Armando b corripio ph d p e , michael newell turning of industrial control systems, third edition international society of automation (2015)

Tuning of Industrial Control Systems Third Edition By Armando B Corripio, Ph.D., P.E Chemical Engineering Louisiana State University and Michael Newell Automation Designer Polaris Engineering Notice The information presented in this publication is for the general education of the reader Because neither 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 particular 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 presented 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 referenced belong to the respective owner of the mark or name Neither the author nor the publisher makes any representation regarding the availability of any referenced commercial product at any time The manufacturer’s instructions on the use of any commercial product must be followed at all times, even if in conflict with the information in this publication Copyright © 2015 International Society of Automation (ISA) All rights reserved Printed in the United States of America 10 ISBN: 978-0-87664-034-0 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 permission of the publisher ISA 67 Alexander Drive P.O Box 12277 Research Triangle Park, NC 27709 Library of Congress Cataloging-in-Publication Data in process Preface to the Third Edition This third edition of Tuning of Industrial Control Systems has been significantly simplified from the second edition with the goal of having the discussion more in line with modern control systems and with language that is less academic and more in tune with the vocabulary of the technicians who the actual tuning of control systems in industry For example, we have eliminated any references to first- and second-order models since these terms are highly mathematical and may discourage some from appreciating the usefulness of the models We have also eliminated the distinction between series and parallel PID controllers since most modern installations use the series version and there is not much difference between the tuning of the two versions We have reduced the tuning strategies to just one; the quarter-decay-ratio (QDR) formulas slightly modified by the Internal Model Control (IMC) rules for certain process characteristics All the tuning strategies are intended for responses to disturbances with a discussion on how to modify these responses to avoid sudden excessive changes of the controller output on set point changes when such changes are undesirable Chapter 10 is new and deals with the auto-tuning feature that has become standard on current process control systems We have successfully used the auto-tuning feature in our tuning work on oil refineries as a reference to guide our selection of the final tuning parameters for the controllers We have kept the previous edition’s discussions on the problems of process nonlinearities and reset windup, and how to compensate for them All of the tuning strategies are demonstrated with computer simulation examples ix Contents Preface to the Third Edition ix – Introduction 1-1 The Goal of Tuning 1-2 Feedback Control 1-3 Other Control Strategies 1-4 Organization of the Book 1-5 Summary References Review Questions 8 8 – The Feedback Controller 11 2-1 The PID Control Algorithm 2-2 Stability of the Feedback Loop 2-3 PID Controller Tuning by the Ultimate Gain and Period Method 2-4 The Need for Alternatives to Ultimate Gain Tuning 2-5 Summary References Review Questions 12 19 21 29 29 30 30 – Open-loop Characterization of Process Dynamics 33 3-1 Open-loop Testing—Why and How 3-2 Process Parameters from the Open-loop Test 3-3 Physical Significance of the Time Constant 3-4 Physical Significance of Dead Time 3-5 Effect of Process Nonlinearities 3-6 Summary References Review Questions 34 36 41 46 50 53 54 54 v vi Tuning of Industrial Control Systems, Third Edition – How to Tune Feedback Controllers 57 4-1 Tuning from Open-loop Test Parameters 58 4-2 Practical Controller Tuning Tips 67 4-3 Reset Windup 70 4-4 Processes with Inverse Response 71 4-5 Effect of Nonlinearities 74 4-6 Summary 74 References 75 Review Questions 76 – Mode Selection and Tuning of Common Feedback Loops 77 5-1 Deciding on the Control Objective 78 5-2 Flow Control 79 5-3 Level and Pressure Control 80 5-4 Temperature Control 87 5-5 Analyzer Control 91 5-6 Summary 92 References 92 Review Questions 92 – Tuning Sampled-Data Control Loops 95 6-1 The Discrete PID Control Algorithm 96 6-2 Tuning Sampled-data Feedback Controllers 103 6-3 Selection of the Sampling Frequency 111 6-4 Compensation for Dead Time 113 6-5 Summary 117 References 118 Review Questions 118 – Tuning Cascade Control Systems 121 7-1 When to Apply Cascade Control 122 7-2 Selection of Controller Modes for Cascade Control 125 7-3 Tuning of Cascade Control Systems 127 7-4 Reset Windup in Cascade Control Systems 136 7-5 Summary 140 Review Questions 140 – Feedforward and Ratio Control 143 8-1 Why Feedforward Control? 144 8-2 Design of Linear Feedforward Controllers 149 Contents vii 8-3 Tuning of Linear Feedforward Controllers 8-4 Nonlinear Feedforward Compensation 8-5 Summary References Review Questions 152 158 165 166 166 – Multivariable Control Systems 169 9-1 What is Loop Interaction? 9-2 Pairing of Controlled and Manipulated Variables 9-3 Design and Tuning of Decouplers 9-4 Tuning of Multivariable Control Systems 9-5 Model Reference Control 9-6 Summary References Review Questions 170 174 186 193 196 198 199 199 10 – The Auto-tuner Application 201 10-1 Operation 10-2 Applications 10-3 Features and Settings 10-4 Summary Review Questions 202 205 209 212 213 Appendix A – Suggested Reading and Study Materials 215 Appendix B – Answers to Study Questions 217 Index 233 Introduction Automation is essential for the operation of chemical, petrochemical, and refining processes It is required to maintain process variables within safe operating limits while maintaining product purity and optimum operating conditions Because all processes are different in their speed of response and sensitivity to control adjustments and disturbances, the parameters of the automatic controllers must be adjusted to match the process characteristics This procedure is known as tuning The purpose of this book is to provide you, the reader, with an understanding of the most commonly used and successful tuning techniques for the various control strategies used in industry This first chapter presents a general discussion of the goal of tuning, a description of feedback control—the most common strategy—and a brief introduction to other common control strategies Learning Objectives — When you have completed this chapter, you should be able to A Define the main goal of tuning a control system B Understand the feedback control strategy C Identify the various components of a feedback control loop Tuning of Industrial Control Systems, Third Edition 1-1 The Goal of Tuning The goal of tuning is to produce a smoothly operating process One common misconception is that every process variable should be brought to its desired value as quickly as possible and closely maintained at that value When a controller is “tightly” tuned to maintain close control of a process variable, it must make large, fast changes in its output, which usually causes disturbances to other variables in the process As the controllers of these other variables take action they, in turn, cause further disturbances that affect other variables Before long the entire process is in a state of continuous change, which is undesirable and may be unsafe in some occasions The situation worsens when the controllers cause oscillatory process responses, because then the process variables will be continuously changing The following heuristics (“rules of thumb”) may prove helpful to those just starting in the tuning of processes: • The variability of the controller output should not be excessive; however, keeping the output variability low must be balanced against the precision with which the process variable is to be controlled • Some variables not have to be maintained at their desired values The most common example of this is liquid levels, which usually only need to be kept within a safe range • The controller cannot move the process variable faster than the process can respond, so the controller speed must be matched to the speed of response of the process Some processes respond in a matter of minutes, while others may take close to an hour or longer to respond Not many processes respond in a matter of seconds One more item to keep in mind is that there is no such thing as fine-tuning a controller, particularly a feedback controller In most cases the tuning parameters need only be adjusted to one, or at most, two significant digits There are two reasons for this One is that feedback controllers are not that sensitive to variations in the third digit of their tuning parameters The other is that the characteristics of most processes—that is, speed of response and sensitivity to changes in controller output—vary with operating conditions, sometimes slightly and other times not so slightly This means that the controller tuning Introduction parameters are usually compromises selected to work in the range of operating conditions, and so their values are not precise Understanding this simplifies the task of tuning because it reduces the number of values of the tuning parameters to be tried For example, it is a lot easier to decide between gain values of 1.0 or 1.5 than to try to find out whether the gain should be 1.276 In practice, all three of these values will work about the same Armed with these heuristics and basic concepts, we are now ready to look at the feedback control strategy 1-2 Feedback Control Feedback control is the basic strategy for the control of industrial processes It consists of measuring the process variable to be controlled (the controlled variable), comparing the measurement with its desired value or set point, and taking action based on the difference between them to reduce or eliminate the difference—that is, to bring the controlled variable to its desired value The action taken results in the adjustment of a process flow, such as the steam flow to a heater, which has a direct effect on the controlled variable, such as the outlet process temperature The three instrumentation components required for feedback control are: • A sensor/transmitter to measure the process variable and send its value to the controller (Measurement) • A controller to compare the value of the process variable to its desired value, determine the required control action and send it to the final control element (Decision) • A final control element, usually a control valve or variable speed drive, to vary the manipulated process flow (Action) A fourth element of the loop is the process itself, through which the manipulated flow affects the controlled variable The controlled variable is also known as the process variable (PV), its desired value is the set point (SP), and the signal from the controller to the final control element is the controller output (OP) Tuning of Industrial Control Systems, Third Edition It is important to realize that a feedback controller does not use a model of the process to compute its output It takes action by trial and error Tuning the controller is the procedure of adjusting the controller parameters to ensure that the controller output converges quickly to its correct value In order to better understand the concept of feedback control, consider as an example the process heater sketched in Figure 1-1 The process fluid flows inside the tubes of the heater and is heated by steam condensing on the outside of the tubes The objective is to control the outlet temperature T of the process fluid in the presence of variations in process fluid flow (throughput or load) F and in its inlet temperature Ti This is accomplished by manipulating or adjusting the steam flow to the heater Fs and with it the rate at which heat is transferred into the process fluid, thus affecting its outlet temperature Figure 1-1 Feedback Temperature Control of a Process Heater SP Steam OP TC Fs PV F Ti TT Process fluid T Steam trap Condensate Index Terms Links H half decoupling 188–189 heat exchanger 45 61 88 44–45 50 81 87 89 93 130 161 3–6 19 21 25 34–36 52 61–62 74 88 90–91 107 115 149 160–161 163–165 144 heat rate controller (QC) heat transfer heater 88 206 efficiency 162 feedforward controller 163 temperature hydrogen/nitrogen ratio hysteresis 106–110 117 134 69 80–81 128 60 62–64 74 80 92 132 92 108–109 13–14 21 25 51 58 66 70–71 78–79 82 85 112 124 126 136 139–140 206 I IMC 222 independent variables integral controller 198 79–80 223 integral mode This page has been reformatted by Knovel to provide easier navigation Index Terms Links integral mode (Cont.) integral time integrating process interaction interaction measure 145 152 172 175 188 218 222 226 230 13 17 19–21 24–25 30 58–60 62–64 66 68–69 74 79–80 82 86–87 92 97 100 105 107 109 112 127–128 132 136–137 196 218 221–225 41 69 112–113 158 169 171–173 175 178–181 184–185 189 193–194 196 199–200 205 174 176–177 199 230 intermediate level control 86 internal model control (IMC) 60 62–64 74 80 92 132 71–74 76 133 173 185 188 194 222 222 intuitive 174 inverse response J jacketed reactor 136 This page has been reformatted by Knovel to provide easier navigation Index Terms Links L lag 41 153 lagniappe 123 lead 153 lead-lag compensation 163 lead-lag unit 144 151–153 155–156 163–164 166 187 72 81–82 84–87 92 96 118 125 182 207 227–228 level control 223–224 linear feedforward controllers liquid storage tank 149 42–44 local override 202–203 local set point 18 loop interaction 152 205 212–213 170 173–174 186 198–199 230 22 60 68 70–71 78 88 149 151–152 160 162 166 174 206 217 225 227–228 M manipulated variable manual output master controller material balance control measured disturbance 34 138 184–185 144 147–148 158–161 163 166 197 227–228 This page has been reformatted by Knovel to provide easier navigation Index Terms microprocessor microprocessor-based controllers Links 96 130 156 228 35 100–101 111 113 118 224 96 205 microprocessor-based PID control algorithm 103 minor disturbances 146 mode 12 152 automatic 205 derivative 14–16 21 25 27 30 58 61 66–67 78 82 87 97 110 126–128 218 223 226 13–14 21 25 51 58 70–71 78–79 82 112 124 126 136 139–140 145 152 171–172 175 188 218 222 226 12 14 16 30 66 76 82 100 125 integral 230 manual 205 proportional 218 rate 14 reset 13 171 selection 77 226 model reference control 196 move-suppression parameters 198 multiple input, single output (MISO) 143 multiplexer 140 This page has been reformatted by Knovel to provide easier navigation Index Terms multivariable control Links 169–170 174 182 190 192–194 196 173 179 181 185–186 188 194–195 199 231 198 N negative feedback negative interaction noise 209 nonlinear feedforward compensation 158 nonlinear gain nonlinearity 101–102 118 224 58 148 221 174 181 12 26–27 78 126 139 226 78 90 174–178 180 199–200 230 33–36 58 61–63 67 74 86 103 110 O off-line offset on-off controllers open-loop gain open-loop test optimizing feedback loops 112 oscillations 203 output 184 4–7 12–14 16 18–20 25–27 30 34–37 39–41 50–51 54 61–63 65–76 80 82 85 95–98 100–104 106 108 111–114 This page has been reformatted by Knovel to provide easier navigation Index Terms Links output (Cont.) overshoot 116 118 122 125 128 131 133 135–137 139 144 146 148–150 152–153 155–159 162–163 169–171 173–174 176–179 181 186 188 190 194 197–199 202 204 218–219 221–222 224 229–230 63 69–71 100 125 127 132 136–137 139 173 188 194–195 210–211 222 226 232 173–180 183 185–186 188–189 195 199–200 69 194 P pairing 230 parallel 45 225 parallel paths 149 PD controller 78 percent controller output 17 perfect control performance 188 144 166 227 28–29 50 57–59 64 66–69 72 74 76 95–96 99–100 107 110–113 116 121 123–125 127 133 140 144 151 This page has been reformatted by Knovel to provide easier navigation Index Terms Links performance (Cont.) 158 165 169 173–174 181 185 188 210 221–222 25–26 30 58 61–63 74 79 85–86 92 97 105–106 109–110 115–116 132 192 223 226 pH control 101 PI controller PID PID algorithm 96 PID controller 11–12 17 21 24–27 30 58 61–64 72 76 92–93 106 110 114 132 223 173 179 185 195 230 80–81 86 92 112 125 136 pneumatic 96 positive interaction pressure control primary controller 122 problematic loops 206 process dead time 37 48 59 76 106 109 111 113 225 33 50 52–53 55 57–59 65 68 109 122 204 208 220–221 process gain 232 process noise 209 This page has been reformatted by Knovel to provide easier navigation Index Terms Links process nonlinearity 33 51 74 process time constant 37 41 60 86 91 106 108 111 222 1–3 5–8 12–19 25 34–36 41 46 64 66–67 71 76 78 82 88–89 91–92 95–98 100 102 106 110 112 115 118 121 125–128 131 136 139–140 144–145 149 151–152 158–159 166 169–174 176 178–179 181 187–188 192–193 197–199 205–206 208–209 212–213 217 219 221–222 224 226–227 process variable 230 processing frequency programmable logic controllers (PLC) 111 96 proportional band 17–18 78 82 proportional controller 21–23 25–26 29 78 85–86 90 92 102 223 12 17 19 22 27–28 58 65 68 78 80 82 85–86 97 101 118 125–126 133 218 221 226 proportional gain This page has been reformatted by Knovel to provide easier navigation Index Terms proportional kick proportional mode proportional-integral controller proportional-integral-derivative controller Links 100 118 224 12–14 16 66 76 82 100 125 218 18 27 128 223 11 18 28–29 87 91 101 224 96–98 100 224 12–17 25–27 30 34 37–39 48 61–64 96–99 102 145–146 149–150 170–172 175–177 187–188 209–210 218–219 23 25 58 74 133 24 26 61–63 132 80 96 proportional-only controller proportional-on-PV pulse PV 82 100 Q QDR response QDR tuning quarter-decay ratio (QDR) response 28 58 R rate time ratio control reactor 14 58 7–8 143–144 148–149 166 179 227 41 54 70–73 122–123 126 This page has been reformatted by Knovel to provide easier navigation Index Terms Links reactor (Cont.) 129 131–133 135 137 139 218 175–180 183 185–186 189 193–196 198–200 174 178 185 138–139 226 reset time 13 58 reset windup 51 57 69–71 74 76 121 128 130 136 138 140–141 222 42–43 45 86 19 217 relative gain 230–231 matrix reset feedback resistance resistance temperature device (RTD) 89 reverse reverse action running away 19 S sample time 106 225 sampling frequency 103 107 111 131 sampling period saturation secondary controller self-regulating 91 104 50–51 69–70 138 188 136 122 35 41 48 60 78 82 87 108 123–124 127–128 145 158 164 206 223 226 205 sensor This page has been reformatted by Knovel to provide easier navigation Index Terms sensor time constant sensor/transmitter Links 87 89 128 34–35 37 58 79 91 45 48–49 98 127–128 226 5–7 12–19 22–24 27 30 61–71 75–76 78–80 82 86 89 92 96–98 100–102 106 113 116 118 122–123 125–128 131–133 135–137 139 144–148 150–152 158–159 162 166 170–171 173 181 186 189–190 192–193 198 203 209–210 213 222–224 226–227 232 80 100 171 series set point set point element 146 shrink 72 simple lags 41 simulation of tuning response 211 simulation response 211 single-mode controller 18 slave 67 slave controller 18 slow sampling 108–109 Smith Predictor 113–114 119 225 11 14 16 19 21 78 80 124–126 stability This page has been reformatted by Knovel to provide easier navigation Index Terms Links stability (Cont.) 165 188 192 205 static compensation static compensator 163–164 187 152 static friction 80 steady-state 52 55 71 78 158 160 165 172 174–175 177–178 228 158 178 180–181 148 160 steady-state gain 195 steam heater 115–116 step size 204 step test 33 35–36 39–41 48 53–54 57 68 103 106 131–132 202 204 207 210 212 19 25 35 70 72 87 90–91 108–110 117 123–124 129 139 160 165 167 223 219 stiction 206 subcritical 46 swell 72 T temperature control temperature-to-flow control scheme three-mode controller 128 18 This page has been reformatted by Knovel to provide easier navigation Index Terms tight control time constant time delay Links 68 81 83 87 92 206 33–34 37–46 48 50 53–55 57–60 62–64 68 74 76 78–80 82 85–86 89 91 93 97 103–106 108 111–113 115 119 128–129 132 135–136 144 151 153 156 189 191 194 207 219–223 227–228 232 46–47 91 145 40 46–48 228 transducer transfer function 113 transportation lag 36 163 tuning tuning for robustness 211 tuning methods 210 tuning parameter 2–3 12–14 17 24–29 61–63 67–68 76 104 107 109 114 116 119 126 193 198 206–207 209 212 222 two-mode controller two-point method 18 181 This page has been reformatted by Knovel to provide easier navigation Index Terms Links U ultimate gain ultimate period uncontrollability uncontrollable process 21 24–25 27 29 34 58 76 103 21 23 25 27 31 76 103 221 59–62 67–69 76 91 104 111 144 221–222 64 67 69 19–21 41 53 86 188 203 74 unstable 232 V vacuum 130 226 valve characteristics 50 conductance 50 gain 50 hysteresis 69 valve position control 112 velocity 139 80–81 W windup 51 57 69–71 74 76 121 128 130 136 138–141 222 226 This page has been reformatted by Knovel to provide easier navigation Index Terms Links Z Ziegler and Nichols 23–24 38 74 This page has been reformatted by Knovel to provide easier navigation 58 ... must be the controller action, direct or reverse? The Feedback Controller The basic concept of feedback control was introduced in the preceding chapter This chapter presents details of the feedback... relationship between the controlled variable and a disturbance can also be determined, provided that the disturbance variable can be changed and measured This chapter considers only the PV/OP variable pair,... step tests, and how to determine the process gain, time constant, and dead time from the results of the step tests These parameters of a simple-lag-plus-dead-time (SLPDT) model will be used to