. References
7 Functional Design of MPC Controllers
1. 7.1 What Is Functional Design?
2. 7.2 Steps in Functional Design
3. References
8 Preliminary Process Test and Step Test
1. 8.1 Pre‐Stepping, or Preliminary Process Test
2. 8.2 Step Testing
3. 8.3 Development of Step‐Testing Methodology over the Years
4. Reference
9 Model Building and System Identification
1. 9.1 Introduction to Model Building
2. 9.2 Key Issues in Model Identifications
3. 9.3 The Basic Steps of System Identification
4. 9.4 Model Structures
5. 9.5 Common Features of Commercial Identification Packages
6. References
10 Soft Sensors
1. 10.1 What Is a Soft Sensor?
2. 10.2 Why Soft Sensors Are Necessary
3. 10.3 Types of Soft Sensors
4. 10.4 Soft Sensors Development Methodology
5. 10.5 Data‐Driven Methods for Soft Sensing
6. 10.6 Open Issues and Future Steps of Soft Sensor Development
7. References
11 Offline Simulation
1. 11.1 What Is Offline Simulation?
2. 11.2 Purpose of Offline Simulation
3. 11.3 Main Task of Offline Simulation
4. 11.4 Understanding Different Tuning Parameters of Offline Simulations
5. 11.5 Different Steps to Build and Activate Simulator in an Offline PC
6. 11.6 Example of Tests Carried out in Simulator
7. 11.7 Guidelines for Choosing Tuning Parameters
8. References
12 Online Deployment of MPC Application in Real Plants
1. 12.1 What Is Online Deployment (Controller Commissioning)?
2. 12.2 Steps for Controller Commissioning
3. References
13 Online Controller Tuning
1. 13.1 What Is Online MPC Controller Tuning?
2. 13.2 Basics of Online Tuning
3. 13.3 Guidelines to Choose Different Tuning Parameters
4. References
14 Why Do Some MPC Applications Fail?
1. 14.1 What Went Wrong?
2. 14.2 Failure to Build Efficient MPC Application
3. 14.3 Contributing Failure Factors of Postimplementation MPC Application
4. 14.4 Strategies to Avoid MPC Failures
5. References
15 MPC Performance Monitoring
1. 15.1 Why Performance Assessment of MPC Application Is Necessary
2. 15.2 Types of Performance Assessment
3. 15.3 Benefit Measurement after MPC Implementation
4. 15.4 Parameters to Be Monitored for MPC Performance Evaluation
5. 15.5 KPIs to Troubleshoot Poor Performance of Multivariable Controls
6. 15.6 Exploitation of Constraints Handling and Maximization of MPC Benefit
7. References
16 Commercial MPC Vendors and Applications
1. 16.1 Basic Modules and Components of Commercial MPC Software
2. 16.2 Major Commercial MPC Software
3. 16.3 AspenTech and DMCplus
4. 16.4 RMPCT by Honeywell
5. 16.5 SMOC–Shell Global Solution
6. 16.6 Conclusion
7. References
Index
End User License Agreement

List of Tables

Chapter 01
1. Table 1.1 Typical Payback Period of MPC
2. Table 1.2 Typical Benefits of MPC Implementation in CPI
3. Table 1.3 Typical Benefits of MPC implementation in Refinery
Chapter 02
1. Table 2.1 Description of CV, MV, and DV in a Simple Distillation Column Shown in Figure 2.4
Chapter 05
1. Table 5.1 Typical Value of Factor β
2. Table 5.2 Average Value and Standard Deviation of Quality Parameters
Chapter 06
1. Table 6.1 Typical Performance of Control Loops in Industry
2. Table 6.2 Ziegler‐Nichols Tuning Parameters
3. Table 6.3 Recommended PID Tuning Parameters
4. Table 6.4 IMC Tuning Parameters
Chapter 08
1. Table 8.1 Difference between Normal Step Testing and PRBS Testing
Chapter 11
1. Table 11.1 Simulation Initial Condition File for MVs
2. Table 11.2 Simulation Initial Condition File for CVs
3. Table 11.3 Controlled Variables with Their Limits for Simulation Studies
4. Table 11.4 Controlled Variables with Their Limits for Simulation Studies
Chapter 14
1. Table 14.1 Retaining Initial MPC Benefits after 12 Months
2. Table 14.2 Commercial MPC Monitoring Tools

List of Illustrations

Chapter 01
1. Figure 1.1 Flow scheme of a simple distillation column using multivariable model predictive controller
2. Figure 1.2 Hierarchy of plant‐wide control framework
3. Figure 1.3 Expected cost vs. benefits for different levels of controls
4. Figure 1.4 Typical benefit of MPC
5. Figure 1.5 MPC stabilization effect can increase plant capacity closer to its maximum limit
6. Figure 1.6 Reduced variability allows operation closer to constraints by shifting set point
7. Figure 1.7 Operating zone limited by multiple constraints
8. Figure 1.8 Opportunity loss due to operator action
9. Figure 1.9 Advance control implementations by one of the major MPC vendors
10. Figure 1.10 Spread of MPC application across the whole spectrum of chemical process industries
Chapter 02
1. Figure 2.1 Optimum operating point vs. operator comfort zone
2. Figure 2.2 Different module of MPC
3. Figure 2.3 A general MPC calculation
4. Figure 2.4 Schematic of distillation column
5. Figure 2.5 Model of distillation column
6. Figure 2.6 CV prediction due to past MV change
7. Figure 2.7 Model reconciliation and bias update
8. Figure 2.8 Operating region of a distillation column with two manipulated variables and six controlled variables
9. Figure 2.9 Revised CV trajectory and steady state error
10. Figure 2.10 Develop a detail plan of MV movement to drive the steady state error to zero
11. Figure 2.11 Controlled variables predictions with and without control moves
12. Figure 2.12 Manipulated variables move plan for distillation column
Chapter 03
1. Figure 3.1 Brief history of development of MPC technology
Chapter 04
1. Figure 4.1 Different steps in MPC implementation project
2. Figure 4.2 Schematics of steps involved in MPC project with vendor
Chapter 05
1. Figure 5.1 Benefit estimation procedure
2. Figure 5.2 Stabilizing effect of MPC and moving of set point closer to limit
Chapter 06
1. Figure 6.1 Various probable reasons of failure of control loops
2. Figure 6.2 Valve sizing problem detection by process gain
3. Figure 6.3 Typical trends when valve stiction presents
4. Figure 6.4 Typical trends of the process having hysteresis and backlash
Chapter 07
1. Figure 7.1 Different steps in functional design
Chapter 08
1. Figure 8.1 Expectation matrix (√ definite response expected, X no response expected, ? response is doubtful)
2. Figure 8.2 Basic concept of step test
Chapter 09
1. Figure 9.1 Advantages and disadvantages of various model structures
2. Figure 9.2 Flowchart of identification process
3. Figure 9.3 System identification structure
Chapter 10
1. Figure 10.1 Types of soft sensors
2. Figure 10.2 Steps involved in developing reliable soft sensors
3. Figure 10.3 Artificial neural network architecture
4. Figure 10.4 Schematic of SVR using an e‐insensitive loss function
Chapter 11
1. Figure 11.1 Different tuning parameters
2. Figure 11.2 Hard and soft limits
Chapter 12
1. Figure 12.1 The schematic of interface of MPC controller and DCS
2. Figure 12.2 Schematic of online commissioning of the controller
Chapter 13
1. Figure 13.1 Effect of move suppression (or MV weight) on CV and MV trajectory
2. Figure 13.2 Effect of CV give up on CV trajectory and CV error
Chapter 14
1. Figure 14.1 Benefit loss over time
2. Figure 14.2 Contributing failure factors of postimplementation of MPC applications
3. Figure 14.3 Strategies for avoiding MPC failures
Chapter 16
1. Figure 16.1 Major linear MPC companies and their products
2. Figure 16.2 Basic structure of MPC software
3. Figure 16.3 Comparison of different MPC identification technology
4. Figure 16.4 DMCplus product package
5. Figure 16.5 MPC project outline: Conventional vs. adaptive approach
6. Figure 16.6 Optimization in adaptive control mode
7. Figure 16.7 RMPCT product package
8. Figure 16.8 History of SMOC
9. Figure 16.9 SMOC product package

This edition first published 2017
© 2017 John Wiley & Sons Ltd

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Sandip Kumar Lahiri to be identified as the author of this work has been asserted in accordance with law.

Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial Office
111 River Street, Hoboken, NJ 07030, USA
The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty
In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Lahiri, Sandip Kumar, 1970– author.
Title: Multivariable predictive control : applications in industry / Sandip Kumar Lahiri, Supra International Private Ltd, Vadodara, India.
Description: First edition. | Hoboken, NJ, USA : Wiley, 2017. | Includes bibliographical references and index. | Description based on print version record and CIP data provided by publisher; resource not viewed.
Identifiers: LCCN 2017010553 (print) | LCCN 2017012540 (ebook) | ISBN 9781119243519 (pdf) | ISBN 9781119243595 (epub) | ISBN 9781119243601 (cloth)
Subjects: LCSH: Predictive control. | Multivariate analysis.
Classification: LCC TJ217.6 (ebook) | LCC TJ217.6 .L34 2017 (print) | DDC 629.8–dc23
LC record available at https://lccn.loc.gov/2017010553

Cover design by Wiley
Front Cover Image: FotoBug11/Shutterstock
Back Cover Image: Paulo Vilela/Shutterstock

Figure List

Figure 1.1	Flow scheme of a simple distillation column using multivariable model predictive controller
Figure 1.2	Hierarchy of plant‐wide control framework
Figure 1.3	Expected cost vs. benefits for different levels of controls
Figure 1.4	Typical benefit of MPC
Figure 1.5	MPC stabilization effect can increase plant capacity closer to its maximum limit
Figure 1.6	Reduced variability allows operation closer to constraints by shifting set point
Figure 1.7	Operating zone limited by multiple constraints
Figure 1.8	Opportunity loss due to operator action
Figure 1.9	Advance control implementations by one of the major MPC vendors
Figure 1.10	Spread of MPC application across the whole spectrum of chemical process industries
Figure 2.1	Optimum operating point vs. operator comfort zone
Figure 2.2	Different module of MPC
Figure 2.3	A general MPC calculation
Figure 2.4	Schematic of distillation column
Figure 2.5	Model of distillation column
Figure 2.6	CV prediction due to past MV change
Figure 2.7	Model reconciliation and bias update
Figure 2.8	Operating region of a distillation column with two manipulated variables and six controlled variables
Figure 2.9	Revised CV trajectory and steady state error
Figure 2.10	Develop a detail plan of MV movement to drive the steady state error to zero
Figure 2.11	Controlled variables predictions with and without control moves
Figure 2.12	Manipulated variables move plan for distillation column
Figure 3.1	Brief history of development of MPC technology
Figure 4.1	Different steps in MPC implementation project
Figure 4.2	Schematics of steps involved in MPC project with vendor
Figure 5.1	Benefit estimation procedure
Figure 5.2	Stabilizing effect of MPC and moving of set point closer to limit
Figure 6.1	Various probable reasons of failure of control loops
Figure 6.2	Valve sizing problem detection by process gain
Figure 6.3	Typical trends when valve stiction presents
Figure 6.4	Typical trends of the process having hysteresis and backlash
Figure 7.1	Different steps in functional design
Figure 8.1	Expectation matrix (√ definite response expected, X no response expected, ? response is doubtful)
Figure 8.2	Basic concept of step test
Figure 9.1	Advantages and disadvantages of various model structures
Figure 9.2	Flowchart of identification process
Figure 9.3	System identification structure
Figure 10.1	Types of soft sensors
Figure 10.2	Steps involved in developing reliable soft sensors
Figure 10.3	Artificial neural network architecture
Figure 10.4	Schematic of SVR using an e‐insensitive loss function
Figure 11.1	Different tuning parameters
Figure 11.2	Hard and soft limits
Figure 12.1	The schematic of interface of MPC controller and DCS
Figure 12.2	Schematic of online commissioning of the controller
Figure 13.1	Effect of move suppression (or MV weight) on CV and MV trajectory
Figure 13.2	Effect of CV give up on CV trajectory and CV error
Figure 14.1	Benefit loss over time
Figure 14.2	Contributing failure factors of postimplementation of MPC applications
Figure 14.3	Strategies for avoiding MPC failures
Figure 16.1	Major linear MPC companies and their products
Figure 16.2	Basic structure of MPC software
Figure 16.3	Comparison of different MPC identification technology
Figure 16.4	DMCplus product package
Figure 16.5	MPC project outline: Conventional vs. adaptive approach
Figure 16.6	Optimization in adaptive control mode
Figure 16.7	RMPCT product package
Figure 16.8	History of SMOC
Figure 16.9	SMOC product package

Table 1.1	Typical Payback Period of MPC
Table 1.2	Typical Benefits of MPC Implementation in CPI
Table 1.3	Typical Benefits of MPC implementation in Refinery
Table 2.1	Description of CV, MV, and DV in a Simple Distillation Column Shown in Figure 2.4
Table 5.1	Typical Value of Factor β
Table 5.2	Average Value and Standard Deviation of Quality Parameters
Table 6.1	Typical Performance of Control Loops in Industry
Table 6.2	Ziegler‐Nichols Tuning Parameters
Table 6.3	Recommended PID Tuning Parameters
Table 6.4	IMC Tuning Parameters
Table 8.1	Difference between Normal Step Testing and PRBS Testing
Table 11.1	Simulation Initial Condition File for MVs
Table 11.2	Simulation Initial Condition File for CVs
Table 11.3	Controlled Variables with Their Limits for Simulation Studies
Table 11.4	Controlled Variables with Their Limits for Simulation Studies
Table 14.1	Retaining Initial MPC Benefits after 12 Months
Table 14.2	Commercial MPC Monitoring Tools

Applications in Industry