Understanding Process Dynamics and Control 1st Edition Costas Kravaris Download PDF
Understanding Process Dynamics and Control 1st Edition Costas Kravaris Download PDF
Understanding Process Dynamics and Control 1st Edition Costas Kravaris Download PDF
com
https://ebookmeta.com/product/understanding-
process-dynamics-and-control-1st-edition-costas-
kravaris/
OR CLICK BUTTON
DOWLOAD EBOOK
https://ebookmeta.com/product/advanced-chemical-process-
control-1st-edition-morten-hovd/
https://ebookmeta.com/product/basic-process-engineering-control-
paul-serban-agachi/
https://ebookmeta.com/product/process-operational-safety-and-
cybersecurity-a-feedback-control-approach-advances-in-industrial-
control-zhe-wu/
https://ebookmeta.com/product/theory-of-applied-robotics-
kinematics-dynamics-and-control-reza-n-jazar/
Dynamics and Advanced Motion Control of Off Road UGVs
1st Edition Yue Ma
https://ebookmeta.com/product/dynamics-and-advanced-motion-
control-of-off-road-ugvs-1st-edition-yue-ma/
https://ebookmeta.com/product/process-control-modeling-design-
and-simulation-2nd-edition-b-wayne-bequette/
https://ebookmeta.com/product/advanced-dynamics-modeling-duality-
and-control-of-robotic-systems-1st-edition-edward-y-l-gu/
https://ebookmeta.com/product/flight-dynamics-simulation-and-
control-for-rigid-and-flexible-aircraft-2nd-edition-ranjan-vepa/
https://ebookmeta.com/product/theory-of-applied-robotics-
kinematics-dynamics-and-control-3rd-edition-reza-n-jazar/
Understanding
Process Dynamics
and Control
Costas Kravaris
Ioannis K. Kookos
Understanding Process Dynamics and Control
Understanding Process Dynamics and Control presents a fresh look at process con-
trol, with a state-space approach presented in parallel with the traditional approach
to explain the strategies used in industry today.
Modern time-domain and traditional transform-domain methods are integrated
throughout and the advantages and limitations of each approach are explained; the
fundamental theoretical concepts and methods of process control are applied to
practical problems.
To ensure understanding of the mathematical calculations involved, MATLAB
is included for numeric calculations and Maple for symbolic calculations, with the
math behind every method carefully explained so that students develop a clear
understanding of how and why the software tools work.
Written for a one-semester course with optional advanced-level material, features
include solved examples, cases including a variety of chemical process examples,
chapter summaries, key terms and concepts, as well as over 240 end-of-chapter prob-
lems, including focused computational exercises.
Series Editor
Arvind Varma, Purdue University
Editorial Board
Juan de Pablo, University of Chicago
Michael Doherty, University of California-Santa Barbara
Ignacio Grossman, Carnegie Mellon University
Jim Yang Lee, National University of Singapore
Antonios Mikos, Rice University
Costas Kravaris
Texas A & M University
Ioannis K. Kookos
University of Patras, Greece
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India
79 Anson Road, #06–04/06, Singapore 079906
www.cambridge.org
Information on this title: www.cambridge.org/9781107035584
DOI: 10.1017/9781139565080
© Costas Kravaris and Ioannis K. Kookos 2021
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2021
Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
ISBN 978-1-107-03558-4 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
Dedicated to our families,
and
1 INTRODUCTION 1
Study Objectives 1
1.1 What is Process Control? 1
1.2 Feedback Control System: Key Ideas, Concepts and Terminology 2
1.3 Process Control Notation and Control Loop Representation 8
1.4 Understanding Process Dynamics is a Prerequisite for Learning
Process Control 9
1.5 Some Historical Notes 11
Learning Summary 15
Terms and Concepts 15
Further Reading 16
Problems 17
3 FIRST-ORDER SYSTEMS 55
Study Objectives 55
3.1 Examples of First-Order Systems 55
3.2 Deviation Variables 58
3.3 Solution of Linear First-Order Differential Equations with Constant
Coefficients 59
ix
x Contents
17 BODE AND NYQUIST STABILITY CRITERIA – GAIN AND PHASE MARGINS 575
Study Objectives 575
17.1 Introduction 575
17.2 The Bode Stability Criterion 576
17.3 The Nyquist Stability Criterion 594
17.4 Example Applications of the Nyquist Criterion 597
17.5 Software Tools 604
Learning Summary 607
Terms and Concepts 607
Further Reading 608
Problems 608
Index 760
Preface
xvii
xviii Preface
invaluable educational tool for the student to understand connections of dynamic systems,
including the feedback loop, and also to calculate and appreciate frequency response charac-
teristics. For this reason, Laplace-domain methods are used in this book, and they are used
in parallel with state-space methods. Whenever a quick manual calculation is feasible, the
student should be able to go to the “Laplace planet” and come back, whenever calculations
are very involved or simulation is needed, the student should be able to handle it computa-
tionally using software.
This book offers a strong state-space component, both conceptually and computation-
ally, and this is blended with the traditional analytical framework, in order to maximize
the students’ understanding. But there is also an additional advantage. Because of its state-
space component, this book brings the process dynamics and control course closer to other
chemical engineering courses, such as the chemical reactor course. A chemical reactor course
introduces local asymptotic stability in a state-space setting and tests it through eigenval-
ues, whereas a traditional control course defines stability in an input–output sense and tests
it through the poles of the transfer function. This gap is nonexistent in the present book:
asymptotic stability is defined and explained in a state-space context, input–output stability
is defined and explained in a transfer function or convolution integral context, and the rela-
tionship of the two notions of stability is discussed. Moreover, there are a number of chem-
ical reactor examples throughout the book that link the two courses in a synergistic manner.
A final comment should be made about the word “understanding” in the title of this book.
It is our firm belief that engineers must have a thorough understanding of how their tools
work, when do they work and why they work. If they treat a software package as a magic
black box, without understanding what’s inside the box, they have not learned anything. For
this reason, special care is taken in this book to explain the math that is behind every method
presented, so that the student develops a clear understanding of how, when and why.
the second part of Chapter 9 on Nyquist diagrams, which is only needed in the second part
of Chapter 17.
The rest of the chapters are on process control. Chapters 10–14 cover the basic feedback
control concepts and analysis methods. Chapter 10 gives a general introduction to feedback
control, and also defines the PID controller in both state-space and transfer function form.
Chapters 11 and 12 do transfer function analysis of the feedback control loop, whereas in
Chapter 13 the same analysis is done in state space. Chapter 14 discusses systems with dead-
time, both open loop dynamics and feedback control. Deadtime is treated separately because
of its distinct mathematical characteristics. Chapters 10–14 provide an absolute minimum
for the feedback control part of the course. From that point, the instructor can choose what
design methods he/she wants to put emphasis on, root locus (Chapter 15), optimization
(Chapter 16), gain and phase margins (Chapter 17) or model-based (Chapter 19). Also, the
instructor has the choice to discuss issues in multivariable control (Chapter 18) or stay SISO
throughout the course. The last chapter (Chapter 20) discusses cascade, ratio and feedfor-
ward control. These control structures are discussed first at a conceptual level, and then
model-based design for cascade and feedforward control is derived. The conceptual part is,
in a sense, a continuation of Chapter 10 and it is essential to be taught; the model-based part
is a continuation of Chapter 19.
The last section of each chapter is about software tools. The use of software for the appli-
cation of the theory of the chapter is explained through simple examples. Two alternative
software packages are used: MATLAB and its control systems toolbox is chosen because of
its strength in numerical calculations, and Maple and its libraries (LinearAlgebra, inttrans,
etc.) because of its strength in symbolic calculations.
The following table gives a sample syllabus for the process dynamics and control course at
Texas A&M University, as it has been taught in the past three semesters. It reflects the per-
sonal choices of the instructor on (i) the design methods for the control part of the course
(optimization and model-based are emphasized) and (ii) the pace of covering the material
(slower at the beginning, faster at the end). Of course, there are many other options, depend-
ing on instructor priorities and students’ background.
STUDY OBJECTIVES
After studying this chapter, you should be able to do the following.
1
2 Introduction
The amount of knowledge that has been generated in the past 100 years in the field of
automatic control is vast. Using this knowledge to design and operate control systems in
practice is vital for maintaining the same pace of development in the years to come. This
book aims at explaining the fundamental principles of process control in a way that makes
it easier for future chemical engineers to comprehend past developments and to develop new
tools that advance engineering practice. We also hope that learning process control methods
and concepts will help future chemical engineers to interact and collaborate with control
engineers of other disciplines.
High
Set Point
Low
• What is the best sequence of actions that the operator needs to take when the liquid
level deviates from the set point?
• Do we need a human operator to control the level? Why not build an “automated
system”?
• Would an “automated system” be more efficient and reliable than the human in con-
trolling the level of the liquid?
• Are there any other measurements that could be used, in combination with the level of the
liquid in the tank, in order to more effectively maintain the liquid level at the desired value?
Some of these types of questions have an easy answer, but some will need further thinking
and elaboration throughout your process control course and even throughout your profes-
sional career.
Most undergraduate courses in chemical engineering consider processes that operate at
steady state. This is a logical and well-documented simplification that allows chemical engi-
neers to design fairly complex processes in a reasonable amount of time. However, in actu-
ality, processes operate in a dynamic environment. Imagine that you design a heat exchanger
that uses sea water as cooling water to cool down a process stream from 100 °C to 50 °C.
At the design stage you have to make an assumption about the temperature of the cooling
water (a unique value) and suppose that you have selected a temperature of 20 °C. Now think
about the chances of the cooling water temperature being exactly 20 °C. Will the system fail
if the actual temperature of the water is 15 °C or 10 °C? The answer is yes, the system will
fail to keep the temperature at the desired value of 50 °C, unless a valve is installed, which
can appropriately adjust the flowrate of the cooling water. In addition, a temperature sen-
sor needs to be installed, to measure the temperature of the process stream exiting the heat
exchanger. Then, using the measured and recorded temperature, an operator can check if
the temperature is at the proper value, and appropriately adjust the cooling water flowrate
to correct any discrepancies, as shown in Figure 1.2. The sea-water temperature can vary
Sensor
Controller
Process
Final
control
element
throughout the day, so the operator will need to perform frequent adjustment of the valve
opening to keep the temperature of the process stream close to the desired temperature. In
addition, the operator can implement changes in the desired temperature, if there are reasons
related to the operation of downstream processes.
The basic elements of the temperature-control system shown in Figure 1.2 are also shown
in the block diagram of Figure 1.3. The blocks are used as a means of representing the com-
ponents of the system and the arrows denote a signal or information flow. The measurement
(the line exiting the sensor and entering the controller) is not, in the case of the control
system of Figure 1.2, an actual signal but an information flow and denotes the reading of
the temperature indication by the operator. The operator/controller is a necessary element
of the loop that processes (using a control algorithm) the information and decides on the
appropriate action to be taken (opening or closing of the valve). The opening or closing of
the valve determines the flowrate of the cooling medium (sea water) and thus the rate of heat
transfer in the heat exchanger. Finally, the temperature of the product stream is measured
by the sensor (operator’s eyes) and the loop is closed. In most cases, the controller is a com-
puter-based system that receives a signal from the sensor, executes the control algorithm and
sends a signal that sets the valve position, as indicated in Figure 1.4. Computers can perform
very complex calculations in a very short time, can handle more than one control system
simultaneously and work continuously and, in most cases, without human intervention.
Controller
Set
Sensor
point
Process
Final
control
element
Let’s now summarize the types of process variables encountered in the primitive control
systems that we have seen so far.
Disturbance variables. Any process is affected by several external influences, and many of
them vary in an uncontrollable and unpredictable manner. These are the disturbances that
cause the operation of the process to deviate from the desired steady state. In the case of the
heat exchanger, potential disturbances are the sea-water temperature, the temperature and
the flowrate of the incoming process stream, or equipment aging known as fouling (which
increases the resistance to heat transfer). Some of the disturbances could be measured in real
time but others are difficult, expensive or even impossible to measure.
Manipulated variables. The manipulated variables are those process variables that are
adjusted by the controller in order to achieve the control objectives. A manipulated variable
is also called control input, to signify that it represents the control action that “feeds” the
process. The most frequent manipulated variable in the chemical process industries is the
flowrate through the installation of a control valve or a pump.
Measured variables. Measured variables are all the variables for which we have installed a
sensor or measuring device that continuously measures and transmits the current value of
the variable. Of course, sensors cost money and need frequent maintenance, therefore their
installation should be well justified. The most common measured variables in the chemical
industries are temperature, pressure, flow and level. Others, such as composition, are more
costly and less frequently used.
A measured variable that the controller is maintaining at a particular desired value is
called a controlled variable. The desired value is called the set point of the controlled variable.
The set point is usually kept constant for a long time, but sometimes a need may arise to
change the set point, and this should be handled by the controller.
When the value of the controlled variable agrees with its set point, it is “in control,” oth-
erwise there is an error. The error is defined as the difference between the set point and the
value of the controlled variable, and the job of the controller is to make it equal to zero.
Figure 1.5 depicts a generic feedback control system. It shows all the types of variables
that come into play, as well as the basic elements of the control system and how they are
connected with each other.
The final control element (usually a control valve) together with the process and the
sensor comprise the physical system or open-loop system. We see from Figure 1.5 that, when
the sensor is connected to the controller, and the controller is acting on the final control
element, the overall system has a circular structure, like a ring or a loop, and it is called
the closed-loop system. It is also called a feedback control system. The idea of feedback
control involves continuous monitoring of the controlled variable and “feeding back”
the information, to make changes and adjustments in the process, through changes in the
manipulated variable. The controller’s action is usually based on the error, i.e. the discrep-
ancy between the set point and the measurement of the controlled output. Depending on
the error (its current value, its history and its trend), the controller takes corrective action.
In simple terms, one can describe the operation of a feedback control system as: monitor,
detect and correct.
6 Introduction
Disturbance
variable
Set Measured/
Manipulated
point controlled
Error variable
variable
ysp e=ysp – ym u Final y
Controller control Process
+ element
–
ym
Sensor
Measurement
Closed-loop system
Figure 1.5 Basic elements of a feedback control system and their interconnections.
Sensors play a critical role is the proper operation of a feedback control system. Sensors
use an electrical or mechanical phenomenon in order to determine the temperature, pres-
sure, level or flowrate. Temperature sensors are based on the expansion of a liquid or gas
(thermometers), on the Seebeck or thermoelectric effect, the creation of voltage between two
junctions at different temperatures (thermocouples), the variation of electrical resistance of
several materials with temperature (resistance temperature detectors and thermistors) and
the thermal radiation emitted (pyrometers). Most pressure sensors are based on measuring
the deflection or strain caused by the pressure when applied to an area (strain-gauge, elec-
tromagnetic, piezoelectric, etc.). Pressure sensors are used in conjunction with an orifice or
a Venturi tube to measure flow, as differential pressure across the orifice or between two seg-
ments of a Venturi tube (with different aperture) is strongly related to flow. Pressure sensors
are also used to calculate the level of a liquid in a tank as the pressure difference between the
top and the bottom of a tank is directly proportional to the height of the liquid. The trans-
mitter is used to convert the primary measurement by the sensor to a pneumatic or electrical
signal. The combination of the sensor and the transmitter is called a transducer.
In a chemical plant, there may be hundreds or thousands of feedback control loops like
the one depicted in Figure 1.5. The need to transmit all information and functionality to a
central “control room” (see Figure 1.6) to achieve continuous monitoring and reduce dras-
tically the manpower required was quickly identified and implemented in the 1960s. This
centralization was really effective in improving the operation of the plant. At that time, the
controllers (one controller for each control loop) were behind the control room panels, and
all control signals were transmitted back to plant. Gradually the structure was modified as
all functionalities were assigned to a network of input/output racks with their own control
processors which could be distributed locally in the plant (and could communicate with the
1.2 Feedback Control System: Key Ideas, Concepts and Terminology 7
Figure 1.6 A control room of the 1960s and a more recent DCS control room (from
Wikimedia Commons, the free media repository).
control room). The distributed control system or DCS was thus born, in which the control-
lers were placed close to the processing units but transmitted all information to a central
location through a central network to minimize cabling runs. Monitoring, interconnection,
reconfiguration and expansion of plant controls were finally easy. Local control algorithms
could be executed by the central units in the case of system failure and thus reliability was
greatly enhanced. Recent advantages such as wireless technology and Internet of Things as
well as mobile interfaces might have a real impact in the near future.
8 Introduction
Temperature Temperature
Controller Transducer
TC TT
Electric signal
Process stream
that is flow-controlled, is cooled using a heat exchanger and cooling water. The temperature
of the outgoing process steam is measured by a temperature sensor and the measurement is
indicated by the temperature transducer (TT) symbol in Figure 1.7. Bubble or circle symbols
are used to indicate instrumentation (measurement) or control function. Inside the circle
symbol a two-letter coding system is used to denote the specific functionality of the block.
The first letter in the two-letter naming system refers to the variable controlled or measured
and the letters commonly used are the following.
T: temperature
F: flow, flowrate
L: level
P: pressure
C (or A): composition
The second letter indicates whether this is a measuring device or transducer (T) or a
control device (C). TC is therefore used to indicate a temperature controller while FC indi-
cates a flow controller in Figure 1.7. The two circle symbols denoted as TT (temperature
transducer) and TC (temperature controller) are connected through a dashed line (– – – –)
which indicates an electrical signal (4–20 mA, 1–5 V or 0–10 V). Other common conven-
tions are the following: a pneumatic signal is denoted by —//—//— (normally in the range
of 3–15 psig) and a data-transfer signal is denoted by —o—o— (usually binary signal).
We will not try to indicate explicitly whether a signal is electrical or pneumatic, as it adds a
complexity that is unnecessary within this book, and we will be using a dashed line to indi-
cate exchange of information between a sensor, a controller and a final control element, as
shown in Figure 1.7.
In the first part of this book we will study the dynamics of an isolated system, to try to
understand its transient behavior. We will see different kinds of transient behavior, and we
will explain the behavior and characterize it in a systematic way. We will see how to calcu-
late these transient responses, analytically and numerically. One of the key concepts that we
will discuss is the concept of stability, and we will derive tests to determine if a system has
stable behavior. We will also introduce the necessary software tools to calculate routinely the
dynamic response of common process systems.
Equipped with these concepts and tools, we will study interconnected dynamic systems,
in a feedback control loop. We will see how the dynamic behavior of all the elements of the
loop can be combined, and we will derive the dynamic behavior of the overall system, and
calculate its transient response.
Typical process systems’ dynamic responses are presented in Figure 1.8. The response
can be fast or relatively slow as shown in Figures 1.8a and b, respectively. A characteristic
commonly encountered in process systems is that of delayed response as shown in Figure
1.8c. These three general responses are the ones usually obtained by chemical processes like
distillation and absorption columns, evaporators and heat-transfer equipment. There are
0
0 t
(c)
y(t)
0
0 t
(d)
y(t)
0
0 t
(e)
y(t)
0
0 t
1.5 Some Historical Notes 11
processes, like chemical reactors, that can exhibit oscillatory (Figure 1.8d) or even unstable
(Figure 1.8e) response.
Central shaft
(connected to the Steam to engine
engine)
Figure 1.9 Basic operation of the flyball governor (engine speed control system).
12 Introduction
slow progress took place in the field in the nineteenth century. Things, however, made a dra-
matic change in the beginning of the twentieth century.
Things were very different in the chemical industry just 100 years ago. We present the fol-
lowing extract from one of the most significant journals in chemical engineering (the Journal
of Industrial & Engineering Chemistry, published by the American Chemical Society or ACS)
that was published in 1918 (I&ECR, p. 133, February 1918, by R. P. Brown from The Brown
Instrument Company, Philadelphia):
Probably no employee causes the average works manager so many sleepless nights as does the
furnace man, on whose shoulders rests the responsibility for the accurate heat treatment of the
steel and the uniformity of the product. This is not only true of a steel plant, but is also equally
true in the chemical industry, where the temperature of numerous processes must be accurately
controlled; … The old furnace man, through years of practice, will endeavor to gauge the tem-
perature of the furnace with his eye. Providing he has not been up all the previous night and his
eye is clear, he will probably judge the temperature fairly accurately. …
But we can pardon the works manager or director for asking, “Suppose John dies, gets sick or
quits his job, how am I to handle the output of these furnaces?” He would like to have an under-
study for the old furnace man, but the latter does not like the idea. So he wonders why someone
does not develop a device to automatically control the temperature of the furnaces, so that he
can cease worrying about them.
This is one reason why a great amount of study has been given, not only to perfection of
pyrometers, but also to the automatic control of temperature. It has, however, been only recently
that real results have been accomplished in automatic temperature control.
Some primitive control systems are described only a few years later. One of these is shown
in Figure 1.10. The so-called vapor-tension system consisted of a metallic bulb partly filled
with a liquid, located in the container under
Valve control, and a metallic capillary leading from
the bulb to an expansion or capsular chamber
Heating in the regulator case. Any change in the tem-
fluid Heating perature of the space surrounding the metallic
in Plug fluid
out
bulb will cause a change in the vapor pressure
Stem of the enclosed liquid and therefore in the pres-
Diaphragm sure below the diaphragm. As the pressure in
Capillary tube the capillary tube and diaphragm changes, this
causes the stem and the plug to move up (when
the temperature increases) or down (when the
temperature decreases). If the plug is used in a
valve to restrict or release the opening through
which a thermal fluid is supplied to the system then a rudimentary control action can be
achieved. When the temperature increases the vapor pressure of the enclosed liquid also
increases, raising the stem and the plug and blocking the flow of the heating fluid. This
causes less heating fluid to be supplied to the system and the temperature drops. This is a
continuous control system that was chosen here purposefully despite the fact that most early
control systems were simple on–off devices.
A number of interesting descriptions of early, basic control systems appeared in the lit-
erature in the late 1920s (see, for instance, I. Ginsberg, Automatic control in the chemical
industries, I&ECR, 21(5), 410, 1929). Real progress was achieved in the 1930s and mainly in
the early 1940s with the publication of the seminal papers by Ziegler and Nichols (Ziegler,
J. G. and Nichols, N. B., Trans. Am. Soc. Mech. Eng., 64, 759–68, 1942, and 65, 433–44,
1943). These papers presented a systematic method for selecting the parameters of process
controllers with minimal process upsetting and were based solely on the response of the
open-loop system to a step change in the process input. The controller parameters were then
determined as a function of the characteristics of the step response. The method proved
to be particularly successful and gained widespread acceptance from the community. The
Ziegler–Nichols method is still selected by many researchers, 80 years later, as a basis for
establishing the advantages of their controller tuning techniques.
The numerous military applications of automatic control, developed and perfected dur-
ing the Second World War, created a strong interest in the chemical process industries. Just
before the beginning of the Second World War, a significant change started to materialize in
the chemical plants as most of the large enterprises, such as E. I. du Pont de Nemours and
Company, realized the benefits of continuous operation as compared with batch operation.
However, it all had to wait until the introduction of microcomputers in the late 1950s and
it was during the 1960s that the process control became an integral part of the operation of
the chemical plants.
In the 1950s, process control theory witnessed significant progress as particular atten-
tion was placed in the asymptotic response of process systems to sinusoidal input vari-
ation. This analysis technique, known as frequency response, was particularly mature at
that time in other areas such as telecommunication systems. At this time the first book in
process control was published by N. Ceaglske (Automatic Process Control for Chemical
Engineers, New York: Wiley, 1956) from the Department of Chemical Engineering,
University of Minnesota, Minneapolis. Significant progress was also reported in using
electrical analogs to simulate the response of common process units such as distilla-
tion columns and heat exchangers (see, for instance, work reported in I&ECR, p. 1035,
June 1956, by J. M. Mozley from E. I. du Pont de Nemours and Company). Based on a
technique that used electrical components to simulate the response of process systems
using electrical analog circuits, this work offered the opportunity to simulate relatively
complex dynamics efficiently when computers were expensive and relatively awkward to
use (judged from today’s point of view). The side effect of this development was that
time-domain analysis techniques were studied by chemical engineers in parallel with fre-
quency-response techniques.
14 Introduction
The theory of chemical process control reached maturity in the 1960s and early 1970s with
numerous applications and theoretical developments and the publication of several books
that facilitated wider understanding of the field:
• P. S. Buckley, Techniques of Process Control, New York: John Wiley & Sons, 1964
• P. Harriott, Process Control, New York: McGraw-Hill, 1964
• D. R. Coughanowr and L. B. Koppel, Process Systems Analysis and Control, New York:
McGraw-Hill Company, 1965
• P. W. Murrill, Automatic Control of Processes, Scranton, PA: Intext Educational
Publishers, 1967
• E. F. Johnson, Automatic Process Control, New York: McGraw Hill Company, 1967
• J. M. Douglas, Process Dynamics and Control, volumes 1 & 2, Englewood Cliffs, NJ:
Prentice-Hall Inc., 1972
• W. L. Luyben, Process Modeling Simulation and Control for Chemical Engineers, New
York: McGraw-Hill Company, 1973.
All these books are mainly concentrated around what is now called classical automatic
control theory, which is mainly characterized by the use of Laplace and frequency-domain
analysis of control systems. These particular methods of control systems analysis and design
were used extensively in electrical engineering and communication systems in the 1940s and
1950s and offered, even in the 1960s, a strong and comprehensive set of tools that could be
used efficiently to analyze and design process control systems. The time-domain approach,
which is based on using sets of first-order differential equations to describe system dynamics
(including feedback control systems), is an alternative method of formulating control sys-
tems theory, which has become known as the modern or state-space approach. Both classical
and modern approaches will be presented in this book in a parallel way and the reader will
soon realize that both approaches have their advantages and disadvantages. In any case, a
process control engineer must be able to understand and use system analysis and controller
synthesis tools from both approaches.
In the past 40 years, process control theory is considered a well-established field of
study and research in chemical engineering and has expanded its areas of application.
Automatic process control systems are in extensive use in the chemical process industries.
Several books were published in the 1980s and 1990s (a list of some of the most influen-
tial textbooks is presented at the end of the chapter) and the field can now be considered
as a fairly mature field. In addition, computer software tools that are now available make
application of the theory for controller synthesis, closed-loop system simulation and con-
troller prototyping an easy task. Technologies such as digitalization and the Internet of
Things (IoT) are expected to have a real impact in the years to come. Nonlinear con-
trol, model-predictive control and real-time optimization are now routinely installed in
advanced control systems. These developments have shifted the focus and the educational
needs in the field of process control. Future chemical engineers must have a solid under-
standing of the underlying theory but more importantly they must be able to combine this
Terms and concepts 15
knowledge with software tools in order effectively to design controllers for large-scale,
interacting and integrated plants.
LEARNING SUMMARY
Automatic control is an interdisciplinary field of study with many common elements between
different engineering disciplines such as electrical, mechanical and chemical engineering.
Chemical process control is a well-established field in chemical engineering and it involves
the application of automatic control in the chemical, petrochemical, food and related indus-
tries. Automatic operation is achieved mainly through feedback control systems, whose main
constituting elements are shown in Figure 1.3.
The controlled variable is measured continuously through a sensor and is fed back to the
controller, which can be an operator or a computer system. Controller actions are imple-
mented to the process through a final control element, which, in the case of chemical process
industries, is usually a control valve.
Chemical process control has a history of roughly one century but is advancing at a fast
pace by taking advantage of innovations in other fields, mainly related to computer soft-
ware and hardware. Advanced control algorithms are routinely implemented at several levels
of the control system. This book presents both approaches to process control theory: the
classical approach that is based on the frequency response and the Laplace domain, and
the modern approach that is based on the state-space approach in the time domain. It also
introduces basic software tools in commercial computer software necessary for the efficient
implementation of theory into practice.
FURTHER READING
The books that follow are the most frequently used textbooks in chemical process control
and presented according to the date of publication of their most recent edition.
Stephanopoulos, G., Chemical Process Control. Englewood Cliffs, NJ: Prentice Hall, 1984.
Luyben, W. L., Process Modeling Simulation and Control for Chemical Engineers, 2nd edn.
New York: McGraw-Hill, 1990.
Ogunnaike, B. A. and Ray, W. H., Process Dynamics, Modeling and Control. New York:
Oxford, 1994.
Shinskey, F. G., Process Control Systems: Application, Design, and Tuning. New York:
McGraw Hill, 1996.
Luyben, M. L. and Luyben, W. L., Essentials of Process Control. New York: McGraw-Hill, 1997.
Marlin, T. E., Process Control: Designing Processes and Control Systems for Dynamic
Performance, 2nd edn. New York: McGraw Hill, 2000.
Chau, P. C., Process Control, A First Course with MATLAB. Cambridge, UK: Cambridge
University Press, 2002.
Brosilow C. and Joseph, B., Techniques of Model-based Control. Engelwood Cliffs, NJ:
Prentice Hall, 2002.
Bequette, B. W., Process Control: Modeling, Design and Simulation. Engelwood Cliffs, NJ:
Prentice Hall, 2003.
Smith, C. A. and Corripio, A. B., Principles and Practice of Automatic Process Control, 3rd
edn. New York: Wiley, 2005.
Riggs, J. B. and Karim, M. N., Chemical and Bio-Process Control, 3rd edn. Boston, MA:
Pearson International Edition, 2006.
Coughanowr, D. R. and LeBlanc, S., Process Systems Analysis and Control, 3rd edn. New
York: McGraw-Hill Education, 2009.
Svrcek, W. Y., Mahoney, D. P. and Young, B. R., A Real-Time Approach to Process Control,
3rd edn. Chichester, UK: Wiley, 2013.
Seborg, D. E., Edgar, T. F., Mellichamp, D. A. and Doyle III, F. J., Process Dynamics and
Control, 4th edn. New York: Wiley, 2016.
Rohani, S. (ed.), Coulson and Richardson’s Chemical Engineering Volume 3B: Process Control,
4th edn. Cambridge, MA: Butterworth-Heinemann, 2017.
Information about standard symbols and convections used in P&IDs and sensor and actua-
tor technology can be found in the following books.
Meier, F. A. & Meier, C. A., Instrumentation and Control Systems Documentation. Research
Triangle Park, NC: International Society of Automation (ISA), 2011.
Hughes, T. A., Measurement and Control Basics, 5th edn. Research Triangle Park, NC:
International Society of Automation (ISA), 2015.
Dunn, W. C., Fundamentals of Industrial Instrumentation and Control. New York: Mc-Graw
Hill Education, 2018.
Toghraei, M., Piping and Instrumentation Diagram Development. New Jersey: John Wiley &
Sons, Inc., 2019.
Problems 17
PROBLEMS
1.1 Draw a block diagram for the control system generated when you drive an automobile.
What are the “sensors” that you use and what are the final control elements? What are
the disturbances?
1.2 Draw a block diagram for an automobile cruise control system. Select some informa-
tion from the internet about autonomous driving and list the sensors that you think
you need to achieve such a task.
1.3 Draw a block diagram for the control system that maintains the temperature in a home
refrigerator. Identify the potential disturbances.
1.4 Draw a block diagram for the control system for a home air-conditioning system. Iden-
tify the potential disturbances.
1.5 Collect information from the internet concerning the basic principles of temperature
sensors.
1.6 Collect information from the internet concerning the basic principles of pressure and
differential pressure sensors.
1.7 Collect information from the internet concerning the basic principles of flow metering
devices.
1.8 Suppose you were to develop the first sensor for measuring the liquid level in a tank
or the pressure in a closed vessel for gas storage. Think about the general principles
on which we can base the development of such a sensor. Search the internet to collect
information about the most used sensors for liquid-level or pressure measurement.
1.9 Suppose you have been asked to develop a system that follows the Sun and positions
a surface always to face the Sun (so that the Sun beam direction is normal to the cap-
turing surface). Think about the characteristics of the sensor that detects the position
of the Sun and the necessary actuators to move the surface. Search the internet, collect
information about the systems used in solar tracking control systems for photovoltaic
panels and prepare a short presentation.
1.10 Compare Figures 1.9 and 1.10 and discuss the common elements.
2 Dynamic Models for Chemical Process Systems
In this chapter we discuss the main principles of modeling the dynamics of chemical pro-
cesses. We start by stating the general inventory rate equation. The conservation of mass
and energy, which are special applications of the inventory rate equation, are then presented
and a number of representative examples are analyzed. The development of an appropriate
dynamic process model is the first step in understanding the underlying phenomena that
result in the transient behavior of a process system. Modeling will be essential in developing
appropriate control laws and in understanding the effect that a controller has on process
transients.
STUDY OBJECTIVES
After studying this chapter you should be able to:
• Possess basic understanding and skills in dynamic model development.
• Apply the appropriate inventory equations to model the dynamics of representative
chemical processes.
• Apply differential equation solvers of software packages like MATLAB or Maple, to
simulate the dynamic behavior of chemical process systems.
2.1 Introduction
A mathematical model is a representation of our knowledge about a physical system, which
is “translated” into a set of mathematical equations. The aim is to use the model to increase
our understanding of the real system’s behavior (model simulation can provide insights), and
also to design and optimize the operation of the process.
Dynamic models provide a quantitative description of the transient behavior of a process,
in addition to its steady-state characteristics. This is very useful in order to select proper
operating conditions for the process, so that undesirable transients are avoided. It is also
very important for the design of controllers, because, as we will see in subsequent chapters, a
controller can modify the dynamic behavior of a process, for better or for worse.
18
Another random document with
no related content on Scribd:
PIED DUCK.
Ardea virescens, Linn. Syst Nat. vol i. p. 238.—Lath. Ind. Ornith. vol. ii. p.
684.—Ch. Bonaparte, Synopsis of Birds of United States, p. 307.
Green Heron, Ardea virescens, Wils. Amer. Ornith. vol. viii. p. 97, pl. 61,
fig. 1.—Nuttall, Manual, vol. ii. p. 63.
Winter.
Tringa helvetica, Linn. Syst. Nat. vol. i. p. 250.—Lath. Ind. Ornith. vol. ii. p.
728.
Tringa squatarola, Linn. Syst. Nat. vol. i. p. 252.—Lath. Ind. Ornith. vol. ii.
p. 729.
Black-bellied Plover, Charadrius helveticus, Wils. Amer. Ornith. vol. vii.
p. 41, pl. 59, fig. 4. Summer.
Charadrius helveticus, Ch. Bonaparte, Synopsis of Birds of United States,
p. 298.
Grey Lapwing, Vanellus melanogaster, Richards. and Swains. Fauna
Bor.-Amer. vol. ii. p. 370.
Black-bellied or Swiss Plover, Nuttall, Manual, vol. ii. p. 26.
In the stomach were several shrimps. The lobes of the liver very
unequal, the right being 2 1/4 inches in length, the other 1 4/12. No
gall-bladder.
The trachea is wide, flattened, membranous, 4 twelfths broad at the
upper part, gradually diminishing to 2 twelfths, its rings which are
very slender, about 100. The lateral muscles exceedingly thin, but
becoming more distinct towards the lower part; the sterno-tracheal
slender. Bronchi of moderate length, of about 20 half-rings.