Advanced Process Control E-Book

Cecil L. Smith

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 651

Veröffentlichungsjahr: 2011

Beschreibung

This book fills the gap between basic control configurations (Practical Process Control) and model predictive control (MPC). For those loops whose performance has a direct impact on plant economics or product quality, going beyond simple feedback or cascade can improve control performance, or specifically, reduce the variance about the target. However, the effort required to implement such control technology must be offset by increased economic returns from production operations. The economic aspects of the application of the various advanced control technologies are stressed throughout the book.

Leseprobe

Table of Contents

Title Page

Copyright

Preface

Chapter 1: Introduction

1.1 Implementing Control Logic

1.2 Control Blocks for Process Control

1.3 PID Controller

1.4 Integrator or Totalizer

1.5 Lead-Lag Element

1.6 Dead Time

1.7 Selector Block

1.8 Cutoff Block

1.9 Hand Station

Chapter 2: Cascade Control

2.1 Jacketed Reactor

2.2 Block Diagrams

2.3 Problem Element

2.4 Cooling Media Disturbances

2.5 Effect of Varying Heat Transfer Rate

2.6 Cascade Control Modes

2.7 Remote Set Point

2.8 Output Tracking

2.9 Control Modes

2.10 Interacting Stages

2.11 Tuning Cascades

2.12 Windup in Cascade Controls

2.13 Integral Tracking

2.14 External Reset

2.15 Inhibit Increase/Inhibit Decrease

Chapter 3: Split-Range Control

3.1 Storage Tank Pressure Control

3.2 Split Range

3.3 Temperature Control Using Liquid Bypass

3.4 Recirculating Jacket with Heat and Cool Modes

Chapter 4: Override Control

4.1 Limit on the Cooling Water Return Temperature

4.2 Example without Windup Protection

4.3 Integral Tracking

4.4 External Reset

4.5 Inhibit Increase/Inhibit Decrease

4.6 Limits on Heat Transfer

4.7 Other Examples

Chapter 5: Valve Position Control

5.1 Polymer Pumping Example

5.2 Terminal Reheat Systems

5.3 Equilibrium Reaction

5.4 Reactor with a Once-Through Jacket

Chapter 6: Ratio and Feedforward Control

6.1 Simple Ratios

6.2 Ratio Control in Digital Systems

6.3 Feedback Trim

6.4 Dynamic Compensation

6.5 Ratio Plus Bias

6.6 Characterization Function

6.7 Cross-Limiting

6.8 Directional Lags

6.9 Feedforward Control

6.10 Feedforward Control Example

Chapter 7: Loop Interaction

7.1 Multivariable Processes

7.2 Issues with the P&I Diagram

7.3 Steady-State Sensitivities or Gains

7.4 Quantitative Measures of Interaction

7.5 Loop Pairing

7.6 Starch Pumping System

7.7 Reducing the Degree of Interaction

Chapter 8: Multivariable Control

8.1 Decoupler

8.2 Dead-Time Compensation

8.3 Model Predictive Control

Index

Copyright © 2010 by John Wiley & Sons, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone]. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, [phone], fax [phone], or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at 877- 762-2974, outside the United States at [phone] or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Smith, Cecil L.

p. cm.

Includes bibliographical references and index.

ISBN 978-0-470-38197-7 (cloth)

1. Chemical process control. I. Title.

TP155.75.s583 2010

660′.2815–dc22

2009045870

Preface

Exactly what is advanced process control? My favorite definition is from an attendee to a continuing education course: Advanced control is what we should be applying in our plants but are not applying, for whatever reason. This definition lacks specificity, but it does reflect the reality that what seems advanced to some does not seem advanced to others.

To be categorized as advanced, a control configuration must have at least one of the following attributes:

It relies on more than one measurement.It drives more than one final control element.It incorporates a process relationship of some form (which may be as simple as a characterization function).It incorporates functions such as constraint control that are intended to optimize process operations.It addresses interaction between process variables.It is beyond the capabilities of a technician (or at least all but the best of them).

One possible definition is anything other than simple feedback control, which is understood to be a configuration consisting of three elements:

A final control elementA PID controller that generates the output to the final control elementA measurement device that provides the process variable input to the PID controller

If simple feedback control provides the required performance, it should definitely be used. Going beyond simple feedback control always incurs costs that must be justified by the returns from the improved performance. Advanced control should be pursued only when the improved performance translates into enhanced process performance.

Cascade is a good example of the difficulty of defining advanced process control. To most, a level-to-flow cascade is only slightly above simple feedback control on the scale of sophistication. Few would consider these to be advanced control. But consider a temperature-to-temperature cascade applied to a process consisting of interacting stages (as are most temperature processes). Most find these quite challenging and beyond the capabilities of all but the most experienced technicians. Given the importance of temperatures to process operations, arguments can be made to include such cascades in the advanced control category.

The term advanced control is sometimes used to refer to some form of model predictive control (MPC) technology. Model predictive control is definitely advanced control; however, other control technologies deserve to be included in the advanced control category.

The focus of this book is process control, not process safety. Process control must operate the process in the most effective manner, which often leads to considerable complexity. Process safety must avoid unsafe process operating conditions, usually by initiating a shutdown or trip. Although these two are largely separate issues, one requirement must be imposed on the process controls: The process controls must not take any action that would necessitate a reaction from the safety system. Such trips are unnecessary trips and must not happen.

In the process industries the P&I diagram is used almost universally to present the control configuration. This representation encompasses all normal control functions. But for smooth operations, the following requirements must be addressed:

Bumpless transfer

For control configurations that generate multiple outputs, an “all or none” option is not acceptable. The operators must be able to assume control of an individual output at any time. This must not in any way disrupt the other functions being provided by the control configuration. When the manually controlled output is returned to automatic control, there must be no abrupt change in the value of the output (or in any other output from the controls).

Windup protection

When the output of a PID controller ceases to affect its measured variable, the reset mode will drive the controller output to a limit. This is windup. Subsequently, the controller must “unwind,” and this is where the consequences appear. A common cause of windup is when a limiting condition has been attained. Limits apply to all process control applications, the simplest manifestation being a fully open or fully closed valve. However, limits can arise within the process, a common example being heat transfer limiting conditions.

Addressing these issues is often as challenging as developing the configuration for the normal control functions. This book gives such topics appropriate attention.

What if these issues are ignored? Consequences that surface during periods of normal control activities are usually considered to be nuisances that the operators can easily handle (we say that the control configuration has some “warts”). Unfortunately, consequences are most likely to appear during process upsets when the operators are very busy. What would otherwise be a nuisance becomes a distraction that takes the operator's attention away from more pressing matters. Given the “right sizing” of operations staffs, such distractions become serious matters.

This is one aspect that commercial model predictive control packages generally address quite well. Most permit operators to assume control of any output without disrupting the remaining functions. Limiting conditions can be imposed on the outputs, on dependent variables, and so on. That such factors have received appropriate attention has certainly contributed to the success of these packages.

This book also reflects the “You have to understand the process” philosophy that dates from my early years in this business. Process control is appropriately a part of chemical engineering, and those with a process background have made important contributions to the advancement of process control. Even though model predictive control relies on certain principles of linear systems theory, those who pioneered the initial applications were firmly rooted in the process technology.

I am a firm proponent of the time domain. Absolutely no background in Laplace transforms is required to understand the presentations in this book. The word “Laplace” is not mentioned outside this preface, and the Laplace transform variable s is not used anywhere. I firmly believe that Laplace transforms should not be taught in a process control course that is part of the undergraduate chemical engineering curriculum.

Cecil L. Smith

Taos, NM

September 2009

The logic statements approach is used herein. The inputs and outputs for the control blocks as used in this book will be explained shortly. This enables the logic statements for the level controller (LC) to be read as follows:

The value for output tracking (input MNI) is the set point of the flow controller.

Output tracking is active when the flow controller is not on remote.

The value for integral tracking (input MRI) is the current value of the flow.

Integral tracking is active when the output of the flow controller is at its upper output limit (the flow control valve is fully open).

Composing the statements for the inputs depends on the nature of the input: