Wade BasicAdvRegCont 3rd Excerpt

Standards
Certification
Education & Training
Publishing
Conferences & Exhibits
Table of Contents
View Excerpt
Buy the Book

Basic and Advanced
Regulatory Control:
System Design and
Application
Third Edition
By Harold L. Wade
Notice
The information presented in this publication is for the general education of the reader. Because nei-
ther the author nor the publisher has any control over the use of the information by the reader, both the
author and the publisher disclaim any and all liability of any kind arising out of such use. The reader is
expected to exercise sound professional judgment in using any of the information presented in a particu-
lar application.
Additionally, neither the author nor the publisher has investigated or considered the effect of any
patents on the ability of the reader to use any of the information in a particular application. The reader is
responsible for reviewing any possible patents that may affect any particular use of the information pre-
sented.
Any references to commercial products in the work are cited as examples only. Neither the author
nor the publisher endorses any referenced commercial product. Any trademarks or tradenames refer-
enced belong to the respective owners of the mark or name. FOUNDATION Fieldbus is a registered
trademark of the Fieldbus Foundation, HART is a registered trademark of the HART Communication
Foundation. Neither the author nor the publisher makes any representation regarding the availability of
any referenced commercial product at any time. The manufacturers instructions on the use of any com-
mercial product must be followed at all times, even if in conflict with the information in this publication.
Copyright 2017 International Society of Automation (ISA)

All rights reserved.
Printed in the United States of America.

10 9 8 7 6 5 4 3 2
ISBN: 978-0-87664-013-5
No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by
any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written per-
mission of the publisher.
ISA
67 T. W. Alexander Drive
P.O. Box 12277
Research Triangle Park, NC 27709
Library of Congress Cataloging-in-Publication Data in process

Dedication
To Mary, my wife of many years, who provides love, support,

encouragement, and criticism when needed.
Contents
About the Author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Preface to the Third Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xv
Chapter 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Chapter 2 Mathematical Background, Diagrams, and Terminology. . . . . . . . . .7

Mathematical Foundation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Functions of a Variable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Transfer Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Diagrams and Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Direct- or Reverse-Acting Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Chapter 3 Process and Control Loop Characteristics. . . . . . . . . . . . . . . . . . . .39

Steady-State Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Control Valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Valve Stem Stiction and Stick-Slip . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Process Dynamic Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Types of Dynamic Response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Control Loop Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
vii
viii Basic and Advanced Regulatory Control, Third Edition
Flow Control Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Temperature Control Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Pressure Control Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Liquid-Level Control Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Chapter 4 PID Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83

Feedback Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Modes of Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Proportional Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Integral Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Derivative Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Chapter 5 Modifications to Standard PID Control . . . . . . . . . . . . . . . . . . . . . .111

Set-Point Softening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Derivative Mode on Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Proportional Mode on Measurement. . . . . . . . . . . . . . . . . . . . . . . . . 114
Set-Point Ramping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Two-Degree-of-Freedom (Set-Point Weighting) Controller. . . . . . . . . . 116
Linear Combination of Inputs to Modes . . . . . . . . . . . . . . . . . . . . . 117
Integral-Only Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Interactive or Noninteractive Controller. . . . . . . . . . . . . . . . . . . . . . . . . . 122
Independent Gains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Internal Filter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Nonlinearization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Set-Point Tracking and Bumpless Transfer . . . . . . . . . . . . . . . . . . . . . . . 132
Bumpless Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Preventing Reset Windup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
External Reset Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Batch Switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Discrete Control Algorithms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Incorporating Engineering Units into Controller Gain. . . . . . . . . . . . . . 148
Process Control Using FOUNDATION Fieldbus . . . . . . . . . . . . . . . . . . . 149
Process Control Using FF Function Blocks . . . . . . . . . . . . . . . . . . . . 152
FF Function Block Subclasses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
FF Basic Control Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Chapter 6 Tuning Feedback Control Loops . . . . . . . . . . . . . . . . . . . . . . . . . . .161

Performance Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Contents ix
Tuning for Self-Regulating Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Trial-and-Error Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Tuning from Open-Loop Test Data . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Tuning from Closed-Loop Test Data. . . . . . . . . . . . . . . . . . . . . . . . . 188
Improving As-Found Tuning Parameters . . . . . . . . . . . . . . . . . . . . 192
Tuning for Integrating Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Liquid Level ControlIdeal Model . . . . . . . . . . . . . . . . . . . . . . . . . 219
Real-World Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Modified Tables for Tuning Parameters. . . . . . . . . . . . . . . . . . . . . . 237
Other Approaches to Liquid-Level Controller Tuning . . . . . . . . . 245
Other Integrating Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Typical Tuning Values for Particular Types of Loops . . . . . . . . . . . . . . 250
Practical Considerations for Loop Tuning . . . . . . . . . . . . . . . . . . . . . . . . 250
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Chapter 7 Self-Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259

Scheduled Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
On-Demand Tuning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Adaptive Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Tuning Aids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Chapter 8 Advanced Regulatory Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269

Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Chapter 9 Cascade Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .275

Cascade Control Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Identifying Candidate Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
Implementation, Operation, and Tuning . . . . . . . . . . . . . . . . . . . . . . . . . 284
Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
Operation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Commissioning and Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Cascade Control Using FOUNDATION Fieldbus . . . . . . . . . . . . . . . . . . . . 288
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Chapter 10 Ratio Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295

Ratio Control Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Automatic Ratio Adjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Scaling the Ratio Control Components. . . . . . . . . . . . . . . . . . . . . . . 303
Ratio Control Using FOUNDATION Fieldbus Function Blocks . . . . 305
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
x Basic and Advanced Regulatory Control, Third Edition
Chapter 11 Feedforward Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .311

Designing Feedforward Control Systems . . . . . . . . . . . . . . . . . . . . . . . . 315
Additive Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
Multiplicative Feedback. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Feedback Adjustment of the Feedforward Controllers
Reference Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Compensation for Process Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Determining A(s) and B(s). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Fine-Tuning the Feedforward Controller . . . . . . . . . . . . . . . . . . . . . . . . . 333
Further Considerations of the Feedback Controller . . . . . . . . . . . . . . . . 340
Feedforward: In Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
Feedforward Control Using FOUNDATION Fieldbus . . . . . . . . . . . . 346
Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Chapter 12 Override (Selector) Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .353

Override Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Real-World Applications of Override (Selector) Control . . . . . . . . . . . . 363
Transition from Start-Up to Normal Operations . . . . . . . . . . . . . . . 363
Operating Near the Limit of a Process Utility . . . . . . . . . . . . . . . . . 363
Prevention of Tower Flooding in a Distillation Tower . . . . . . . . 364
Pipeline Industry, Compressor Control . . . . . . . . . . . . . . . . . . . . . . 366
Other Methods of Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
Pass-Through Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
Forced Manual for the Nonselected Controller . . . . . . . . . . . . . . . . 372
Velocity (Incremental) Mode Control Algorithms. . . . . . . . . . . . . . 372
Pseudo-Velocity Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
Selection Based on Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Override Control Using FOUNDATION Fieldbus . . . . . . . . . . . . . . . . . . . 376
Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Chapter 13 Control for Interacting Process Loops . . . . . . . . . . . . . . . . . . . . . .383

Variable Pairing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Decoupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Forward Decoupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
Inverted Decoupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Partial Decoupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
Decoupling Application Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Petroleum Refinery Heater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Spray Water Temperature and Flow Control. . . . . . . . . . . . . . . . . . 411
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Contents xi
Chapter 14 Dead-Time Compensation and Model-Based Control . . . . . . . . . .419

Smith Predictor Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Algorithm Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
Internal Model Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
Nonlinear, Model-Based Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
PMBC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Provisions for Dead Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
Chapter 15 Multivariable Model Predictive Control . . . . . . . . . . . . . . . . . . . . . .449

Real-World Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
Unconstrained MPC for SISO Processes. . . . . . . . . . . . . . . . . . . . . . . . . . 452
Process Model Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
Calculating Control Moves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
Incorporating Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Summary Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Unconstrained MPC for MIMO Processes . . . . . . . . . . . . . . . . . . . . . . . . 465
Constrained MPC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
Variations in MPC Vendor Offerings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
MPC in Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
Chapter 16 Other Control Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .479

Split-Range Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
Cross-Limiting Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
Configuration Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
Provisions for Changes in Required Fuel-to-Air Ratio. . . . . . . . . . 487
Floating Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
Floating Pressure Control for Distillation Columns . . . . . . . . . . . . 490
Floating Pressure Control for Steam Systems . . . . . . . . . . . . . . . . . 492
Hot or Chilled Water Supply Systems . . . . . . . . . . . . . . . . . . . . . . . 494
Cooling Tower Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
Increasing Valve Rangeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
Small and Large Valves Operating in Parallel . . . . . . . . . . . . . . . . . 496
Small and Large Valves Operating in Sequence . . . . . . . . . . . . . . . 498
Time Proportioning Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
xii Basic and Advanced Regulatory Control, Third Edition
Appendix A Derivation of Equations for Installed Valve Characteristics . . .507
Appendix B Signal Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .515
Appendix C Exercise Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .523
Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .553
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .559
About the
Author
Harold L. Wade, PhD, has had over 50 years of experience in designing,

applying, and installing process control systems in such industries as petro-
leum refining, chemical processing, textiles, waste and water treatment facili-
ties, and others. He has a BS in mechanical engineering and an MS and PhD in
systems engineering. He has held technical positions with Honeywell, with
Foxboro, and with his own consulting engineering firm. He is a Fellow of ISA,
a Life Member of IEEE, and a licensed professional engineer in Texas. Dr.
Wade was a 2002 inductee into Control magazines Process Automation Hall
of Fame and was the 2008 recipient of the Donald P. Eckman Award pre-
sented by ISA.
For almost 20 years, Dr. Wade taught courses for ISA on process control sys-
tems design. He has also presented process control and control tuning semi-
nars for many companies worldwide. Dr. Wade is the developer of the
process control training program, PC-ControLAB.
xiii
Preface to the
Third Edition
This book presents a practical approach to process control for the chemical,
refining, pulp and paper, utilities, and similar industries. It is the result of
many process control seminars presented in the United States and abroad. A
typical participant in one of those seminars is an engineer, currently
employed by a processing company, who may have had formal training in an
undergraduate process control course, but who may not be able to fully relate
the material from that course to his or her work experiences. This book aims to
meet this need by explaining concepts in a practical way with a minimal
amount of theoretical background. The book serves both the beginning and
the experienced control systems engineer. For the beginning engineer, it ini-
tially presents very simple concepts. For the experienced engineer, it develops
these initial concepts to provide deeper understanding and/or new insights
into familiar concepts. The purpose is to provide everyone, beginner or expe-
rienced engineer, with something they can put to beneficial use in their plant.
Although this is intended to be a practical how-to book, readers should not

infer that it is devoid of mathematical concepts. Where such concepts are uti-
lized, it is their application to practical situations, rather than the theory
behind the concepts, that is emphasized. A theme of the first edition was that
wherever a choice had to be made between providing mathematical rigor or
promoting intuitive understanding, preference was given to understandabil-
ity. This theme has been carried forward into later editions, and this practical-
ity distinguishes this book from many academic texts. End of chapter
exercises have also been added to the present edition.
xv
xvi Basic and Advanced Regulatory Control, Third Edition
The book is organized in three parts. The first three chapters present back-
ground information, including a brief non-rigorous mathematical review, a
discussion of symbols and terminology, and a description of general charac-
teristics of processes and of selected types of control loops.
The second partChapters 4 through 7addresses feedback control. The

objective is to provide the reader with a thorough intuitive grasp of feedback
control behavior and all its nuances. Examples of current commercially avail-
able products, DCSs and PLCs, have been updated. In Chapter 5, the brief
mention made in the previous edition about two degree of freedom control-
lers (called set-point weighting by some manufacturers) has been expanded to
give proper emphasis to this recent (within the digital era), very important,
and often overlooked addition to PID capabilities. This modification is also
mentioned in the following chapter, in relation to controller tuning. In Chap-
ter 6, on feedback controller tuning, the discussion on improving as-found
tuning (also called intelligent trial-and-error tuning) was expanded consider-
ably in the second edition, and is further expanded in this edition, to consider
overshoot ratio, as well as decay ratio when formulating controller-tuning
objectives. This tuning technique has been proven in practical applications
and has been well accepted in training classes in which it has been presented.
An additional technique is presented in this edition for improving-as-found
tuning; this technique considers process dynamics as approximated by a first-
order-lag plus dead-time model. Also in this chapter, new material is included
on tuning liquid-level control loops. The tuning of these loops, which have a
completely different characteristic from most other process control loops, has,
in general, received very little specific attention in process control literature.
The third portion of the book, Chapters 8 through 16, covers advanced regula-
tory control topics. Chapter 8 defines the penalty that must be paid if feedback
control alone is used. This leads into a discussion of advanced regulatory con-
trol techniques that can significantly reduce this feedback penalty. Additional
chapters provide a detailed discussion of the technology and application
examples of cascade (Chapter 9), ratio (Chapter 10), feedforward (Chapter 11),
override (Chapter 12), decoupling (Chapter 13), model-based (Chapter 14),
and model-predictive control (Chapter 15). The chapter on feedforward con-
trol covers both additive and multiplicative forms of feedforward. The chap-
ter on override (selector) control includes additional application examples for
this technique, as well as an assessment of the performance of several alterna-
tive techniques. The chapter on the control of multiple-input multiple-output
Preface to the Third Edition xvii
(MIMO) processes (Chapter 13) contains additional coverage of inverted

decoupling. This MIMO technique was introduced in the first edition; new
material previously available only in technical journals is presented here.
The chapter on model-based control in the first edition has been divided into
two chapters. Chapter 14, devoted primarily to dead-time compensation, cov-
ers Smith predictor control and internal model control. The other chapter,
Chapter 15, contains an introductory discussion on model predictive control.
Chapter 16 covers topics on process control application that do not readily fit
into any of the other chapters. In addition to cross-limiting control for fired
heaters, which was covered in the first edition, these new topics include float-
ing control, techniques for increasing valve rangeability, and time proportion-
ing control.
Also contained in this section are illustrations of the use of FOUNDATION

Fieldbus (FF), which permits the control strategy to be distributed directly
into field devices. The network architecture, communication, and implemen-
tation aspects of FF are briefly summarized in Chapter 5. The chapters on
modifications to feedback control, cascade, ratio, feedforward, and override
(Chapters 5, 9, 10, 11, and 12) all conclude with an example in which that
chapters strategy is implemented using FF function blocks.
I would like to express gratitude to the many students who, by asking probing
questions, have enabled me to revise and sharpen my presentation and come
up with examples that are more meaningful.
I would also like to express my thanks to Dr. R. Russell Rhinehart for many
helpful comments and suggestions, to my longtime friends Greg Shinskey
and Victor Wegelin for their mentoring and helpful suggestions, and to my
friend Vu Van Phin in Viet Nam (with whom I have had the privilege of meet-
ing only via email) for many enlightening technical exchanges.
Lastly, I have special thanks to my editors, Liegh Elrod and Scott Bogue, for
their diligence, which has helped me to produce a better final product.
1
Introduction
The term process control implies that there is a process for which there is a
desired behavior and that there is some controlling function that acts to elicit
that desired behavior. This broad concept can embrace everything from soci-
etal processes governed by some regulatory control authority to automated
manufacturing processes. In practically all cases, however, a common thread
is that some measure of the actual process behavior is compared with the
desired process behavior. This feedback action then generates a control policy
that acts to minimize or eliminate the difference between desired and actual
behavior.
This book focuses on a particular segment of automated process control that is

applied to chemical, refining, pulp and paper, power generation, and similar
types of processes. Even within this limited scope of applications, the discus-
sion is restricted primarily to processes that are operated continuously for
long periods of time and within a narrow region of the operating variables. In
other words, the book excludes such important operating modes as batch pro-
cessing, start-ups, and grade changes. Many of the control techniques to be
presented here, however, can be adapted to these other modes of operation.
The behavior of the processes in this book is often characterized by measured

values of process variables, such as temperatures, flow rates, and pressures.
The desired behavior is stated to be the set points of those process variables.
Until fairly recent times, most applications of industrial process control used
simple feedback controllers that regulated the flow rates, temperatures, pres-
1
2 Basic and Advanced Regulatory Control, Third Edition
sures and other process variables. These controllers required a form of adjust-
ment called tuning to match their controlling action to the unique
requirements of individual processes.
As long as most of the control systems were implemented with analog hard-
ware, applications were limited to simple regulatory control. This was due to
the cost of additional components and additional interconnections that more
advanced control required, along with the burden of additional maintenance
and the vulnerability to failure of many devices in the control loop. With the
advent of digital control systems, however, more sophisticated loops became
feasible. Advanced regulatory control techniques, such as ratio, cascade, and
feedforward control, as well as additional forms such as constraint (selector)
control and decoupling, could readily be implemented simply by configuring
software function blocks.
With this additional capability, however, a need developed for a systematic

approach toward using it. This is called control strategy design. In order to
design a technically successful and economically viable control strategy, the
control system engineer must be well grounded in the techniques of feedback
control as well as the tools of advanced regulatory control. The requisite
knowledge includes both how to implement and how to tune both the basic
and the advanced regulatory control strategies. Even before that, however, the
control system engineer must be adept at recognizing when to use (and con-
versely, when not to use) certain control strategies, as well as projecting the
expected benefits.
Using advanced regulatory control provides multiple benefits. One of the

most important is simply closer control of the process. It will become clear
later in this book that with basic regulatory (i.e., feedback) control, there must
be a deviation from a set point before a control action can occur. We will call
this the feedback penalty. While the feedback penalty also applies to advanced
regulatory control, the primary objective of advanced regulatory control is for
the control action to be taken while incurring only a minimal feedback pen-
alty. The reduction in feedback penalty may be stated in a variety of ways,
such as a reduction of the maximum deviation from the set point, or as a
reduction in the amount of off-spec product produced. This reduction in feed-
back penalty can provide several forms of economic benefit, such as improve-
ment in product quality, energy savings, increased throughput, or longer
equipment life.
Chapter 1 Introduction 3
Process control is but one part of an overall process control and information
system hierarchy (see Figure 1-1). It extends downward to safety controls and
other directly connected process devices and upward to encompass optimiza-
tion and even higher levels of business management, such as scheduling,
inventory, and asset management. Indeed, corporate profitability may be
enhanced more significantly as a result of these higher-level activities than
from improved process control per se. However, since each layer of the hierar-
chy depends on the proper functioning of the layers beneath it, one of the pri-
mary benefits of advanced regulatory control is that it enables the higher
levels, optimization and enterprise management and control.
&/5&313*4&-&7&-$0/530-4
"DDPVOUJOH 4DIFEVMJOH "TTFU.BOBHFNFOU FUD
015*.*;"5*0/
"EWBODFE1SPDFTT$POUSPM
"%7"/$&%3&(6-"503:$0/530-
$BTDBEF 3BUJP 'FFEGPSXBSE 0WFSSJEF
%FDPVQMJOH %FBE5JNF$PNQFOTBUJPO
#"4*$3&(6-"503:$0/530-
'FFECBDL
4"'&5:$0/530-4
130$&44
Figure 1-1. Overall Process Control and Information System Hierarchy

Symbols
Many mathematical symbols are used throughout this book. Some symbols
are used only for the discussion of a particular topic; these symbols are there-
fore defined in that discussion and are not listed here. Some symbols may
have multiple meanings that depend on the context; additional meanings are
provided as needed for clarity in the appropriate sections of this book. Chap-
ter 2 discusses the graphical symbols used in control system documentation.
Chapter 15 uses a unique set of symbols that are defined at the beginning of
that chapter.
The following symbols are used throughout this book:
b = bias value (manual reset) on proportional-only controller

output
e = error (deviation between the set point and process variable)1
E = when capitalized, refers to (Laplace) transform of error
K = steady-state gain of first-order lag
KC = controller gain (noninteractive and interactive control
algorithms)
KD = derivative gain (independent gains control algorithm)
KI = integral gain (independent gains control algorithm)
KP = proportional gain (independent gains control algorithm)
Kp = process gain (change in process variable/change in controller
output)
m = manipulated variable, controller output
M = when capitalized, refers to (Laplace) transform of manipulated
variable
PB = proportional band
PI = control algorithm with proportional and integral modes
PID = control algorithm with proportional, integral, and derivative
modes
PV = process variable (see also symbol x)
SP = set point (see also symbol xSP)
1. The symbol e can also be used as the basis of natural logarithms, for example, when
expressing the Laplace transform of dead time.
Chapter 1 Introduction 5
TD = derivative time (noninteractive and interactive control

algorithms)
TI = integral time (minutes/repeat) (noninteractive and interactive
control algorithms)
x = process variable (see also symbol PV)
xSP = set point (see also symbol SP)
u = disturbance variable
= derivative gain (when a derivative filter is used with a
noninteractive or interactive control algorithm)
= dead time
= first-order lag time constant
Exercises
1.1 An incongruous illustration of feedback control is that of driving an
automobile by looking only in the rear-view mirror. You can see where
you have been, but you cannot see what disturbances (hills, curves or
bumps in the road, other vehicles, etc.) are coming. Nevertheless, an
analogy can be made between this example and components of an
ordinary feedback control loop. Complete the following table:
Set point Center of your lane
Process variable __________________
Error (or deviation) __________________
Controller __________________
Final control element Steering mechanism
1.2 Another, more realistic, example of a feedback control loop is speed

control. (Assume that what you are driving does not have automatic
cruise control, which would be a true feedback loop.) Here, if you see
on the speedometer that the vehicle is slowing down, you apply more
pressure to the accelerator pedal; if you see that it is traveling faster
than your desired speed, you reduce the pressure on the accelerator
pedal. Complete the following table:
Set point Your desired speed
Process variable indicator __________________
Controller __________________
Final control element __________________
Manipulated variable __________________
Constraint Speed limit
1.3 Another analogy between driving and process control is driving while
looking forward. For example, if you see a hill ahead, you depress the
accelerator pedal just as you approach the base of the hill, and you
reduce pressure on the accelerator pedal as you approach the crest of
the hill. If these actions are taken at the right times, and in the right
amounts, there will be no deviation from the desired speed. This type
of control is called: ________________________.
1.4 There are many important topics in an overall automation project. As

important as they are, which of the following topics are not within the
scope of this book?
Enterprise integration
Safety instrumented systems
Sensor/analyzer selection and installation
Alarm management
Abnormal situation management
Human-machine interface (HMI) design
Cybersecurity
Communications technology (for plant instrumentation and

control systems)
3
Process and Control
Loop Characteristics
In order to design, analyze, or commission a process control system, one must

be familiar with the characteristics of the process itself. Although it is highly
beneficial for the control engineer to have a good understanding of the physi-
cal and chemical phenomena that govern the process, his or her view of the
process will usually differ from that of the process design engineer. The dis-
cussion in this chapter is meant to develop the thought processes of a control
engineer. Although some of the following points may seem to overstate the
case, they will enable us to highlight the differences in the ways the control
engineer and process engineer think, and to call attention to the process
details that the control engineer should thoroughly understand.
The process engineer is concerned with meeting production rate and quality
specifications, which are often called the design conditions. The control engi-
neer is concerned with operating an existing process outside of design condi-
tions, often with reduced throughput, variations in feedstock, or other
abnormal conditions.
The process engineers objective is often to minimize the initial cost (or the
life-cycle cost) of the processing equipment. The control engineers objective is
to make the most efficient use of the equipment that is already installed.
The process engineer considers those design parameters that can be specified
as independent variables. Other parameter values that are derived from these
are dependent variables. For example, the pressure of a saturated steam sys-
39
tem might be an independent variable that can be specified during the design
process; the temperature then becomes a dependent variable. The control
engineer considers as independent variables the control points (for instance,
valve positions or flow rates) that can be manipulated to affect the process.
The steam pressure then becomes a dependent variable that results from those
valve positions or flow rates.
The process engineer is usually concerned with the steady-state conditions of

the process. The control engineer must necessarily take into consideration
both the steady state and the dynamic, or transient, behavior of the process.
The characteristics of each process will be different. Even so, from the control
engineers viewpoint, certain characteristics are similar from process to pro-
cess. It is these characteristics that will be emphasized here.
Steady-State Characteristics
When all inputs and external influences are held constant, most, but not all,
processes come to a steady state. (Liquid level is different. Unless the inflow to
and outflow from a liquid-level process are equal, the process will not come to
a steady state, even though the inflow and outflow themselves are constant.)
We will use the heat exchanger depicted in Figure 2-7 to illustrate the nature of
the steady state. This is redrawn as Figure 3-1 here, but in this case the inde-
pendent and dependent variables are identified. Let us assume that we have a
liquid phase process stream that must be heated to a specified temperature. Let
us also assume that we have a liquid phase heating medium, such as hot water
or hot oil. (If this process unit is just a heat exchanger, with no chemical reac-
tion taking place, then both the inlet and outlet temperatures of the process
stream will be at a lower temperature than the inlet and outlet temperatures of
the heating medium. We will designate the process side as the cold side, the
flow rate as Fc, and the inlet and outlet temperatures as Tc-in and Tc-out. Simi-
larly, on the hot side, the flow rate of the heating medium will be designated as
Fh, with its inlet temperature as Th-in and its outlet temperature as Th-out.)
From a process control viewpoint, the independent variable is the valve posi-
tion, or equivalently, the controller output. The dependent variable of interest
is the process outlet temperature, Tc-out. Other dependent variables are the
flow rate of the heating fluid, Fh, and the outlet temperature of the heating
medium, Th-out. In a typical operating plant, these may be monitored to detect
abnormal operation, but from a control viewpoint they are inconsequential.
Chapter 3 Process and Control Loop Characteristics 41
7KLQ )K
7FRXW
7FLQ )F
)F
7FLQ
FLQ
',6785%$1&(6
7KLQ
2WKHU 7KRXW
Figure 3-1. Disturbances to a Control Loop
Other important variables to be considered are the disturbances to the process.

These are sometimes called load changes. They can be considered as external,
random influences on the process. It is the purpose of the control system to
counteract the effects of these disturbances. Some of the disturbances that
could affect the heat exchanger are:
Changes in the process flow rate, Fc
Changes in the process inlet temperature, Tc-in
Changes in the source temperature of the heating medium, Th-in
Changes in the upstream or downstream pressure of the heating fluid

(this would change the hot stream flow rate, Fh, even though the valve
position did not change)
Scaling of the heat exchanger tubesthus affecting the heat transfer

coefficient
Environmental effects, if the heat exchanger is not perfectly insulated

For the purposes of illustration, we will disregard the latter three of these dis-
turbances (i.e., we will assume that they are constant) and concern ourselves
only with Fc, Tc-in, and Th-in. For the time being, we will also consider that
these three variables are also being held constant. In other words, the only
independent variable is the valve position, which uniquely sets the heating
medium flow rate. With this consideration, we state an important principle:
If all external influences on a process are held constant, then each value of the control
signal (independent variable) produces a specific and unique measurement value
(dependent variable). (This is somewhat of an idealized statement, since it
ignores the very real probability of valve stick-slip and hysteresis, which is
described in a subsection below.) There are rare cases, such as the discharge
pressure of a centrifugal compressor versus suction flow or index of refraction
versus composition, where this unique relationship may not be true.)
This one-to-one relationship can be depicted in graphical form, as shown in

Figure 3-2. We call this relationship the process graph. Keep in mind that the
process graph depicts the steady-state relationship between the controller out-
put (valve position) and the measurement for a particular combination of the
disturbance variables.
4H IN
4C OUT
4C IN
#ONTROLLER/UTPUT
&H
Figure 3-2. Process Graph

If any of the disturbance variables change in value, then we will have a new
process graph. Figure 3-3 shows the process graphs for three combinations of
disturbance variables. The process graph for the original values of Fc, Tc-in,
and Th-in is shown by the dotted curved line. The upper line is the process
graph for an increase in Tc-in. The lower line is the process graph for an
increase in Fc.
4H IN
2ISEIN4C IN
4C OUT
)NCREASEIN&C
4C IN
#ONTROLLER/UTPUT
&H
Figure 3-3. The Shifting of a Process Graph as a Result of Disturbances
The process graphthe steady-state relationship between the controller out-

put and measured (dependent) variables for a particular combination of dis-
turbance variablesis an important concept for understanding control loop
behavior. However, it is not something that we need to determine in actual
practice. Indeed, it would be impractical to determine the infinite number of
process graphs that would result from all combinations of the disturbance
variables.
Nevertheless, we can deduce that if we wish to control the measurement to a

particular value (set point), the process graph determines the required value
of the controller output. If there are load changes on the process that cause the
process graph to shift, we will need a new value for the controller output. It is
the duty of the controller to determine the precise point on the process graph
that brings the measurement to the desired value, as shown in Figure 3-4.

X
0ROCESS6ARIABLE

M
#ONTROLLER/UTPUT
Figure 3-4. The Process Graph Determines the Controller Output Required to
Bring the Measurement to the Desired Value
Although we will normally not have a precise process graph available for
even one combination of disturbance variables, there are certain attributes of
the process graph that we must know. First, we must know whether the pro-
cess graph slopes upward or downward. This is equivalent to saying that we
must know whether the process is direct-acting or reverse-acting. An upward
slope represents a direct-acting process (an increase in controller output
causes an increase in measurement); a downward slope signifies a reverse-act-
ing process. Recall from Chapter 2 that to avoid positive feedback, the control-
ler must be of opposite actionreverse-acting for a direct-acting process and
vice versa.
We must also know, either explicitly or implicitly, the slope of the process
graph, at least in the vicinity of the most probable operating point. The slope
can be defined as the change in measurement divided by the change in con-
troller output. This is called the process gain. Process gain, Kp, is defined by the
following equation:
Change in measurement
K p = ----------------------------------------------------------------
Change in valve signal
x
= -------- (3-1)
m
The process gain often varies with the operating point. This is equivalent to
stating that the process, and hence the process graph itself, is often nonlinear.
Except for some rare misbehaved processes, however, the process graph is
monotonic. That is, the direction of the slope (upward or downward) does not
change, so there is a unique relationship between each value of the controller
output and the measurement.
Process nonlinearities can be caused by a number of conditions, including

physical or chemical factors inherent in the process itself. In one frequently
encountered situation, the process variable responds linearly to changes in the
ratio between two variables, such as the manipulated variable and a distur-
bance variable, whereas it would respond nonlinearly to each of the two vari-
ables considered separately. To illustrate this, suppose that a process heater
can be modeled by a simple heat-balance relationship:
F p C p ( T out T in ) = F g H v E ff (3-2)
where
Fp = heater feed rate (the disturbance variable)
Cp = specific heat of the processed material
Tout = outlet temperature (the process variable)
Tin = inlet temperature
Fg = fuel rate (the manipulated variable)
Hv = heating value of the fuel
Eff = heater efficiency
Equation 3-2 can be rearranged to show the outlet temperature on the left-
hand side of the equation and all other terms on the right-hand side:
H v E ff F g
T out = T in + --------------- ------ (3-3)
C F p p
This demonstrates that the outlet temperature responds more or less linearly
to the fuel rate-to-heater feed rate ratio, Fg/Fp. If a temperature controller
directly manipulates the fuel rate, then the process gain seen by the controller
is the sensitivity of the outlet temperature to changes in fuel rate. Specifically:
T out H v E ff 1
--------------- = --------------- ------ (3-4)
F g Cp Fp
In other words, the process gain of the control loop is inversely proportional
to the process flow rate. At a low process flow rate (such as during start-ups),
the process gain will be high; at higher flow rates, the process gain will be
lower.
If the ratio itself were the manipulated variable, rather than simply the fuel
rate, then the process gain seen by the control loop would be the following:
T out H v E ff
- = --------------
------------------------ - (3-5)
( Fg Fp ) Cp
As long as the fuel heating value, heater efficiency, and specific heat of the
process fluid remain fairly constant, then the control loops process gain will
remain constant. This strategy will be used in relation to ratio control in Chap-
ter 10 and in multiplicative feedforward control in Chapter 11.
Control Valves
Nonlinearity in a control loop may also be caused by the nonlinear character-
istic of the valve. There is also the friction in the packing gland of the valve
that will affect stem movement, and contribute to the dynamics of the control
loop. Therefore, before continuing with dynamic characteristics, we will dis-
cuss control valves.
There are many types of control valves; these differ by the type of valve body,
by the type of valve actuator, and by whether or not a valve positioner is used.
The discussion below will focus on a common type of control valve: a globe
valve with a sliding stem plug and seat (the closure mechanism), and a spring
and diaphragm actuator, with and without a valve positioner as depicted in
the elementary diagram, Figure 3-5.
6
Tuning Feedback
Control Loops
The power of the PID controller is that it can be adjusted to provide the
desired behavior on a wide variety of process applications through the judi-
cious choice of one, two, or three parameter values, and with only modest
knowledge about the process. Determining acceptable values for these param-
eters is called tuning the controller.
In the process industries, those who tune the controllers often face a number
of adverse factors.
The process dynamics are usually not well known.
There are often nonlinearities in the process that cause the process
response to change with operating conditions.
There is often an unwanted signal component (called noise) on the

measurement.
The loop may be subject to random load changes.
Frequently, the interaction between control loops makes it difficult to

discern the tuning effects of a particular loop from the interactive
response with other loops.
If the final actuator is a valve, it may also contribute to the problem:
o It may be the wrong size (often oversized).
161
o It may contribute to the nonlinear response, as described in

Chapter 3.
o Due to valve stiction (described in Chapter 3), the process variable

may be driven into a limit cycle, even if the set point is being held
constant. Depending on the type of valve and positioner, if one is
being used, it may or may not be possible to reduce this limit cycle
by controller tuning.
In addition to these problems, the loop tuner often must work on an

ongoing process, which allows for only minimal or no experimentation
or testing. Given these adversities, it is a wonder that so many PID
loops provide more or less satisfactory performance. On the other
hand, it is probably true that the tuning could be improved for a
significant number of all control loops.
Also contributing to the difficulty the tuner faces is the fact that there is no
general agreement as to what constitutes good tuning. Therefore, before dis-
cussing the tuning procedures, we will discuss performance criteria, both
informal and formal. The tuning procedures that will then be presented can be
grouped into the following categories:
Tuning for self-regulating processes
o Trial-and-error tuning
o Tuning from open-loop tests
o Tuning from closed-loop tests
o Improving as-found tuning
Tuning for integrating processes
o Idealized liquid-level control loops

o Real world considerations for liquid-level control
o Other approaches to liquid-level control tuning
o Other integrating processes
Following this, we will present typical tuning values for some of the more
common types of control loops, the chapter will close with general admoni-
tions and recommendations for the controller tuner.
Chapter 6 Tuning Feedback Control Loops 163
Performance Criteria
If an operator were asked their preference of the various responses to a set-
point change shown in Figure 6-1, they would probably designate a as the
first choice, b as acceptable, c next, then e, and f last. Response d
would probably be unacceptable, due to the undershoot of the first peak.
Unfortunately, however, preferred choice a provides a poor choice for rejec-
tion of a step disturbance at the process input, whereas f probably provides
the best choice for disturbance rejection. One behavioral fact about any type of
traditional PID controller is that if the loop is tuned to give a desirable
response to a set-point change (say, both a low decay ratio and a low over-
shoot ratio), the response to a disturbance may be too sluggish. On the other
hand, if the controller is tuned for a more aggressive response to minimize the
response to a disturbance, the response to a set-point change may be overly
aggressive.
6(732,17
6(732,17
&+$1*(
&+$1*(
D 5LVHVWRVHWSRLQW E 2QHRYHUVKRRW
ZLWKQRRYHUVKRRW WKHQVHWWOHV
6(732,17
6(732,17
&+$1*(
&+$1*(
F 2QHRYHUVKRRW G )LUVWSHDNXQGHUVKRRW
RQHXQGHUVKRRW /RQJVHWWOLQJWLPH
$
%
6(732,17
6(732,17
&+$1*(
&+$1*(
%
'(&$<5$7,2
$
H $FFHSWDEOH I /DUJHRYHUVKRRW
/RQJVHWWOLQJWLPH $SSUR[GHFD\UDWLR
Figure 6-1. Examples of Various Forms of Response to a Set-Point Change

For many control loops, the set point is rarely changed. The purpose of these
loops is to minimize the effect of disturbances. Even so, because set-point
changes are usually made more easily than load changes, in actual practice
many loops are tuned for a suitable response to a set-point change. Then, the
resulting response for a load upset is accepted even though it may not be the
best. A preferable tuning strategy would be to tune the controller for the best
response to a load change, then use one of the set point softening techniques
or the two-degree of freedom controller described in Chapter 5 to ameliorate
the effect of occasional set-point changes. If actual load changes cannot be
made, the effect can be simulated by placing the controller in Manual, chang-
ing the controller output, and then quickly returning the loop to Automatic.
Another alternative is to initially tune the loop for a quarter-decay response to
a set-point change, because tuning for that response provides a very accept-
able disturbance response. You would then apply one of the set point soften-
ing techniques mentioned in Chapter 4 (derivative mode on measurement,
proportional mode on measurement, set point ramping, or two-degree of free-
dom controller) to obtain an acceptable set-point response.
The performance criteria can be refined beyond the qualitative figures by

quantifying the requirements for certain parameters. For example, values may
be designated for maximum and minimum amounts of overshoot. A value
may also be specified for the decay ratio as shown in Figure 6-1f. This figure
illustrates the traditional definition of decay ratio as the ratio of the deviation
from the set point at the second peak after a set-point change to the deviation at the
first peak. This is also depicted in Figure 6-2a. Occasionally, the set-point
response is such that this definition is not useful. A better definition of decay
ratio is the ratio of the difference between the second peak and its succeeding valley to
the difference between the first peak and its succeeding valley.1 This is depicted in
Figure 6-2b. This definition, though more cumbersome, will work in all cases.
Most of the time, however, the simpler and more widely used definition
depicted by Figure 6-2a will suffice.
The decay ratio can also be defined for a disturbance or load upset. For a step
change in load, the behavior depicted in Figure 6-2c is typical. Here, the decay
ratio must be determined by the ratio of peak-to-valley differences. A load-
1. This second definition, although not exactly correct, is more defensible theoretically because the
set-point response is the composite of a filtered exponential rise and a damped sinusoidal sig-
nal. If the rise time of the exponential is sufficiently fast, then the two definitions are essentially
the same. This is why the first definition is valid in most circumstances.
$ %
%
6(732,17
6(732,17
$
&+$1*(
&+$1*(
% %
'(&$<5$7,2 '(&$<5$7,2
$ $
D 5HVSRQVHWRD6HW3RLQW&KDQJH E 5HVSRQVHWRD6HW3RLQW&KDQJH
7\SLFDO%HKDYLRU 8QXVXDO%HKDYLRU
%
',6785%$1&(
',6785%$1&(
'(&$<5$7,2
5(63216(
5(63216(
$ %
'(&$<5$7,2
$
% $
%
F 5HVSRQVHWRD'LVWXUEDQFH G 5HVSRQVHWRD'LVWXUEDQFH
7\SLFDO%HKDYLRU 8QXVXDO%HKDYLRU
Figure 6-2. Quarter-Amplitude Decay Responses
upset response like that depicted in Figure 6-2d is somewhat unusual for most
processes but is typical of the load-upset response of level control loops.
One well-known criterion for controller tuning is a decay ratio of one-fourth

following a set-point change. This is also called quarter-wave decay, quarter-
wave damping, and quarter-amplitude decay. This criterion states that if a loop is
oscillating, each peak deviation should be only one-fourth of the previous
peak deviation on the same side of the set point. This is equivalent to stating
that on each half cycle, the amplitude of deviation should be decreased by
approximately one-half, making the total decrease one-fourth for a full cycle.
The decay ratio, however, may not provide a useful performance specifica-
tion. The overshoot ratio may be a preferable performance criterion. Even if a
loop exhibits quarter-decay response to a set-point change, it probably will
not be considered as being acceptably tuned if there is excessive overshoot.
Most people would probably prefer the overshoot to be no more than 510%
of the magnitude of the set-point change. A lower limit, such as 5%, may also
be specified for the overshoot ratio; if however, there were no overshoot, the
rise time to the set point may be excessive. These are soft limits, as will be seen
when we discuss improving as-found tuning.
Other criteria that are sometimes used to measure control loop performance
include rise timethe time between a set-point change and the first crossing of
the set point; and settling timethe time, following a set-point change or dis-
turbance, that it takes for the oscillation to become so small that the deviation
does not exceed some specified amount. These will have little use in this book,
due to their relative infrequent use in normal process control tuning activities.
The criteria mentioned so far have been primarily judgement-based. That is,
there is no formal justification for them, other than they appear to be satisfac-
tory. Formal criteria that can be used to evaluate loop tuning are based on
minimizing the integral of some function of the error. The following four inte-
gral-error criteria can be considered:
Integral of the absolute value of the error (IAE):
IAE = e dt (6-1)
Integral of the square of the error (ISE):
2
ISE = e dt (6-2)
Integral of time absolute value of the error (ITAE):
ITAE = t e dt (6-3)
Integral of time square of the error (ITSE):
2
ITAE = te dt (6-4)
Note that the simple integral of the error is not a valid criterion because the
integration of a positive error would be canceled out a half-cycle later by the
integration of a negative error. The four criteria just listed avoid this, either by
taking the absolute value of the error or by squaring the error.
Minimizing each of these integral-error criteria will produce a different

response. For example, a plant that uses the ISE criterion pays an increasingly
large penalty as the magnitude of the error increases. Therefore, for a given
loop, the ISE criteria will result in a smaller maximum deviation value than
the IAE criterion, but it may cause the oscillation to persist longer.
The rationale for the last two criteria listed (ITAE and ITSE) is that the longer
an error persists after a set-point or load change, the more heavily it should be
penalized. Thus, the ITAE criterion will permit a greater initial deviation than
an IAE criterion, but it will force the oscillation to die out sooner.
If there is noise on the process variable, any of these criteria will increase with-
out bounds. To be valid as measures of performance, the same time span
should be used for integration in any cases to be compared.
Integral-error criteria are also useful for academic, theoretical, and control
simulation studies for providing insight into the tuning process. They may
also be used in control loop audits. They are rarely used in actual control loop
tuning, however. For a typical process, there can be quite a variation from the
point of optimum PI tuning parameters (those that minimize the IAE) without
a significant increase in IAE and, consequently, without a significant change
in the observed response.
Figure 6-3 depicts graphically the typical amount of change in IAE as the PI
tuning parameters are varied from their optimum values. The data for this fig-
ure was obtained by simulating a typical process. Because comparing actual
numbers for two different examples is meaningless, normalized parameter
values are used in Figure 6-3. Table 6-1 shows both actual and normalized
parameter values at both the minimum IAE point and at point A. The point
the figure and table illustrate is that there can be a considerable variation of
tuning parameter values from the so-called optimum values without incur-
ring a significant change in the IAE metric.
1.5
0
1.4 1.5
5
1.3
1.2 A
0
1.2
1.1
05
1.
5
1.1
02
Normalized KC
1.
01
Minimum IAE
1.0 1.
01
1.
5
02
0.9
1.
05
1.
0.8
10
1.
25
0.7
1.
50
0.6 1.
5
1.7
0.5
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Normalized TI
Figure 6-3. A Typical Relation between Normalized IAE and Normalized PI Tuning
Parameters
Table 6-1. Actual and Normalized Values for IAE, KC, and TI at Minimum
IAE and Point A
Minimum IAE Point A
Actual Normalized Actual Normalized

Parameter
Value Value Value Value
IAE 68.7 1.0 73.51 73.51/68.7 = 1.07
KC 1.57 1.0 1,98 1.98/1.57 = 1.26
TI 8.53 1.0 7.42 7.42/8.53 = 0.87
To summarize this section, what constitutes good tuning is often a subjective

matter that can vary from application to applicationas well as from person
to person tuning the loop.
Tuning for Self-Regulating Processes

Because self-regulating and non-self-regulating processes have a different
character, the tuning procedures that are applicable to one may not be appli-
cable to the other. In this section, we will restrict our comments to self-regulat-
ing processes. The next section is devoted entirely to tuning controllers for
non-self-regulating processes, primarily liquid-level control.
Trial-and-Error Tuning
Most loops are tuned by an experimental technique. Even when a formal
technique, such as open-loop or closed-loop testing, is used to determine the
initial tuning, a final bit of fine tuning may be in order. Trial-and-error tuning
requires the user to observe the response of the loop to a previous event,
either a set-point change or load change, and then decide what tuning
parameter (or parameters) should be changed, in which direction, and by how
much. Experience helps in interpreting the response. The user must also have
a thorough understanding of the effect of changing each of the tuning
parameters.
Response to Various Tuning Parameter Combinations

For a self-regulating process controlled by a PI controller (interactive or non-
interactive), the tuning map on Figure 6-4 depicts the response to a set-point
change for various combinations of proportional and integral tuning. Move-
ment from top to bottom of the map represents an increase in control algo-
rithm gain or a decrease in proportional band. Movement from left to right on
the map represents an increase in integral action. The left-hand column of
graphs represent proportional action only. Moving to the right represents
adding integral action, starting with minimum integral action as determined
by the largest possible value for minutes per repeat, or the smallest possible
value for repeats per minute.2
2. Minutes per repeat and repeats per minute are the most widely used measures of integral
action. In general, we will speak of minutes per repeat in this book. Some manufacturers, as
well as the Fieldbus Foundation, use seconds per repeat for integral action and seconds for deriva-
tive time.
Bibliography
ANSI/ISA-5.1-2009. Instrumentation Symbols and Identification. Research Triangle Park,

NC: ISA (International Society of Automation).
ANSI/ISA-75.05.01-2000 (R2005). Control Valve Terminology. Research Triangle Park,

NC: ISA (International Society of Automation).
ANSI/ISA-75.26.01-2006. Control Valve Diagnostic Data Acquisition and Reporting.

Research Triangle Park, NC: ISA (International Society of Automation).
strm, K. J., and T. Hgglund. Advanced PID Control. Research Triangle Park, NC:
ISA (International Society of Automation), 2006.
PID Controllers: Theory, Design and Tuning. Research Triangle Park, NC: ISA
(International Society of Automation), 1995.
Ayral, T. E. We Need A Bigger Surge Tank. Pearland, TX. In Control XI, no. 1, 1998.
Badgwell, T. A. and S. J. Qin. Model Predictive Control in Practice. Encyclopedia of

Systems and Control (2014). London: Springer-Verlag.
Bennett, S. A History of Control Engineering, 1800 ~ 1930. New York: Institution of

Electrical Engineers, 1979.
553
Berge, J. Fieldbuses for Process Control: Engineering, Operation, and Maintenance.

Research Triangle Park, NC: ISA (International Society of Automation), 2002.
Blevins, T. L., G. K. McMillan, W. K. Wojsznis, and M. W. Brown. Advanced Control

Unleashed. Research Triangle Park, NC: ISA (International Society of Automa-
tion), 2003.
Bode, H. W. Network Analysis and Feedback Amplifier Design. New York: D. Van
Nostrand Co., 1945.
Boger, H. W. Flow Characteristics for Control Valve Installations. ISA Journal 13,
no. 11: 50-54 (November 1966).
Box, G. E. P., G. M. Jenkins, and G. C. Reinsel. Time Series Analysis, Forecasting and
Control. 3rd ed. Upper Saddle River: Prentice-Hall, 1994.
Bristol, E. H. On a New Measure of Interaction for Multivariable Process Control.

IEEE Transactions on Automatic Control (January 1966).
Camancho, E. F., and C. Bordons. Model Predictive Control. London: Springer-Verlag,

1999.
Caro, R. M. The Consumers Guide to Fieldbus Network Equipment for Process Control. 4th
ed. Chestnut Ridge, NY: Copperhill and Pointer, Inc., 2007.
Cohen, G. H., and G. A. Coon. Theoretical Considerations of Retarded Control.

Taylor Instrument Companies Bulletin TDS-10A102.
Connell, R. Basic Math for Process Control. Research Triangle Park, NC: ISA
(International Society of Automation), 2003.
Corripio, A. B., and M. Newell. Tuning of Industrial Control Systems. 3rd ed. Research
Triangle Park, NC: ISA (International Society of Automation), 2015.
Cutler, C. R., and B. L. Ramacher. Dynamic Matrix Control A Computer Control Algo-
rithm. Paper No. 51b, AIChe 86th National Meeting, April 1979.
Dahlin, E. B. Designing and Tuning Digital Controllers. Instruments and Control Sys-
tems (June 1968).
Index
ABB 179 anti-reset windup 140141

absquare 130131 protection 136
actuator 68, 132, 154, 161, 498 techniques 139
adaptive control 259 apparent
adaptive gain 260 control loop 423424
adaptive tuning 259260, 264, 266 dead time 5960, 71, 128, 181182, 185186
additive 309, 325, 421 time constant 182, 186
disturbance 431, 433 approximate process model 422
feedback 318, 320, 325, 341 approximate time constant 262
feedback-feedforward 323 automatic 64, 83, 135, 142, 154, 164, 174, 188, 191,
feedforward 334335, 346347 201, 263, 265, 271, 285, 287288, 376, 386, 390,
stream 296 392, 401, 403, 482
advanced control 269 ratio adjustment 299
advanced regulatory control 23, 152, 252, 269 reset 96, 107
270, 273, 275, 396, 473474 set ratio control 300, 305, 307, 517
air-fuel ratio control 295, 299, 333 switch 25, 132, 156, 285
air-to-fuel ratio 296, 299302, 324, 487, 515516 to manual 133134, 146
air-to-open 32 auto-tuning 259
algorithm 101 auxiliary variable 449, 453, 469470
synthesis 423, 429 averaging liquid-level control 247
analog
control 144, 285, 297, 426, 482 back pressure regulator 72
controllers 70, 106, 117, 122, 126128, 134, 136, backward link 155
146 basic regulatory control 152, 156
equipment 332 basis weight 421
hardware device 316 batch 139
input 144, 152, 284, 290 chemical process 343
output 32, 144145, 152, 285, 482, 502 controller 139
system 112, 287 digester 363
world 332 switch 138140
analog-to-digital (A/D) converter 144 bias 4, 8890, 95, 152, 246, 297
annealing furnace 343 blending 296, 434, 441442
559
Bode plots 21 386, 394, 396, 398, 401402, 404, 408, 419420,
boiler-drum control 346 423, 431, 438439, 449, 451, 461, 481482, 491
box car integration 142 control modes 106, 191
Bristol 386 control move 453, 456, 458462, 464, 468469, 472
bumpless transfer 132134, 146, 155156, 284, 287 473
288, 304, 372, 404 control strategy 2, 26, 28, 144, 150152, 154, 156,
bumpless tuning 135, 261, 375 252, 263, 275, 279, 340, 353, 363, 385, 396, 411,
423, 427428, 443, 473474, 488489, 497
calculus 78, 1011 control system engineer 2, 31, 270, 449, 452
cascade 284 control systems designer 405
control 6768, 71, 114, 116, 148, 269, 275284, control valve 46, 64, 66, 68, 7172, 219, 275, 278,
288, 290 413414, 490, 494, 498, 510
loop 114, 116, 121, 228229, 284285 control variable 453, 456, 458, 461
mode 288 controlled flow 295297, 299, 303
cascade-local switch 285 controlled variable 28, 453
catalytic cracking 451 controller gain 4, 67, 8890, 9293, 96, 98, 100101,
characterization 130, 132, 247 123, 126, 130, 135, 137, 148, 171, 182, 184186,
chart recorder 8, 97, 99, 102 189, 192, 194, 201, 246, 249, 260, 271, 280, 321,
chemical 1, 3940, 45, 64, 76, 343, 395 426
reactor 28, 114, 249 controller graph 90, 9294, 96
closed-loop controller output 8, 2932, 40, 4245, 5052, 8486,
response 170, 187188, 425426 8892, 94, 96, 98104, 106, 112117, 120121,
test 162, 169, 177, 188192, 212, 264 127, 132, 135137, 139141, 143146, 148, 164,
tuning 263 171, 173175, 177178, 180181, 189, 217, 246,
Cohen-Coon 178179 251, 260, 262264, 270272, 285, 301304, 321,
coincidence point 434 326, 341, 345346, 354, 357358, 361, 365366,
combining feedback and feedforward 318, 325, 370372, 375376, 398399, 402, 414, 419, 422,
327, 342, 414 426, 428, 435, 450, 453, 456, 479, 481, 488, 490,
combustion control 300 498501, 515516, 519
commissioning 151, 282, 288, 451 controller tuning 20, 52, 54, 9798, 119, 124125,
composition 128, 135, 162, 165, 178, 182, 186, 190, 192193,
analyzer 64, 316, 342343 201, 212, 217218, 251, 262263, 333, 357, 419,
control 74, 296, 383, 428 423, 461, 503
controller 283, 343, 373, 385 controlling function 1
loop 106 control-loop behavior 194
compressor station 366, 368 convolution model 472
configuration option 111, 113114, 135, 153, 314 cooling 479
constraint 2, 223, 419, 428, 449453, 469471, 473 cooling tower systems 494
474 Coriolis meter 70
control algorithm 45, 26, 29, 75, 106107, 111, 116, coupled 5761, 408
121, 123, 126127, 134, 139, 141142, 169, 174, critically damped 223, 227, 234
186, 223, 225, 234235, 247, 260, 262, 314, 321, cross-limiting 295
372, 420, 422424, 426, 429, 440, 474, 497 control 300, 479, 482
control engineer 912, 26, 29, 3940, 192, 260, 269 crude switch 343
270, 353354, 434, 481482
control graph 91 damped frequency 231
control horizon 453, 456, 462, 465 damping characteristics 174175
control law 29, 473 damping factor 222, 224, 231234
control loop 2, 7, 10, 26, 2831, 43, 46, 50, 52, 55, 67, data vector 453
71, 74, 85, 88, 9192, 96, 105107, 112, 121, 125, DCS 151152, 156, 265, 299, 404, 426, 438, 452, 482
127128, 133, 140141, 146148, 161162, 164
167, 171, 173, 177178, 183, 185, 192, 199, 217
218, 222, 224, 231233, 245246, 250251, 260,
262, 265, 278, 282283, 295, 315, 333, 343, 383
Index 561
dead time 45, 1921, 60, 7071, 74, 128, 153, 178, distillation column 7172, 343344, 348, 383, 428,
180181, 184186, 188, 191, 199201, 212, 215, 490
217, 220, 228229, 245, 331334, 336337, 340, control 342343, 395
346, 399402, 404, 419, 422423, 428, 434, 441, distillation tower 28, 74, 245, 247, 283, 296, 329,
443444, 449, 453 364, 373, 409
dead zone 132 distributed control systems 26, 141
dead-band 51 disturbance 5, 10, 2930, 4145, 76, 112, 119121,
dead-time compensation 21, 186, 270 163164, 166, 171, 176, 178, 182, 188, 192193,
dead-time-to-time-constant 186, 199 197, 199, 220, 222224, 226228, 230231, 233
Decay Ratio 228 235, 241, 243244, 246, 252, 265, 272273, 275,
decay ratio 163165, 180, 193194, 197, 200, 222, 278284, 291, 312313, 315317, 321, 323329,
224225, 227, 230, 233234 332335, 341342, 371375, 408, 422, 426427,
decoupler 396, 404405, 408 431432, 435, 438, 440442, 445, 449450, 453,
decoupling 2, 384385, 396, 400, 403, 409, 411412, 456, 461, 464465, 487488, 503
420, 427, 450, 473475 drum level 6566, 219
control 252, 270, 406, 420, 450, 475 dynamic 7, 12, 40, 68, 76, 93, 107, 217, 220, 327328,
element 396, 398399, 402403, 407408 332333, 398, 408, 411, 434435, 438, 441442,
dependent variable 1617, 3940, 42, 90 470, 474, 507
deriv. time 107 behavior 12, 15
derivative 89, 11, 67, 388 characteristics 46, 52, 56, 67, 329, 395
action 74, 127, 173174, 188, 249 compensation 325, 328329, 332334, 336337,
contribution 103, 113 339341, 346348, 399, 441
gain 4, 128, 176 response 12, 56
derivative mode 4, 55, 70, 85, 101103, 105106, system 18, 231232
112115, 118, 121, 127128, 153, 164, 172173,
175176, 178, 341, 355 effective controller gain 123, 130, 248
contribution 102 effective integral time 248
derivative on error 114, 118, 249 efficiency 4546, 300, 316317, 342, 357
derivative on measurement 114, 116, 118, 121, 144, electric motorized valve 132
180, 249 electrical stepping motor 147
derivative spike 112115, 128 elements 56
derivative time 5, 101, 121, 123, 125, 128, 169, 176, energy efficiency 299
179, 190 engineering units 88, 142, 144, 148149, 154, 485
design conditions 39 486, 515520
design ratio 303 equal percentage 499500, 513
detune 384 valve characteristic 498
deviation 2, 4, 101, 111, 130, 132, 134, 164167, 170, valves 498, 507, 509
172, 180, 192, 221, 223224, 226, 228, 231, 234, ERF 358359, 365
270, 272273, 282, 332, 336337, 341343, 371, error 4, 2932, 52, 88, 9498, 101, 112115, 117118,
376, 386, 393, 432, 434435, 437, 516 125, 128, 130136, 138139, 142144, 153, 166
device description 153 167, 173, 178, 191, 217, 248, 260, 271272, 301
differential equations 7, 12, 1415, 84 302, 315, 317, 334, 336, 357, 359362, 369372,
digital control 88, 127 374375, 396, 407, 414, 428, 450, 459, 469, 497,
digital control system 2, 32, 331 503
digital-based control system 331 error signal 8, 10, 98, 102, 130, 247, 375
digital-based controller 70 error-squared 130
digital-to-analog (D/A) converter 145 algorithm 131, 247
direct digital control 147 EXACT 264265
direct effect 29, 386, 389, 391, 395 exothermic reactor 5455, 106, 248
direct-acting 3132, 44, 88, 102, 113, 115, 142, 221, external feedback 288
301 external reset 137
discharge temperature 282 external reset feedback 137138, 358, 365368, 372
discrete algorithm 148
distillation 434, 491 fail-closed 3233, 287, 356, 480
fail-open 3233, 301, 480 forward 396

feedback and feedforward control 318, 321, 325, decoupling 398, 401402, 404
327, 329, 341 link 155
feedback control 2, 21, 54, 8385, 107, 119, 144, 182, FOUNDATION Fieldbus 149150, 251, 288, 305, 346
272273, 275, 281, 288, 301, 313314, 325, 333, 347, 376, 404
340, 342343, 345, 398, 419, 423, 439, 474 Foxboro EXACT 264
feedback controller 1, 74, 76, 8486, 106, 132, 136, frequency response 21, 221
252, 275, 279280, 291, 299, 314, 318320, 323 fuel-to-air ratio 482483, 485489
327, 334, 340341, 343, 345, 353, 414, 423, 429, function block 2, 30, 32, 120, 144146, 148, 150,
431 152156, 269, 284, 287, 290, 297, 299, 301, 305,
feedback loop 54, 133, 273, 280, 301302, 315, 320, 314, 316, 331, 334, 337, 346, 353, 376, 404, 423,
323, 327, 406, 424, 429 426, 452, 473, 480, 482, 497, 500, 518519
feedback penalty 2, 273, 279, 302, 313314, 334, functional diagramming 25
340, 342343, 345346
feedback trim 319321, 345346 gain 5, 70, 74, 90, 93, 96, 100, 107, 121122, 126, 135,
feedforward control 2, 270, 284, 302, 312, 314315, 148, 152, 169174, 176, 179, 185, 188, 190191,
317, 321, 325329, 331335, 340343, 346, 348, 194, 196197, 218, 233234, 299, 332, 341, 359,
398399, 413, 439, 441, 443, 450 368369, 371, 384, 393, 399, 402, 424, 442
feedforward controller 312314, 316, 318, 325328, gap 497
331334, 337, 341342 gap action 132, 497
gain 334, 336 gas transmission pipelines 74
feedforward gain 330, 335336, 341, 346 gas-pressure loop 74, 106
fieldbus 149, 269, 480, 482 generic model control 434
Fieldbus Foundation 150, 152, 156, 169, 305, 346 graph 19, 8991, 9394, 99, 170173, 194, 196, 201,
filter 5, 112, 121122, 127128, 154 212, 215, 217, 338, 340, 499
filter time constant 128
filtering 70, 128, 186, 501, 503 Hgglund 107, 263264, 345
final control element 2829, 31, 6869, 74, 86, 112, headbox 421
134, 143, 145148, 172173, 280, 312313, 345, heat and mass balance equations 412
383384, 396, 401402, 405 heat exchanger 22, 2829, 4041, 58, 71, 275276,
fine-tuning 175, 186, 333334, 337 278, 280, 283, 492, 527
first-order lag 45, 1718, 20, 5658, 6061, 63, 68, heater 46, 316317, 326327, 329, 355357, 359,
71, 74, 76, 128, 137, 182, 187, 246, 359361, 368, 361362, 409, 411, 483, 485, 501
422, 424, 426, 428, 443, 488489 heating 479
coupled 61 heating, ventilating, and air conditioning (HVAC)
uncoupled 5960 280, 383
first-order plus dead time (FOPDT) 63, 180181, Heaviside transforms 15
262263, 329, 426, 449 Heaviside, Oliver 15
flag passing 146 holdup time 246
flashing liquid 245 Hurwitz 84
floating control 479, 489, 491, 494 hydraulic analogy 53, 57, 6061
flow chart 194197, 199 hysteresis 42, 51, 68, 84
flow control loop 6870, 72, 220, 312, 489
flow controller 12, 53, 68, 70, 72, 149, 219, 275, 278, IAE 166168, 178180, 199, 228
282283, 287, 297, 299, 303, 315, 319, 328, 362 ideal 396
365, 385, 392, 398, 402, 411, 413414, 455, 482 decoupling 407
483, 485, 487489, 516 PID 101, 107, 117, 122123
flow loop 68 identification 25, 453, 455, 472473
flyball governor 83 improved control 151
food process 28, 343 improving as-found tuning 165, 192193, 199201,
FOPDT 2021, 63, 71, 177178, 180181, 183, 185, 252
187, 190, 199, 329330, 333334, 399, 402, 406, incorporating feedforward control 464
426, 431, 442 incremental 144
incremental algorithm 146
Index 563
incremental form 144, 146147 ISA symbols 25, 275, 295

independent gains 4, 126127, 174, 176, 434 ISE 166167
independent variable 8, 1617, 3940, 42, 90 ITAE 166167
indeterminate process gain 182, 184 ITSE 166167
indirect effect 387, 390, 395
industrial process control 1, 1819, 22, 6364, 84, lag time 61, 331, 336338, 340
383, 451 Laguerre polynomials 473
inherent 45, 4849, 75, 441 Lambda tuning 187188, 199, 266, 424
initialization 153, 155156, 287288, 304305, 404 Laplace notation 97, 101, 122, 128
inner loop 278283, 288 Laplace transform 4, 15, 2021, 106, 118, 221, 422
installed characteristic 49 Laplace, Pierre 15
installed characteristics 4950, 496, 512513 lead-lag 121, 153, 295, 330334, 338, 346, 399, 402,
integral 8, 67 431, 482, 487
action 7475, 98, 136, 138140, 169172, 174, lead-lag ratio 331, 336338, 340
188, 191, 247, 249, 356, 387, 426, 491, level arrest time 226, 230
497 level controller 66, 72, 76, 219220, 231, 233, 245,
gain 4, 174, 176 343
mode 4, 74, 96, 101102, 104106, 115, 121, 130, level loop 106, 247
133136, 138139, 172, 217, 247, 419 level-control
of the error 142, 166 loop 75, 187, 219220, 226
response 98, 112, 171172 system 66, 346
time 5, 7071, 97, 121, 123, 125, 137, 171, 174 linear controller 130
176, 179, 184185, 190, 193, 201, 228, linear valve 49, 6970, 509, 512
233235, 248, 280, 341, 369, 384 Liptk 52, 369, 494
tuning 135136, 169 liquid-level
integral-error criteria 166167 control 53, 75, 131, 169, 219, 229, 247
integral-mode contribution 135, 139 control loop 76, 221
integral-only controller 75, 130, 245
control 121, 491, 497 load 2930
controller 121, 496 load change 29
mode 121 load changes 41
integrating process 5354, 65, 7376, 162, 183, 187, loop diagrams 26
218, 246, 428 Lopez 178, 199
integrator 96, 9899, 134, 136, 138, 146, 171, 359,
431, 461 magnetic flow meter 70
interact 383 magnitude ratio 21
interacting 66 manipulated variable 4, 28, 32, 4546, 324325,
interaction 66, 68, 153, 161, 279, 383384, 387, 392, 343, 366, 394, 434435, 438440, 443, 449, 453,
394396, 398, 401, 411 469
interactive 45, 122123, 127, 153, 161, 169, 176, manual 30, 32, 64, 72, 133134, 154, 164, 171, 174,
179180, 201, 451 177, 248, 251, 262, 265, 285, 287288, 305, 334,
controller 122, 124125, 172 346, 362, 372, 386, 389, 401404, 450
form 123, 125126, 128, 174, 178 manual reset 4, 89, 96, 246
form of PID 174 manual to automatic 10, 133134, 288
interchangeability 151 manual/automatic switching 133, 146, 284, 287
internal model manufactured characteristics 509
control 187, 420, 428, 432, 451 maximum rate of change 147, 227228, 231
controller 188 McAvoy 388, 407
interoperability 151 measurement 28
inverse response 6566, 419, 428, 449 microprocessor-based
inverted 396 control 90, 114, 126, 141, 144, 146, 269
decoupler 406407 controllers 106, 128
decoupler configuration 411 system 98, 111, 134, 144, 148, 188, 284, 287, 297
decoupling 397, 401404, 407 min./repeat 107, 197
minutes per repeat 9799, 124, 169170, 174176, oscillation 64, 7475, 93, 107, 166167, 170171,
179, 188, 197, 222 173174, 185, 188, 190194, 197, 201, 218, 223,
model predictive control 269270, 396, 420, 427, 226, 232, 234, 246, 251, 264, 279
450, 474 outer loop 278279, 283
model-based outflow arrest time 226227, 231
control 187, 270, 420, 425, 434 output bias 8889, 9294, 96, 99, 141
controllers 419 adjustment 94
modes of control 85 overdamped 223
monotonic 45 response 170
move suppression 462463 override control 137138, 270, 353, 355356, 362
multi-level cascade control 282283 364, 366369, 372376, 427428, 450, 469, 471,
multiple-input 449, 467 494
multiple-input multiple-output processes 66, 76 overshoot 116, 126, 138141, 163164, 170171,
multiple-level cascade 287 180, 187, 193, 200201, 212, 215, 217
multiple-output 449, 467 overshoot ratio 165, 199, 226, 233
multiplicative
feedback 321, 323325, 327 P 8586, 107, 115, 137, 175, 178179, 190, 193, 197,
feedforward 335 201, 212, 215, 226, 453, 460, 517, 519
feedforward control 46, 325, 334, 347 P/TI ratio 199, 201, 217
feedforward controller 341 P&ID 2224, 275, 295
pairing 384386, 388, 391392, 394396, 412
natural frequency 232233, 240 paper machine 421, 428
negative feedback 3132 partial decoupling 401, 405, 408
negative overshoot 141 pass-through 368370, 372, 374
noise 7071, 75, 105106, 122, 127, 161, 167, 172, performance criteria 162164, 266
175176, 178, 185, 190, 234, 247, 265, 455 period 144, 174175, 189, 191, 194, 197, 201, 217
noninteracting 218, 226, 230232, 234, 247, 264
controller 124 petrochemical industry 395
PID 175, 180 petroleum refinery 343, 451
noninteractive 45, 126127, 169, 174, 179, 201 heater 409
controller 122, 124125, 172, 178 pH control 131, 434
form 123, 125, 153, 180 phase lag 121
form of PID 174 PI 4, 25, 8586, 97, 100, 107, 137141, 167169, 172,
nonlinear 45, 68, 162, 245, 259, 474 175176, 178179, 190, 193, 197, 201, 217, 220,
control 247 225, 233235, 247, 355356, 358, 361362, 365,
nonlinearity 46, 69, 71, 299 388390, 423, 425426, 428, 434
nonlinearization 130132 PID 4, 7, 22, 25, 8486, 96, 101, 106108, 111119,
nonrealizable 401, 405 121, 126127, 130, 136137, 141, 143148, 150
nonselected controller 357359, 368372, 374, 376 154, 156, 161163, 174176, 178180, 190, 271,
non-self-regulating processes 53, 169 285, 288290, 297, 299, 305, 346347, 355, 358,
normalized value 142, 148149, 168, 304, 517519 372, 374376, 404, 419420, 423, 426, 474, 502
Nyquist diagram 21 pipeline industry 366
piping and instrumentation diagram 22
offset 86, 107, 191, 247, 297 plant 29
on/off 84, 102, 263, 280, 311, 501 plastic extrusion 501, 503
on-demand tuning 260, 262263, 266 position 99, 143
one-quarter decay 193, 227 position-mode 318, 372
one-twentieth decay 227 algorithm 261, 319
open-loop test 178, 183, 185, 187, 201, 262263 positive feedback 31, 44, 395
tuning from 162, 177 power generation industry 26, 30
open-loop unstable 52, 55, 106 pre-act 107
operating point 16, 4445, 9495, 136139, 177, 188, predicted value 104105, 437, 450, 453, 456, 461,
251, 259261, 263, 329, 343, 363, 469470, 474 463, 470
optimization 3, 269, 450451, 473, 496 prediction horizon 453, 456, 458, 462, 465
Index 565
predictive functional control 434 processing cycle 142

predictive nature 105 process-model-based control 434
predictor 332, 428 product quality 2, 342, 411, 491
pressure control 54, 72, 74, 489492, 494 prop. band 107
loop 72, 7475 proportional
pressure controller 7273, 149, 295, 363365, 367 band 4, 67, 7071, 7475, 90, 99100, 135, 138
368, 483, 485, 488, 490494, 496 140, 169170, 174175, 189, 194, 246
pressure reducing station 72, 493 247
pressure regulator 72 contribution 115
primary 275 control 86, 102103, 174
primary controller 68, 114, 149, 278279, 283285, controller 8688, 105
287288, 303305, 396, 402 gain 4, 174176
principal disturbance 315 kick 115
process 1, 29 proportional mode 8586, 98, 107, 115, 117118,
behavior 1, 66 121, 130, 217, 247
characteristics 54, 66, 71, 189, 260, 335 contribution 102, 117
control 1, 3, 78, 1819, 22, 3940, 6364, 83 on error 121
84, 100, 106, 126, 149, 152, 166, 249, on measurement 112, 114, 153, 164
270, 363, 383, 385, 451, 489, 496, 503 proportional on error 117
control engineer 8, 14, 250, 451 proportional on measurement 117
controller 84, 96, 107, 363364, 485, 487488, proportional response 98, 112, 115116, 128, 180
492, 494, 497 proportional-band 90
design engineer 39 proportional-only
dynamic behavior 76, 342 control 246
dynamic characteristics 52, 71, 327 controller 4, 89, 92, 100, 105, 246
dynamics 75, 128, 161, 316, 327328, 333, 342, proportional-on-measurement configuration
395, 400 option 121
engineer 22, 3940, 107 proportional-plus-derivative control 103
flow diagram 26 proportional-plus-integral controller 100
gain 4, 20, 4546, 56, 63, 69, 71, 7576, 178, 180, pseudo dead time 63
182183, 186, 262, 266, 320323, 327, pseudo time constant 63
333, 386, 389, 391, 393, 422, 428, 431, pulp and paper industry 188, 363
453, 481482 pulse-response model 471473
graph 4245, 76, 9092, 9496 pure dead time 56, 59, 421
heater 45, 71, 299, 315, 324, 327, 331, 334335, pure delay 331
354, 470, 482 PV tracking 133
input 455
model 21, 56, 63, 177, 180, 187, 190, 217218, quadratic programming 470
221, 259, 262, 315, 317, 327, 334, 392, quarter-amplitude decay 165, 174, 191, 197, 222
419422, 425426, 428, 432, 434435, quarter-wave damping 165
437, 440441, 449450, 453, 455, 460, quarter-wave decay 165
471, 473474 quick-opening 48
response 1920, 63, 161, 171, 176178, 184, 262, quick-opening valve 69
340, 399, 434
time constant 178, 182, 186, 262, 407, 422, 426 ramp generator 102
variable 12, 45, 8, 2832, 45, 52, 6970, 75, ramped 116
8890, 99, 103, 116, 120, 132, 136, 142, ramping 98
148149, 156, 162, 167, 187, 189, 192, rangeability 413, 479, 496, 498499, 507
245, 251252, 260, 263264, 270, 272, rate action 101, 107
278, 280, 283, 285, 287288, 297, 299, rate of change 8, 15, 17, 86, 101102, 104, 116, 147,
312, 327, 329, 342, 353, 375376, 383 173, 227228, 234, 246, 436, 469
384, 386, 391, 394395, 420, 423, 426, rate of change of variable 449
431, 433437, 440, 449450, 453, 461,
464, 469, 483, 490, 492, 515516, 519
ratio 2, 18, 21, 4546, 49, 61, 152, 164, 178, 186, 191, switch 117, 353, 360, 373
193194, 199, 231, 233, 264, 280, 296297, 299, self-regulating process 5253, 5657, 6364, 76,
303305, 321323, 331, 333, 342343, 347, 364 162, 169, 172, 177, 187, 195, 218
365, 387, 393, 395, 400, 414, 421, 482, 485486, self-tuning 259260, 262, 265
496 sensor-transmitter 29
ratio control 46, 71, 152, 269, 275, 295297, 299300, set point 12, 45, 2931, 43, 52, 68, 75, 84, 86, 88,
302305, 324325, 333, 385, 482, 517 90, 9295, 97100, 104, 112114, 116, 118, 120
realizability 400, 404, 406408 121, 128, 130, 132136, 138142, 144, 148149,
realizable 400401, 404, 406, 429 156, 162, 164166, 170, 172174, 187188, 190
decoupling elements 401, 404 192, 194, 200, 219, 223, 226, 231, 233234, 246,
reboiler 245, 364, 373, 383, 492 251, 260, 266, 270, 272, 275, 278, 282, 284285,
reference trajectory 453, 462463, 473 287288, 295297, 299300, 304, 312, 315, 318,
reflux 283, 342, 383 322, 326328, 343, 345346, 354359, 361, 364
relative gain 387388, 391394, 408 365, 371372, 386, 392, 395, 398, 402, 411, 424,
analysis 386, 395396 426, 431, 433435, 440, 443, 450451, 458, 461,
array 387388, 391, 393395 463, 469, 473, 483, 485, 488492, 494, 496497,
matrix 394 503
relay method 263 of the feedforward controller 325
repeats per minute 9899, 107, 169, 171, 174, 188 change 100, 112116, 119121, 126, 163166,
reset 96, 98, 107, 171172, 174175, 191, 250, 260, 169170, 172, 174, 176, 180, 187188,
341 192193, 197, 199201, 220224, 226,
action 192 231, 251, 420, 423, 428, 431432, 440,
feedback 137138, 358361, 365366, 368, 372 445, 455
preload 141 filtering 121
windup 136137, 139, 151, 154 ramping 112, 116
reverse action 32, 287, 356, 419 response 119, 164, 187, 224, 266
reverse-acting 3132, 44, 88, 94, 142, 153, 301, 371, softening 112
480 tracking 132134, 288
reverse-acting controller 3132, 89, 102, 104, 113, weighting 121
115, 118 weighting controller 117
RGA 387, 395396 settling time 166, 172, 426, 435
rise time 164166, 180, 182 Shinskey 67, 74, 125, 139, 365, 385, 396, 490, 492,
robustness 407 500
root locus 21 shrink-and-swell 65, 245
Routh 84 side lag 6163
runaway 5455, 76, 248249 signal generator 97
signal selector 353354, 358, 488489, 492493
SAMA 2527, 275, 295 signals 8
SAMA symbols 2526, 275, 295 simplified 396
sample period 371, 456, 460462 sinusoidal disturbance 231, 234
sampled data step-response model 453 SISO process 452453, 461, 465
sampling period 186 Smith predictor 420423, 426429, 434, 438
sampling time 136, 439 softwiring 144
scheduled tuning 248, 260261, 263, 375, 427 split-range control 479
Scientific Apparatus Makers Association 25 S-shaped curve 5859, 61, 63, 177
secondary 275, 295 stability 84, 171, 249, 266, 404408, 434, 451
secondary controller 114, 116, 121, 148149, 249, stabilizing direction. 172
278279, 283285, 287288, 304, 404 states 84
seconds per repeat 169 status bits 146, 155156, 288
selected controller 138, 369, 371372, 374, 376 steady-state
selector 2, 137, 354355, 363, 374, 376, 483, 485, 487, characteristics 40, 76
494 error 171, 341
control 353 gain 4, 18, 329, 331, 386388, 394395
device 138 offset 74, 9295, 100101, 105, 107, 170, 246
Index 567
steam generators 65, 482 ultimate period 189, 191, 264

stem position 48, 5052, 68, 156, 413 ultimate proportional band 189
step- and pulse-response 471472 uncoupled 5760
step response 59, 63, 177, 456 undamped natural frequency 222, 231232
stick-slip 5051 underdamped 64, 76, 222223, 225227, 231, 503
stiction 50 undershoot 200
stoichiometric 299 unrealizable 401
superposition 337 usual controller tuning 423
sustained oscillation 189191
valve 28
tank holdup time 220, 224, 230 actuator 46, 5051, 71, 290
tank time constant 220 characteristics 48, 69, 76, 251, 283, 413, 507
Taylor Instrument Co. 179 position 4042, 49, 51, 61, 86, 103, 144, 146
temperature control loop 22, 31, 7071, 74, 139, 147, 156, 186, 276, 278, 284, 288, 311
282, 411, 479 312, 346, 355, 359362, 413, 490, 496
temperature controller 46, 71, 106, 114, 139, 249, 497, 499, 507, 509510, 512
275276, 278, 282283, 287, 302, 315, 321323, position controller 364, 490, 492, 494, 496497
343, 356357, 361362, 410411, 414, 481, 494, positioner 32, 46, 52, 68, 150, 154, 156, 219, 251,
496, 501 283, 480
thermostat 84, 94, 280, 282, 311312 rangeability 479, 496, 498, 507
throughput 2, 39, 71, 192, 252, 321, 327, 342 variable pairing 384385, 392
time 8 velocity (incremental) 372, 376
time constant 5, 1821, 56, 5861, 63, 71, 123, 128, velocity mode 144
137, 180182, 185186, 191, 220, 246, 262, 332 vortex meter 70
333, 368, 399, 407, 419, 421422, 424426, 428,
431, 438, 440, 442443, 453, 463, 489 water supply systems, hot or chilled 494
time delay 19, 59, 217, 332, 371, 426 wild flow 295297, 299, 303
time proportioning windup 136138, 140141, 146, 156, 287288, 356
control 479, 501502 357, 359, 366, 372
controllers 502503 Wobbe index 316
tower flooding 364, 373
transducer block 154, 156 Ziegler-Nichols 178179, 184, 245, 426
transfer function 1518, 2021, 30, 63, 118120,
137138, 179, 220221, 224, 328329, 331, 333,
398399, 406, 424425, 428, 433, 472
transmitter 29, 32, 70, 88, 144, 149150, 154, 156,
230, 278, 290, 303, 384, 403, 405, 485487, 515
520
transport lag 20, 56
trial-and-error tuning 162, 169, 173, 192, 218
tube temperature 354357, 359362, 469470
controller 354357, 362
tuning 2, 161
aids 265
log 197, 212, 215, 251
map 169172
the decoupler 408
values 124, 162, 171, 178, 184, 186, 250, 261
262
two-degree-of-freedom controller 117
two-position control 84
ultimate effect 301, 387, 390391, 395

ultimate gain 189, 191, 264

Wade BasicAdvRegCont 3rd Excerpt

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Wade BasicAdvRegCont 3rd Excerpt

Caricato da

Copyright:

Formati disponibili

Standards

Education & Training

Conferences & Exhibits

Buy the Book

Copyright 2017 International Society of Automation (ISA)

Printed in the United States of America.

Library of Congress Cataloging-in-Publication Data in process

To Mary, my wife of many years, who provides love, support,

About the Author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Preface to the Third Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xv

Chapter 2 Mathematical Background, Diagrams, and Terminology. . . . . . . . . .7

Chapter 3 Process and Control Loop Characteristics. . . . . . . . . . . . . . . . . . . .39

Flow Control Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Chapter 4 PID Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83

Chapter 5 Modifications to Standard PID Control . . . . . . . . . . . . . . . . . . . . . .111

Chapter 6 Tuning Feedback Control Loops . . . . . . . . . . . . . . . . . . . . . . . . . . .161

Tuning for Self-Regulating Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Chapter 7 Self-Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259

Chapter 8 Advanced Regulatory Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269

Chapter 9 Cascade Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .275

Chapter 10 Ratio Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295

Chapter 11 Feedforward Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .311

Chapter 12 Override (Selector) Control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .353

Chapter 13 Control for Interacting Process Loops . . . . . . . . . . . . . . . . . . . . . .383

Chapter 14 Dead-Time Compensation and Model-Based Control . . . . . . . . . .419

Chapter 15 Multivariable Model Predictive Control . . . . . . . . . . . . . . . . . . . . . .449

Chapter 16 Other Control Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .479

Appendix A Derivation of Equations for Installed Valve Characteristics . . .507

Appendix B Signal Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .515

Appendix C Exercise Answers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .523

Harold L. Wade, PhD, has had over 50 years of experience in designing,

Although this is intended to be a practical how-to book, readers should not

The second partChapters 4 through 7addresses feedback control. The

(MIMO) processes (Chapter 13) contains additional coverage of inverted

Also contained in this section are illustrations of the use of FOUNDATION

This book focuses on a particular segment of automated process control that is

The behavior of the processes in this book is often characterized by measured

With this additional capability, however, a need developed for a systematic

Using advanced regulatory control provides multiple benefits. One of the

Figure 1-1. Overall Process Control and Information System Hierarchy

The following symbols are used throughout this book:

b = bias value (manual reset) on proportional-only controller

TD = derivative time (noninteractive and interactive control

Set point Center of your lane

Process variable __________________

Error (or deviation) __________________

Final control element Steering mechanism

1.2 Another, more realistic, example of a feedback control loop is speed

Set point Your desired speed

Process variable indicator __________________

Final control element __________________

Manipulated variable __________________

Constraint Speed limit

1.4 There are many important topics in an overall automation project. As

Safety instrumented systems

Sensor/analyzer selection and installation

Abnormal situation management

Human-machine interface (HMI) design

Communications technology (for plant instrumentation and

In order to design, analyze, or commission a process control system, one must

The process engineer is usually concerned with the steady-state conditions of

Figure 3-1. Disturbances to a Control Loop

Other important variables to be considered are the disturbances to the process.

Changes in the process flow rate, Fc