Instrumentation System Fundamentals

Instrumentation Systems
-Fundamentals and Applications-
Tasuku Senbon, Futoshi Hanabuchi (Eds.)
Instrum entation
Systems
Fundamentals and Applications
Springer-Verlag
Berlin Heidelberg GmbH
Instrumentation Systems-Fundamentals and Applications.
All rights reserved. No part of this publication may be

reproduced, stored in a retrieval system or transmitted in any
form or by any means, electronic, mechanical, recording or
otherwise, without the prior written permission of the copyright
owner.
First published in Japanese in 1987 by Ohmsha, Ltd. Tokyo as
Shisutemu no Kiso to OuyouJ
1 Keisou
1991 Springer-Verlag Berlin Heidelberg

Originally published by Springer-Verlag Berlin Heidelberg New York in 1991
Exclusive worldwide distribution by:
ISBN 978-3-662-12091-0
ISBN 978-3-662-12089-7 (eBook)
DOI 10.1007/978-3-662-12089-7
PREFACE
This book, though small, contains a wealth of technical information on control engineering and instrumentation engineering for industrial quantities, on control-system component elements (sensing, conversion, control, monitoring, and actuation), and on the system-design
approaches (system engineering) used in process automation (PA) and
factory automation (FA), discussing them based on examples of their
applications, and covering everything from basics to applications.
Process automation has a long history, with automation of individual functions having begun as early as the 1920's. The feedback control
techniques that constitute its basis grew into an indispensable core
technology along with the rapid advance of control theory and control
devices from the 1960's onward. Today we are progressing further towards system-scale optimal control technology. One of the influences
that spurred major innovation along the way was the birth of microprocessor-based digital computer control in the 1970's. This enabled the
realization of batch and sequential control together with feedback control in the same processor thus allowing an intimate interlinkage
among them all. Technology for communication between multiple processors was also introduced, fostering rapid advances in functional
sophistication and installation density. Moreover, this did not stop
with process automation, but also spread to total factory automation
covering entire plants. This included factory automation aimed at discrete processes.
This book begins with a discussion of control theory. It moves on
to discuss the product hardware and software that implement the theory, and then proceeds to describe instrumentation examples and the
system-design approaches (system engineering) suitable for a variety
of production processes. Thus, we believe it to be ideally suited for use
as a college-level textbook on instrumentation and automation for
undergraduate or graduate students, or as a reference book for practicing instrument engineers in industry.
Since the subject matter deals with extremely specialized technoloPreface
gy, the responsibility for the authorship has been undertaken by Yokogawa Electric experts continually involved in these areas. The Yokogawa Electric Training Center has undertaken the task of editing and
compiling these writings into a text.
At the same time that we express our gratitude to the authors of
the many works used for reference, we would also like to offer our deepest thanks to the staff of our publisher, Ohmsha, Ltd., for their hard
work and earnest cooperation. We hope that this book will be of assistance to our readers in their study of instrumentation and control systems.
September, 1987
Hisashi Tamura, Senior Vice President

Director, SBU Administration
Yokogawa Electric Corporation
vi
Preface
ABOUT THE ENGLISH-LANGUAGE EDITION
Since its publication in 1987, the original Japanese-language edition of lnstrumentataion Systems has already gone through several printings. This is due to its wide readership among those responsible for
instrumentation and control in Japan. There is a significant relationship between the expanding number of readers of this book and the
continuing rapid growth of Japan's industry and economy, with process
automation and factory automation as two of its driving forces.
Today as the barriers between East and West crumble away, we
hope that an even wider international dissemination of this book will
lend support to the world's movement toward global industrial and economic development. The authors and editors have felt this to be one of
their missions. A necessary condition has been the creation of this English-language edition.
This opportunity to carry the English-language version to realization with the full cooperation of Ohmsha, Ltd. and Springer-Verlag
has been a source of great pleasure to the authors and editors. We
wish to extend our thanks for the assistance of those who undertook
the translation and editorial supervision.
It is the hope of all those involved that this book will be widely
read and found useful by members of the instrumentation and control
community all over the world.
Akio Yamamoto, General Manager
Yokogawa Electric Training Center
About the English-Language Edition
vii
LIST OF CONTRIBUTORS
EDITORS
Tasuku Senbon
Futoshi Hanabuchi
AUTHORS (alphabetical order)
Naoki Asakawa
Yoshio Fukai
Katsuhiro Hikasa
Kiyokazu Ishii
Tadamichi Kai
Isao Katsuoka
Hiroshi Kawai
Toshio Kimura
Takane Kudo
Hidesada Kurioka
Tetsuro Matsumoto
Kiyoshi Matsunaga
Teruyoshi Minaki
Yoshiaki Murakami
Yoshio Nagasaki
Shinobu N agase
Eizo Oku
Kazuo Omori
Yoshikatsu Sakai
Makoto Sekiya
Jun Shiozawa
Fuso Takamura
Shin-ichi Takigishi
Akira Tanaka
Yukio Tanaka
Katsuaki Tokunaga
Masahito Tsukamoto
Hideo Tsurumaki
Masahiko Ushioda
Sadahito Watanabe
Shigehiko Yamamoto
Shigeru Yamamoto
Michio Yoshioka
Tsuneo Zeniya
EDITORIAL ASSISTANCE
Akio Yamamoto
Sumiaki Nishikata
List of Contributors
!X
CONTENTS
Preface
v
About the English-Language Edition
List of Contributors
ix
vii
Chapter 1 INDUSTRY AND INSTRUMENTATION

1.1 The Word "Instrumentation" ...................................... 1
1. 2 The Development of Instrumentation ............................... 2
1. 3 Trend toward Total FA .............................................. 4
1. 4 Classification and Use of Instruments .............................. 6
References .................................................................. 9
Chapter 2 PROCESS CONTROL
2 .1 Fundamentals of Feedback Control.. .............................. 11
2 .1.1 Configuration of a control system ............................ 11
2 .I. 2 Characteristics of a control system ........................... 13
2 .I. 3 Feedback control and stability ................................ 19
2 . 2 Process Characteristics .............................................. 23
2. 2 .1 Process degrees of freedom and controlled and
manipulated variables ......................................... 23
2 . 2. 2 Process characteristics ......................................... 25
2. 2. 3 Process models .................................................. 26
2. 3 Control Formats for Various Types of Processes ................ 32
2 . 3 .1 Single loop control systems .................................... 32
2. 3 . 2 Compound loop control system ................................ 39
2. 4 Optimal Adjustment of Control Systems ......................... .45
2. 5 Sequential Control ................................................... 50
2.5.1 Meaning of "sequential control" ............................. 51
2. 5. 2 Types of sequential control.. .................................. 51
2. 5. 3 Sequential control description ................................. 52
2. 5. 4 Devices for sequential control ................................. 57
Practice Questions ........................................................ 58
Answers to Questions .................................................... 59
References ................................................................. 59
Contents
xi
Chapter 3 DETECTION AND CONVERSION OF

INDUSTRIAL VARIABLES
3 .1 Measurement of Industrial Variables .............................. 62
3 .1.1 Methods of measurement ....................................... 62
3 .1. 2 Accuracy of measurement ..................................... 66
3.2 Measurement of Temperature ...................................... 71
3. 2 .1 Thermoelectric thermometers ................................. 72
3. 2 . 2 Resistance thermometers ....................................... 84
3.2.3 Protective tube .................................................. 91
3. 2. 4 Thermistor thermometers ...................................... 92
3. 3 Measurement of Flow ............................................... 98
3. 3 .1 Differential pressure flowmeters .............................. 99
3. 3. 2 Float-type area flowmeters ................................... 105
3. 3. 3 Volumetric flowmeters ........................................ 110
3. 3. 4 Turbine flowmeters ............................................ 113
3. 3. 5 Magnetic flowmeters .......................................... 117
3. 3. 6 Vortex flowmeters ............................................. 125
3. 3. 7 Ultrasonic flowmeters ......................................... 130
3. 4 Measurement of Pressure .......................................... 135
3 .4 .1 Pressure transmitters .......................................... 136
3. 4. 2 Types of pressure detectors .................................. 138
3. 5 Measurement of Liquid Level ..................................... 141
3.5.1 Float liquid-level meters ...................................... 141
3. 5. 2 Pressure differential liquid-level meters .................... 141
3.5.3 Displacer liquid-level detectors .............................. 144
3. 5. 4 Purge-type liquid-level meters ............................... 145
3.5.5 Ultrasonic liquid-level meters ................................ 146
3.5.6 Capacitance liquid-level meters .............................. 147
3. 6 Measurement of Displacement and Angle ........................ 148
3. 6.1 Resistance potentiometer methods ........................... 148
3. 6. 2 Electromagnetic induction methods .......................... 148
3. 6. 3 Magnetic balance method ..................................... 152
3. 6. 4 Magnetic strain method ....................................... 153
3. 7 Measurement of Rotation .......................................... 153
3. 7 .1 Measurement using tachometer generators ................. 153
3. 7. 2 Pulse output sensors ........................................... 155
3. 7. 3 Digital counting tachometers ................................. 156
3. 8 Measurement of Composition ..................................... 158
3. 8.1 Gas chromatography .......................................... 158
3. 8. 2 Infrared analyzers ............................................. 163
3. 8. 3 Oxygen analyzers .............................................. 166
3. 8. 4 pH meters and ORP meters .................................. 169
xii
Contents
3. 8. 5 Moisture/humidity meters .................................... 172

3.8.6 Turbidity meters ............................................... 174
3. 8. 7 Conductivity meters ........................................... 176
3. 8. 8 Other composition measuring devices ....................... 179
3. 9 B/M Systems ........................................................ 188
3. 9.1 Basis weight sensor (B sensor) ............................... 189
3. 9. 2 Moisture sensors (M sensors) ................................. 191
3. 9. 3 Calipers (paper thickness gauges) ............................ 192
3. 9. 4 Moisture sensor for thick paper .............................. 193
3. 9. 5 Color sensors ................................................... 194
3. 9. 6 Ash sensors ..................................................... 195
3.10 Signal Converters ................................................. 195
3.10 .1 The purpose of signal converters ........................... 195
3.10. 2 Thermocouple signal converters ............................ 197
3.10. 3 Resistance signal converters ................................ 199
3.10. 4 Two-wire signal transmission .............................. 200
3.10. 5 Pulse signal converters ...................................... 201
3.10. 6 Computer input equipment. .................................. 202
Practice Questions ....................................................... 205
Answers to Questions ................................................... 205
References ............................................................... . 205
Chapter 4 RECORDERS AND CONTROLLERS
4 .1 Recorders ............................................................ 209
4 .1.1 Types of recorders ............................................ 209
4 .1. 2 Recorder functions ............................................. 210
4 .1. 3 Pen recorders .................................................. 212
4 .1. 4 Multipoint recorders ........................................... 218
4. 2 Controllers........................................................... 222
4. 2 .1 Pneuma tic and electronic controllers ........................ 222
4. 2. 2 Analog electronic controllers ................................. 223
4. 2. 3 Digital controllers ............................................. 226
4. 2. 4 Programmable controllers .................................... 231
4. 2. 5 Batch controllers and blending controllers ................. 235
4. 3 Computing Stations and Set Stations ............................. 241
4. 3.1 Alarm set stations ............................................. 241
4. 3. 2 Programmable computing units .............................. 242
4. 3. 3 Manual set stations and manual operating stations ....... 243
References ............................................................... . 245
Chapter 5 SYSTEM CONTROL EQUIPMENT
5.1 Overview of System Control Equipment ......................... 248
5 .1.1 Development .................................................... 248
Contents
xiii
5 .1. 2 Configuration of a total FA system ......................... 252

5. 2 Distributed Control System ........................................ 256
5. 2 .1 Concept of the distributed control system................... 256
5. 2. 2 Configuration of the distributed control system and
its functions ................................................... 259
5. 2. 3 Feedback control .............................................. 267
5. 2. 4 Sequential control ............................................. 271
5.2.5 Man-machine interface ....................................... 276
5. 2. 6 Communication with other systems .......................... 284
5. 2. 7 Engineering ..................................................... 285
5. 3 Production Line Control System .................................. 291
5. 3 .1 Summary of production line control systems ............... 291
5. 3. 2 Types of production line control systems ................... 292
5. 3. 3 FA computer systems ......................................... 295
5. 3. 4 FA computer system hardware............................... 298
5. 3. 5 FA computer software ........................................ 303
5.4 Computer System Equipment for Production Management .... 306
5. 4.1 Computer components and configuration ................... 306
5. 4. 2 Software for production management computer systems.. 316
5. 5 Data Communication and Equipment............................. 325
5. 5 .1 Data communication and standards ......................... 325
5. 5. 2 Methods of data communications ............................ 327
5.5.3 The IEEE-488 instrument bus ................................ 329
5. 5. 4 The RS-232 C interface and modems ........................ 331
5. 5. 5 Local area networks .......................................... 334
5. 5. 6 Optical communications ....................................... 335
5. 6 Basic Components of Digital Control.. ........................... 336
5. 6 .1 Microprocessors................................................ 336
5. 6. 2 Memory elements and storage equipment ................... 343
5. 6. 3 Display elements and devices ................................. 346
5.6.4 Analog/digital conversion .................................... 351
5. 6. 5 Optical communication elements ............................. 353
References ................................................................ 354
Chapter 6 FINAL CONTROL ELEMENTS
6.1 Types of Control Valves ........................................... 355
6 .1.1 Pneumatic control valves ..................................... 355
6 .1. 2 Electrical control valves ...................................... 355
6 .1. 3 Hydraulic control valves ...................................... 356
6 .1. 4 Self-powered control valves .................................. 356
6. 2 Choice of Control Valves .......................................... 356
6. 2 .1 Various conditions affecting choice.......................... 356
6.2.2 Sizing ........................................................... 360
xiv
Contents
6 02 03 Flow characteristics ........................................... 361

6 02 04 Rangeability .................................................... 363
60205 Materials ....................................................... 364
6 03 Control Valve Bodies .............................................. 367
6 03 01 Characteristics of various types of valves .................. 367
60302 Rating ........................................................... 373
6 03 03 Connection to piping .......................................... 374
6 4 Control Valve Actuators ........................................... 374
6 04.1 Conditions under which an actuator should be installed .. 37 4
6. 4. 2 Power sources .................................................. 37 4
6. 4. 3 Types of actuators and their characteristics ............... 376
6. 5 Positioners and Accessories ....................................... 384
6. 5.1 Positioner functions ........................................... 384
6. 5. 2 Pneumatic pressure positioners .............................. 384
6. 50 3 Current-to-pneumatic positioners ............................ 384
6 050 4 Current-to-current positioners ............................... 386
6 05. 5 Accessories ..................................................... 386
6 06 Self-powered Valves ............................................... 388
6. 6 01 Pressure-regulating valves ................................... 388
6 06. 2 Temperature control valves .................................. 389
6. 6. 3 Flow control valves ........................................... 389
6. 6. 4 Float valves .................................................... 389
References ................................................................ 390
0
Chapter 7 SYSTEM ENGINEERING

7 .1 System Engineering Basics ........................................ 392
7 01.1 Plant construction overview .................................. 392
7 0102 System design considerations................................. 395
7. 2 Instrumentation System Design ................................... 399
702.1 Job planning .................................................... 399
7 02 02 System specifications .......................................... 403
7 02. 3 Device and function specifications ........................... 407
7.2.4 Instrumentation work specifications ......................... 430
7 02 05 Related work ................................................... 434
7. 2. 6 Instrumentation drive system design ........................ 436
7. 2 7 Other system functions (safety, failsafe and
redundancy measures) ........................................ 444
7 03 Control Room and Man-Machine Interface ...................... 453
7. 3 .1 Human engineering and control panel design .............. 453
7 03 02 Control room engineering ..................................... 457
7. 4 Instrumentation Work and Startup ............................... 460
0
Contents
XV
7. 4. 1 Overview ........................................................ 460

7. 4. 2 Instrumentation work planning .............................. 460
7. 4. 3 Instrumentation work design ................................. 463
7. 4. 4 Startup execution .............................................. 467
7. 4. 5 Startup operations ............................................. 469
7. 5 Quality Assurance .................................................. 470
7. 5 .1 Engineering quality ............................................ 470
7. 5. 2 Design review (DR) ............................................ 471
References ................................................................ 482
Chapter 8 ADVANCED CONTROL
8 .1 Control Theory Considerations Control ......................... 483
8. 2 Feedforward Control ............................................... 486
8.2.1 Feedforward control in a heat exchanger ................... 486
8. 2. 2 Combining feedforward control and feedback control ..... 488
8. 2. 3 Determination of feedforward elements ..................... 489
8. 2.4 Feedforward control application examples ................. 490
8.3 Control of Dead-Time Processes ................................ .492
8. 3.1 Dead-time processes ........................................... 492
8. 3. 2 Smith controllers .............................................. 494
8. 3. 3 Sampling PI controller ........................................ 500
8. 4 Non-interacting Control ........................................... 502
8. 4.1 Interaction between process variables ....................... 502
8. 4. 2 Influence exerted by mutual interaction..................... 504
8. 4. 3 Expressing the degree of interaction ........................ 504
8.4.4 Controlled variable and manipulated variable
combination ................................................... 508
8. 4. 5 Non-interacting control ....................................... 509
8. 4. 6 An example of non-interacting control ..................... 511
8. 5 Self-tuning Controller .............................................. 511
8.5.1 Overview........................................................ 511
8. 5. 2 Gain-scheduling control. ...................................... 514
8. 5. 3 Self-tuning controller (STC) .................................. 515
8. 5. 4 STC based on the expert method ............................ 517
8. 5. 5 STC application considerations .............................. 521
8. 6 Optimal Control. .................................................... 521
8. 6.1 The meaning of "state" ..................................... 521
8. 6. 2 Integral optimal regulator .................................... 522
8. 7 Kalman Filter ....................................................... 524
8. 7 .1 Kalman filter formula ......................................... 524
8. 7. 2 Application to the parameter estimation problem ......... 525
8.8 Other Forms of Advanced Contro1.. .............................. 527
References ................................................................ 527
xvi
Contents
Chapter 9 CONTROL OF PROCESS UNITS (Application I )

9. 1 Overview ............................................................ 529
9. 2 Control of Fluid Transport Processes ............................ 530
9. 2 .1 Pump control ................................................... 530
9. 2. 2 Compressor control ............................................ 533
9. 3 Control of Heat Transfer Processes .............................. 540
9. 3.1 Control of heat exchangers ................................... 540
9. 3. 2 Heating furnace control. ...................................... 546
9. 4 Control of Distillation Processes .................................. 550
9. 4 .1 Binary-component distillation column control ............. 550
9. 4. 2 Multi-component distillation column control.. ............. 566
9. 5 Control of Reaction Processes .................................... 573
9. 5.1 Control of a stirred-tank polymerization reactor .......... 573
9.5.2 Control of a gas-phase solid-catalytic reactor ............. 580
9. 6 Other Process Control. ............................................. 590
9. 6 .1 Control of refrigeration equipment .......................... 590
9. 6. 2 Evaporator control ............................................ 592
9. 6. 3 Drying process control ........................................ 595
References ................................................................ 602
Chapter 10 INSTRUMENTATION TO MANUFACTURING INDUSTRIES

(Application IT)
10 .1 Instrumentation Application in the Petroleum Industry ....... 604
10 .1.1 The petroleum industry and instrumentation .............. 604
10 .1. 2 Topping unit instrumentation ............................... 606
10 .1. 3 Off-site instrumentation ..................................... 614
10.2 Instrumentation Applications in the Iron- and
Steel-Industry .................................................... 621
10.2 .1 Overview of instrumentation in the iron-and
steelmaking process ......................................... 621
10.2. 2 Blast furnace instrumentation ............................... 624
10.2. 3 Continuous casting equipment instrumentation............ 635
10.2. 4 Instrumentation for an electrolytic galvanizing line ...... 642
10.3 Instrumentation Applications in the Power Industry .......... 648
10.3 .1 Overview ...................................................... 648
10. 3. 2 Thermal power plants ....................................... 648
10.3. 3 Boiler control ................................................. 649
10.3. 4 Turbine control ............................................... 663
10.3. 5 Power plant system control ................................. 667
10.3. 6 Nuclear power plant overview .............................. 669
Contents
xvii
10.3. 7 Pressurized water reactor control system ................. 675

10.4 Instrumentation Applications in the Food Processing
Industry ............................................................ 687
10.4 .1 Overview ...................................................... 687
10.4. 2 Whiskey distillery instrumentation ......................... 689
10.4. 3 Sugar refinery instrumentation ............................. 693
10.5 Instrumentation Applications in the Paper Manufacturing
Industry ........................................................... 706
10.5 .1 Overview of an integrated paper mill. ..................... 706
10.5. 2 Pulp plant instrumentation .................................. 709
10.5. 3 Instrumentation applied to the papermaking process ..... 719
10.6 Waterworks Instrumentation Applications ..................... 726
10.6.1 Overview of waterworks facilities .......................... 726
10.6. 2 Water treatment-related detectors ......................... 727
10.6. 3 Filtration equipment instrumentation ...................... 731
10.6. 4 Chemical injection equipment instrumentation ............ 732
10.6. 5 Instrumentation for water-supply and
distribution facilities ........................................ 737
10.6. 6 An integrated control system for large-scale,
wide-area waterworks facilities ........................... 738
10.6. 7 Water distribution information management system ..... 739
10.6. 8 Wastewater system overview ............................... 745
10.6. 9 Overview of activated-sludge processes ................... 745
10.6.10 Wastewater treatment instrumentation ................... 747
10.6.11 Sludge treatment instrumentation ......................... 749
10.7 Instrumentation Application in the Automobile Industry ..... 751
10.7 .1 Overview of automobile industry instrumentation ........ 751
10.7. 2 Production management at a painting factory ............ 752
10.7. 3 Storage control ............................................... 755
10.8 Product Control in Batch Processing ........................... 759
10.8.1 Batch process recipe management .......................... 760
10.8. 2 Batch process control ........................................ 763
10.8. 3 Recipe management and operation methods ............... 769
References ............................................................... . 773
APPENDIXES
App.1 Reference Thermoelectromotive Force Tables ............... 778
App.2 Reference Resistance Value of Pt 100 ......................... 782
App.3 Tables of Laplace Transform .................................. 784
Index ............................................................... ....... 785
xviii
Contents
Chapter 1
INDUSTRY AND INSTRUMENTATION
1.1
The Word "Instrumentation"
Within the various topics covered in. this book, a number of com-
pound words and expressions appear utilizing the word "instrumentation," such as "Instrumentation System," "Instrumentation Engineer"
and "Instrumentation Technology." Although the usage here is purely
technical, it's interesting to note that dictionaries also define "instrumentation" as a musical term meaning the "composition of musical instruments in an orchestra," or in other words, the technique of selecting an appropriate musical instrument makeup for an orchestra so as
to achieve optimum results for performing a particular piece of music.
This definition, as it turns out, can serve as a fitting analogy to
the industrial use of "instrumentation." If we replace the words musical instrument, orchestra and piece with industrial instrument, manufacturing plant and manufacturing process, we can define industrial
instrumentation as the technique of selecting appropriate measurement
devices for a manufacturing plant so as to achieve optimum results for
a particular manufacturing process. In this case, the results include
quality of products, cost of production, ease of operation, and so on.
However, as words very often seem to have a life of their own,
their meaning can change from generation to generation, and technical
terms in particular seem to evolve quite rapidly. A case in point is the
word "instrument." As applied to instrumentation within American industry of the 1950's, it then referred to relatively simple measuring instruments, but with time has come to include very complex and sophisticated industrial instruments as well. Moreover, with the advent of information processing tools based on computer and communication technology, it has also become necessary to include computerized systems
when talking about instrumentation. In addition, the range of object
1.1
The Word 'Instrumentation'
processes to which instrumentation is applied has also quickly expanded over the years.
Definition of Industrial Instrument:

According to JIS (Japanese Industrial Standard) Z 8104, "industrial
instrument" is defined as "measuring/controlling equipment used in
production processes in industry." Here, "measuring/controlling equipment" is in turn defined as apparatus which indicate and/or record
quantities or physical properties, as well as having computing, controlling or alarm functions, thus including detectors, transmitters, and the
like.
1.2 The Development of Instrumentation

The automatic control of the measurement of industrial quantities,
such as temperature, flow and pressure, first began in the 1920's in
American oil refining processes. This period was characterized by local
instrumentation in which large-size mechanical controllers were installed in the process area. The subsequent development of instrumentation technology in following periods came about as the needs of various industries and the advancement of industrial instruments became
closely intertwined. The development of instrumentation technology in
postwar Japan and corresponding background events in society are listed in Table 1.1.
[1] The 1950's and 1960's
During the 1950's, instrumentation technology experienced major
development together with Japan's economic restoration centered in its
petroleum, steel and textile process industries. Instrumentation in this
period was mainly characterized by control systems consisting of a
number of controllers each of which performed analog operation processing for one loop. In the beginning, pneumatic controllers driven by
air pressure were used, but with the advancement of electronics and
the shift toward largescale processes, control equipment progressed
from pneumaic-operated to electronic-operated models.
On entering the 1960's, computers for use in process control first
appeared in the field of instrumentation. At first, they were mainly
used for the monitoring and record taking of process operations (data
logging). In addition, by making use of the computer's computational
ability, they were used for computing optimum process conditions or
safe operation conditions with calculated setpoints passed on to control2
Chap.l Industry and Instrumentation
Table 1.1 Development of instrumentation technology in Japan

Year
Social events
Instrumentation Technology
1950
Recovery period after World War ll

Technology introduction from abroad
Prosperity of synthetic textile
industry
Germanium transistor
Standardization of transmission
signal (3 to 15 psi)
Practical use of electronic tube
self-balancing instruments
1955
Construction of petroleum industrial

complex
Construction of power station
Computer
Silicone transistor
Practical use of electronic control

systems
Graphic panel
1960
Construction of blast furnace

Investment for labor saving
Computer control systems
1965
IC
Scalp up and integration trends of
plant construction
DDC
1970
Microprocessor
Investment for environmental
maintenance
(Environmental pollution problem)
First oil crisis
Standardization of transmission
signal in electronic control
systems 4 to 20 rnA DC (IEC)
Hierarchy systems
CAD
Robot
Office automation
Automation of machine tool
1975
Second oil crisis

Popularization of microcomputer
Distributed control systems

Package type control systems
1980
Low economical growth period

Diversification of social needs
Single loop DDC

Factory automation (FA)
lers. This is known as Supervisory Process Control or Setpoint Control

(SPC).
In time, however, thought was given to replacing the functions performed by analog controllers and computational units by using the increasing computational power of the computer. Accordingly, direct
digital control by computer, or DDC, came to be realized. At this time,
centralized DDC was employed in which many loops up to several hundred were controlled by one computer unit. As a consequence, however, since an unexpected computer problem could bring plant operations
to a halt, the incorporation of redundant design elements such as CPU
duplication, backup devices, etc., came to be necessary, resulting in increased costs. As a result, due to economic considerations at this time,
centralized DDC did not fully replace analog control systems.
1.2 The Development of lustrumentation
The 1970's to the present

The introduction of the microprocessor in the 1970's brought about
revolutionary changes to many areas, and the instrumentation field
was no exception. With its lower price and higher performance, control systems, which up to then could only employ the one-computer
centralized-control technique due to processor cost, could now implement a "distributed instrument control system." In these systems,
microprocessors are distributed amongst each function or process area,
and each microprocessor communicates with centralized CRT -based
terminals for process monitoring and operational control. From 1975
onward, manufacturers around the world, including japan, expanded
the application of the microprocessor to batch and sequence control.
Another major point related to this development of instrumentation technology is the adoption of a standard interface. When connecting multiple devices to each other and configuring an instrumentation
loop or constructing a large scale instrumentation system, the standardization of interface signals between the devices is extremely important. Work on this standardization first began in 1950 with SAMA *1 in
the United States, which established a standard pneumatic signal of 3
to 15 psi (0.2 to 1.0 kg/cm 2); this standard eventually came to be used
worldwide.
In addition, in 1970, an electric current signal of 4 to 20 rnA DC
was standardized by IEC* 2 With this standardization, analog industrial
instruments produced by different manufacturers could be interconnected freely, thus contributing to the overall development of instrumentation. However, as modern industrial instruments are rapidly
becoming digitalized and as the interface between devices moves from
analog signals to mass-information-carrying digital signals, further
standardization becomes even more important. In this regard, IEC has
been working on a bus standard for distributed instrumentation control systems called "PROW AY"; however, the current situation in industry still has many independent bus systems from various manufacturers. In addition to PROW A Y, MAP (Manufacturing Automation
Protocol) and other field busses are currently in the process of standardization, but more time is needed before these standards are implemented throughout the instrumentation field.
[2]
1.3
[1]
*1
*2
Trend toward Total FA

From continuous processes to discrete processes
The general process flow of manufacturing operations from raw
SAMA (Scientific Apparatus Manufacturers Association)
IEC (International Electrotechnical Commission)
materials to finished products is shown in Fig. 1.1. As shown here, going from upstream to downstream, the objects of control may change
from fluids to solids, and likewise the corresponding production processes may change from continuous processes handling fluids to discrete (non-continuous) processes handling solids.
Process control
PA(Process automation)
FA(Factory automa tion)
Fig. 1 1 "Upstream" and "Downstream" processes
For example, in the overall paper manufacturing process, the processes employed from pulp cooking up to paper making are continuous
processes, while those employed after paper finishing and cutting,
such as rolling the paper into reels and packaging for shipping, are noncontinuous discrete processes.
If we apply this concept to a single factory, we can see that even
for process industries whose main production is performed by a continuous process, there usually are still discrete processes required such as
product inspection, packaging, transport, storage and shipping.
When speaking of process instrumentation or automation, we generally refer to upstream continuous-process control as process automation (PA). On the other hand, the automation of downstream discrete
processes has come to be called factory automation (FA). Since the
"tools" required here are different than those for PA, FA has been considered up to now as being in a different dimension than instrumentation. However, in order to optimize sequential production processes,
discrete process instrumentation should also be standardized.
The above has described a point of view involving upstream and
downstream processes when considering the factory as a flow of materials or products. If we take a point of view based on the flow of infomation, we can consider a hierarchical division of levels from the upper to the lower, as shown in Fig. 1.2. For example, at the management level, production plans are developed for the work of each manufacturing department within the factory, and specific manufacturing instructions are sent to each control system of these departments.
Accordingly, the control systems of each manufacturing department control their various production processes according to the manu1.3 Trend toward Total FA
Control level
Fig. 1 2
Instrumentation system hierarchy
facturing instructions received and produce products according to the

upper level production plans. In addition, information such as produc
tion status are transmitted on-line to the management level. Based on
production data, new plans or decisions can then be fed back to the control systems.
Although the prior description relates to manufacturing, it should
be mentioned that in addition to manufacturing processes in a modern
plant there are also automated processes within the departments of
technical and business affairs, namely, LA (laboratory automation) and
OA (office automation).
[2] Achieving total FA
Today, in order to survive the intensified market competition between various enterprises, it has become necessary to accurately determine customers' diversified needs as well as to plan for quality improvement, labor savings, energy savings and more efficient multiproduct small-volume production. For this reason, a total FA system
must be developed in which individually developed PA, OA, LA and
FA processes are combined "organically." The enterprises manufacturing systrm can then be advanced to achieve optimal production in
terms of the whole factory. In this context, instrumentation in manufacturing industries must realize a real time, flexible, consistent production system extending from the entrance of the factory to its exit.
1.4 Classification and Use of Instruments
As part of what is generally termed "instruments," this section

considers the roles and features of industrial instruments.
Along with the expansion of the instrumentation field and the development of instrumentation technology, the types of instruments
have been increasing and their classification has been changing as well.
6
Figure 1.3 shows instrument classification. This particular example

describes items related to instruments as taken from the classification
included in the Machinery Statistical Annual Report 1 l issued by the
Ministry of International Trade and Industry (MITI).
Within the electric measuring instruments field, electric instruments and electrical measuring instruments are most often used for
laboratory development and product inspection on a one-unit basis.
In contrast, industrial instruments are characterized by use in factory production processes for interconnecting a system composed of
multiple units. These units consist of sensors, transducers, controllers
and actuators, and system operation is based on mutual relationships
between the component units. Note that computer related equipment
Electric measuring instruments
Electric instruments (indicating instruments, recorders, watt-hour meters etc.)
Electrical measuring instruments (voltmeters, am meters and wattmeters,
IC test instruments, wavefrom measuring instruments etc.)
Industrial instruments (process use industrial instruments, process use
analytical instruments, process monitoring control systems etc.)
Measuring equipment (calipers, micrometers, water-meters, analytical

instruments, instruments for pollution measurement etc.)
Testing equipment (metallic material testing machine, Non-metallic
material testing machine etc.)
Surveying equipment (transits, levels etc.)
Electronic application equipment
X -ray equipment (medical use etc.)
Ultrasonic application equipment (fish finders, washers, welders etc.)
Computers and related equipment (digital computers, control computers,
peripheral devices etc.)
Fig. 1 3 Classification of measuring instruments
Industrial instruments
Process use industrial instruments

temperature instruments, flow instruments otl)er transmitters,
indicators, recorders, controllers, auxiliary instruments,
manual stations
Process use analytical instruments
Process monitoring control system
Other industrial instruments
Fig. 1 4 Classification of industrial instruments
1.4 Classification and Use of Instruments
are classified under electronic application equipment in Fig. 1.3. In

addition, process analysis instruments for use in testing and inspection
of products and materials, as well as process-monitoring control systems are also included in industrial instruments, as shown in Fig. 1.4.
Since industrial instruments are used in a much more continuous
fashion compared with electric instruments and electrical measuring instruments, their availability factor must be high. Because of this, as
well as a sometimes harsh usage environment, durability and reliability
are particularly required. Moreover, when industrial instruments are
installed in a dangerous environment containing combustible materials,
special consideration must be taken to prevent accidental explosions
that can originate in faults or breakdowns of the instrumentation system.
The number of electric measuring instruments produced for the period 1980 to 1985 according to the Machinery Statistical Annual Report0 issued by MITI is shown in Fig. 1.5, and that for types of industrial instruments for the year 1985 is shown in Table 1.2.
400
,--,
...c
300
--
Electric measuring instrui~/
200
;.---'
"5
"2
't:l
c..
~-~
~---
100
____ _,
0
1980
:r----
1981
339
/"_,J
___.
243
llm""l -.......
_.,t;>
'
>--+~:f-~-
1982
1983
1984
40
1985
Year
Fig. 1 5 Trends in production of measuring instruments
Table 1.2 Production of industrial instruments (1985)
100 (%J
Total industrial instruments
243 (Billion YenJ
Process use industrial instruments
117
48
Process use analitical instruments
Process use monitoring control systems
60
25
Others
61
25
REFERENCES
1) MITI, Research and Statistics Department: Machinery Statistical Annual Report

(1986) (in Japanese).
References
Chapter 2
PROCESS CONTROL
The use of "automation," whether it be factory automation (FA),

laboratory automation (LA), office automation (OA), or home automation (HA), has come to pervade almost every major field in modern society. Moreover, process automation (PA), as employed by process industries such as the petroleum, chemical, petrochemical, and steel industries, has seen the introduction of various forms of real automatic control both early and recently in its history, thus reflecting the development of process control through the years. This chapter describes the
essential elements of process control beginning with feedback control,
the core of process control, and leading up to sequential control, which
has come to perform a fusion with feedback control via recently developed distributed control equipment.
2.1
Fundamentals of Feedback Control
2.1.1 Configuration of a control system

(a) Feedback control
"Feedback control" is defined as "control in which a comparison is
performed, based on feedback, of a controlled variable and a desired
value, and the subsequent corrective action taken so as to make the
two values agree."
For example, in an air-conditioning system, the room temperature
(controlled variable) is detected and compared with the set temperature (desired value). If a difference (deviation) between the two exists,
it must be brought to 0 by turning the power (manipulated variable) on
or off, or in other words, by rotating the compressor or stopping it
(corrective action).
A block diagram of this air-conditioning system is shown in Fig.
2.1. As can be seen, air-conditioning equipment is configured in a
closed loop.
2.1
11
CD
~--------- - ---- -- -----,I

I
I
Setting temperature
Comparison
@ Corrective action
@ Temperature detection
'
@Room
@ Outdoor temperature
'
----------- ~
L--------conditioner
Air
Fig. 2 1
Block diagram for air conditioning
Although the above is an example of the "automatic control" of

room temperature, even manual control in which people measure the
room temperature themselves, compare it with the disired temperature
and turn the air- conditioning power on or off, can be called a feedback
control system, since a control loop is formed by a human-assisted
closed loop.
(b) Feedback control system
A block diagram of a generalized feedback control system is
shown in Fig. 2.2.
Figure 2.3 shows a block diagram of a generalized feedback control system as used in process control. In process control, the value of
the controlled variable is obtained at a detection element and is passed
<D
CD
Desired value
Comparison
@ Controlled deviation
@ Control device
@ Control unit
@ Manipulated variable
@ Disturbance
Fig. 2 2
Controlled system
Controlled variable
Feedback control system
<D
CD
Setpoint value (SV)

Comparison
@Controller
@ Computation
Fig. 2 3
12
@ Manipulated variable (MV)
Control valve
Process
Disturbance
Controlled variable
Process variable (PV)
@ Detection Conversion
Block diagram for feedback process control
Chap.2
Process Control
to the controller as a process variable (PV). The controller in turn

compares the value of this variable with the setpoint value (SV) and
the deviation is computed by a control element to produce a manipulation variable (MV), which is forwarded to a control valve corresponding to a final control element. The control valve then opens or closes
depending on the value of the signal, thus adjusting the manipulated
variable. In the above fashion, control deviation generated by outside
disturbance can be corrected. Note here that the above valve is opened
or closed to decrease the amount of deviation, or in other words, negative feedback is performed.
2.1.2 Characteristics of a control system
(a) Block diagram
Control systems are evaluated by considering such factors as stability and speed of response, and in order to understand these factors, it
is necessary to investigate the characteristics of control systems. In
this regard, block diagrams are utilized in which the constituent elements of a control system are indicated by blocks and their interconnection or signal flow by lines and arrows. Components of these block
diagrams are shown in Fig. 2.4 and described below.
Inp'!
I I
AtA-B
Block
Summing point
Out~ut
::
Pick off point
Fig. 2 4 Symbols used for block diagram
(1) Block
Indicates a signal transfer element having particular transfer characteristics between input and output; blocks are used in conjunction with arrows to indicate input and output.
(2) Summing point
Indicates summation of two signed signals; the
dimension of the two signals must be the same.
(3) Pick off point
Indicates signal branching.
(b) Transfer function
These are commonly used to express a signal transfer element,
and are defined according to the Laplace transform. Specifically, taking the Laplace transform of input signal x(t) and output signal y(t) at
an initial value of 0 and obtaining X(s) and Y(s) respectively, G(s) is
then expressed as the ratio of the input and output signals: Y(s)/X(s),
and is called a transfer function. This relationship is shown in Fig.
2.5(a).
A transfer element can also be expressed as a time function, as

2.1
13
Laplace transform
~)
(a ) Tran fer
fu nction
Y(s) = G( s) X ( s)
Convolution integral
(b) Weighting
fu nction
(c) Frequency
transfer
function
x~t)
X ( j w)l G (j w)
u.."''
y ( t ) = ); g ( r ) x ( t - r ) dr
Y(jw) = G(jw) X ( jw)
Fig. 2 5 Various transfer elements
shown in Fig. 2.5(b). In this case, output signal y(t) is expressed as the
convolution integral of input signal x(t) and transfer element g(t). Since
g(t) is considered to be a weighting function here, g(t) is equal to y(t)
when x(t) is a unit impulse o(t)*. To put it another way, the convolution integral of two time functions can also be expressed as the product of the Laplace transform of these two functions.
Convolution integral:
As shown in Fig. 2.6, input signal x(t) is approximated by time
interval partitions .Jr. At '= 0, or in other words, at the point where
f--4t-r, if the input signal has a width of .Jr and a height of 1/.Jr, the
output signal at t= t nearly becomes g(t), since g(t) is a weighting function. Since the height at r = 0 is x(t- r), the output signal is
g(r)x(t- r).Jr. Accordingly,
t
y(t)= L:g(r)x(t- r).dr

r=O
and as .Jr--40, we obtain the integral

y(t)= fg(r)x(t- r)dr
* Unit impulse
o(t) is defined as the following function with o(t) =0 when t0:
f)<t)dt=l
14
Chap.2 Process Control
x ( l)
y (l)
;=0
-1
Fig. 2 6 Convolution integral
(c) Laplace transform
Since transfer functions express the characteristics of transfer elements in the frequency domain, elements possessing specific characteristics in the time domain also exhibit specific characteristics in the frequency domain.
To perform a conversion between a time function f(t) and a frequency function F(w), the well known Fourier transform is used, as
shown below.
Fourier transform:
(2.1)
Fourier inverse transform:
11"'F(w)eiwtdw
/(t)=21Z"
(2.2)
-0>
Since control engineering deals with transient responses in which

t < O is not considered, the Laplace transform which treats all signals
as having an initial value of zero is quite appropriate. If time function
f(t) for t < 0 is 0 and if ["'it(t)idt < oo, then Equation (2.1) becomes
F(w)= ["' /(t)e-iwtdt

2.1
15
If we now let jw = s, we can transform f{t) to a function of s by the

following formula:
.[/(t)]=F(s)= [~ f(t)e-stdt
(2.3)
where .C[f{t)] denotes the Laplace transform. In addition, the inverse

transform of the above function for converting back to f{t) is denoted
as .c-1[F(s)] and is expressed as the following formula:
.- 1(F(s)]= /(t)= 21
l+i~ F(s)etds
lr}
(2.4)
-J~
Laplace inverse transform:

In the case that ~1/(t)ldt >co
convergence factor e-ct must be
multiplied in. The value C here is made larger than the value a in
which jt(t)e-<ftdt< co . At this time, s becomes a complex frequency expressed as s = C + jw, and the Laplace inverse transform becomes
..L'- 1[F(w)]=__!_,.
[C+Jw
2JrJ Jc-Jw
F(s)etds
Based on the Laplace transforms and theorems for representative

signals and transfer elements listed in Appendix 3, the s-transforms
for basic electric circuit elements are shown in Fig. 2. 7.
Elcctroric circuit elements
S-Transform
R
----'VVWv--
Z (s) = R
Capacitance
--tl--
Z(s ) = C.
Inductance
Resista nce
Fig. 2 7
c
L
Z ( s ) = Ls
S-Transform for basic electronic circuit elements
(d) Transient response
In the event that a steady state condition as possessed by an input

signal changes to another state, the characteristics present until the
output signal attains a steady state are called transient characteristics,
and they appear as transient response. Various forms of transient re16
sponses, such as step response, impulse response and ramp response,

can be obtained depending on the type of input signal; however, the
most often used signal is a unit step signal u(t)* for generating a step
response.
Figure 2.8 shows an example in which the form of a transient response is sought when a voltage unit step signal is applied to a first-order lag circuit consisting of a capacitor C and resistor R (initial voltage
of the capacitor is 0).
Unit step signa l
Transient response
RC fir t-order
lag circuit
It's laplace
transform
Fig. 2 8 RC first-order lag circuit
Although the Laplace transform for a unit step signal is listed in

Appendix 3 as 1/s, this can be easily obtained by directly solving the
Laplace transform formula for this function. Since the unit step signal
is 1 for t > O, we get:
F(s)= (""e-stdt= _l[e-st];=l
Jo
Next, we determine the Laplace transform for the CR first-order

lag circuit by using the s-transforms for electric circuit components in
Fig. 2.7 and solving, as follows:
Y(s)=G(s)X(s)
1
G(s)= 1+CRs
1/Cs/csX (s )
R+
1
1
1+Ts
Here, T=CR (Time constant).

* The unit step function is equal to 0 for t < O and to 1 for t > O, i.e., it is non-continuous at t = 0.
2 .1 Fundamentals of Feedback Control
17
In addition, this first-order lag circuit can be expressed as a differential equation, where x(t) represents the input signal, and y(t) the output
signal:
T~ +y(t)=x(t)
where again T= CR.
If we now apply the Laplace transform of a derivative as indicated
by the theorems listed in Appendix 3, we get
TsY(s)+ Y(s)=X(s)
_ Y(s) _
1
G(s)-)((Sf- 1 + Ts
which is the same transfer function as above.

Accordingly, the output signal is then
1
1
Y(s)=G(s)X(s)= 1 + Ts s
1/T
s(s+1/T)
Although the transient response can now be determined by taking

the inverse transform of the above, direct solving of Eq. (2.4), the formula for finding an inverse, can be time consuming. For the example
here, however, we can again refer to Appendix 3 and use the function
corresponding to the transform 1/s(s+a) to obtain the following expression for the transient response:
y(t)=1-e-ttT
In this way, the inverse transform of function F(s) having an algebraic function denominator as shown above can be found by expanding
F(s) into partial fractions, and then summing the inverse time functions obtained from the Laplace transform table for each factor.
(e) Frequency response
The above has described how to obtain the transient response of a
transfer element through use of the Laplace transform. Also of importance is knowing the frequency response of a transfer element. An output signal in steady state resulting from the application of a sine wave
input signal exhibits the characteristics of amplitude gain and phase
shift with respect to the input signal. These characteristics which
change depending on the frequency of the input signal are called frequency characteristics, and they appear as frequency response. In order to express frequency characteristics, frequency transfer functions
are used, and specific frequency transfer functions can be obtained by
18
Chap.2
Process Control
substituting the s in ordinary transfer functions by jw. A transfer element expressed as a frequency transfer function is shown in Fig.
2.5(c).
As an example, to find the frequency response of a first-order lag

circuit, the frequency transfer function G(jw) can first be expressed as
follows:
G(jw)
1+jwT
J1 + w2T2 L
-I
- an w
And for methods of expressing the resulting frequency response,

vector locus plots and Bode diagrams are commonly used.
In vector locus plots, frequency w is used as a parameter on a complex coordinate two-dimensional surface, and the frequency response
is indicated by the locus drawn out by the tip of a vector whose value
changes along with w.
In Bode diagrams, the frequency w is assigned on a logarithmic
scale to the horizontal axis, and the gain and phase shift are plotted
separately along the vertical axis so as to produce a set of two plots;
the units for plotting the gain are usually decibels (dB: 20 log 10n), and
those for phase shift are degrees or radians.
Table 2.1 summarizes the transient response and the frequency response as expressed by vector locus plots and Bode diagrams for the
main types of transfer elements with respect to unit step signals.
2.1.3 Feedback control and stability
(a) Loop transfer function gain
A block diagram obtained by simplifying the feedback control system for processes and by treating transfer elements as proportional elements is shown in Fig. 2.9. If we now solve for process variable PV
and control deviation DV (omitting the s symbol), we get:
Pv
KpKc SV +
KP DS
1+KcKP
1+KcKP
DV= 1+kcKPSV- 1+1j(cKPDS

Term for
setting
Term for
disturbance
As can be seen, the influence of disturbance due to feedback control turns out to be 1/(1 + KcKp). Here, the product KcKp is called the
overall transfer function, and if this value becomes large enough, we
2.1 . Fundamentals of Feedback Control
19
::t
.....
~
a"'
"tl
t._,
e-u
Dead time
element
Derivative
element
Tos
T.S
1
(1 + T,s)(l + T.s)
Second order
lag element
1
1+ Ts
Fist-order
lag element
Proportional
element
Integra l
element
Transfer function
Transfer
element
Table
--I
-- t
-- t
-- t
I~
0 L
II
I~
-t
k::=
Kl
Step response
=0
ofw = O
+00
-$-
w=~~w=O
Vector locus
2 1 Characteristics of transfer elements
1/ T,
---1/T,
1/ T
- 20dB/
-w
:J
dB
- 90'
dB
0
- 180'
O'
1/ T ,___..,!!dB/ dec
-- w
_.----
- 20dB /de~ w
............... 1/ T,
--w
~~dB/dec
0 -z~--w
dBI
- 90'
dB
0
( dB )
. . ,F
Bode diagram
Di sturbance OS
ontroller
------------j
I
L __
PV
Fig. 2 9 Simplified proces control loop
obtain:
1
PV=SV+ KcDS
1
DV=- KcDS
Since in process control, controller gain Kc is made large if the overall
transfer function gain is large, the above becomes
PV=SV
DV=O
In other words, if the controller gain is sufficiently large, the control

variable agrees with the desired value, and the influence of disturbance becomes 0. However, due to an actual delay in processes, stability
becomes a problem and insufficient rise in controller gain occurs frequently. The controlled variable is called stable in the case of damped
oscillation, is considered to be at stability limit in the case of fixed amplitude continuous oscillation, and is called unstable in the case of
divergent oscillation.
(b) Characteristics equation
A block diagram for a generalized feedback control system is
shown in Fig. 2.10.
Y ( s)
Fig. 2 10 Generalized feedback control loop
This system is expressed as

G(s)
( )
( )
Y s = 1+G(s)H(s)X s
The equation
2.1
21
1 + G(s)H(s)=O
(2.5)
is called the characteristics equation of the control system, and its

roots (characteristic roots) can be used as stability criterion. If we indicate these roots as s~> s 2 , ... , sn. the real part of each characteristics
root must be negative for the system to be stable* 1
Finding the roots of
(1) RouthjHurwitz stability tktermination method
and Routh and
simple,
always
not
is
equation
the above characteristics
stabilidetermining
for
methods
proposed
Hurwitz have independently
methtwo
The
equation.
characteristics
the
of
ty from the coefficients
staRouth/Hurwitz
the
as
known
now
are
and
ods have been combined
below.
described
as
method,
bility determination
If we denote the characteristic equation as a0 sn + a 1sn-J + ...
+ an_ 1s +an= 0, then the following conditions must hold for stability:
(1) a 0 ,a~> ... ,an must all exist and be positive (in the case of first and
second orders, only this condition is required)
(2) For higher orders, the following expressions must exist and be
positive:
Third order: a1a2- a0a3
Fourth order: a3(a 1a2- a 0a 3 ) - a~a 4
Fifth order: a1a2- a0 a3, (a1a2- a0 a3)(a 3a4 - a 2a 5) - (a 1a4 - a 0 a 5) 2
This method makes use of
(2) Nyquist stability determination method
the vector locus plot of the overall transfer function [G(jw)H(jw)]. If,
when incrementing from w = 0 to w = oo, the point ( -1,0) is to the left
of the vector locus, the system is stable, while if the point is to the
right of the vector locus, the system is unstable*'. Figure 2.1l(a)
shows a vector locus plot indicating stability, and Figure 2.1l(b) one indicating instability.
(c) Gain margin and phase margin
In the Nyquist stability determination method, when the vector
locus is drawn with the point ( -1,0) to its left, the closer it approaches
this point, the closer it approaches instability. Furthermore, if the
locus intersects this point at the frequency w 0 , it has reached the limit
*1 Characteristics equation and stability determination: System response is expressed
as c0+ c,es,' + ... + cne 5n', where c0, c,, ... , Cn are constants determined by initial conditions and the input signal. In order for the system to be stable, each of the factors e8 ' '
e8 ' ' , ... , e8 n' must decrease with time.
* 2 Nyquist stability criterion: In vector locus plots, the vector with origin at ( -1, 0) is called the vector of the characteristic equation. In the s-plane, when w changes from 0 to oo,
then in a clockwise direction from -oo to 0, if a root is located on the right side plane, its vector rotates around a characteristic root in a clockwise direction. This corresponds to rotation
of the vector of the characteristic equation around the point ( -1, 0).
22
.f
(- 1,0)
i{
(( - 1, 0) lies on the
right of trajectory]
[( - I. 0) lies on the
left of trajectory]
( b) In tability
Fig. 2 11
Nyquist stability criterion
of stability and the gain becomes 1 and phase delay 180 (positive feedback) with continuous oscillation of the frequency w 0 Accordingly, the
degree of stability can be indicated by the distance from the above
limit of stability.
Gain Margin (GM): The value indicating the margin between a
gain of 1 and the gain at the frequency in which the phase becomes -180 .
Phase Margin (PM): The value indicating the margin between a
phase of -180 o and the phase at the frequency in which the gain becomes 1.
Gain margin and phase margin are shown in Fig. 2.12 using a vector locus plot and a Bode diagram.
Gain margin
L------==::;-w--' - 210"
( a ) Vector locus
( b ) Bode diagram
Fig. 2 12 Gain margin and phase margin
2.2
Process Characteristics
2.2.1 Process degrees of freedom and controlled and manipulated variables

For processes which are in an equilibrium state, the number of process variables which can be changed independent of each other is called
the degrees of freedom of the process. For example, the process shown
2 .2
23
Flow rate
(a)
Pressure
Flu id process
Degree of
freedom : I
(b)
Liqu id level
process
Degree of
freedom : 2
Output temperature
of water
(c)
Heal exchange
process
Degree of
freedom : 4
Drainage
Fig. 2 13
Process and degree of freedom
in Fig. 2.13(a) characterized by the flow rate and pressure of a fluid

flowing in a conduit has one degree of freedom, while that in Fig.
2.13(b) characterized by liquid level has two degrees of freedom (inflow
and outflow), and while that in Fig. 2.13(c) characterized by water output temperature has four degrees of freedom (steam flow, steam temperature, water flow and water temperature). In other words, the degrees of freedom is equivalent to the number of available manipulated
variables in the control system.
In order to control a controlled variable, a manipulated variable obviously is required. In the example of Fig. 2.13(a), the controlled variable may be either flow rate or pressure, while the manipulated variable
must be flow rate. In this kind of a situation, it is possible to control
only one controlled variable; using the same manipulated variable to
configure more than one control loop leads to mutual interference. In
response to such a situation, override control (see Sec. 2.3) can be used
in which, depending on process conditions, any one of controlled variables is selected and controlled. In the Fig. 2.13(b) example, where the
Chap.2
Process Control
controlled variable is liquid level and the manipulated variable is either

inflow or outflow, control can be performed through either of the latter. However, if we introduce a buffer tank here so that the inflow is
no longer a manipulated variable and the degrees of freedom becomes
1, override control becomes necessary to control either outflow or liquid level. In Fig. 2.13(c), the controlled variable is the output water
temperature, and although the manipulated variable may be selected
from any of four, in practice, steam flow is made the manipulated variable. Note here that in cases where the number of controlled variables
is less than that of manipulated variables, variation in the non-selected
manipulated variables creates disturbance in regards to the selected
controlled variable, and should thus be kept constant.
2.2.2 Process characteristics
Process characteristics can be largely divided into static characteristics and dynamic characteristics, as described below.
(a) Static characteristics
This refers to steady-state characteristics when step signals of
various sizes are applied as input signals. In self-regulating processes,
the property describing the size of the controlled variable with respect
to that of the manipulated variable is called the process static gain.
Since the controlled variable and the maniplated variable are not necessarily of the same dimension, the dimension of the transfer element
would be adopted in such cases.
Self-regulation, or self stabilization, is possessed by many processes such as first-order lag systems and heat-exchangers. In contrast, an example of a process not having self-regulation is the constant outflow process shown in Fig. 2.14. Here, if the inflow=outflow,
the liquid level does not change, but if the inflow > outflow, the liquid
level continuous to increase, and inversely if the outflow > inflow, it
continues to decrease, resulting in an integrative process. In addition
<lo Ou1now
Tank area C
Fig. 2 14
Am example of the process without self-regulation

(Liquid level process with constant outflow)
2.2 Process Characteristics
25
to integrative processes, another example of non-self-regulating processes are those involving exothermic chemical reactions such as polymer reactions. In order to drive such processes, control is absolutery
required, and in comparison to self-regulating processes, it is more difficult to perform. Moreover, though many processes possess nonlinearity, there are many times in which they can be treated as linear at a
point about the equilibrium point.
(b) Dynamic characteristics
Since dynamic characteristics ave characterized by an output signal y(t) corresponding to an input signal x(t), signal relationships can be
expressed by a convolution integral in the time domain, and by a transfer function or frequency transfer function in the frequency domain.
In addition, as for ways of portraying dynamic characteristics, transient response for time and frequency response for frequency can beeffectively used as previously described.
(c) Disturbance
In processes, external influences upsetting the state of a system,
i.e., disturbance, typically exist. There are various forms, sizes and
entry points of disturbance, and they can often be represented by a
unit step signal applied to the process.
In the heat-exchanger example of Fig. 2.13(c), examples of disturbance would be changes in any of the three quantities other than the
steam flow selected as the manipulated variable: steam temperature,
water temperature, and specifically water flow corresponding to
changes in the load. Factors such as outside air temperature must also
be considered as external disturbance.
2.2.3 Process models
A process can also be viewed as one transfer element. Although
representative transfer elements and their characteristics have been
shown in Table 2.1, the characteristics of process models are investigated below.
(a) Proportional element
In the example of Fig. 2.13(a), the fluid flowing in the conduit can
be regarded as a proportional element if we ignore the delay caused by
inertia. The amount of flow rate change K occurring when the valve is
opened or closed by a unit amount is referred to as proportional gain
(Table 2.1(a)).
(b) Dead-time element
The conveyor shown in Fig. 2.15 possesses a dead-time element. If
we specify the distance between the hopper outlet and the weighing
scale as l, and the speed of the conveyor belt as v, the time from when
certain particles are released up to the point when they reach the
26
Dead time L =
..!...
v
Fig. 2 15 An example of dead time process

(Belt conveyer)
weighing scale is L = lfv, where L is the symbol for dead time. The Laplace transform for dead time L is e-Ls .
As can be seen from the Bode diagram in Table 2.1(d) for
dead-time elements, the gain is always constant and is unrelated to frequency. The phase, on the other hand, is delayed infinitely with increase in w. It can thus be seen, without considering gain and phase
margin, that control of a dead-time element is difficult.
(c) Integral element
The liquid level system in Fig. 2.14 consisting of the constant outflow tank is an integral element process. If we designate the inflow as
Q;, the outflow as q 0 , the liquid level as y(t), and the tank cross-sectional area as C, we obtain the expression
If we let
Qo
be constant, and
y(t)=
Q;- Qo =x(t),
we then get
~ jx(t)dt
where C= T (Time constant).

If we now express the above as a transfer function, we get
Xhl=_1
XfsT Ts
As shown by the Bode diagram in Table 2.1(e), the gain in this
case is 1 at a frequency where wT= 1, becoming a -20 dB/dec* negative slope straight line; moreover, it is infinite at w = 0. The phase has
a 90 delay and is unrelated to the frequency.
* -20 dB/dec: this signifies -20 dB (deci-Bel) per decade, or in other words, if the fre
quency is increased by 10 times, the gain turns out to be 1/10. Likewise, - 6dB/oct
siginifies -6 dB per octave, or if the frequency is doubled, the gain becomes 1/2.
2.2
27
(d) First-order lag element
The liquid level system shown in Fig. 2.16 characterized by natural down flow is assumed to be a first-order lag element process.
--l.)
Infl ow
q,( = z ( t))
II\
Liquid r
Level
Y (t )
Tank a~
Outflow
Fig. 2 16 An example of first-order lag process

(Liquid level process)
It is expressed as
=q -q
C dy(t)
dt
0
I
where inflow qt =x(t).

If we now designate the outflow resistance as R and let the
amount of outflow be proportional to the liquid level, then
q o_JIJJl
- R
cd~~t) =x(t)- y~)

If we let CR = T(Time constant), then we get
d~/) +y(t)=Rx(t)
Expressing the above as a transfer element:
~=l+RTs
The step response of a first-order lag element is shown in Fig.
2.17. At t=O, if we take a tangent to the response curve and extend it
to the curve's final value, the corresponding time interval is called the
time constant. In addition, the value of the curve at time constant T is
63.2% of the curve's final value. As shown by the Bode diagram in Table 2.1(b), the gain is characterized by a crossover point at the frequency where wT= 1. At frequencies below this point , the gain is constant
along the asymptotic line, while at frequencies above this point, the
gain drops off to the right as -20 dB/dec. The gain at the crossover
point is -3 dB; the phase at the crossover point frequency has a 45 de28
Chap.2
Process Control
Ti me t
Fig. 2 17 Step response for first-order lag
lay, and approaches a 90 delay with increasing frequency.

(e) Second-order lag element
As shown in Fig. 2.18(a), a second-order lag element can be formed by connecting two first-order lag elements in a two-stage sequential fasion, and its transfer function is simply the product of the transfer functions for each first-order lag element. Such a sequential structure, in which following stages have no influence over previous stages,
is called a cascade connection.
~
~-~~ L~
n
-
(a) An example of non-interacting

second-order lag process
(Dependent two tanks in series)
( b ) Block diagram
Fig. 2 18 Cascade connection of tanks
The transfer function of the first - stage-tank has been previously

determined as
Yihl
XTsT
29
Moreover, since the input Xls) of the second-stage tank is the outflow of the first-stage tank, X2(s)= Y 1 (s)/R 1 The entire transfer function can then be expressed as
_Y(U
XTsY
The block diagram for this process is shown in Fig. 2.18(b).
The step response for a second-order lag system is shown in Table
2.1(c). For lag systems of second or higher orders, the slope of the response curve at t= 0 is zero, and since the response curve has an inflection point, the delay can no longer be expressed by a single time constant. Looking at the gain diagrams, there are crossover points at the
1/TI and 1/T2 frequencies, and at frequencies above the higher crossover point, the gain becomes a 2 x (- 20 dB/dec) straight line. The
phase is delayed up to a maximum of 2 x 90 = 180.
If two tanks are connected as shown in Fig. 2.19, the liquid level
of the second-stage tank here does influence the first-tank, which
means a non-cascade connection. The transfer function here becomes
IM2_
XTsY
The new time constant R 2 C1 in this second- order lag system indicates
interference between the two tanks.
PtM
Fig. 2 19
An example of interacting second-order lag process

(Interacting two tanks in series)
(f) Higher-order lag systems
A system formed by connecting multiple first-order lag systems

sequentially is called a higher-order lag system. Step responses for cascade- connected first-order lag systems having identical time constants
is shown in Fig. 2.20. As can be seen, as the order of the lag system increases, the time until the rising edge begins lengthens, indicating similar characteristics as the addition of a dead-time element. Figure 2.21
illustrates a characteristic of higher- order lag systems in which a tangent taken at the inflection point is often used to approximate systems
with dead time and a time constant. Here, the dead time is called the
equivalent dead time, and the time constant the equivalent time constant . Expressing this as a transfer function, we get
30
Chap .2
Process Control
Fig. 2 20
10
12
14
Time
Fig. 2 21
16
ti T
Step response for II/( I + Ts)l"
Simulation of higher-order lag response
Ke-Ls
G(s)= 1+ Ts
The value L/T which increases as the order of the lag system increases
is used as an indicator of the relative difficulty of control. The values
of equivalent dead times and equivalent time constants for the first
five higher-order lag systems with identical time constants are shown
in Table 2.2.
Table 2.2 Equivalent dead time and equivalent time constant
Order of lags
First-order lag
Second-order lag
Third-order lag
Fourth-order lag
Fifth-order lag
Equivalent
dead time
0
0.28 T
0.81 T
1.43 T
2.10 T
Equivalent
time constant
T
2.7T
3.7 T
4.5T
5.1 T
Equivalent dead
time/Equivalent time
constant
0
0.103
0.218
0.321
0.410
In regards to the frequency response, if T 1 > T 2 > ... > T n. then the
gain asymptote for 0 < w < 1/T1 is 0 dB/dec, for 1/T1 < w < 1/T2 it is - 20
dB/dec, and for 1/Tn < w it is - 20n (dB/dec). For the phase, successive
31
delays of up to n x 90 occur.
2.3 Control Formats for Various Types of Processes
2.3.1 Single loop control systems
(a) On-off control
The on-off controller shown in Fig. 2.22(a) is called a two-position
controller since its output is turned on or off in response to the polarity of the deviation. In addition, since the manipulated variable at on
time or off time will be either too big or too small for maintaining the
controlled variable at the desired value, cycling occurs.
Manipulated m1
variable
-e
ffll
+e
Deviation
-e
0
ffl2
( a) Theoretical ON-OFF
controller
+e
Differential
gap
( b) Practical ON-OFF
controller
( c ) Three-position
controller
Fig. 2 22 ON-OFF controllers
In actual on-off controllers, however, a differential gap (hysteresis) is employed, as shown in Fig. 2.22(b). Without such a differential
gap, the controller will rapidly turn on and off continuously in the area
of the desired value, and the lifetime of the on-off mechanism will be
greatly shortened. In some cases, as in a bimetal thermostat, a differential gap is inherent, while in others, such as through an on-off controller, a differential gap may be purposely implemented. In Fig. 2.22(c), a
three-step controller is shown, in which an intermediate step has been
designed in.
If there is a differential gap, the cycling period will become longer, and the amplitude as well will become larger. In addition, in an
on-off control system, if the value of the manipulated variable just between that at on time and off time does not bring the controlled variable to the desired value, the average value will then deviate from the
desired value, and offset (see Proportional control below) results.
(b) P ID control
(1) Proportional control
Control action in which the size of output is
propotional to that of the input is called proportional action (P-action),
and is expressed by the basic equation:
(2.6)
Y(s)=KcX(s)
32
Chap.2
Process Control
where Kc is the proportional gain. In actual controllers, however, the

following property called the proportional band (PB) is used instead of
Kc:
PB=
kc
X 100
(%)
In other words, as shown in Fig. 2.23, the proportional band indicates in percentages the width of change in input necessary to produce
an effective change in output from 0 to 100%.
Process variable (PV)
Fig. 2 23 Proportional band
Note that the transient response and frequency response of proportional action is the same as that for the proportional element shown in
Table 2.l(a), in which proportional action is a variable gain proportional element.
For step response in a proportional control system, controlled
deviation whose value after a sufficient amount of time becomes fixed
is called an offset or steady-state deviation. In proportional control,
since the controller output value takes on a specific value corresponding to the control deviation, offset will occur as long as the controlled
variable resulting from the manipulated variable (output value) when
the deviation is zero is not equal to the desired value.
In Eq. (2.6) above, since the output becomes 0 with an input of 0, a
bias M is addid as follows:
Y(s)=KcX(s)+M
If this bias is made variable, controller output can then be changed

regardless of the control deviation and the controlled variable can be
33
manually reset to the desired value. This manual reset function is normally provided in proportional controllers.
(c) Integral action
In integral action, also called !-action or reset action, the size of
the output is proportional to the time integral value of the input, as follows:
1 X(s)
Y(s) =,..J.IS
where T1 is the reset time.

Since the output continues to be increased or decreased until there
is no more deviation, the offset generated in proportional control can
be eliminated-thus the term reset action. The strength of integral action is indicated by the reset time: the shorter the reset time, the
stronger the integral action. Time units normally employed are minutes or seconds.
The transient response and frequency response of integral action
is the same as that for the integral element shown in Table 2.1(e). At
w = 0, the theoretical gain becomes oo. Although in proportional control, the offset decreases with rise in proportional gain, the gain increase across the entire frequency band will cause instability. Accordingly, it is considered that in integral action, offset can be eliminated
by raising the gain only in the lower frequency band. On the other
hand, since a phase delay of 90 across the entire frequency band is
not preferable to control stability, integral action is normally combined
with proportional action (PI-action) as follows:
Y(s)=Kc( 1+ i.s )x(s)
The step response and Bode diagram for PI-action are shown in
Figs. 2.24 and 2.25. In the step response, the time up to the point when
the output due only to proportional action and that due only to integral
Step input
Process variable PV 1'--<---'-----
Fig. 2 24 Step response for proportional plus integral action
34
~Od~;:,_
'~
Gaint
Kc[dB]
1/T,
--(J)
o-------==----
Phasej45.
____ _ _ , - - -
-90-~==-----------
Fig. 2 25
Bode diagram for proportional plus integral action
action become equal is the reset time. From the phase diagram, it can
be seen that at w = 1/T1 the phase delay decreases to 45, and it approaches zero with increasing frequency.
Reset Windup:
In integral action, since the time integral value for the deviation is
output, if a state having deviation continues for a long period of time,
the output due to integral action becomes saturated, similar to a control halt period in batch control. This situation is called reset windup
(reset action saturation). In Fig. 2.26, control for a reactor batch process is shown. After the process is started, the deviation continues to
decrease with the rise in reactor temperature, and eventually becomes
zero, while the output due to integration remains saturated. Next, as
the deviation polarity begins to change, the output likewise begins to
decrease. However, since in general, controller output will exceed the
0 to 100% range to some extent (to ensure complete closing of the controller valve), a further delay will occur until the control valve begins
PV
Batch start
-time
Fig. 2 26 Batch process control
35
to function. Consequently, a large reset overshoot occurs. In response

to this situation, batch control has come to use controllers with a batch
switch for switching output when needed as well as controllers performing proportional control in which integral action can be temporarily stopped.
(d) Derivative action
In derivative action, also called D-action or rate action, the size of

the output is proportional to the time derivative of the input value (the
rate of change of input), as follows:
Y(s)= TosX(s)
where T 0 is the rate time.

As derivative action is characterized by output which is proportional to the speed of change in input, it cannot be used exclusively by itself but must be combined with proportional action or proportional and
integral action. The ramp response and Bode diagram for combined proportional and derivative action are shown in Figs. 2.27 and 2.28. In the
ramp response, the time up to the point where the output due to proportional action and that due to derivative action become equal is the
rate time, and the longer the rate time the stronger the derivative ac-
Derivative time
Fig. 2 27
Ramp response for proportional plus derivative action
Gaint
_ _ _ __
--"'
Kc[dB]-----~-t:.__
1/To
90"---------=:;;;;---
Phaset
Fig. 2 28
36
4~:--- ~~
-"'
Bode diagram for proportional plus derivative action
Chap.2
Process Control
tion. The time units normally used are minutes or seconds.

As shown by the phase diagram, the phase advances to 45 at
w = 1/T0 , and advances to 90 with increasing frequency. Consequently,
the process phase delay can be compensated for and control stability
improved. On the other hand, above the crossover frequency, the gain
increases along the 20 dB/dec asymptote. As a result, the output can
undesirably change significantly in response to a small change in the
setpoint value, measured value, or noise. Thus, actual controllers employ an incomplete derivative, in which a first-order lag is added to a
derivative system. Extracting the differential term, we get
where the first-order lag time constant Td is called the derivative time
constant and Tv/Td is the derivative amplitude. A value of about 10 is
normally chosen for the derivative amplitude.
The step response for incomplete derivative action and the Bode
diagram for combined proportional and incomplete derivative action
are shown in Figs. 2.29 and 2.30. In the step response, the ratio bePV
a
Step input
MV
n-a
Derivative time constant

n=
~: (Derivative amplitude)
rdec
Fig. 2 29 Step response for incomplete derivative

nKc[dB)------------,.....---Gaint
Kc[dB)-------=~-~---:-!=::-----
Phase
Fig. 2 30
1/Tn
I/Td - w
90'------------
o-......,~=---------""'"-
Bode diagram for proportional plus incomplete derivative
37
tween the amplitude of the input step signal and the maximum amplitude obtained from derivative action is the derivative amplitude, and
the time constant indicated by the response curve is the derivative
time constant. Note also that as shown by the Bode diagram, the phase
never reaches 90, though the gain levels out at a multiple of the derivative amplitude.
PV Derivative: Up to now, although we have been tacitly assuming a controlled deviation for X(s), a configuration in which the measured value (PV) is applied to the input signal is called a PV Derivative
or a Derivative Ahead. A block diagram for PV Derivative type PID is
shown in Fig. 2.31. In regards to disturbance, this type of configuration functions in the same way as the deviation derivative type, while
the output, in response to a change in the setpoint value (SV), does not
change suddenly; this enables changes in the setpoint value to be performed easily.
Controller
Fig. 2 31
Block diagram for derivative ahead
In addition, digital controllers employ algorithms in which the

measured value can be applied not only to the derivative term but to
the proportional term as well. Moreover, a PID with two degrees of
freedom is adopted in which effectiveness of the above is adjustable. In
regards to a change in the setpoint value, optimal algorithms in conjunction with the control mode (auto or cascade) are selected. (Refer to
Subsec. 4.2.3 (b) for more details.)
(e) P ID action
The combined three type of actions, proportional, integral and derivative, are referred to as PID action. The basic equation for PID action is expressed as follows:
(2.7)
An example of a step response for a PID controller is shown in Fig.

2.32.
38
Chap.2
Process Control
MV
Fig. 2 32 Step response of PID controller
2 .3 .2 Compound loop control system

(a) Cascade control
In a feedback control system, the control which is performed by
having the output signal from one control device (primary controller)
determines the desired value of another control device (secondary con-
( a ) Single loop control sy tern
( b ) Cascade loop control ystem
Fig. 2 33 Furnace control system
(!) Desired value

Primary controller
Secondary controller
@ Disturbance
Secondary process
Primary process
Process
Controlled variable
Fig. 2 34 Block diagram for cascade control system
39
troller) is referred to as cascade control. Figure 2.33 shows examples

of a furnace in a single-loop control system and one in a cascade-loop
control system. Block diagrams for a cascade control system are shown
in Fig. 2.34. In particular, Figure 2.34(b) is a modification of (a), in
which the secondary control system becomes one part of the process,
as seen from the primary control system.
The purpose for configuring a cascade control system is to eliminate, through use of a secondary controller, the influence of disturbance on the primary process, which enters through the secondary control loop. In this furnace example, the fluctuation in the amount of
fuel is suppressed by the secondary control system. In addition, since
the phase delay of the secondary process is improved by the secondary
control system, the response of the primary control system becomes effectively faster and the settling time shorter. Moreover, the non-linearity of the secondary process decreases.
In order to construct an effective cascade control system, the period of natural frequency of the secondary control system shoud be 1/3
or less than that of the primary control system. Figure 2.35 shows an
example of temperature-temperature cascade control for a polymerization reactor. Since there is not much difference between the primary
and secondary periods here compared with temperature-flow control,
the secondary controller, tuned by mainly proportional control action,
makes the response of the control system as fast as possible.
Hot water
(split range)
Fig. 2 35 Cascade control of polymerization reactor
If the secondary control system possesses non-linearity, the loop

transfer function gain in the primary control system will undesirably
fluctuate. For example, if the flow rate is measured with an orifice,
the transmitter output signal is proportional to the second power of
the flow rate. Thus, since convertors and transmitters are located on
the feedback side of the control loop, the amount of feedback at small
40
Chap.2
Process Control
Flow F.
Fig. 2 36 Concept of ratio control
flow rate decreases, and the gain of the secondary control system increases. Accordingly, when the flow rate at times like startup is small,
the control system will tend towards instable; a square-root extractor
device to provide linearity is required.
(b) Ratio control
Control which attempts to preserve a ratio relationship between
two or more variables is called ratio control. A typical example would
be air-fuel ratio control in a combustion furnace, in. which the ratio between two flow values is controlled. The fundamental concept of ratio
control is shown in Fig. 2.36. As shown, the ratio between two flow
values is determined by a divider and passed to a controller in which
the desired ratio has been set. The controller in turn manipulates one
or the other flow value.
However, in this method, since the divider is situated within the
loop, if we place the control valve on the flow line corresponding to
the ratio numerator, the loop gain due to change in PV corresponding
to the denominator also changes, and alternatively if we place the control valve on the flow line corresponding to the denominator, the loop
gain due to change in MV changes creating an undesirable non-linearity. In actuality, the configuration shown in Fig. 2.37 is employed.
Here, one of the flow values is multiplied by the ratio as set by the
manual ratio setting device, and the result is used as the setpoint value
for the controller governing the other flow rate. In contrast to this se-
Fig. 2 37
Rati o control (series)
41
Fig. 2. 38 Ratio control (Parallel)
Fig. 2 39 Remote ratio setting
ries setting format, a parallel setting format is shown in Fig. 2.38, in

which the flow delay on the controlled side in a series setting can be
eliminated.
The setting range of a ratio set station for an analog unit is about
0.3 to 3.0. When using orifices or flow meters having square-law characteristics in which the square of the set ratio will become the actual
ratio, a scale range from about 0.6 to 1. 7 is used. In addition, ratio set
stations incorporating microprocessors provide a wide setting range
from about 0.0 to 8.0, as well as a square-root extraction function.
Note that the ratio to be set on the ratio set station is not an actual
flow rate ratio but that when each flow range is supposed to be 0 to
100%.
For ratio set stations employing microprocessors, external ratio
setting is possible. In the example of Fig. 2.39, where the calorie value
needs to be constant, the ratio in the set station is being remotely set.
In addition to ratio control of instantaneous flow, an integrating
flow ratio control system can be used, called blending control. Here,
flow meters such as PD meters and turbine meters which can accurately generate pulses are used (see Subsec. 4.2.5).
(c) Selector control
This form of control includes selection of measured values as well as selection of output from either of two controllers sharing a common manipulated variable, and having different con42
Chap.2
Process Control
(a)
(j) An alyzer
(b)
Fig. 2 40 Selective control of process variables
Liquid level setpoint
Fig. 2 41
Override control of buffer tank
~ 10~
~
105fo
PV
.,. 50
;
s
:]~-
~
~
~
"
~v
-Time
'C.J
PV
u:
OL----------------------T-ime
Fig. 2 42 Override control
trolled variables.
Figure 2.40 shows examples of measured value selection control.
In (a), the object is to protect the furnace, while in (b), the object is to
43
avoid the influence of analyzer failure by incorporating redundant analyzers.

Figure 2.41 shows an example of buffer tank override control. The
object here is to maintain a constant feed to the next process as long
as the tank is not empty. If the liquid level is above the setpoint value,
the flow control system is selected and the feed amount is kept constant, while if the liquid level falls to the setpoint value, the liquid
level control system is selected, and control is performed to prevent
the liquid level from dropping below the setpoint value. This process is
illustrated in Fig. 2.42. In this example, a "reverse operation" (air to
open) control valve is used, so that the flow controller is to be "reverse
action" and the liquid level controller "direct action," and the selector
becomes a low selector (if the control valve is of direct operation, the
selector is then a high selector).
Since there is deviation in the non-selected control loop, reset windup will occur due to integral action of the PI controller. To prevent
this, an external feedback type of controller can be used, as shown in
Fig. 2.43. As indicated by the following equations, the integral term of
the non-selected controller can be substituted by the output from the
selected controller, and the non-selected controller can thus function
as a proportional controller.
Selected controller: YF(s)=-
Kc( 1+ is )xF(s)
Non-selected controller: YL(s)=KcXL(s)+ YF(s)
Manipulate signal
External feedback
Fig. 2 43
External feedback for override control
Now, given that the flow controller is selected, since a plus deviation (SV < PV) arises in the direct-action liquid level controller, YL
(s) > YF(s). However, if the amount of intake flow decreases and the liquid level drops to the setpoint value, the deviation for the liquid level
controller becomes 0, so that YL(s) = YF(s). Moreover, if the liquid level
drops below the setpoint value, the polarity of the deviation reverses,
44
Chap.2
Process Control
(a)----(b)--1---(c)
Responses
(d)
(e)
I[VV""
Ideal response
Quick response but large overshoot
.--_,.-........
.....--- Compromise between response speed
_y
and overshoot
CO?"::
No overshoot but slow response
Fig. 2 44 Control responses for setpoint change
i.e., YL(s)<YF(s), and a "bumpless" transfer is made to the liquid level

controller.
2.4
Optimal Adjustment of Control Systems
(a) Control system evaluation

Control systems are evaluated by such characteristics as stability,
speed of response and size of offset. In Fig. 2.44, four examples of responses to a step change in the setpoint value are shown. The step
change is shown in (a), and the four responses in (b) through (e). Although (b) is the ideal response, it cannot actually be realized. Response (d), which lies between responses (c) and (e), is usually the most
sought after; evaluation criteria, however, may differ depending on
the type of process.
(b) Evaluation method
Response waveforms and their related terminology are shown in
Fig. 2.45. Although there are various criteria for evaluating such waveforms, they can be roughly divided into the following three types:
(1) Amplitude damping ratio
This is a value indicating the way in
which the response waveform undergoes damping, and is determined
as shown in terminology example 5.
(2) Control area
Three ways of integrating the control area are available, as shown in terminology example 4.
(3) Transient overshoot and response time
Since the transient overshoot (terminology example 1) and the response time (terminology example 7) possess opposing properties, response time for a given overshoot can be used as evaluation criteria.
(c) Effects of P, I and D
Figure 2.46 shows the resultant control responses when applying
external disturbance in the form of a step signal, equivalent to 30% of
an input signal, to a P, PI, PD or PID controlled third-order lag system having three first-order lags with a 10 second time constant and a
2.4
45
.!! ><
""'"o-><
="
... .D
C-"'
0 ...
""'
2 :0"
c0 E
u
"'
(j) Specified band
u~
>
( a ) For set point change
Time
( d ) Oscillatory
';;[X 100
[%]
2.
3.
2 to 5 % of span
Specified band
Time required for the output to enter and remain
Settling time
within a specified band centered
4.
Control area
5.
Settling time
Period
Time
( c) Non-oscillatory
Terminology : 1. Overshoot
( b ) For disturbance
Shaded portion of figures (c) and (d)
Integral of absolute value of error
A,=}~ix(t) -x(oo)
Integral of square of error
A,=}~{x(t)-x(oo)}'dt
Integral of time multiplied by absolute error
A 3 =}~Kix(t) -x(oo) ldt
ldt
Amplitude damping ratio
6.
Period
7.
Response time
(J)
Fig. 2 45 Terminology for transient response
gain of 2. From these results, the indivisual effects of P, I and D actions can be summarized as follows.
(1) Effects of proportional action: As the proportional band narrows,
-Offset decreases
-Amplitude damping ratio becomes large (oscillatory), eventually diverging
-Oscillation period becomes short
(2) Effects of integral action: As the reset time shortens,
-Offset disappears, and the time until reset becomes short
-Amplitude damping ratio becomes large, eventually diverging
(3) Effects of derivative action: As the rate time lengthens,
-Amount of offset remains unchanged
-Amplitude damping ratio becomes small, and then large again
-Oscillation period becomes short
46
Chap.2
Process Control
"
PB=200[%)
30
~~
c
~.._~
"> "
.!;!
""' ""
""
Oil
.0:
<.>
<=0
"
~,---,
100%
20
""' "
0 "
"
b
"u
.!;!
PB=70[%)
Oil
.0:
<.>
10
Setpoint ...
"~ 20
>
10
Setpoint ...
0
0
17% 1
Time [min)
Time [min)
-10
T1 =0.5 [min)(constant)
TD=0.1 [min)(constant)
( d ) PID Control with varios PB
( a) P Control with various PB
PB =30 [%) (constant)

T 1= Z[min) To=O.l [min)(constant)
0.5
0.2
-10
Time [min)
Time [min]
(e) PID Control with various T 1
( b ) PI Control with various T 1
PB =30 [%](constant)
PB=30[%) (constant)
T1 =0.5 [min](constant)
To= 0 [min]
3
0.5
Time [min]
Time [min)
(f) PID Control with various T 0
( c ) PD Control with various T 0
..
"
:0
;::: ~ 20
;>~
P Control
PI Control
PD Control
PID Control
~.
] ~
PB=50[%]
PB=70[%) T1 =0.5 [min]
PB=20[%) To=0.05 [min)
PB=30[%) T1 =0.5 [min], To=0.1[min)
g~
"
0
Setpoint ... ~L0-~~:;:t:::::::~!I;;~~-='""F=--
Time [min)
( g ) Control with optimum settings
Fig. 2 46
2.4
Responses for PID control (for disturbance)
47
From the above effects and from the comparison of PID values for
control under optimal conditions in Fig. 2.46(g), we see that for proportional action, offset remains; if we add in integral action, however we
can eliminate the offset, although stability is degraded to some extent.
If we then add in derivative action, stability increases, and the response becomes faster as well.
(d) Optimal adjustment
As control responses vary depending on the above PID values, determining those PID values which satisfy evaluation criteria for a particular process is referred to as optimal adjustment, optimal setting, or
tuning. Two main methods are used in this regard:
(1) Determination from closed loop characteristics: PID values are
determined from the response in a closed control loop, with amplitude damping ratio as common evaluation criteria.
(2) Determination from process characteristics: PID values are determined from investigating the process response in an open control
loop, with control area or response time frequently used as evaluation criteria.
In the optimal setting methods described below, derivative and proportional action are assumed to operate on deviation for either case of
disturbance or setpoint change. When PD derivative or PID with two
degrees of freedom is employed, the optimal setting for the setpoint
change may differ.
(1) Determination method from closed-loop characteristics
The most
well-known method is the Ziegler-Nichols ultimate-sensitivity method. Under proportional control conditions, as the proportional band is
gradually narrowed from a sufficiently large value, the response determines 1) the proportional band causing fixed-amplitude continuous
oscillation (ultimate proportional band PBu) and 2) the corresponding
oscillation period (ultimate oscillation period Pu). From these values,
PID values can then be calculated using Table 2.3. Although these PID
values in effect provide 25% damping with respect to disturbance or a
change in the setpoint value, some adjustment is necessary depending
on the process.
Table 2.3 Ultimate sensitivity method
by Ziegler-Nichols
Control
PB
T,
p
PI
2 PBu
2.2PBu
1.7 PBu
0.83 Pu
0.5 Pu
PID
48
0.125 Pu
Chap.2
Process Control
"'
"'"'~
-....
::!
....
....
"'~
"'....
::!
;::"
I':>
~-
.s-
~
....
l\)
Chien
Hrones
Reswick
(Notes)
Yoshikawa
L/T"2':.1
{L/Ts1
L/T"2':.1
PID {L/Ts1
PI
PI
PID
PI
PID
PI
PID
PI
PID
PI
PID
PI
PID
Control mode
To
0.125(T+L)
0.25 L
0.42 L
2L
T+L
2L
0.5(T+L)
L
0.4 L
2.4 L
2.3 L
0.47 L
1.35 T
4L
0.5 L
T
T
0.45 L
1.2 T
3.3 L
2.2 L
0.5 L
3.3 L
2L
Optimum setting
T,
100 KPL/T
167 KPL/(T+L)
250 KPL/(T + 2 L)
133 KPL/(T + Y3 )L)
200 KpLf(T + L)
143 KPL/T
143 KPL/T
83 KPL/T
333 KPL/T
167 KPL/T
105 KPL/T
143 KPL/T
167 KPL/T
105 KpLfT
333 KPL/T
286 KPL/T
167 KPL/T
110 KPL/T
110 KPL/T
77 KPL/T
100 KPL/T
110 KPL/T
83 KPL/T
PB
., Type A: for setpoint change; Type B: for disturbance.

T, L, KP: obtained from transient response.
J Refer to Fig. 2.45.
Chien
Hrones
Reswick
Fujii
Chien
Hrones
Res wick
Chien
Hrones
Reswick
A,B
Type*'
Takahashi
Ziegler
Nichols
(1942)
Proposer
Table 2.4 Transient response method (Trans. ASME, etc.)
Minimum control area
20 % overshoot and
minimum response time
No overshoot and
20 % overshoot and
No overshoot and
Minimum control area
25% damping
Optimum condition*'
(2) Determination method from process characteristics

(i) Transient response method: If we apply a step signal to a process,
then as shown in Fig. 2.47, we can measure the process gain Kp, the
equivalent time constant T, and the equivalent dead time L. From
these values, PID values can then be calculated. Since control valves,
detectors, transmitters and the like are used as input-output devices to
measure process transient response, their characteristics are inevitably
included in the process characteristics. Table 2.4 shows various transient response methods.
MV
~I
-I
-I
Fig. 2 47 Step response of process
(ii) Frequency response method: If the frequency characteristics of a pro-
cess are known, the frequency characteristics of a loop transfer function including a controller can be easily obtained. Since the frequency
characteristics of the controller are determined by PID setting values,
it is possible to set PID values so that, for instance, gain margin and
phase margin become favorable values.
(e) P ID for representative control systems
Process control systems such as in temperature control and pressure control each have their own particular characteristics, which is
necessarily reflected in their PID settings. Table 2.5 list PID characteristics for some representative control systems.
Table 2.5 PID settings for common process control loops (Shinsky 5 >)
Property
Flow and
liquid pressure
100-500
*50-200
Integral
Essential
Derivative No
PB(%)
Liquid
level
Temperature and
vapor pressure
Composition
0-5
5-50
10-100
100-1000
Unnecessary
Unnecessary
Seldom Yes
Essential
No
Gas
pressure
Essential
If possible
* Liquid pressure
2.5 Sequential Control
Although the material discussed up to now has been concerned
50
Chap.2
Process Control
with continuous control processes, there are many cases in which

sequential control forms a part of process control along with continuous control. This section describes the meaning and elements of
sequential control.
2.5.1 Meaning of "sequential control"
Sequential control is defined as "control which successively advances to each control level according to a previously determined order." Examples of sequential control in our daily life are fully automatic washing machines and elevators, while in actual process control,
some examples are polymerization and crystallization batch processes,
and cleaning of a water supply filter bed. In addition, even in continuous processes, sequential control must be performed at times like startup and shutdown, regardless of whether the operation is done manually or automatically.
2.5.2 Types of sequential control
Sequential control is devided into the following two types:
(1) Program control* (process control)
In this case, control proceeds
according to a previously determined program (from process to process).
(2) Conditional control (monitor control)
Here, internal and external
conditions are monitored, and control is performed in response to these
conditions.
An example of program control would be the fully automatic washing machine mentioned above. After placing laundry and soap into the
washing machine and then opening the water faucet and pushing the
start button, the processes of water filling, washing, rinsing and draining are all performed sequentially according to a previously set program. This program advances in response to water level, timer and
counter signals, and finally ends by sounding a wash-over chime.
For conditional control, the elevator is a fitting example. External
conditions would be calling for an elevator and specifying the desired
floor, while internal conditions would be those concerned with the
elevator cage, i.e., the current floor, cage stopped or moving, existence
of passengers, etc. In addition, in the event of a multi-elevator system, the cage conditions of another elevator must also be considered,
and in response to all of these conditions, an elevator can then be
driven up or down.
There are also many cases in which sequential control as a form of
Program control: In addition to program control narrowly defined as "control in

which the desired value undergoes a previously established change," we also have this
more broader form of program control as one type of sequential control.
51
process control will be a combination of both program control and conditional control.
2.5.3 Sequential control description
The following five formats are commonly used for describing the
actions of sequential control.
(1) Relay circuit: Since sequential control relay circuits were traditionally realized by such elements as relays and timers, relay circuit diagrams are still used as a descriptive aid.
(2) Logic circuit: This form of description makes use of logic signals,
in particular, those circuit signals as standardized by ]IS and
MIL.
(3) Flowchart: This description format makes use of the flowchart
concept as applied to computer programming.
(4) Time chart: In this kind of chart, the mutual interaction between
sequential control devices is shown according to the passage of
time.
(5) Decision table: This format shows the operations corresponding
to different conditions in a matrix table fashion.
Each of the above description formats has its advantages and disadvantages. In general, flowchart and time chart formats are appropriate for program control, and relay circuit and logic circuit formats are
appropriate for conditional control, while decision tables can be applied
to both. In fact, it is common to adopt a particular description format
according to the program format of the sequential control device being
used.
.-----o~ PB1 Start push button
l-------o---~1 ON to repeat
Sequence repetition switch
v1
- 3 5 1 (ON when VI closed)

.
From prevtous
process
ON at level high limit
V2
u
I
ON at level low limit
-~------~~
ON at temperature. TA
low limit
To the next process
Fig. 2 48 A simple cooling process
52
Chap.2
Process Control
Next, taking as an example the simple cooling process shown in

Fig. 2.48, the relay circuit, flowchart and decision table description formats will be described in more detail.
The cooling process is performed as follows:
(1) Sequential control start: If no liquid is in the tank, i.e., lower
level limit switch LA 2 is on, sequential control can begin by pushing the PB 1 start button.
(2) Tank filling operation: After confirming that solenoid valve V 3
is closed, solenoid valve V 1 is opened and the tank is filled with
liquid from the previous process until upper level limit switch
LA 1 turns on, at which time solenoid valve V 1 is closed. In the
event that solenoid valve V 3 is initially open, a valve closing
operation is first performed before proceeding as described
above.
(3) Cooling operation: After completing the tank filling operation, solenoid valve V 2 is opened and cooling is performed until lower
temperature limit switch T A turns on, at which time solenoid
valve V 2 is closed.
(4) Liquid transfer operation: After completing the cooling operation, solenoid valve V 3 is opened, and cooled liquid is supplied to
the next process until lower level limit switch LA 2 turns on, at
which time solenoid valve V 3 is closed again.
(5) Sequential control repeat or termination: If, at the point of liquid
transfer completion, repeat SW 1 is on, sequential control is repeated again from the tank filling operation. If repeat SW 1 is
off, however, sequential control is terminated.
Operation conditions and corresponding actions are shown for this
cooling sequential control in Table 2.6. Since this sequential control is
a program control type, if we implemented it with relay circuits, we
Table 2.6 Operations for cooling process
Operation
Description
Conditions for operation
Filling tank
Open V 1
PB 1 ON, LA 2 ON and LS 3 ON
or SW 1 ON, LA 2 ON and LS 3 ON
Cooling
Close V 1,
Open V2
LA10N
Discharging
liquid
Close V 2,
Open V3
TAON
Liquid
discharge end
Close V3
LA20N
53
Relay power supply
Rl
LS3
.---~--~==~-----4
R2
R3
~--~
~-----------1
Fig. 2 49 Relay ladder diagram for cooling process
would need process memory relays R 1 to R 4 for recording the occurrence of each process, as shown in Fig. 2.49. In addition, as shown in
Fig. 2.50, if we describe the same with the use of a flowchart, sequential control can be easily expressed in terms of each separate process.
Next, let's try expressing the above in terms of a decision table. A
typical format for a decision table is shown in Fig. 2.51. Conditions are
listed in the upper half of the table and actions in the lower half, with
the right side of both halves divided into rule number columns. To establish a certain condition, Y is entered to indicate that sequence elements should be ON, and N for OFF; if the condition has nothing to do
with the elements, the entry is left blank. Likewise for action description, Y is indicated if an ON action signal is output, and N is indicated
if an OFF action signal is output, if no signal is output, the entry is
left blank. Figure 2.52 shows a relay circuit described with such a decision table. In this figure, the symbols in the SYMBOL column are
sequential control elements used by decision tables. Process memory relays R 1 toR 4 are substituted for internal switches which are sequential control elements. Based on the flowchart described in Fig. 2.50,
the sequence is indicated in the decision table shown in Fig. 2.53. As
can be seen, the internal switches for process recording are substituted
by step numbers and step transfer conditions.
54
~-;:::::==:;:t--~- ~
Fig. 2 50
Flow chart for cooling process
55
(
(
32
c
E
0(,)
"'
32
(D
(D Table number
Rule number
@Symbol
@ (Step number)
Condition signals
Condition rules
(J) Action signals
Action rules
(Next step number)
I
I
r-
f--(J)
f.-'
.....__ 32 ---------------Row -----------Fig. 2 51 An example of decision table
T AB L E NO.
Ioo I
01
02
II
12
S T 002
DES C Rif'TION
I CIJ Cooling
process
~UL NO
SYMBOL
COM~EPNo.
N 111016
NS W 2U4
13 NS W 211~
I< Nlli!MIS
15 N S W203
16 N SW2U I
17 NI>IOOI
18
19
20
PBI
R
R2
SWI
N N N
y
y
LAI
TA
LA2
LSJ
rn-......_-..........
Ill Ill
y y
Rl
N0 1004
Ill Ill
y
y
N Ill
y y
.......
......-:::
..---
L.---~
-..........::
Rl
Ill :N SW2UIL
112 NSW 202L
Ill NSW203L
Ill I NS W204 L
115 Nl)(lUIOl.
116 NllOOIIL
117 N I 100121.
y
y
R2
VI
V2
VJ
y
y
:::::-.....
rfll
y
y
RJ
~t;:-
~~
12
~
192
01 02 oJ o OS 06 07 08 09 10 II
RJ
N lli ()(M I
N 1)1002
Ia
NE XT STEI'
--;..I
II I
II I
..-
..............
I'
II
THEN
ELSE
C O~"T INUF.
TABLE
1931
~o .
Fig. 2 52
56
I
Entry example of decision table corresponding
to the relay ladder diagram
Chap.2
Process Control
TAII U : NO
Ioo I
STIKJ:I
IIESt:RII'TION
Cooli ng process l b
I (T)
~NO. 01 02 03 04 o; 06 01 08 09 10 II
1' No
I 2
2 3 4 ;
y
01
02 SYMBOL
I I Nllllllfi
12 N lllll02
13 N !>1003
14 NJ>IOO I
15 N l >IOOO
16 N f1100~
11
t:O M
I'R I
LA2
V N
LS3
LAI
TA
SW I (ON to repeat)
y
y
18
1-lJ>.
41
42
Ill
Vt
V2
NDOOIO ll
112 NDOOII H
113 NDOOI2 11
114
115
116
117
V3
142
NEXT STEP
CONTINUE
TABLE NO
1193 1
y N
y N
JJ!
rfii
g!192
I
I
THEN
ELSE
'-"
..2
Shows correspondence to
the relay l adder di agram I a
J 4
; e
LL
L RJ
R2
Rl
NDOnnnH : Latched type switch
Fig. 2 53 Entry example of decision table using step numbers
2.5.4 Devices for sequential control

Sequential control devices originally consisted of relay control panels made up of relays, timers and the like. With the coming of the
transistor age, however, such devices came to be substituted by those
employing logic circuits made up of transistors and diodes. Also used
at that time were sequencers in which a control program previously
fixed by wiring could be set as needed through a pinboard.
The above period was in turn followed by developments in semiconductor technology, which gave rise to the stored-program PLC (Programmable Logic Controller) characterized by the use of ICs and memory elements. This stage was soon followed by the development of the
microprocessor, which along with the parallel development of peripheral devices such as CRTs, enabled an even further expansion of functions, referred to as PC (Programmable Controller). Although descrip2 .5 Sequential Control
57
tive formats such as flowcharts or logic circuits are used in programming PLCs or PCs depending on the particular kind of machine, the relay circuit format is the most generally used.
On the other hand, in DDC systems (see Chapter 1) employing
minicomputers, the computational ability available could be applied to
incorporate sequential control programs as well as continuous control
programs as needed. With distributed control systems made possible
by the development of the microprocessor, however, a high level of automatic process control, which is a fusion of continuous control and
sequential control, came to be realized. In the DDC era, sequential control programs were written in either assembler or list formats based
on flowcharts, but with distributed control systems, the application
range has broadened, system comprehensibility has become easier, and
documentation simplified with the use of descriptive formats such as
decision tables.
In addition, through the use of digital input/output and the logic
processing of programmable single-loop controllers incorporating
microprocessors, simple sequential control has been made possible.
PRACTICE QUESTIONS
Determine the step response to a unit step signal for the circuit shown in Fig.
2.54, for the three cases:
2.1
(a)
G(s)= 1 +Ts' H(s)=1,
(b)
G(s)=n>1. H(s)= l+Ts'
(c)
G(s)=l, ll(s)=-1+ Ts,
positive feedback
negative feedback
positive feedback
1/s
Fig. 2 54
For the case of a constant outflow liquid level process (integral element) under proportional control, determine the offset for both a change in the liquid level setpoint value and a change in the amount of outflow.
2.3 Explain any problems when using a PI controlJer in a batch control system.
2.4 In the buffer tank selection control example, devise a 3-value override control system in which an upper liquid level limit is incorporated to prevent overflow.
2.2
58
ANSWERS TO QUESTIONS
Setting up the problem with the expression, G0 (s) = G(s)/1 G(s)H(s)

(a) G0(s) = 1/Ts or in other words, integral characteristics; transient response is the
same as shown in Table 2.1(e).
2.1
(b)
Go(s)=(l+Ts)/(I:n+
~s)~Cl+Ts)/(1+ ~s)
or in other words, a proportional gain of 1, and consisting of proportional and derivative action with derivative amplitude n; response is as shown below.
(c) G0 (s) = (1 + Ts)/Ts or in other words, proportional and integral action with a proportional band of 100%; see Fig. 2.27 for the transient response.
2.2 Indicating tank level as H(s), level setpoint value as H.(s), inflow as Q,(s), outflow
as Q.(s), tank cross-sectional area as A, proportional gain as Kc and deviation as E(s),
the following block diagram can be drawn.
Expressing E(s) in terms of H 0 (s) and Q0 (s):
1
1
1+ Kc/sA H.(s)+ Kc+sA Q.(s)
E(s)
For a unit step change, this becomes:

E( 5 )
1
s(1 + Kc/sA)
s(Kc +sA)
As t--+=, the deviation, or offset can thus be expressed as:
SE(s)
s~o
1
1 + Kc/sA
Kc+sA
0+-1-
Kc
or in other words, in regard to a change in the setpoint value, offset is zero, while for a
change in outflow, offset appears.
2.3 Please refer to the discussion on integral action and reset windup.
2.4 For the override control system shown in the figure:
If the flow rate control system is selected: Mu<MF<ML
If the upper level limit control system is selected: MF<Mu<ML
If the lower level limit system is selected: Mu<M L<MF
If, however, air-to-close is used for the control valve, the above relationships will
be reversed. In either case, an M (intermediate value) selector is used.
REFERENCES
1) T. Ishii: Process Control Instrumentation Systems, Denki Shoin (1973) (in Japanese).
2) The Society of Instrument and Control Engineers: Automatic Control Handbook,
Devices and Applications Volume, Ohm Sha (1983) (in Japanese).
References
59
3)
K. Matsunaga: "Documenting Process Control Sequences by Decision Tables,"

Yokogawa Tech. Rep., 26, 3 (1982) 44-50 (in Japanese).
4) ]. G. Holbrook: Laplace Transforms for Electronic Engineers, Pergamon Press
(1959).
5) F. G. Shinsky: Process-Control Systems, 2nd Edition, McGraw-Hill (1979).
6) N. A. Anderson: Instrumentation for Process Measurement and Control, 2nd Edition, Chilton Book (1972).
7) T. Tohyama: "Guide to Understanding of Control Engineering," Instrumentation
and Automation, 3, 4-9 (1975) (in Japanese).
8) Y. Kasai and M. Terao: Electrical Automatic Control, Denki Shoin (1975) (in Japanese).
9) Y. Ohshima: Encyclopedia of Terms used in Automatic Control, Ohm Sha (1971)
(in Japanese).
10) ]IS B 0155-1986: Glossary of Terms Used in Industrial Process Measurement and
Control (in Japanese).
60
Chap.2
Process Control
Chapter 3
DETECTION AND CONVERSION OF

INDUSTRIAL VARIABLES
In the petroleum, chemical, iron, textile, and paper industries, raw

materials are supplied to production facilities, energy is added, and the
raw materials are physically and chemically processed into finished
products. In these processes, a variety of variables related to each part
of the process, such as temperature, pressure, flow rate, liquid level,
composition, and quality are measured and controlled.
In machine industries like automobile and electrical equipment production, which stress mechanical processes such as material processing, assembly, inspection and so on, the processed objects are solids,
and variables such as location, shape, dimensions, and position are
measured and controlled. These types of measurements made in the
production process and in related activies such as acceptance of raw
materials and product shipping inspection are called industrial measurement. The variables of industrial measurement and control are called
industial variables, and can be broadly classified according to their
characteristics in the following way.
(1) Process state variables such as temperature, pressure, flow rate,
liquid level, humidity, heat content, viscosity, and density
(2) Mechanical variables such as length, angle, displacement, position, and shape
(3) Dynamic variables such as mass, energy, time, revolutions,
speed, and vibration
(4) Composition variables of gases, solutions and solids
(5) Electrical variables such as voltage, current, wattage, frequency,
and magnetism.
There are various kinds detecting elements for these industrial
variables. In this chapter we will describe some typical examples of
devices currently in use.
Chap.3 Detection and Conversion of Industrial Variables
61
3.1
Measurement of Industrial Variables 1>-s>
3.1.1 Methods of measurement

There are a number of methods of measurement, that is to say,
methods of comparing measured variables with some reference value.
The major methods are described below.
(a) Direct and indirect measurements
Direct measurement involves comparing the measured variable with
a reference value of the same type. Examples of direct measurement include measuring length with a ruler, and measuring electrical resistance by comparison, using a Wheatstone bridge circuit for reference
values.
Indirect measurement, on the other hand, is done by measuring
some other variable that has a fixed relationship with the variable to
be measured. For example, by applying an electrical voltage E to a resistance Rand measuring the current I, the resistance can be obtained
by using the relationship R=E/I. Determining the temperature in an
oven with a thermocouple, and obtaining flow rate from the pressure
differential across an orifice in a pipe are other examples of indirect
measurement.
There are a great many industrial variables, which range from
temperature, flow rate, and pressure to solution concentration and gas
composition. Measurement of these variables employs detection methods that make use of various kinds of physical law and effects. Consequently, indirect measurement methods are most common by far.
(b) Deflection methods and zero methods
When using a spring balance to measure weight, as in Fig. 3.l(a),
the measured weight is read from the position of a pointer which indicates the displacement of a spring resulting from placing the measured
object on the balance. A measurement method that in this way directly
0
II/
( a ) Spring balance
( b ) Zero balance
Fig. 3 1 An example of deflection method and zero method
62
translates something like a displacement or angle which has a fixed relationship to the variable to be measured into a measurement value is
called a deflection method. The Bourdon tube pressure gauge and the
movable coil voltmeter and ammeter are deflection method measuring
devices. The deflection method is direct and the measuring mechanism
is simple, so this method is widely used. However, since movement of
the indicator takes enegy from the measured object, accurate measurement cannot ordinarily be hoped for.
As shown in Fig. 3.1(b), weight can also be determined with a zero
balance. In this method, the measured weight is balanced with a
known weight(in this case pieces of brass). By adjusting the known
weight such that the measuring instrument indicates zero, the weight
of the object being measured can be known. This type of method is
called a zero method. Typical examples of this method are electrical potentiometers for measuring voltage, and methods for measuring electrical resistance and impedance that use a Wheatstone bridge.
When the manual balancing operation in zero method measurement is replaced with automatic balancing by servo-motors, it is called
a self-balancing method. The zero method has the special feature that if
the balancing is not complete, the difference between the measured
variable and the reference value can be detected, and a feedback operation can correct the reference value.
Measurement of industrial variables mostly uses self-balancing
methods for the following reasons.
(1) The balancing operation uses a separate energy supply, so the
operation is strong and measurement error from friction and so on
is avoided.
(2) Measurement of good accuracy can be expected bacause the
measured quantity is compared with an exact reference quantity.
(3) The variable being measured is not subjected to outside disturbances, since measurement is accomplished in equilibrium.
(c) Potentiometer and Wheatstone bridge
In industrial measurement, the variables to be measured are often
converted into electrical quantities such as voltage resistance and impedance. These converted quantities are often measured by potentiometer or Wheatstone bridge circuits.
(1) Direct current potentiometer circuit
When D.C. voltage is measured with a moving coil voltmeter, as in Fig. 3.2, the current IM flows
in the measurement circuit, a voltage drop /MRs occurus in the measured object side of the circuit, so the value Ex cannot be measured exactly. To measure D.C. voltage accurately, the zero method potentiometer mentioned above can be used. The principle is illustrated in
3.1
Measurement of Industrial Variables
63
r--
CD--,I
r-I
--,
I
I
I
I
I
:I
l
CD
Measured object
Voltmeter
I
I
_ _ _ _ _ _ .JI
L-------J
Ex : Unknown voltage
Rs : Internal resistance
RM : 'Internal resistance of voltmeter
lM: Circiut current
lM=Ex/(Rs+RM)
Fig. 3 2
Voltage measurement by voltmeter
Fig. 3 3
Potentiometric circuit
Fig. 3.3.
Reference voltage Es produces a fixed current flow Is through a
variable resistor. An unknown voltage Ex and galvanometer G are connected, and point C is sought, such that the galvanometer reads zero.
If the resistance between points C and B is expressed by R, then the
equation Ex=lsR holds. Thus, the value Df Ex can be read directly
from a voltage scale placed on the variable resistor. When the circuit is
in balance, no c;urrent flows through Ex, and the internal resistance of
the object of measurement is not affected. Thus, if the voltage Es and
the variable resistor are exact, precise measurement of voltage can be
made.
The Wheatstone bridge circuit is often used
(2) Wheatstone bridge
for zero method measurement of resistance and impedance. In the
Wheatstone bridge circuit shown in Fig. 3.4, Rx is the input side unknown resistance. If the bridge is balancd by moving the brush of the
variable resistor so that the galvanometer G indicates zero, the relationship RxR 2 = R 1 R 3 holds. Since R 1 ,R2 is determined by the position of
the variable resistor brush, if R 3 is a known value, the value of Rx corresponds to the brush position.
64
Chap.3
Detection and Conversion of Industrial Variables
Fig. 3 4 Wheatstone bridge circuit
(d) Self-balancing methods
In the self-balancing method, the operation of balancing the measurement circuit so as to obtain a zero reading (such as in the methods
illustrated in Figs. 3.3 and 3.4) is done automatically. The principle of
a self-balancing potentiometer is shown in Fig. 3.5. Here, the input
measurement voltage Ex and the voltage across the variable resistance
up to the brush position Es are compared and the voltage difference e
is detected by an amplifier rather than the galvanometer. Depending
on the polarity of the voltage difference e, the output of the amplifier
either causes a clockwise or counterclockwise rotation in the servo-motor. Since the servo-motor and the brush are mechanically linked, the
servo-motor will move the brush until the unknown voltage Ex and the
voltage at the brush Es are equal, and then automatically stop at this
equilibrium position. At the time Ex= E 5 , the brush position indicates
the measured value. If the brush is mechanically linked to an indicator,
the value of Ex will be displayed. Figure 3.6 shows the principle of a resistance bridge self-balancing measurement system.
Ec
Balancing (Servo)
motor
(f)
Amplifier
L--------~----------J
Fig. 3 5 Principle of self-balancing potentiometer
3.1
65
Balancing (Servo)
Rx
b
Amplifier/
____________
.../I
Fig. 3 6 Principle of self-balancing resistance bridge
3.1.2. Accuracy of measurement

The degree of correctness of the results of measurement is generally called accuracy. Looking at it another way, accuracy is an index
representing how little error there is in the measurement, and is usually represented by limits of error under some appropriate conditions.
(a) Errors and their causes
Measured values always include some error. If M represents a
measured value, and T the true value, then the error of measurement e
represents the difference between the measured value and the true value, or, expressed symbolically, e = M- T. ') The percentage of error can
be expressed by (e/T) x 100. The true value is an ideal value that cannot actually be known. So in order to express measurement error as a
numerical value, the value measured by a standard device or instrument of the best accuracy is taken as the true value.
The sources of error can be classified in the following way. 3)
(1) Theoretical errors
These are errors related to the principle or
theory of measurement. For example, when measuring the level of a
liquid in a tank with a level gauge that works on the principle of a pressure differential, there is a difference in measured pressure that is
caused by a difference between the density under actual use conditions
and the design density. The error of measurement resulting from this
pressure difference is an example of theoretical error. It is possible to
eliminate this error by correcting for the difference in density.
(2) Instrument error
This type of error originates in the measuring
instrument itself, and results from mechanical or electrical deviations
in the components of the instrument such as mechanical clearance, error in resistance values, and so on. For this class of error, the measured value can be corrected by calibrating the instrument.
(3) Individual human error
These errors result from the manner in
which the measurer habitually makes the measurements. For example,
a person might consistently take low readings when reading a scale.
66
Chap.3
These errors can be eliminated to a certain extent by care on the part

of the measurer.
(4) Environmental error
These are errors caused by influence of the
environment or the conditions under which a measurement is made,
such as ambient temperature or pressure, the position in which the
measuring device is set, improper insertion of a temperature sensor,
etc. If these environmental factors change, it will exert a generally irregular effect on the measurements.
(5) Anomalous error
These errors result from defective measurements caused by abnormalities in the measuring instrument, differences in readings taken by the human measurer, and so on.
(b) Accuracy
Accuracy has always been basic term used to express the characteristics of measuring instruments. Going by the Japan Industrial
Standard measurement terminology (]IS Z 8103-1984), the term accuracy is defined as "a quality that includes the correctness and precision
of a measurement or a value displayed on a measuring instrument."It
is used in a statistical sense, assuming that errors in measurement fall
into a normal distribution.
However, the term accuracy is used differently by manufacturers
of measuring instruments. What they call ordinary accuracy, or accuracy rating in instrument specifications is a definition of the limits of error in output (indicated values) permitted according to the specifications for that model, and based on the specified conditions. This is
shown in Fig. 3.7. These specified conditions are ranges within which
effects from fluctuations in outside factors to which the instrument is
exposed, such as temperature, humidity, power supply, and vibration
Output
maximum actual
positive deviation /
Actual downscale
'
calibration curve ,
t,
Accuracy rating
Specified
characteristic
' curve
Measured accuracy
t - - - - Span ~-----1
Fig. 3 7
3.1
Accuracy
67
can be neglected. For the instrument whose operating characteristics

are shown in Fig. 3. 7, accuracy includes the effects of hysteresis
(difference in output according to the direction of change in the input),
dead zone (the largest change in input that fails to produce any output), repeatability (the degree to which measurements of the same object made by the same method, under the same conditions, and repeated within a relatively short period of time, produce the same measured values; see Fig. 3.8), all of which are causes of error.
Output
Repeatability
I
Downscale
l-- -- - -+,.,.___ _
Span _ _ ........,
Fig. 3 8
Input
Repeatabi lity
The accuracy rating of industrial measuring instruments is expressed in terms of error limits, as in the following examples.
(1) A percentage of the span (e.g., 0.2% OF SPAN)
(2) A percentage of the scale length (e.g., 0.5% OF SCALE
LENGTH)
(3) A percentage of the upper range value (e .g., 0.5% OF UPPER.
RANGE VALUE)
(4) A percentage of the reading (e.g., 1% OF READING)
The relationship between error and accuracy rating as expressed
in terms of span and reading is shown in Fig. 3.9. In the case of percentage of span, the error becomes larger in proportion to the reading
as the measured value decreases. Thus, when deciding on the measurement range of the instrument, use of the low part of the range should
be avoided as much as possible.
Reproducibility is an expression of the stability of an instrument
over time. It is the degree of consistency of readings (output) when
the same instrument is used to measure the same quantity at different
times, and thus represents the change in accuracy resulting from a
change or deterioration of the parts of the instrument with time. Such
deterioration requires that the instrument be calibrated with a frequency appropriate to the conditions of use.
68
Chap.3
\
2.0
1.5
1.0
0.5
..,
=
"C
.e
~
0
0.5
1.0
1.5
2.0
Fig. 3 9
"<
'
lr-.....
50
Accuracy : 0.5 % of read ing
25
I
Accuracy : 0.5% of spa n
75
I Mea
100
uring range
( %)
1---
Relation between accuracy rating and error
(c) Sensitivity
Sensitivity is the degree to which an instrument can detect a

change in the quantity being measured. Its value is expressed by a sensitivity coefficient defined as
.. .
ff' .
change in reading
measured quant.t
sensitivity coe ICient = c hange m
Iy
The smallest quantity that an instrument can normally detect is
called discrimination or resolution. These limits of sensitivity are simply
called sensitivity. The smallest amount that an instrument can detect
is also expressed as a percentage of the measuring span (e.g., SENSITIVITY: 0.05% OF SPAN). Although good sensitivity is desirable,
greater sensitivity increases the influence of noise, vibration, and other external conditions on accuracy. Thus it is necessary to achieve a
proper balance of accuracy and sensitivity.
(d) Traceability
According to standard measurement terminology (JIS Z 8103), traceability is "a system linked to a national standard established by calibrating all instruments according to some standard or instrument of
the highest quality." The accuracy of a measuring instrument is influenced by the conditions under which it is used, that is, the measurement environment, the method of use, the measurement range, and so
on. Also, since the instrument will wear with use over time, accuracy
will decline. Thus, to preserve the accuracy of the instrument in use
and maintain the reliability of the readings, the instrument must be
properly used, inspected, and calibrated as often as is appropriate to
the conditions under which it is used.
Calibration involves using a standard or standard sample in place
of the measured quantity as the input to the instrument, and obtaining
the relationship between the output (reading) and the true value. Natu3.1
69
rally, the standard device must be of the highest accuracy, but a standard of even higher accuracy is required in order to calibrate the standard itself. In Japan, the national standards are set according to the
Weights and Measures Act, which conforms to international standards.
Responsibility for maintenance and provision of the national standards
rests with the Electrotechnical Laboratory (ETL), National Research
Laboratory of Metrology (NRLM), Radio Research Laboratory (RRL),
and other such national organs. Figure 3.10 shows the electrical standards traceability system in Japan.
CD
~~~
.OD
1:.
Electricity, light, sound, ray
@ Length, temperature, mass, time
CD ,International standards (Conveution du metre)

International bureau of. weights and measures
National standards laboratories of each country
@ Radio research laboratry (RRL)
Standard radio wave
National standards
Electrotechnical laboratory (ETL)
National research laboratory of metrology (NRLM)
Frequency
@ Astronomical observatory
@Time
Public standards agencies
@ Japan electric meters inspection
corporation (JEMIC)
D.C. and low frequency
Japan machinary and metals
inspection institute (JMI)
@ High frequency
@ Regional laboratories of JEMIC
Public industrial laboratories
Working standards user departments
@ Standards laboratory in the company
@ Standards laboratory in the factory
Each workshop
Fig. 3 10 Electrical standards traceability system in Japan
Figure 3.11 is an example of a measuring instrument manufactuer's basic measurement standard control system. In the standards laboratory, primary standards that have been calibrated by the national
standards are maintained. These primary standards are used to calibrate the calibration standards used in every section of the company.
The calibration standards are in turn used by each section for periodic
independent calibration of the instruments used in ordinary work,
which are called working standards.
70
Chap.3
Public
standards
agencies
Standards
laboratory
Working
standards
user
departments
&)
( Factories
subsidiaries
Responsibilities of each department

Determination of calibration intervals
Accuracy control by register
Validity term control by label
Storage of certificates and calibration reports
Treatment of non-conforming standards
Fig. 3 11
Basic system for measurement control (Typical)
3.2 Measurement of Temperature 2l,GJ-!G)
Measurement of temperature, a frequent form of industrial measurement, uses various forms of sensing. The most often used are the
thermoelectric thermometer and the resistance thermometer. For measurement without contact to the measuring instrument itself, there is
also the radiation thermometer.
As opposed to these electrical sensors, there are also mechanical
sensing devices, based on the expansion of a liquid or solid (e.g., bimetal sensing element). These are not suited for transmission over a long
distance, so they are most often used in local applications. Leaving mechanical sensing to other literaturel 33l, this section describes temperature measurement using electrical sensors.
3.2 Measurement of Temperature
71
3.2.1 Thermoelectric thermometers

(a) Measurement principle
As shown in Fig. 3.12, if two dissimilar metals, A and B, are joined at point n, and are left free at the other end, point m, the difference in temperature at the two points (T2 - T 1) generates the electromotive force EAs(T2 , T 1). The metals A and B are called a thermocouple,
point m is called a cold junction, point n is called a hot junction, and
the electromotive force EAs(T2 , T1) is called the thermoelectro motive
force. Because there is a fixed relationship between the thermocouple
materials, the cold junction temperature, the hot junction temperature,
and the thermoelectro motive force, for a given thermocouple and a
fixed cold junction temperature, the hot junction temperature can be
determined by measuring the thermoelectro motive force.
T,
Emf measuring
instrumeht
\~/ Conductor C
Cold junction
Fig. 3 12 Temperature measurement using thermo couple
To measure the thermoelectro motive force with almost no current

flow, it is necessary to use a measuring instrument with a high input
resistance.
(b) Law of successive temperatures and intermediate conductors
For a set of thermocouples , given a temperature distribution t 11 t 2 ,
and t 3 in which t 3 > t 2 > t 11 if the thermoelectro motive force resulting
from hot junction temperature t 3 and cold junction temperature t 1 in thermocouple elements A and B is EAs(t3 , t1 ), then the following relation
holds.
(3.1)
13
12
1
I
i
I
I
I
-~
(13,11)
II
-r---9
A~ :
~
I
II
: E(t,,t,)
+
----.1B
-?A~
~
I
T
I
E(t,,t.)
Fig. 3 13 Law of successive temperatures
72
Chap.3
That is, the then:noelectromotive force between t 3 and t 1 is the

sum of the forces between fa and t 2 , and t 2 and t 1 This relationship is
called the law of successive temperatures. If, for example, fa= 500C, t 2
=20C, t 1 =0C for a K-type thermocouple, then the values
are obtained from the JIS C 1602 Reference Thermoelectromotive
Force of K. Accordingly, from Eq. (3.1),
EAs(ta, tz)=EAs(fa, t1)- EAs(tz, fJ)=19 842 (t-~V)
The law of intermediate conductors applies to the effect on the thermoelectromotive force when a third intermediate conductor is inserted
between thermocouples. For the hot junction temperature fa and the
cold junction temperature t 1 as shown Fig.3.14, the following relationship holds.
(3.2)
k----1
lJ
fl
8-i
L--+---1
c-:::.:::"'1
c
'
k.----
___
--B - - -~I
1
I
!~
k::-'"C
i
:
-.;:.:. l2
t<.:"-C
:
-....;;_____
B
Fig. 3 14
:
I
-b
Law of intermediate conductor
When both junction temperatures of the intermediate conductor

are the same (in this case, fa), the thermoelectromotive force EAs(ta, t 1)
is not changed.
(c) Types of thermocouples
There are many kinds of thermocoples suited for various uses, JIS
C 1602 defines seven types. Table 3.1 lists their type codes, the materials from which they are constructed, measurement ranges, accuracy
classes, and tolerance values. Thermocouples are broadly classified
into noble metal thermocouples (type codes B, R, and S), and base metal thermocouples (type codes K, E, ], and T). The noble metal therm3.2 Measurement of Temperature
73
"""'
""""
.,
~:;
.....
"'......
;;
"'
"';::;c:;
<::
""~
;::;
.,;::;
;::;
c:;
......
"'
"'......
'"'
tJ
(.,_,
Symbol of
thermocouples
Iron
Alloy, main
components of
which are nickel
and chromium
Alloy, main
components of
which are copper
and nickel
Alloy, main
components of
which are copper
and nickel
composition
-Leg*'
+Leg* 1
Platinum rhodium
Platinum rhodium
alloy containing
alloy containing
6% rhodium
30% rhodium
Platinum
Platinum rhodium
alloy containing
13% rhodium
Platinum
Platinum rhodium
alloy containing
10% rhodium
Alloy, main
Alloy, main
components of
components of
which is nickel
which are nickel
and chromium
Table 3.1
Accuracy class
Class 0.4
Class 0. 75
ooc to 750"C excl.

ooc to 750"C excl.
Class 1.5
Class 0.75
oac to SOO"C excl.

- 200C to O"C excl.
Class 0.4
Class 1.5
Class 0.75
Class 0.4
Class 0.25
ooc to soooc excl.
- 200"C to ooc excl.
O"C to 1 200"C excl.
O"C to 1 OOO"C excl.
O"C to 1 600"C excl.
600"C to 1 700"C excl. Class 0.5
Temperature range
values
Thermocouples (]IS C 1602)
1.5"C or 0.4% of measured

temperature
temperature
2.5oC or 1.5% of measured
temperature
temperature
2.5"C or 0. 75% of measured
temperature
2.5C or 1.5% of measured
temperature
temperature
2.5C or 0. 75% of measured
temperature

temperature
4"C or 0.5% of measured

temperature
Tolerance*'
0l
;;;
~;s
;:;:
""
.....
;::!
~
"';;;
t-v
w
Alloy, main
components of
which are copper
and nickel
ooc to 350oC excl.
tion.
The value expressed in oc or %, whichever is larger, is taken as the tolerance.
Copper
Class 0.4
0.5C or 0.4% of measured

temperature
1 oc or 0. 75% of measured
Class 0. 75
ooc to 350oC excl.
temperature
1 oc or 1.5% of measured
Class 1.5
- 200C to ooc excl.
temperature
thermoelect romotive force, and -leg means op* 1 The +leg means a leg to be connected to +terminal of an instrument to measure
posite side one.
the measured thermoelect romoz Tolerance means the maximum allowable limit of such value that the temperature converted from
temperature measuring juncthe
of
temperature
tive force using the reference thermoelect romotive force table minus the actual
Table 3.2 Element wire of thermocouple (JIS C 1602)

Symbol of
composing
material
B
Diameter of
element wire
(mmJ
Normal operating
temperature lmit* 1
CCJ
Elevated operating
temperature !mit*'
CCJ
0.50
1500
1700
0.50
1400
1600
0.65
650
850
2.95
1.00
750
950
1.25
1.60
850
1050
0.49
2.30
900
1100
0.24
3.20
1000
1200
0.12
0.65
450
500
3.56
1.00
500
550
1.50
1.60
550
650
0.59
2.30
600
750
0.28
3.20
700
800
0.15
0.65
400
500
1.70
1.00
450
550
0.72
1.60
500
650
0.28
2.30
550
750
0.14
3.20
600
750
0.07
0.32
200
250
6.17
0.65
200
250
1.50
1.00
250
300
0.63
1.60
300
350
0.25
Resistance
(!l/mJ
1.75
1.47
1.43
Normal operating temperature limit means the limit of temperature within which
thermocouples can be continuously used in the air.
*2 Elevated operating temperature limit means the limit of temperature at which thermocouples can be used for a short time in the case of unavoidable needs.
*I
76
Chap.3
ocouples generally have excellent resistance to corrosion, but produce

small thermoelectromotive forces compared with the base metal types.
Table 3.2 lists the element wire diameter, normal operating temperature limit, elevated operating temperature limit and resistance of
each type of thermocouple. As the diameter of the element wire increases, the normal operating temperature limit and elevated operating
temperature increase, and the resistance decreases.
Practical use in thermocouples requires the following qualities.
(1) Good corrosion resistance, robustness against gases, etc.
(2) Large thermoelectromotive force.
(3) Good heat resistance and ability to maintain mechanical strength
at high temperatures.
(4) Stable themoelectromotive force even over long periods of use,
with small thermocouple loss.
(5) Interchangeability: thermocouples of the same type normally
have the same characteristics.
Figure 3.15 shows the thermoelectromotive force for each type of thermocouple.
>
8
"~
70
60
..2 50
.::"
0 40
E
2
ti 30
"
-.;
'"
20
i;
..c
f-
Temperature ["C]
Fig. 3 15 Thermal electromotive force of thermocouple
(d) Feature of each type of thermocouple

(1) Features common to the noble metal thermocouples (B, R, and S)
(1) Can be used continuously in an oxidizing environment, but are

not suited for use in a reducing environment
(2) Have good resistance to corrosion and chemicals
(3) Are easily degenerated by hydrogen or metal vapors
(4) Are suited for measurement of high temperatures
(5) Have small thermoelectromotive force (that of type B, is especially small)
77
(6) Thermoelectromotive force has good linearity.

(2) Features of base metal thermocouples (K, E, ], and 1)
(a) K-type (chromel-alumel)

(1) Can be used in an oxidizing environment
(2) Deteriorates slowly in a reducing environment
(3) Strong against metal vapors
(4) Good linearity of thermoelectromotive force
(b) E-type (chromel-constantan)
(1) Good resistance to corrosion and oxidation
(2) Deteriorates slowly in a reducing environment
(3) Large thermoelectromotive force
(4) High electrical resistance
(5) Little measurement lag
(c) ]-type (iron-constantan)
(I) Can be used in an reducing environment
(2) Relatively large thermoelectromotive force; good linearity
(3) Rusts easily
(4) Quality is not consistent
(5) Exhibits hysteresis thermoelectromotive force at high temperatures (800C)
(d) T -type (copper-constantan)
(1) Suited for use at low and extremely low temperatures
(2) Suited for use in reducing environments
(3) Large heat transfer error
(e) Standard thermocouples
]IS C 1602 defines standard thermocouples which produce a fixed
thermoelectromotive force (reference thermoelectromotive force) for a
given hot junction temperature.
(/) Thermocouples with protective tube
The ends of two thermocouple element wires are joined by gas
welding, arc welding, resistance welding and so on, to form the hot
junction. In this state it is called a bare thermocouple. To insulate the
two element wires electrically, they are led into an insulating tube as
shown in Fig. 3.16. At this stage it is called a thermocouple with insulating tube. Because it is susceptible to corrosion and mechanical damage, it is put into a protective metal or non-metal tube, and the two element wires are connected to a terminal at the head of the tube. Fig.
3.17 shows how the thermocouple with protective tube is constructed.
(g) Sheathed thermocouples
In a sheathed thermocouple, the space between the element wires
and the metal sheath is filled with a powdered inorganic insulator and
sealed. It is then processed to produce a single-unit construction. JIS C
1605 defines four types of sheathed thermocoples (type symbols SK,
78
Hot junction
Fig. 3 16 Structure of thermocouple with insulating tube

Terminal case
Terminal
Terminal cover
Gasket
Ceramic insulating tube
Thermocouple element wire

(Hot junction)
Fig. 3 17
Structure of thermocouple with protecting tube
SE, SJ, and ST). The element wire materials of these thermocouples
are the same as the corresponding K, E, J, and T thermocouples.
The metal sheath is made of either austenite stainless steel (symbol A), or nichrome heat resistant alloy (symbol B). The hot junction
can be either the grounded type (symbol G) or the ungrounded type
(symbol U) (Fig. 3.18). The overall unit can either have a terminal
plate or not (Fig. 3.19). Table 3.3 lists the symbol, measured temperature range, accuracy class, tolerance, and normal operating temperature limits for each type of sheathed thermocouple.
Metallic sheath
Mineral insulation
~it+::;rr;!; ~'~lH;9J4f II
Element wire
(a) Grounding type (Symbol G)
(b) Non-grounding type (Symbol U)
Fig. 3 18 Hot junction structure
79
Metallic sheath
~i
.
-U,Oh
( a ) Thermocouple with terminal

Compensating lead wire
Metallic sheath
~===E3 I
LLength_j
( b ) Thermocouple without terminal
Fig. 3 19 Sheath thermocouple
Sheathed thermocouples have the following characteristics.

(1) Fast response. The metal sheath has a small diameter and the
wire elements are fine, so the heat capacity is small. Also, since
the sheath is bonded to the insulating material, which is in turn
bonded to the element wires, heat resistance is low. Thus the response to heat is exceptionally good, and the temperature of objects having low heat capacity can be measured.
(2) Flexible: Because of the small outside diameter of the sheath,(l
mm to 8 mm) it can be curved to fit into small places.
(3) Low mechanical strength: For measurement of fast flowing liquids, or when the sheath is installed horizontally, reinforcement
such as a protective tube is needed.
(4) Insulator deterioration: The insulating material (MgO) is highly
hydrophilic. If it is not completely sealed, it will absorb moisture,
causing deterioration of its insulating properties.
(h) Compensating lead wire
If the thermocouple is connected to the measuring instrument
over a long distance using the element wires as they are, induction
interference will lower accuracy, and values will become high. If a
different conductor is used for the connection to the measuring instrument in place of the element wire, the junction between the conductor
and the element wire will produce a contact potential difference, and
changes in the temperature at that junction will generate measurement
error. To avoid this, compensating lead wire is used to connect the thermocouple to the measuring instrument. Compensating lead wire is a
pair of insulated conductors that, in the appropriate temperature
range, including ordinary temperatures, has about the same thermoelectric characteristics as the thermocouple. Using compensating lead
wire has about the same effect as using element wire to make the ex80
:;:
~
;:;""
......
""~
~
;;!
;:
"'
.,~
t-v
--
--
--
--
- -
----
----
Tolerance means allowable maximum of the remainder of the temperature converted from thermoelectromotiv e force using the table of the refer
ence thermoelectromotiv e force subtracted by the temperature of measuring junction.
Moreover, either one of "C or % whichever is larger is allowed as the tolerance.
Normal operating temperature limit means the limit of temperature within which thermocouples can be continuously used in the air.
l"C or 1.5%
of measured
temperature
Class 1.5
-200"C to
O"C excl
-----
350
350
350
350
350
350
350
300
l"C or 0.75%
of measured
temperature
Class 0.75
O"C to
350"C incl.
ST
750
750
750
750
750
750
650
900
450
800
B
1 050
2.5"C or 0.75%
of measured
temperature
900
A
900
Class 0.75
800
B
1000
O"C to
750"C incl.
900
800
SJ
800
900
2.5"C or 1.5%
of measured
temperature
750
800
8.0
Class 1.5
650
650
750
Normal operating temperature limit CCJ *'

Sheath diameter (mm)
6.0, 6.4
4.5, 4.8
3.0, 3.2
1.0,1.5,1.6,2.0
Sheath material
A
B
A
A,B
A,B
- 200"C to
O"C excl
2.5"C or 1.5%
of measured
temperature
2.5"C or 0.75%
of measured
temperature
Torelanceu
2.5"C or 0.75%
of measured
temperature
Class 1.5
Class 0. 75
Accuracy
class
Class 0. 75
-200"C to
O"C excl
O"C to
1050"C incl.
Measuring
temperature
O"C to
900"C incl.
SE
SK
Symbol of
sheath
thermocouple
Table 3.3 Sheath thermocouples (]IS C 1605)
"'
!;::)
'"\
!;::)
._
-
~
-
~
"'.....'"\
"';::!c:;
'"\
""
;::!
;::!
1:).
!;::)
;::!
c:;
"".....
""....."'
t:l
..
Heat-proof
normal class
General use
normal class
WX-H
VX-G
General use
normal class
Heat-proof
precision class
KX-HS
WX-G
Heat-proof
normal class
General use
precision class
General use
normal class
KX-H
KX-GS
KX-G
Heat-proof
normal class
General use
normal class
RX-G
SX-G
RX-H
SX-H
General use
normal class
Classification
by use and
tolerance
BX-G
Symbol of
Symbol
thermocouple
to be used in
combination
with
Grey
Surface color
Alloy; main
components are
copper and
nickel
Alloy; main
components are
copper and
nickel
Iron
Copper
Alloy; main
Alloy; main
components are component is
nickel and
nickel
chromium
Blue
Alloy; main
Black
components are
copper and
nickel
Copper
Copper
Copper
-Leg
+Leg
Composing materials
Table 3.4 Compensating lead wires (JIS C 1610)
-20 to 100
-20 to 150
0 to 150
0 to 100
Tolerance
Temperature of
connected point
with
thermocouple
[oCJ
2.5
3.0
1.5
2.5
1.5
2.5
+3
-7
[oCJ
Tolerance of
compensating
lead wire
e;
;;
;;
.,~
"'
""'t-v
1.0
Heat-proof
precision class
TX-HS
1.0
2.0
2.0
-20 to 150
2.5
Heat-proof
normal class
Alloy; main
Brown
components are
copper and
nickel
-20 to 150
2.5
TX-H
Copper
Yellow
Alloy; main
components are
copper and
nickel
-20 to 150
General use
precision class
General use
normal class
TX-G
Iron
Purple
Alloy; main
Alloy; main
components are components are
copper and
nickel and
nickel
chromium
TX-GS
Heat-proof
normal class
General use
normal class
JX-G
JX-H
Heat-proof
normal class
General use
normal class
EX-H
EX-G
Remarks 1. Cores of KX-G, KX-GS, KX-H, KX-HS, EX-G, EX-H, JX-G, JX-H, TX-G, TX-GS, TX-H and TX-HS are made
of the same material as the thermocouple which is to be used in combination with.
2. Cores of BX-G, RX-G, RX-H, SX-G, SX-G, SX-H, WX-G, WX-H and VX-G are different materials from
those of the thermocouples which are to be used in combination with, however, their thermo-electric characteristics in
operating temperature range (Table 2) are the same as those of thermocouple.
3. Compensating lead wire of normal class is suitable for use in normal temperature measurement.
4. Tolerance of compensating lead wire of precision class is about il half of that of normal class and is suitable for
temperature measurement which requires special precision.
5. As for the characteristics, WPR-G and WPR-H of the Former ]IS may be used in combination with RandS in almost
all cases.
tension. However, the thermoelectric characteristics are not completely the same, so a very small error will be generated in the compensating lead wire connection. The difference between the thermoelectromotive force of the compensating lead wire and the standard thermoelectromotive force of the corresponding type of thermocouple at the same
temperature, converted into a temperature difference, is called compensating lead wire error. In Table 3.4, compensating lead wires classified
according to the type of thermocouple with which they are used according to JIS standards.
Compensating lead wires are also classified according to working
temperature: G for general use ( -20oC to 90C), and H for heat-resistant applications (OoC to 150C). The ordinary accuracy class is suited
for normal temperature measurements, but where precision is needed,
the precision class is appropriate, since its tolerance is half that of the
ordinary class. The type is identified by the color of the surface covering, and the polarity, is indicated by the color of the core insulation
(red for positive, and principally white for negative).
(z) Special thermocouples
(1) Thermocouples for high temperature use
The Pt40%Rh-Pt20%
Rh thermocouple, which is not covered in JIS standards, can measure
temperatures near 1 900C, but has an unusually low electromotive
force. The W-W 26%Re and the W 5%Re-W 26%Re are representative high melting point metal thermocouples. Tungsten is oxidized in
air or in an oxidizing atmosphere, so these thermocouples are used in
an inert or reducing atmosphere. They have a maximum working temperature of about 2 800oC, and a normal limit of about 2 400C.
(2) Dipping thermocouples
Continuous measurement of the temperature or molten metals is difficult. Thus a short-time measurement is
used in which the tip of the thermocouple is opened and placed in contact with the molten metal.
3.2.2 Resistance thermometers
(a) Measurement principle and characteristics
Because there is a fixed relationship between the electrical resistance of a metal and temperature, it is possible to determine a temperature by measuring electrical resistance. Platinum (Pt), nickel (Ni), and
copper (Cu) are used in this type of thermometer. The temperature
and electrical resistance characteristics of these metals are shown in
Fig. 3.20.
Thermometer materials must satisfy the following conditions.
(1) The relationship between temperature and electrical resistance
must be continuous and have unique values within the working
temperature range.
84
3.0
--;; 2.5
Ni
<>::
::. 2.0
.g
~
i,c'Y
1.5
5v v
~ 1.0
~
0::
0.
-200-100 0
100 200 300 400 500 600
Temperature [c]
Fig. 3 20
Temperature-resistance characteristics
(2) There is no change in the resistance over time or according to

other factors.
(3) There is no hysteresis effect in the relationship between resistance and temperature.
(4) The material is resistant to corrosion, thus stable.
(5) The material has a high intrinsic resistance and resistance temperature coefficient.
(6) Interchangeable.
(7) Easily worked.
Platinum has the best combination of the above properties, and the
platinum resistance temperature detector is the ]IS standard. Its measuring range is from - 200aC to 650aC (PtlOO) or -200aC to 500aC
(JPtlOO).
The resistance temperature measurement method has the following characteristics compared with the thermocouple method.
(1) High sensitivity.
(2) Stable over a long period of time when used in a low-vibration environment.
(3) Response is slow because of its large size.
(4) Has a low maximum working temperature.
(5) Susceptible to shock and vibration because a fine resistance wire
element is used.
When a current flows in a resistor, the voltage between the ends
of the resistor is given by the following relationship, according to
Ohm's law.
V=IR
In this equation, V represents voltage (V), I is current (A), and R is re3.2 Measurement of Temperature
85
sistance (!1). If I is fixed and Vis measured, we can determine R, and

thus measure the temperature of the resistance temperature detector.
(b) Plantinum resistance temperature detector
The relationship between the resistance of a platinum resistance
temperature detector and temperature is standardized in ]IS C 1604.
The resistance value of a resistance element at ooc is called the nominal resistance. At present nominal resistance is 100!1. The current that
flows continuously through the resistance element is called the standard current. The present standard currents are 1 rnA, 2 rnA, and 5 rnA.
There are three working temperature grades: L for low temperature use ( -200oC to 100C), M for medium temperatures (OoC to
Table 3.5 Tolerance (JIS C 1604)
Tolerance
Normal
resistance
(DJ
Operating
temperature
range
Measured
temperature
COCJ
100
H
50
H
86
Chap.3
Class 0.15
Class 0.2
Class 0.5
Temperature Temperature Temperature

COCJ
COCJ
("CJ
-200
0.45
0.55
1.3
-100
0.30
0.35
0.8
0.15
0.15
0.3
100
0.30
0.35
0.8
200
0.45
0.55
1.3
300
0.60
0.75
1.8
350
0.68
0.85
2.05
400
0.95
2.3
500
1.15
2.8
-200
0.45
0.55
1.3
-100
0.30
0.35
0.8
0.15
0.15
0.3
100
0.30
0.35
0.8
200
0.45
0.55
1.3
300
0.60
0.75
1.8
350
0.68
0.85
2.05
400
0.95
2.3
500
1.15
2.8
350oC), and H for high temperatures (OoC to 500oC or ooc to 650C).

(c) Internal wirs
The conductors used to connect the resistance element to the terminals of the resistance bulb are call internal wires. As shown in Fig.
3.21, there are three types of internal wire configurations: 2-wire,
3-wire, and 4-wire. In the figure, s represents the resistance element,
and A and B are the terminals.
::
~[______
(a) 2-Wire system
i
Fig. 3 21
(b) 3-Wire system
(c) 4-Wire system
::
Internal wires of resistance temperature detector
The 2-wire configuration reduces the cost of the external wiring

for connecting the resistance bulb to the conversion instrument, but introduces a large error. This is because the same wire is used to supply
the electric current and to detect the voltage, and as a result, the temperature of the wire directly affects the resistance. If, in order to
make the influence of the external wiring resistance small, the conversion instrument is installed near the resistance temperature detector,
making the instrument heat-proof and protecting it from explosion becomes a problem. For industrial measurement, the 3-wire type is most
often used. If the resistances of the three wires are the same, measurement free from error is possible even if the ditstance between resistance temperature detector and conversion instrument is long and the
resistance is subject to change by external temperatures. To completely eliminate this kind of error, the 4-wire configuration is used. In the
4-wire configuration, the current supply wires are independent of the
voltage detecting wires, so no current flows in the detecting wires.
Thus it is possible to measure the resistance of the resistance element
exactly. This configuration is used in standards and for precision measurement.
(d) Reference resistance element
If Rt is the resistance of the resistance element at t CC), and R 0 is
the resistance at ooc, then RtfRo is called the resistance ratio. The resistance ratio of a metal is reduced by lower purity of the metal and by
working. The purity of platinum is quite high (99.99% or more), and
87
the standard of that purity is represented by R100/ R 0 In JIS C 1604, an

ideal resistance element that has a constant resistance ratio RtfRo for
given temperatures is called the reference resistance element.
(e) Construction of the resistance temperature detector element
(1) Mica-insulated element
An example of a mica-insulated element
is shown in Fig.3.22. The element wire is wrapped around a long, thin
mica plate with a width of about 3 mm to 10 mm serrated on both
sides. Rather wide mica plates are mounted on both sides for insulation, and then two metal sheatR.s are attached. The metal sheath is
bonded to the inside of a protective tube, which has the effect of guarding the element against vibration and shock, and providing good heat
transfer between the protective tube and the element.
Pl ati num element wire

Mica plates
Fig. 3 22
Mica-i nsulated element
The mica-insulated element is not completely fixed in position, so

distortion from heat is small and the temperature-resistance characteristic is stable. However, the unit is relatively large and thus not suitable for use in small or narrow places. Also it responds slowly to
changes in temperature.
(2) Glass sealed element
The construction of a glass sealed element is
shown in Fig. 3.23. The platinum element wire is welded to platinum alloy wires and melted into a glass spool. The diameter of the resistance
element is about 1 mm to 4 mm. The glass softens at about 450 C, so
the maximum working temperature of this element is around 400 C.
t CW
JJH!lf!mWiffllffllln?
W~re
Gla
Platin um element wire
Fig. 3 23
spool
Gla s ealed element
(3) Ceramic-sealed element

The construction of a ceramic sealed element is shown in Fig. 3.24. This is a rodshaped element which uses ceramic spool. Its maximum working temperature is approximately
800C. It responds quickly to temperature changes because of its small
outer diameter (1.6 mm to 3 mm), and has good insulation characteristics.
88
Chap.3
Ceramic Platinum
element wire
Ceramic pool
Fig. 3 24 Ceramic sealed element
(f) Resistance temperature detector with protective tube
A mica- insulated element inserted into a protective tube, with internal wires connected, is called a resistance temperature detector
with protective tube. An internal wire material is used which does not
generate a thermoelectromotive force, and which does not oxidize or
degenerate at the maximum working temperature. For low to medium
temperature use, silver wire covered with fluorocarbon resin or a glass
tube or for high temperature use, nickel wire covered with a porcelain
tube, is use. For the external wiring, the 3-wire configuration is mostly used. Moisture entering the protective tube degrades the insulation,
so for low temperature use, the protective tube is packed with paraffin
or filled with dry air and tightly sealed.
Figure 3.25 shows the construction of a resistance temperature detector with protective tube. JIS C 1604 specifies protective tubes with
outside diameters between 3.2 mm and 15 mm, and lengths between
250 and 1 000 mm.
Gasket
Insulating tube
Protecting tube
Fig. 3 25
Resistance temperature detector with protecting tube
3 .2 Measurement of Temperature
89
(g) Other resistance temperature detectors
Other than platinum, other metals such as nickle or copper are

used for the resistance material. Nickel is less expensive than platinum, and its temperature coefficient is 0.6% at room temperature.
However, because the thermal characteristics vary with the element
meterial, the temperature coefficient is adjusted to the standard value
by combining the material with constantan wire or other such material. The working temperature range is- sooc to 300C.
Copper wire of high purity can be obtained, and its temperature
characteristics are uniform. Furthermore, it has better accuracy, linearity and interchangeability than nickel. However, because its intrinsic
resistance is low, a long coil is needed. In addition, at around 250 C,
oxidation begins to change its resistance. Thus the working temperature range is ooc to 120C.
(h) Sheathed resistance bulbs
Sheathed resistance bulbs are specified in JIS C 1606. The space inside the sheath and between the resistance element wires is filled with
a powdered inorganic insulator and processed to produce a single-unit
construction, and platinum is used in the resistance temperature detector.
Ordinary stainless steel having sufficient heat resistance and durability to prevent penetration by environmental elements is used for
the metal sheath.
The inorganic insulation protects the resistance element, the internal wires and the metal sheath. This material must be stable in the appropriate temperature range, and must not affect electrical characteristics. Magnesium oxide (MgO) is often used.
The internal wire must not generate a thermoelectromotive force,
and must not be susceptible to problems arising from evaporation or oxidation . Thus nickel is largely used.
The 3- wire and 4- wire configurations of internal wires are illustrated in Fig. 3.26.
The resistance element either a ceramic sealed element or a glass
sf~-========::: ~ ::~,:e)
: B (Wh ite)
(a ) 3- Wire
ystem
~~~~~~e)
B (White)
Resistance element
Lead wir~___.-) n orgauic insulation
/
Fig. 3 26
90
r ========iA( Red)
(b) 4- Wire sy tern
2' %<t=1
Metallic sheath
st
Chap.3
ac
14
Internal wires
sealed element. These elements are themselves hermetically sealed,

and buried in the inorganic insulation inside the sheath. Thus they are
not affected by the environment. They also have fast response and are
strong against mechanical vibration.
3.2.3 Protective tube
Thermocouple and resistance temperature detectors are put into
protective tubes during use in order to protect them from the environment. Protective tubes must statisfy the following requirements.
(1) Able to withstand the temperatures and pressures of the environment.
(2) Chemically stable.
(3) Mechanically strong.
(4) Airtight
(5) Not produce any gases harmful to thermocouples.
It is difficult to choose a material that satisfies all of these requirements, but a material that is suitable for each particular set of use conditions can be selected.
Protective tubes are classified into metal types and non-metal
types.
(a) Metal protective tubes
Metal tubes are widely used because they are easily manufactured
and have great mechanical strength. Typical metal tube materials are
listed in Table 3.6.
Ordinary protective tubes are cut to the required length from
seamless tubing. One end is sealed by welding, and the other end is
threaded so that the terminal box can be attached. If a tube produced
Table 3.6 Material of typical metal protective tube
Material
Maximum operating
temperature
["CJ
Brass
Mild steel
300
600
Calorized
mild steel
13Cr steel
20Ni 30Cr steel
Stainless steel
Monel
800
900
1000
850
500
Inconel
1000
Feature
Easy to work, low temperature resistance
High mechanical strength, cheap, oxidized
with damp
Improved mild steel on heat and corrosion resistance, low reducing gas resistance
High acid resistance
High sea water resistance, high dilute sulfuric acid resistance
High strength under high temperature, applicable to oxidizing environment
91
by this method is not strong enough, a thick-walled tube can be made

by boring out the center of a piece of rod stock. This is called a
well-type or bored-type protective tube. Although well-type tubes are
heavy and have bad response characteristics, they are widely used
where high pressures or high-speed flows are present. Sheathed thermocouples are very often in well-type protective tubes in applications
where there is severe shock or vibration.
(b) Non-metal protective tubes
Non-metal protective tubes are not as mechanically strong as metal tubes, but they have excellent resistance to chemicals, and there are
many materials of use with high maximum working temperatures. Typical non-metal protective tube materials are listed in Table 3. 7.
Table 3.7 Material of typical non-metal protective tube

Material
Maximum operating
temperature
Features
CCJ
Fluorocarbon
resin
250
Good corrosion resistance
Hard glass
500
Antioxidation, preventing gas penetration,

low resistance for quick heating and cooling
Quartz
1000
High resistance for quick heating and cooling

for continuous use at high temperature ..
Becomes brittle if exposed to high temperature.
Ceramic (PT-2)
1300
Low resistance for quick heating and cooling
Ceramic (PT-1)
1450
Higher heat resistance compared with PT-2
Recrystallized
alumina (PT-0)
1600
Most applicable to high temperature, low resistance for quick heating and cooling
3.2.4 Thermistor thermometers

Thermistors are sintered bodies of the oxides of metals such as
manganese, nickel, and cobalt, whose resistance varies with temperature. The resistance-temperature coefficient of thermistors is negative; an increase in temperature lowers the electrical resistance.
Compared with platinum resistance temperature detectors, thermistors, have the following characteristics.
(1) Large electrical resistance
(2) High sensitivity
92
(3) Poor linearity of resistance

(4) Fast response due to small size
Measurement principle
The resistance Rr of a thermistor at temperature Tis represented
theoretically as
(3.3)
Where Ra is the resistance (il) at temperature Ta, and B is the thermistor constant (K).
If a is taken as the thermistor temperature coefficient, then
because a =(1/Rr)(dRr/d1), and from Eq.(3.3),
(3.4)
From Eq,(3.4), we see that thermistor sensitivity varies greatly
with the temperature. The value of B is usually 3 OOOK to 5 OOOK, but
actually depends on temperature. The value of a is as large as- 3%tC
to- 6%tC at room temperature.
The standard temperature characteristic of thermistors is shown
in Table 3.8.
3.2.5 Radiation thermometers
(a) Black body radiation
Physical objects emit all wavelengths of electromagnetic waves,
and the strength of the radiation increases with the temperature of the
object. An object which absorbs completely all wavelengths of electromagnetic waves from external sources is called a black body. The
spectral radiant emittance for a black body is represented, according to
Planck's law, by
MB(A, T) =
~~ exp( cJ,.\ T) -1
(WI m3)
(3.5)
where, Ma(A, 1) is the spectral radiant emittance of a black body,

A= wavelengh (m),
T=absolute temperature (K)
c1 =first radiation constant,(3. 7418 x 10- 16 W m2) and
c2 =second radiation constant (1.4388x I0- 2 mK).
The spectral radiant emittance for a black body is shown in Fig.
3.27.
(b) Emissivity, luminance temperature, and effective wavelength
The thermal radiation energy emitted by a physical object at a
given temperature is greatest for a black body. For a wavelengh Ae,
the relationship between the spectral radiance of an ordinary object
93
"'
"'
<:to
~
;::;
"
";:;
......
....
"'
~
~
;:;
<:;
;;;""
;:;
<:e
;:;
;:,
;:;
!:>..
""~<:;
~
(.,:)
150
to
300
8.000
k!1
200
to
350
-50
to
100
0
to
100
200
175
150
125
100
75
330.5
2.583 765.7
n
k!1
2.007 661.0 3.238
k!1
n
k!1
7.111
k!1
4.000
k!1
550.0
330.6
913.3
538.2
0.9702 3.332 865.6

n
k!1
k!1
13.33
k!1
1.510
k!1
1.267 4.179
k!1
k!1
0.8659 1.718 4.971

k!1
k!1
k!1
1.190 2.328
k!1
k!1
912.1
3.046
k!1
1.601
k!1
560.7
6.281
k!1
3.000
k!1
6.081
k!1
215.6 1.017
k!1
n
421.2 2.000
k!1
n
100
to
250
150
to
300
200
to
350
1.504 2.265 3.415
1.840 2.797
1.094 2.290 3.415
1.378 2.837
1.094 1.841 3.415
1.360 2.535
1.094 1.841 3.415
2.123 1.748 2.997

k!1
k!1
k!1
904.2 4.356
n
k!1
50
to
150
50
1.841 3.415
2.677
3.415
-50
to
50
1.328 2.585
200
to
350
2.853 2.548
k!1
k!1
150
to
300
2.179 10.67
k!1
k!1
100
to
250
25
50
to
200
Ratio
4.610 3.357
k!1
k!1
0
to
150
Combined resistance
6.000 30.00
kh
k!1
8.041
k!1
100
to
250
19.53
k!1
13.57
k!1
50
to
200
-25
0
to
150
11.86
k!1
-50
to
100
8k!1
6k!1 30k!1 3k!1 0.55k!1 4k!1
(OoC) (OoC) (100"C) (200"C) (200"C) (200"C)
75.36
k!1
Operating
temperature
range ('C)
Nominal
resistance
Element replaceable
-50
Temperature
COCJ
Connecting
type
Table 3.8 Standard temperature characterisitics for thermister (]IS C 1611)
~
.....
l':!
;::
""
.....
:i!
l':!
oo
t-.:,
350
325
300
275
250
225
230.0
347.8
602.4
920.1
kil
1.445
kil
2.356
468.0
690.7
kil
1.030
kil
1.619
kil
2.634
kil
4.481
kil
kil
kil
1.000
kil
1.120
kil
kil
335.8 1.299
kil
396.8 1.564
1.316 472.0 1.953
kil
1.598 561.0 2.505
1.094
1.274
1.094 1.517
1.273 1.841
1.094 1.515 2.274
1.264 1.841 2.809
Wavelength [pm]
Fig. 3 27 Spectral radiant emittance for black body
L(J..e, T) and a black body LB().e, T) is represented, using emissivity

c:(J..e, T), as
L(J..e, T) = c(.-le, T)LB(Ae, T) =La(tie, S)
(3.6)
Here, Ae is called the effective wavelengh, S is the luminance temperature, T is the true temperature.
(c) Optical pyrometers
Optical pyrometers are specified in ]IS Z 8706, measuring the
luminance temperature for an effective wavelength of 0.65 ,urn. The
operating principle is illustrated in Fig. 3.28. The current to the internal electric lamp is adjusted until the operator determines by sight
that the filament luminance matches the emission luminance of the object being measured. The current to the lamp filament at that time is a
measure of the luminance temperature. A red filter is used to restrict
the measurement wave band to a narrow region which falls within the
visual sensitivity of the human eye. The optical pyrometer is widely
used because it has a simple construction and is relatively accurate.
For luminance temperature S (K), true temperature T (K), and efObjective lens
Grey filter
Red filter
Eye-piece
Temperature indicator
Fig. 3 28 Operation principle for optical pyrometer
96
fective wavelength Ae = 0.65 (.urn), the following equation holds.
-1 =4.52
(3.7)
10- 5 logec
An example of emissivity for an effective wavelength of 0.65 .urn

Table 3.9 Emissivity (wavelength; 0.65.um) (]IS Z 8706)
Material and its stage
Temperature [oCJ
Carbon
Emissivity
0.85 to 0.95
Chrome oxide
900
0.81
Alumina
900
0.18
Ceramics
1200
0.25
1500
0.32
1800
0.38
Iron
Melted iron
1100 to 1 900
0.4
Melted slag
1 400 to 1 830
0.65
800
0.98
1200
0.92
1200
0.37
Solid iron oxide
Iron non-oxide
Melted steel
Bare metal surface
0.35
Oxide film
0.5 to 0.8
Nickel oxide
800
0.96
1300
0.85
Nickel
Copper
0.37
Melted
Non-oxide
0.15
Liquid
0.15
Solid
0.11
Oxide
Tungsten
Platinum
0.6 to 0.8
700
0.46
2 000
0.43
3 000
0.41
1 000
0.29
1300
0.30
1 700
0.32
97
is shown in Table 3.9.

(d) Silicon radiation thermometers
Figure 3.29 shows an example of a silicon radiation thermometer

system. Light from the object of measurement is condensed by a variable focal point objective lens, and a part is sent through a half- silvered
mirror to a finder for determining the position of the object and focussing. After passing through the mirror, the light is restricted by a slit
to control the field of vision to a very small portion of the beam for
temperature measurement. To heighten accuracy and stability, it is
necessary to avoid the influence of external disturbance by visible
light. For this purpose, a filter with narrow band characteristics is positioned in front of the photocell. This filter passes light of wavelengths between 0. 7 ,urn and 1.1 ,urn. The silicon photocell is sensitive
to light in the range of 0.4 ,urn to 1.2 ,urn. By changing the standard
current of the electrical circuit, the effective emissivity is corrected.
The measurement range is 600C to 1 600C.
Protecti ng filter
Sca le gla s
~ Objective lens
Fig. 3 29 System for sil icone radiation thermometer
3 .3
Measurement of Flowl7)-J2)
Flow measurements are often used in industrial processes. Of the

methods of measuring flow, the constriction-type differential pressure
flowmeter has been used from early times. There are many other
forms of flow meaurement having various theoretical bases, such as
the area flowmeter, the positive displacement flowmeter, the electromagnetic flowmeter, and the ultrasonic flowmeter. The type of
flow measuring instrument is selected according to the required accuracy, magnitude of flow, type of fluid being measured (liquid, gas, etc.),
and the properties of the fluid (temperature, pressure, viscosity, corrosiveness, etc.). The principles and characteristics of each type of
flowmeter are described in the following sections.
98
Chap.3
3.3.1 Differential pressure flowmeters

Differential pressure (head) flowmeters, also called constriction
flowmeters, are suitable constriction devices placed in the pipeline.
The flow causes a difference in the pressure in front of and behind the
constriction. This pressure differential corresponds to the amount of
flow, so by measuring this pressure differential, the flow rate can be
determined. These devices have a simple construction and are suitable
for use with liquids, gases, and steam, so they have long been the most
commonly used type of industrial flowmeter. For proper measurement
of flow using constriction devices, the following JIS concerning constriction devices is widely applied.
JIS Z 8762 Measurement of Fluid Flow by Means of Orifice Plates,

Nozzles and Venturi Tubes
(a) General equations for constriction device flow calculations
Figure 3.30 illustrates a constriction device installed in a horizontal pipe that has a circular cross section. The constriction device is a
flat plate (called an orifice plate) having a circular hole that is concentric with the pipe. It is mounted perpendicular to the flow . If the fluid
is incompressible and not affected by viscosity, with a steady-state
flow which completely fills the pipe, then between the upstream cross
section (CD in Fig. 3.30) and the downstream cross section, called the
vena contracta (<]) in Fig. 3.30), the flowing two equations hold:
1
zPVt
1
2
+ Pt =zPV2
+ P2
(3.8)
(3.9)
p,
Perma nent pr
ure lo s
~-p,- ( Exaggerated)
p,
Fig. 3 30
Differential pressure profile with orifice plate
3.3 Measurement of Flow
99
Here, v is the mean stream velocity (m/s), p is the pressure (Pa abs), p
is the density of the fluid (kg/m3 ), F is the sectional area of the flow,
and the subscripts 1 and 2 refer to cross sections CD and @ in Fig.
3.30. These equations are derived from Bernoulli's equation and the
flow equations of continuity.
Equations (3.8) and (3.9) lead to the following relationship between
the volume rate of flow Q (m 3/s) and the pressure differential (p1 - P2)
(Pa).
(3.10)
In a practical equation for real fluids which have viscosity, as well as
for compressible fluids (gases), F2 is replaced by the sectional (rr/4)tP,
and the volume rate of flow Qv (m 3/s) becomes
Qv=cc (rr/4)d
~
1-P
j2
rr d 2jl(
-(p,-p2)-ca-.p p, _ P2)
0
(3.11)
Also, from Qm =Qvp1 , the mass rate of flow, Qm (kg/s) is,

(3.12)
Here, p 1 is the fluid density (kg/m3) for the upstream pressure measurement section, {3 is the ratio of the constriction orifice diameter d (m) to
the pipe diameter D (m) (/3 = d/ D), c is the coefficient of discharge, a is
the flow coefficient (a =cf ~. and E is the expansion correction
factor (for an incompressible fluid, E = 1).
Equations (3.11) and (3.12) are the general equations for flow rate
calculation, and are suitable for use with any type of constriction
device, such as a nozzle or venturi tube, in the same way as with the
orifice plate. The coefficients in these equations are determined empirically. The flow coefficient a is given as a function of the constriction
device diameter ratio {3 and the Reynolds number of the flow for each
set of conditions including the type of constriction device and the tap
position at which the differential pressures are measured (pressure tap
method). The expansion correction factor E corrects for the change in
density of a compressible fluid (gas) after passing through the constriction device, and is determined by the type of constriction device, the
specific heat capacity ratio of the gas, p" P2 and {3. The Reynolds number is a characteristic value of the flow conditions, and the Reynolds
number Reo with reference to the inside diameter D of the pipe is a
dimensionless number given by Eq.(3.13).
100
Chap.3
vD
Rev=-
(3.13)
1/
Here, v is the mean stream velocity (m/s), D is the inside diameter of

the pipe (m), and 11 is the kinematic viscosity of the fluid (m 2/s).
Figure 3.31 shows the relationship between the flow coefficient a and
the Reynolds number Rev for a typical constriction device. Corresponding to each constriction area ratio, /3 2 = (d/ D) 2 , for the respective values
above which they have a Reynolds number, the flow coefficient a takes on a more or less constant value. Generally, flow measurement using constriction devices is done in this region of Rev in which a can be
regarded as fairly constant (that is, can be treated as a mathematical
constant).
""
C'
-;~
c
~ --
0
8c..,
0]
0 .85 10' 2 3 5 1 IO' 2 3 5 1
2 3 5 1
ws
2 3 5 1 1o
..........
0.80
LP' = 0.64
~""- -
0.75
~- -
0.70
- ~--
0.65
0.6
0.5
0.4
0.2
0.60
....... 0.05
1.15
"']a'
~
..-.o
"'c
4:
<1)-
-~
,8' = 0.60
1.10
0.50
1.05
0.20
.l
1.00
-" 1
"
" ..
.g6
0']
0.30
1.00
1.05
c Jl
0.40
,......
......-r.
0.95
0.90
0.85
v
0.80
0.75
I-"'
,8' = 0.490
0.440
0.391
0.360
0.250
0.160
0.050 6
Fig. 3 31 Correlation between flow coeffiuent a and reynolds number R,v
As can be seen from Eqs.(3.11) and (3.12), the magnitude of the

pressure differential (p1-P2) is proportional to the square of the flow
rate. When the flow rate is 1/10 of its maximum, the pressure differential is 1/100 of its maximum. Thus caution is required when measuring
low flow rates, because error in measuring the pressure differential exerts a large influence on the accuracy of the indicated flow rate.
(b) Temperature and pressure correction for gas flows
As shown by Eqs.(3.11) and (3.12), flow rate is principally obtained
3 .3 Measurement of Flow
101
by measuring the pressure differential created by a constriction

device, taking the density of the fluid to be constant. However, the
density of a gas changes with change in pressure and temperature. Accordingly, if measurement is done of a flow at temperatures or pressures different from the design values, the fluid density will also be
different, causing measurement error. If we assume that the fluid is an
ideal gas (taking the compressibility coefficient to be 1), then from the
law of gases, the density under actual use conditions p 1 is
(3.14)
where p is the gas pressure (Pa abs), t is the temperature of the gas expressed as an absolute temperature (K), and the subscript n represents
the design condition.
From Eqs.(3.12) and (3.14),
Qm =
ca{-d2 2Pn~ ~: (pi- P2)
(3.15)
This is to say, by measuring the changing pressure P1 and temperature t1 and the pressure differential (p1 - P2 ) at the same time and then
correcting the results by multiplying by the ratio with the design values for the constriction device, the correct flow rate can be obtained.
(c) Constriction device shape
There are many kinds of constriction devices, but the most typical
among them are orifice plate, nozzle, and venturi tube specified by JIS.
(1) Concentric orifice
A concentric orifice is illustrated in Fig. 3.32.
The orifice plate has a simple construction and is inexpensive. For
these reasons, it is the most generally used among the constriction devices for a wide range of applications. However, because the flow is abruptly dammed up, the orifice plate produces a relatively large pres-
<D
Vent hole
Diameter of orifice
Drain hole
(!) Outer diameter of pipe
Sharp edge
Thickness of edge
(f) Thickness of orifice plate
@ Inner diameter of pipe
Fig.
102
Concentric orifice plate
Chap.3
(a ) Corner tap
( b) Ven a contracta taps
(c) Flange taps
Fig. 3 33 Pressu re tap
sure loss compared with the other constriction devices, and wear from
abrasion at the edge of the orifice is a cause of measurement error.
Thus it is necessary to avoid adulteration of the fluid by solid objects.
Figure 3.33 illustrates methods of tapping the pressure differential
(pressure tapping) based on JIS.
(i) Corner taps: This type is mostly constructed of an orifice plate sandwiched between two annular chambers (carrier rings). The pressure
differential can be tapped by reading the pressure immediately in front
of and behind the orifice plate, averaged by the chambers.
(ii) Vena contracta taps : The pressure is tapped upstream at a distance
equal to the inside diameter of the pipe (lD), and downstream at the
vena contracta, where the cross-sectional area of the flow is the smallest. The vena contracta position has an empirically confirmed fixed relationship with {3 . Depending on the value of {3, the vena contracta position is between 0.34 D and 0.84 D downstream from the orifice plate.
(iii) Flange taps : Pressure tap holes are placed 25.4 mm from the orifice
plate surface on both upstream and downstream sides of the oriffice
plate. With this configuration, flanges with pressure tap holes are required.
(2) Nozzles
There are two types of nozzles: the JIS standard nozzle
(ISA 1932), and the long-radius nozzle (Fig. 3.34). The flow coefficient
of a nozzle a is near 1. It has a strong construction and excellent durability, and is thus suited for use with high temperature and fast moving fluids.
(3) Venturi tubes
The JIS nozzle venturi and classical venturi are illustrated in Fig. 3.35. The constriction is gradual, so this type of constriction device has the advantages that it is difficult for solids contained in the flowing liquid to accumulate on the upstream side of the constriction, and that permanent pressure loss is small. Also, as the shape
is smooth, it has good durability against abrasion. Figure 3.36 compares permanent pressure losses of venturi tubes, orifices, and nozzles.
(4) Quadrant edge orifices
Although not specified in JIS, the quadrant edge orifice is suited for low Reynolds number flows. The flow
103
(a ) Venturi nozzle
( b) Long-radius nozzle
(b)
Fig_ 3 35
Fig. 3 34 Flow nozzle

100
....
90
80
60
'\
r-.."""-
Orifice p la[e
'\ "\
Flow nozzle " \
._ a so
0
~ -~ 30
~~
o': ~ 20
10
"\
'\1'\
-~ ~.. 40
Vemuri tube
' 'I~
70
~
lassical venturi [ube
'
i' \. '\
~en[uri
nozzle
r- -1::-+-
Classical venturi [ube

I
I
I
J
0.2
0.3 0.4
0.5 0.6 0.7
'
0.8
fl
Fig. 3 36 Comparision of permanent pressure loss
among primary flow elements
coefficient and characteristics of this device are given in Fig. 3.31, and
the configuration is illustrated in Fig. 3.37.
(d) Standards for constriction flowmeters
As the flow coefficient of a constriction device a and the expansion correction factor c of a gas cannot be calculated by theory, there
is no other way of obtaining these values than by experiment. By making a constriction device geometrically equivalent to one on which ex104
Chap.3
Fig. 3 37 Quadrant edge ori fi ce
periments have been conducted, and using it under dynamically equivalent fluid conditions (the same Reynolds number), the values for a and
c previously determined by experiment can be used without having to
determine them anew. Standards are experimental data reports which
provide the empirical values of a and c, and the geometric and dynamic conditions under which they were obtained. Some of the main specification organizations are DIN 1952 (Deutsche Industrie Norm, Germany), ASME (American Society of Mechanical Engineers, U.S.A.),
ISO R541 (International Organization for Standardization), and AGA
(American Gas Association, U.S.A.). In Japan, there is the JIS Z 8762
established on the authority of ISO.
Although at a glance there appears to be a large number of experimental data reports issued, the basic data is from DIN and ASME; other standards are nothing but converted forms of this basic data. Constriction flow meters are generelly used without calibration, using the
given a and c values as they are, to obtain the flow rate indirectly by
measuring the pressure differential. Thus it is necessary to pay attention to the details of the standards.
(e) Differential pressure converters
The pressure differential generated by the constriction device is
lead to a transmitter, where it is converted to an electrical or pneumatic signal and sent to a receiving measuring device. When it is not necessary to transmit the signal, the flow rate is displayed by a U-shaped
tube manometer or diaphragm manometer installed at the measurement site. See Sec. 3.4 regarding transmitters.
3.3.2 Float-type area flowmeters
The position of a float placed in a vertical taper tube in which the
fluid to be measured is flowing from bottom to top will move up and
down according to changes in the rate of flow. The float-type area
flowmeter measures the flow rate by detecting the position of such a
float. Figure 3.38 illustrates this principle. The float is in equilibrium,
stationary in the tube. At this time, the annular area A between the
tube and the float (the flow-through area), and the volume flow rate
105
Taper tube
Float
(Equivalent den ity PI Volume VI )
Q,p
A : Flow-through area
A 1 : ectional area of float
Fig. 3 38
Measuring principle of float type area flowmeter
through that area Q are related by the Bernoulli equation, as described

earlier for the pressure differential flowmeter.
Q=AaJ~(Pt-H)
Here, a is the discharge coefficient,
(3.16)
p
is the density of the fluid, and
(P 1 - P 2 ) is the upward and downward pressure differential of the float.
If the float is stationary in the flow, its effective weight is balanced by

the pressure differential applied to the float surface. Thus, because
(3.17)
and
Pt-H= Vf(pf-p)
Af
(3.18)
the pressure differential (P 1-P2) becomes a constant value. From Eqs.

(3.16) and (3.18),
(3.19)
is obtained. The discharge coefficient a is a constant obtained empirically from the shape of the float and the Reynolds number of the
flow. Also because Vf> PJ> p, and Af> are also constants, Equation
(3.19) shows that the flow-through area A (that is to say, the float position) is proportional to the flow rate.
(b) Types of area flowmeters and their structures
(1) Transparent taper tube area flowmeter
Figure 3.39 shows a transparent taper tube area flowmeter constructed simply from a tapered
transparent hard glass tube and a float. The flow rate can be measured
106
Chap.3
Outlet connection
Stopper
Ga ket
Support
Float guide
Float
T aper tube
~~=1~~jtr G asket
Stopper
Inlet connection
Fig. 3 39 Transparent taper tube type area flowmeter
by reading the float position directly on a flow rate scale printed on

the taper tube. The float is made of a material such as stainless steel,
Teflon, or PVC, selected according to the fluid being measured.
(2) Metal taper tube area flowmeter
In this type of flowmeter, the taper tube is made of metal, and the float and its spindle are made as one
movable piece. The spindle then serves as an extension of the float. As
seen in Fig. 3.40, the flow rate can be displayed directly by the position of the moving part viewed through a transparent straight tube
mounted on the upper part of the flowmeter. Alternatively, the motion
of the float extension can be detected by a magnetic coupler and transmitted to an external indicator through a lever or linkage mechanism.
In either case, the metal tube has excellent mechanical strength
against heat and pressure compared with the glass tube, and is also
suitable for use with opaque fluids.
(3) Metal taper tube area flowmeter with converter
As shown in Fig.
3.41, the motion of the moving part inside a metal tape tube is detected by a magnetic coupler. In addition to displaying the flow rate on
an indicator, by attaching a position transducer, the flow rate signal
can be transmitted to a distant location. The signal output by the transducer can be an electrical signal of from 4 rnA to 20 rnA DC, or a pneumatic signal of from 0.2 kgfcm 2 to 1 kg/cm 2
The structure of a magnetic induction position sensor and position-to-electrical current converter used in an electrical position transducer is shown in Fig.3.42. The input is a degree of rotation that is pro3 .3
Measurement of Flow
107
Air bleeder
Support
Indicating cale
T ranspa rent straight tube
Outlet connection
Ga ket
Float g uide
Taper tube
Movable pa n
Stopper
Inlet connection
Fig. 3 40
Metal taper tube type flowmeter

Transducer assembly
with indicator
Float exten ion

Ta per tube
Float
Fig. 3 . 41
Metal taper tube type flowmeter with converter
portional to the rate of flow. A ferrite core moves in correspondence

with the rotation of the input stem, changing the connection between
an alternating current exciting primary coil and a secondary coil. The
exciting coil (primary coil) and receiving coil (secondary coil) are formed on a printed circuit board, and connected electromagnetically by
the U-shaped core. As the core is rotated counterclockwise, the area
of connection between the receiving and exciting coils increases, so the
108
Chap.3
v.
Exciting coil
Printed circuit b o a . d
Detecting coil
Position sensor
0 cill ator
A Amplification
@ Synchronous rectification
(i> Voltage regulator
Position sensor
Fig.
Voltage to current converter
<D
Ferrite core
...
3. 42 Block diagram of posi tion transducer
voltage induced in the receiving coil also increases. This voltage is amplified and rectified, and output as a direct current signal between 4
rnA and 20 rnA that is proportional to the degree of rotation.
(c) Features of area flowmeters
(1) Area flow meters are more able to measure small flow rates and
highly viscous fluids than pressure differential flowmeters. Figure
3.43 shows the empirically determined relationship between the dischage coefficient and the Reynolds number of the flow path for
typical float shapes. For shapes @ and @, the discharge coefficient a is more or less constant until the Reynolds number becomes small.
1.00
Type of float
;::;
"
0.80
"~
0.60
~
u
The discharge coefficient is smaller

than <D. but little influenced by
viscosity.
"'
.s::.
~
Features
The dischange coefficient is large,
but subject to influenoes from
viscosity.
0.40
10
5000 10'
500 IO'
Reynolds number
Fig. 3 43
rh
The discharge coefficient is smaller

than <2), but steady until very
low Reynolds number.
Relation between discharge coefficient and Reynolds number
(2) The flow rate scale is a theoretically uniform scale, so the effective range of measurement can be from 10% to 100% of the maximum flow rate.
(3) Area flowmeters are theoretically influenced by changes in density, so when the density in actual use p' differs from the design
value p, it is necessary to multiply the reading Q on the flowmeter
109
by the following correction factor to obtain the correct flow rate

Q' .
3.3.3 Volumetric jlowmeters

These devices measure volume flow rate by using the space of
fixed volume formed between a rotor, piston, or other such moving
part and its case as a measuring container, that is by continuously filling it up and emptying it again and counting the times this is done.
This type of device is called a positive displacement flowmeter (PD meter).
(a) Operating principle
There are various types of volumetric flowmeters such as rotor
types, piston types, and rotating vane types, each having different shapes measurement space formed by the moving part and its case. Here,
as examples of the rotor type construction used in industrial measurement, the principle of the rotating oval gear type and the rotaing lobe
type is described (Fig. 3.44). In both types the volumetric chambers
are two halfmoon shaped areas formed by the space between the case
and two rotors of identical dimensions. The pressure differential in the
fluid between the intake and outlet sides exerts a rotary torque on the
rotors, which are continuously driven in the direction shown by the arrows. The fluid trapped in the volumetric chambers is discharged
through the outlet, so that with each complete revolution of the rotors, four times the volumetric chamber capacity flows out. Accordingly, by counting the revolutions, the integrated flow rate can be deter-
Volumeric
chambers
Timing gear
(a)
(b)
( a ) Rotating oval gear type
Fig. 3 44
110
( b ) Ro ta ting lobe type
Pictoria l operating principle of P.O. nowmeter
Chap.3
Shafl
( a) Measuring principle
( b ) Spiral rolors
Fig. 3 45 P.O. flowmeter with spiral rotors
mined. Alternatively, from the speed of revolution, the instantaneous

flow rate can be determined.
Figure 3.45 shows a spiral rotor flowmeter. Two spiral gears are
used for rotors, so they rotate at a uniform speed. Thus the outflow
does not pulsate, and little vibration or noise is generated from this
flowmeter.
Oval gear flowmeters have capacities ranging from 0.2 I/h to 1 000
m3/h, and are used for measuring water solutions, oils, etc. The spiral
rotor type, making the most of their special properties, are used for
management of relatively large volume loading and unloading of petroleum products. As their precise construction allows little seepage between the rotors and the case, both types are capable of accurate flow
mesurement over the range from small flow volume to large volume.
(b) Construction and characteristics of oval gear flowmeters
As shown in Fig. 3.46, the rotation of the oval gears forming the volumetric chambers is transmitted by an electromagnetic coupling that is
completely sealed off from the liquid to a transduccer located outside
the case. This transducer is in turn connected to a counting mechanism
(register).
A reduction gear train converts the rotation into a unit of volume
(1/, 10 l, etc.) convenient for display on the register (indicator). The
register consists of a counter representing the measured amount and a
Uniform motion gears
To reduction gear train , pulse transducer a nd register
Fig. 3 46 Typical magnetic coupling assembly
111
display mechanism on which the counter value can be read directly, on

site.
For transmitting the measured value to a distant location, the
transducer is fitted with a pulse generator which outputs a pulse signal
whose frequency is proportional to the flow rate. The output signal
can be sent to any kind of meter or control device. An example of a
pulse generator is shown in Fig. 3.47. A disk with slits rotates between a high-frequency osillating coil and a detecting coil. The rotation of the disk corresponds to the rotation of the oval gears in the
flowmeter. The induced magnetic field between the two coils is intermittently cut, producing a square-wave pulse signal with a frequency
proportional to the number of slits in the disk and the number of rotations of the flowmeter rotors.
Output signal
Wave form
CD
<D
High frequency oscillator

Detector
Fig. 3 47
Rectifier
@ Wave- haping
Oscill at ing coil

Detecting coil
Principle of pu lse Lransducer, type PG30 {Ova l engi neering Co.)
Figure 3.48 shows the error and pressure loss characteristics for
representative medium capacity oval gear flowmeters (Type 55 and
Type 56). For highly viscous fluids, error values are nearly constant
with respect to flow rate. However, for fluids of low viscosity, error
can be seen to vary with flow rate as a result of leakage between the
rotors and the flowmeter case. Pressure loss is high for highly viscous
fluids because of viscous resistance, mechanical friction of the rotors,
and fluid loss.
(c) Features of oval gear volumetric flowmeters
(1) These are theoretically a volume counting type of flowmeter.
They are especially accurate for highly viscous liquids ( 0.5% of
the displayed value; 0.2% is possible with the precision grade),
and can be used as a commercial or standard flowmeter.
(2) The flowmeter does not have to be attached to a straight part of
the pipe.
(3) Because there are moving parts in the fluid stream, it is necessary to place a strainer (filter) at the flowmeter intake to prevent
112
Chap.3
Type 55 O
Type 56
12
Flow rate (m3/ h)
8
(D Gasoline 0.4 cP
Water lcP
o.2
a~
0.1
.t
10
16
20
@ Light oil 3 cP
@ Heavy oil 100 cP
Fig. 3 48 F low characterist ics for viscosity
solid objects from being ingested by the rotors.

(4) Pressure loss is relatively large with the filter attached.
3.3.4 Turbine flowmeters
A turbine rotor placed in a round pipe will rotate at a speed that is
proportional to the flow rate of a fluid flowing through it. The turbine
flowmeter measures flow by detecting the rotation speed of such a turbine. Generally, in industrial measurement a highly accurate turbine
flowmeter is used, which has a low friction, low flow resistance axial
flow rotor (the rotor revolves at a right angle to the flow, with the center of flow as its axis).
(a) Structure of the turbine flowmeter
An example of the structure of a turbine flowmeter is shown in
Fig. 3.49. Inside a pipe- shaped housing, the rotor rides on bearings
held in place by supports fitted in the housing. The rotor diameter is
about the same as the inside diameter of the housing. From four to ten
blades (vanes) are attached to the rotating spindle at the fixed angle.
The blades are magnetized to allow detection of their rotation. Low
friction, durable carbide sleeve bearings are used. The upstream support has radial flow conditioning blades to direct the flow to strike the
vanes of the rotor.
The number of rotor rotations is detected by a pickup coil that is
electromagnetically coupled to the rotor blades, separate by the
non-magnetic housing. That is, as shown in Fig. 3.50, when the rotor
vane passes under the pickup coil, .the magnetic flux density of the coil
changes, generating an induced alternating voltage. A preamplifier amplifies the coil voltage and shapes the waveform, converting it into a
113
Inlet now cond itioner

and rotor support
Fig. 3 49 Cutaway view of turbine flow meter (Example)
0. C . C y
rnr
Magnet
CV\J\
JUUL
ore Amplification Schmiu circuit

--- -0
~ Power supply
--~To
ind icator
Regulated voltage
Rotor
Fig. 3 . 50
Measuri ng principle of turbine slow meter
pulse signal, which is then transmitted to a display device.

Within a certain range of flow rates, the frequency of output pulse
is proportional to the volume rate of flow of the fluid being measured.
The constant of proportionality is called the flow coefficient. This value is accurately measured by calibration with the actual flow rate, and
is recorded on the flowmeter housing in pulses per liter.
(b) Flow rate characteristics
A turbine flowmeter calibration curve is shown in Fig. 3.51. The
ideal characteristic maintains a constant value as the flow rate
changes, as shown by the broken line in the figure. However, the actual characteristic, represented by the solid line, varies somewhat with
the flow rate, because there is a minimum flow rate required to move
114
Chap.3
Ideal characteristic
~ ---------~----------_1
~
f 0.5% of rate
'5
Actua I characteristic
u
1:::
!!
"3:
0
u __ _ _ _ _ _ _ _ _ _ _ __ _ _ __ _
Flow [m'/h)
Fig. 3 51
Typical calibration curve of turbine meter
the rotors due to mechanical friction in the bearings. Normally an accuracy of 0.5% is equivalent to 0.5% of the width at most of flow
coefficient variation. The flow coefficient is more or less constant
when the kinematic viscosity of the fluid is low, but for fluids of high
kinematic viscosity, the change in the flow coefficient becomes exceedingly large. Lets consider why this is so.
Let the angle between the rotor vanes and the direction of flow be
8. If a fluid of density p moving at a velocity of v strikes the vanes, a
force proportional to p Qvtan ()(where Q is the volumetric flow rate) is
generated in the rotational direction of the rotor. The rotor rotates
with the angular velocity w, thus the rotor torque T is
T=cpQ(v tan 8-wr)r
(3.20)
where r is the mean radius of rotation, and c is the constant of proportionality.

The rotor rotates in balance with a mechanical counter-torque T m
resulting from bearing friction, and a fluid counter-torque T f caused
by fluid viscosity, turbulence and so on, thus
cpQ(v tan () - wr)r= Tm+ Tf
(3.21)
Solving Eq.(3.21) for w/Q (this value is equivalent to the flow-coefficient in Fig.3.51 and remains constant in the appropriate use range of
the flow rate, and can be referred to as the ideal characteristic) we obtain the following equation.
w
v tan()
rQ
(3.22)
If A is the cross- sectional area of the pipe, then v = QjA, and Equation
(3. 22) becomes
w
Q
tan()
rA
(3.23)
115
The counter-torque values T m andTf in the second term on the right

side of Eq.(3.23) are zero, and w/Q reduces to a constant value determined solely by the physical dimensions of the flowmeter, thus becoming the ideal flow characteristic. In reality, neither T m nor Tf are zero,
so both exert an influence on the flow characteristic.
As the flow rate increases and regains of turbulence develop, a
counter-torque due to flow resistance becomes dominant. This counter-torque Tf is proportional to the density p and Q2 , and so can be represented by Tf=c 1 pQ2 Accordingly, assuming Tm=O, we obtain
e ___f!__
.!!!...= tan
Q
rA
(3.24)
cr 2
Thus, wjQ is a constant value not influenced by flow rate or fluid viscosity. Hewever, in laminar flow areas, the influence of a counter-torque due to viscosity appears, and at extremely low flow rates, a counter-torque due to friction becomes large, and w/Q changes greatly.
(c) Features of turbine flowmeters
(1) Turbines flowmeters are most suitable for pure liquids of low
viscosity and little fluid resistance. They have good linearity in
turbulent flows, thus their accuracy is stable. Accuracy is generally 0.5% of the displayed value, but 0.2% are also available as
commercial flowmeters. The flow rate range has been standardized as between 0.8 m3/h (8 mm nominal bore) and 7 500 m3/h (600
mm nominal bore).
(2) The rotor response is fast, producing a high resolution pulse signal that is proportional to the flow rate.
(3) It is necessary to have a level straight pipe at least 20D (D: nominal diameters of pipeline) long on the upstream side of the flowmeter for use with a flow conditioner. Also, a strainer (filter) must
be placed in front of the straight pipe section to prevent foreign
objects from damaging the rotors. Figure 3.52 shows an example
of such an installation. Circular flow and an uneven flow speed distribution affect the torque on the turbine vanes, and lower meas-
..
Example of
now conditioner
5D min.
Flow conditioner A
20D
Fig. 3 52
116
~~---~~~++=11~~:::E~l l~-/--~11------ll cross-section

Flow met.er
SD
AA "
d = approx. ~D to
lv
Length and assemblage of each part of meter run
Chap .3
urement accuracy. To guard against this, it is necessary to provide

a fixed flow conditioner upstream from the flowmeter to make the
flow velocity profile symmetrical around the pipe axis.
3.3.5 Magnetic flowmeters
Magnetic flowmeters apply Faraday's law of electromagnetic induction. They measure the flow rate of an electrically conductive fluid
flowing through a magnetic field by detecting the voltage induced in
the fluid. In industrial measurement, magnetic flow meters having
nominal bores ranging from a very small 2.5 mm to a huge 2 000 mm
are in practical use. Figure 3.53 shows the external appearance of
some typical magnetic flowmeters.
Fig. 3 53 Detector and converter of magnetic flowmeter

(Yokogawa :Type YEWMAG)
As shown in Fig. 3.54, a measuring tube having an inner diameter
d (m) whose inner wall is coated with an insulting material crosses at a
right angle through a magnetic field of flux density B (T). If a conductive fluid fills this tube and flows with a mean velocity of v (m/s), and
a pair of electrodes are fitted opposite each other on the wall of the
pipe at right angles to the magnetic field and the flow, then an electromotive force E (V) will be generated between the electrodes. This
electromotive force is expressed as
E=kBdiJ
(V)
(3.25)
where k is a constant intrinsic to the detector and dependent on the

magnetic field conditions.
For volumetric flow rate Q (m 3/s),
3 .3
Measurement of Flow
117
Detector
~ - -- - --C-or~--
Measuring tube ( d in inner di a.)
i
Exciting current
r,~=r:::::;;-1
Power supply
Output ignal
Magnetic fie ld
( B in magnetic nux den ity)
"1\c:i-''"'
Electrode
Jo.... ~------ - ___j ( in signal electromotive force)
<(\0..,
~~~
IJ'I
Fig. 3 54 Principle of measmemenl
v= (tr/~dz = !~z
Cm/s)
(3.26)
Thus from Eqs.(3.25) and (3.26), the electromotive force E becomes

E=k.i..JJ.Q (V)
7r
d
(3.27)
Furthermore, from Eq.(3.27), flow rate Q becomes

Q=l_ trd. E Cm 3/s)
k 4 B
(3.28)
As shown by Eq.(3.28), if the flux density B is constant, then the

flow rate Q is proportional to the electromotive force E. This electr
omotive force is amplified and converted to a universal signal (4 to 20
rnA, DC) or pulse signal by a converter and then output.
Further, it can be proven theoretically that the generated electr
omotive force E is proportional to the mean flow velocity iJ in Eq.(3.25)
if the flow velocity profile in the measuring tube is symmetrical with
respect to the central axis of the pipe, and the magnetic flux density is
uniform.
(b) Excitation method
If a direct current magnetic field is used, the electrodes become
polarized. To avoid this, an excitation method using commercial fre
quency alternating current had always been used in the past. With this
method, the flow rate signal voltage generated between the electrodes
also has the commercial frequency. Influenced by eddy currents gener
ated by changes in the magnetic flux, and other induced noise, the
zero point of the signal fluctuates. Nowadays, to eliminate the influ
ence of various kinds of noise on the flow rate signal voltage and im
prove the stability of the signal zero point, a square wave of lower fre
quency than the commercial frequency is often used. The following are
the main reasons.
118
Chap .3
(1) Using a square wave current having a low frequency that is a

fraction of the commercial frequency as the exciting current, and
detecting the flow rate signal at a stationary magnetic flux value,
the noise based on the rate of change of the magnetic flux (magnetic flux differential noise) can be eliminated, making the flow rate
signal zero point more stable. The noise is generated when eddy
currents flowing in the fluid are not symmetrical about the electrode axis, and when the magnetic flux of the impressed magnetic
field becomes interlinked with the signal detection loop formed by
the fluid, the electrodes, and the electrode signal wires. Under
these conditions, the noise is generated even when the flow rate is
zero, and is thus a source of error.
(2) In the low-frequency excitation method, the signal frequency is
different from the commercial frequency, so the signal is not easily affected electrostatic or electromagnetic noise induced in the
lead wires carrying the signal from the detector to the converter.
(c) Detector structure
Examples of the various types of detectors are listed in Table 3.10.
Detector construction varies somewhat according to the nominal bore
and the application, but it basically consists of a measuring tube lined
with an electrical insulator, electrodes for tapping the generated electromotive force, and a coil and core for generating a magnetic field. Figure 3.55 shows the structure of a detector of small nominal bore.
Nonmagnetic stainless steel pipe, which easily passes magnetic flux to
the fluid, is used for the measuring tube (spool pipe). The lining on the
inner surface is either fluororesin (PF A) or polyurethane rubber, chosen according to the nature of the fluid being measured and the temperTable 3.10 Detectors of magnetic flowmeter (Examples)
[Yokogawa Tech. Rep., 28, 2, p.145 (1984)]
Model
Nominal
diameter
[mm]
ss
s
YM100
2.5~15
YM200
25~100
tI
YM300
150~400
I
I
YM400
500~1
000
LL
YM500
1100~2
600
Symbol
for size
General use
type
Submerg ~anitary Explosion-proof

ible type
type
type
200
I
I
I
I
I
Remarks: Liner materials

--~ ; PF A (Tetra fluoroethylene perfluoro alkoxy resins)
<--- ---> ; Polyurethan rubber
119
Terminal box
Signal lead wire
Housing
Fig. 3 55 Small size detector (Size : S)
ature. The measuring tube, housing, terminal box, and flanges are all
welded together, and the inside of the detector is protected by sealing
with resin or inert gas, thus it is mechanically strong and moisture- proof. For large nominal bores, after the core and coil are attached to the inside of the steel tube, polyurethane rubber is applied
by centrifugal casting to produce an integrated, one- piece construction. Table 3.11 lists the characteristics of lining materials.
The electrodes must be made of non-magnetic material that is also
resistant to corrosion, so SUS 316L plantinum-iridium can be used. Depending on the fluid, hastelloy C, titanium, tantalum or other such
corrosion resistant metal can also be used. An example of electrode
Table 3.11 Characteristics of lining
Lining
PFA
Characteristics
Mechanically strong and smooth, so
free from pipe wall deposits. Resist
ance to heat and corrosion.
Temperature limits: - l0C to + 160
oc (14 oF to + 320F)
Measured liquid
Permeable liquids such as hydrofluoric acid, hydrochloric acid, acetic acid.
Liquids will a tendancy to solidify
will adhere. Corrosive liquids (electrolytes, caustic soda, sulfuric acid, etc)
Polyure- Good abrasion resistance to fluids con- Suitable for water supply, waste wathane
taining solids. Resists weak acids and ter including sewage industrial water,
rubber weak alkalies. Unsuitable for liquids and sea water.
containing organic solvents.
Temperature limits: -10oC to + 40C
(l4 Fto +104F)
120
construction for a PF A resin lined detector is shown in Fig. 3.56. The

measuring tube lining and the electrode seals are molded as one piece,
surrounding the electrode with insulating PF A resin and ensuring the
insulation of the electrode from the grounding steel tube, and improving the strength of the seal against pressure from the fluid. As shown
in Table 3.10, in addition to general use detectors, there are explosion-proof models that can be installed in explosive environments, sanitary models which use sanitary joints that allow easy removal for the
periodic cleaning required in the food industry, and submersible
models coated with tar epoxy that can be installed in places like pits
where there is danger of the device falling into water.
Output leadwire
Tighting nul
In ulati ng sleeve (Tenon)
Washer
Spring
Fig. 3 56 Sealing and insulating construction for electrode
(d) Converters
The basic functions of the converter are to provide and exciting

power source for the detector, and convert the flow rate signal voltage
generated between the electrodes into a proportional output signal (4
to 20 rnA DC), or a pulse signal. However as described above, while the
electromotive force E generated at the electrodes is proportional to the
volumetric flow rate Q, it is at the same time proportional to the flux
density B. Accordingly, it is necessary to either stabilize the exciting
current and make the flux density B constant, or to eliminate the influence of fluctuations in magnetic flux density by means of dividing the
electromotive force E by a signal that is proportional to the exciting
current (i.e. , proportional to the flux density). Furthermore, the electromotive force E is a low-level alternating voltage, and it is also necessary to remove the influence of induced noise as described in Subsec.
(b) above.
As an example, the circuitry of a high accuracy, multi-function
converter incorporating a microprocessor and used in combination with
a low frequency square wave excitation method detector is shown in
Fig. 3.57. The signal waveforms of each part are shown in Fig. 3.58.
The exciting coil of the dectector is driven by a three-valued (positive, zero, negative, zero) constant current from the exciting circuit
121
}:c~______:::._-1---+----1--=:-IV>-...., ,r ,_"""- 'r--r-----~ Status input
L __J----4 Range selecting input
Power
supply
High / Low limit

alarm contact
CD
Amplifier
Watch-dog timer
Exciting circuit
@) Ragulator
Fig.
Timer
A/ D Converter
Power supply
frequency counter
CV
Setting display
D/ A Converter
Pulse output
@ Status output
@ Status input
@ Alarm output
3 57 Circuit diagram of microprocessor-based convener
Commercia l power supply

(a) Exci ting . current
Signal volt age
o ~!:~~~~2:~~~~=t:J~~
(Flow rate measuring mode)
-,-- e, ,,~-.a--
( b) Signal voltage
sampling value
0 ~~__.!iii---E=!--"'1----=-;~_l~-'+---
(Cali bration mode)

(c) Exciting current , 0 _ ...._NI._.._.""-",.__ _;_,_oa,,;r-_;_-""~-
reference voltage
ampling va lue
Fig. 3 58
Signal waveforms of convener main parts
(Fig. 3.58 (a)) . The exciting frequency is synchronized with the frequency of the commercial power supply at either 50/8 Hz or 60/8 Hz. A
signal voltage es that is proportional to the average flow velocity and
the exciting current lex is generated between the electrodes. The sig122
Chap.3
nal voltage is sampled by the multiplexer switch (Fig. 3.58 (b)), converted to a digital signal, and read into the microprocessor. The exciting
current is also read into the microprocessor, and the following calculation is performed.
Here, K is a constant.
Using this caluculation, the influence of fluid noise (noise cause by
fluctuation in contact potential on the surface of the electrodes) and
magnetic flux differential noise can be prevented. Also, the result of
digital filtering of e0 is sent to the digital-analog converter, which converts it to a pulse width signal, and generates a universal 4 to 20 rnA
DC signal. In addition to an electrical current output, a pulse output
that is proportional to the flow rate is also generated.
The power supply uses an insulation-type swiching regulator.
This regulator has the merits of using both alternating current and direct current, and having a wide control range of 20V DC to 130V DC
and 80V AC to 138V AC. Figure 3.59 shows the circuit diagram of the
power supply, which is the heart of the exciting circuit. A three-valued magnetic flux timing pulse controlled by the microprocessor turns
transistor swiches Q1 and Q2 alternately on and off. While the transistor switch is off, the counter electromotive force generated in the exciting coil is stored in a condenser. When the switch is turned on, the
stored energy is reused as an exciting current supply, and power consumption is reduced. Regarding the exciting current, the microprocessor controls the absolute value of the amplitude, and at the same time
controls the flatness r to a value of one, as expressed by in the following equation, thus eliminating magnetic flux differential noise.
r
Every few cycles of the exciting current, the standard voltage
is read and span self-checking is performed (Fig. 3.58 (c)).
For slurries and fluids of low conductivity, the electrochemical potential of the electrodes is constantly changing, which can appear as a
noise in the flow rate signal. This type of noise decreases as the exciting current frequency increases. Thus, although it is accompanied by a
small loss of accuracy, the noise can be reduced by raising the exciting
current frequency. For this purpose, this converter has a 50/2 Hz and
60/2 Hz exciting mode in addition to the normal 50/8 Hz and 60/8 Hz
setting, and the exciting mode suitable for the application can be chosen.
VsTD
123
Timing pu lse
for excitation
Conversion
Fig. 3 59 Scheme of exciting circuit
Using the keyboard and LED display on the converter control panel, parameters such as flow rate range, nominal bore of the detector,
and damping constant can be set, and various functions such as multiple range switching, forward and reverse flow rate switching, and exciting mode switching can be selected.
In addition to the flow rate indication, . the display also indicates
the upper and lower limit flow alarm values and the results of self-diagnosis (exciting circuit short or open circuit, abnormal input signal,
abnormalities in the A/D converter or microprocessor, etc.), and at the
same time outputs an external interface signal.
Figure 3.60 shows an example of an electromagnetic flowmeter
having an integrated construction. This flowmeter is designed to be
small and lightweight , with the converter and detector integrated
into a single unit. It also has the advantage of being easy to use,
because of simplified function . The converter's signal processing circuit is designed for stability of the signal zero point, using a two-valued excitation method of low frequency. It is contained in an airtight
hybrid IC and mounted in a compact converter case.
124
Chap.3
ircuit board assemb ly
Fig. 3 60 Magnetic flowmeter with integrated converter (Typical)
(e) General characteristics of electromagnetic jlowmeters

(1) Electromagnetic flowmeters are not affected by fluid temperature, pressure, density or viscosity, and can output a signal proportional to the volumetric flow rate over a wide measurement range.
(2) The detector has no moving parts and does not obstruct the flow,
so if a suitable lining is used, measurement of slurries or corrosive
liquids is easy.
(3) There is no pressure loss, and a shorter lengh of upstream
straight pipe is need compared to other flowmeters.
(4) Flow in both directions can be measured, with no response delay .
(5) These flowmeters cannot measure gases, oils, or non-conductive
fluids.
3.3.6 Vortex jlowmeters
As shown in Fig. 3.61, if a columnar object (vortex shedder) is inserted perpendicular to the flow, vortices revolving in mutually opposite directions are created alternately on either side of the object.
Downstream, what is called a Kalman vortex street is formed . The number of vortices created per unit of time (vortex frequency) is, within a
certain range of Reynolds numbers, proportitional to the flow velocity.
Thus flow rate can be measured by detecting the vortex frequency .
.~~
v
Vortex shedder
Fig. 3 61
Kalman vortex
Ka lman vortex street
125

Iff is the vortex frequency, dis the width of the vortex shedder,
and Vis the flow velocity, then the following relationship exists.
(3.29)
/=Sr([
Here, St is a non-dimensional constant called the Strouhal number determined by the shape and dimensions of the vortex shedder. With an appropriately chosen vortex shedder, the Strouhal number is constant
over a wide range of Reynolds numbers. This means that within this
range of Reynolds numbers where St is constant, the vortex frequency
f is proportional to the flow velocity V, and is not affected by factors
such as the fluid density and viscosity. Further, the Reynolds number
Re for a typical value of d can be expressed as Re = Vd/ll where ll is the
kinetic viscosity.
The relationship in Eq.(3.29) is also established for the flow in a
measuring pipe of inner diameter D, so for flow rate Q and flow velocity in the constricted part V = Qj {(rc/4)D 2 - dD},
(3.30)
where K=Stf{(rc/4)D 2 -dD}. Accordingly, if St is known in advance,
the flow rate Q can be measured by measuring f.
The vortex shedder is a dominant constituent of the flowmeter. It
must produce strong, stable vortices, and at the same time, the proportional relationship between Q and f shown in Eq.(3.30) must hold over
a wide range of flow rates. For this reason, much effort is concentrated on experimentally determining the most appropriate shape for the
vortex shedder.
Figure 3.62 shows the cross-section shapes of several vortex shedders currently in practical use. What is common to all of these is that
the surface facing the flow has sharp edges, so that the point at which
the vortices are shed is well-defined, improving the linearity of the
Strouhal number St( St remains constant even if the Reynolds number
fluctuates). Also, the width of the vortex shedder facing the flow d,
taking advantage of the vena contracta effect near the walls of the
pipe, acts to average the flow velocity distribution in the longitudinal
direction of the vortex shedder, so by choosing the right shedder
width d in relation to the bore of the pipe D, a stable Karman vortex
street can be formed even in a round pipe.
Figure 3.63 shows the relationship between St and the Reynolds
number for a flowmeter using the vortex shedder having a trapezoidal
cross section described below. For Reynolds numbers between 2 x 104
126
Chap.3
~0~
D
D
(d)
(a)
(b)
(c)
Fig. 3. 62
(e)
D~
(f )
Section of vortex shedders

(Example)
Measura bl e range
~
0.3
.D
E
:l
'"
0.2
....__
Normal operating range
------
..c
:l
0
!:::
"'
0.1
5Xl0 3
Fig. 3 63
2XIQ 4
Reynolds number
7 Xl0 6
Relationship between Strouhal number and Reynolds number
and 7 x 106 , St is more, or less constant, and this represents the normal
operating range of the flowmeter.
Various types of vortex frequency detectors are in use, in combination with vortex shedders. The major types of sensors are described in
Table 3.12. They are broadly classified into those that detect changes
in flow velocity caused by vibrations in the fluid resulting from vortex
shedding, and those that detect changes in pressure.
(b) Flowmeters that use the stress detection method
As an example of an actual device, a flowmeter in which the alternating stress created in the vortex shedder by vortex formation is detected by a piezoelectric element is described. The external appearance
of this flowmeter is shown in Fig. 3.64.
Figure 3.65 shows the vortex frequency
(1) Stress detection principle
detection principle. If the fluid flows perpendicular to the page, an
alternating lift is applied to the vortex shedder that has the same frequency as the vortices shed from it. This alternating lift produces a
change in stress inside the vortex shedder. The frequency of this
127
Table 3.12 List of sensors

Quantity detected
Detection system
Flow velocity change
cooling the heating element
Sensor
Hot wire
Thermistor
-----------------------------------Detecting sound velocity change

Ultrasonic wave
Diaphragm +piezoelectric
element
Detecting differential pressure
Diaphragm+ capacitance
Diaphragm+ inductance
Pressure change
Distinguish movement
Flag+ strain gauge

Ball+ inductance
Distortion produced in vortex

shedder
Strain gauge
Stress produced in vortex

shedder
piezoelectric element
Convener assembly
Body a sembly
Fig. 3 64 Vortex flowmeter (Yokogawa: Type YEWFLO)
change in stress (that is to say, the vortex frequency) is detected by a

piezoelectric element mounted inside the vortex shedder.
The magnitude of the lift is proportional to the square of the flow
velocity V and the density p, and the peak value of the alternating lift
FL is expressed by
FL=
~CLpV2 dD
where cL is a dimensionless constant, D is the pipe bore and d is the

width of the vortex shedder.
128
Chap.3
#I
Piezoelectric element
ompression
Tension
Section of vortex
hedder
nFlow direction
(a) Section
Fig. 3 65
( b ) Stress d i tribution in
piezoelectric element
Sensing principle of vortex
hedding frequency
The average stress aM created in the piezoelectric element, and

electric charge q induced in the piezoelectric element are expressed by
the following equations.
Here, K is a constant determined by the shape of the vortex shedder
and support method, d0 is the piezoelectric constant, and S is the area
of the piezoelectric element. This electric charge is processed by an
electronic circuit, and the vortex frequency is detected.
(2) Structure
As shown in Fig. 3.64, this flowmeter consists of a detector body assembly made up of a section of pipe and vortex shedder
with a built- in piezoelectric element, and a converter assembly with a
built-in amplifier. Figure 3.66 shows the basic mechanism of a vortex
flowmeter.
(3) Characteristics
Figure 3.67 presents an example of actual airflow
measurements. A volumetric flowmeter of 0.2% accuracy was used as
a standard. The broken line represents the Strouhal number obtained
for water flow, and is in agreement with the characteristics of airflow.
In addition, the linearity of the Strouhal number is good.
This flowmeter has a solid construction that integrates the vortex
shedder and detector into a single unit, has no moving parts. The detector (piezoelectric element) does not come in contact with fluid, so it is
highly reliable, and can be applied to the measurement of liquids,
gases, and stream.
(c) General characteristics of vortex flowmeters
(1) The output signal of a vortex flowmeters is directly proportional
to the flow rate, and a pulse signal can be obtained.
(2) The output signal indicates the volumetric flow rate, and is not
influenced by temperature, pressure or the type of fluid.
129
Piezoelectric element
Body
Vortex shedder a sembly

Flow d irection
Fig. 3 66
Ba ic mechanism of vortex now meter

Pressure tap
Pressure air n ow
-
Okgf/c m'
-o- 2kgf/c m'
-o- 4kgf / cm'
0.170
~ 6kgf/c m'
~------:--~0.~:-~_:;::
Water
0.165
Fig. 3 67
R.o
Characteristics for air flow
(3) The applicable Reynolds number range is wide, and accuracy is

high.
(4) The pressure loss is relatively low, but is affected by flow velocity distribution, so a straight section of pipe is required upstream,
15D to 40D, depending on pipe conditions (D: nominal bore of the
flowmeter), and downstream, 5D.
3.3.7 Ultrasonic flowmeters

With ultrasonic flowmeters, ultrasonic waves are generated in the
130
Chap .3
fluid from outside the pipeline, and the flow rate is determined by detecting externally the change in the transmitted waves or the reflected
waves caused by the flow velocity. At this time, they are mainly used
to measure the flow of water. There are two typical methods in practical use, classified broadly by the principle of measuremant they employ. One, called the transit time differential method, obtains the flow
rate by measuring the difference in arrival times of ultrasonic pulses
between detectors set facing each other obliquely to the pipe axis on
the outside pipe walls. The other method uses the Doppler effect. By
measuring the difference in frequencies of the transmitted waves and
those reflected back from foreign objects in the fluid, the velocity of
those objects, and thus the velocity of the fluid, can be determined.
(a) Transit time differential method
Figure 3.68 shows the basic configuration of the measuring device.
A pair of detectors, P 1 and P2 are mounted facing each other on the exterior of the pipe at an angle to the pipe axis, and in contact with the
outside walls.
Flow
Fig. 3 68
Block diagram of ultra ionic now meter

(Tran it time differential method)
Figure 3.69 shows an example of the structure of a detector. An

ultrasonic vibrator is attached to a plastic shoe. When the vibrator receives ultrasonic excitation signals from the converter, it emits ultrasonic wave pulses. These pulses travel through the plastic shoe and
through the walls of the pipe and arrive at the facing detector. As the
ultrasonic vibrator, a piezoelectric element that has a resonance frequency of between several hundred kilohertz and about one megahertz
is used. The material of the vibrator is usually lead zirconate.
Detectors P1 and P2 incorporate piezoelectric vibrators and can
both transmit and receive ultrasonic waves. Ultrasonic pulses are transmitted along the paths shown in Fig. 3.68. The transmission is controlled by the opening and closing of the switches in the converter.
The transit times T2 (from P1 to P2) and T2 (from P2 to P1) are ex3.3 Measurement of Flow
131
Compression spring assembly
Fig. 3 69 Typical con tructi on of detector
pressed by the following two equations.

D/cos e
c+ Vsin 8
D/cos 8
v sine
c-
(3.31)
Here, D is the nominal bore of the measurement tube, c is the speed of

sound through the stationary fluid, V is the mean flow velocity, and e
is the angle of incidence of the ultrasonic pulse. However, Equations
(3.31) disregard the transit time through the pipe walls and the detector shoes.
For liquids, the speed of the sound pulse c (about 1 500 m/s) is nearly three orders of magnitude lager than the flow velocity V (several
m/s), and is affected by changes in the temperature and density of the
fluid. Thus correct values of the transit times T 1 and T 2 cannot be obtained using Eq. (3.31) as they are. To eliminate the term c from the
equations, the inverses of the transit times (that is, the frequencies)
are taken, and their difference Jf becomes.
Ll/=_1___1_= V sin 28
T, T2
D
(3.32)
There are various signal processing techniques available for taking the
inverse of the transit times T, and T 2 , but methods using the phased-locked loop (PLL) are often used. As shown in Fig. 3. 70 the values of T, and T 2 are converted into the output signals of two voltage
controlled oscillators (VCO). These two output signals have the frequencies of / 1 and / 2 The VCOs control their outputs such that T- N/f
is equal to zero, thus / 1 = N/T, and / 2 = N/T2 Increasing the value of N
improves the resolution of Jf ( = / 1 - / 2 ). To obtain the frequency difference Jf, we use the following equation.
NVsin28
D
For example,
132
if D=1m,
Chap.3
V=2m/s,
(3.33)
8=70 , and N=1 000,
then
AI
<D
Receiver
Fig. 3 70
Synchronizer
Block diagram of ultrasonic flowmeter (Using PPL method)
.df= 1 285.5 Hz. From Eq.(3.33), flow velocity Vis given by
D
V = N sin 2 8 (11 - /z)
(3.34)
This equation does not include the term c, and demonstrates that given
nominal boreD and the angle of incidence of the ultrasonic wave 8, the
flow velocity V can be measured. The volumetric flow rate of a full
pipe Q is obtained by multiplying the measured flow velocity by the
cross-sectional area of the pipe. However, the velocity V representing
the ultrasonic flow measurement averages the flow velocities along the
path of the beam, and is different from the flow velocities averaged
over the pipe cross section. If V is the flow velocity as measured by
the ultrasonic method, and V is the average flow velocity over the pipe
cross section, then the ratio of these two values K = VjV can be called
the flow rate correction factor. The following empirical formula is
widely used to express the relationship between the flow rate correction factor and the Reynolds number Reo for a pipe of inner diameter
D.
K=l.119-0.0lllog Reo
(3.35)
Here, Reo= VDjv, and vis the kinetic viscosity of the fluid.
According to Eq.(3.35), even a ten-fold change in the flow velocity, that is, Reo. the value of K changes by a mere 1% or so. Accordingly, from Eq.(3.34), the flow rate Q that we wish to determine is finally
obtained from
1r 2 V
1rD3
(
)
(3.36)
Q=4D K= 4KNsin28!1 -!2
Flowmeters using the transit time differential principle become problematical when there are particles or bubbles in the fluid that scatter
133
ultrasonic waves, so care must be taken regarding this.

(b) Doppler method
In Fig. 3. 71, assuming that minute particles in the fluid flow together at the same flow velocity Vas the fluid itself, if transmitter P 1
emits an ultrasonic wave of frequency It at an angle of 8 to the flow,
then the particles approach P1 at a velocity of Vcos 8, and at the same
time approach the receiver P 2 at the same velocity. Thus the frequency
of the signal /r received by P 2 , given c>V, can be expressed by the following equation.
Ir
v cos 8 j, ~
c- V cos 8 t
c+
(1 + 2v cos
8 )tt
c
(3.37)
Here, c is the speed of sound through the stationary fluid, and 8 is the
angle of incidence to and reflection from the reflecting particles.
Tra n miner
Fig. 3 71 Operating principle using doppler effect
From Eq.(3.37), the Doppler frequency fd is

/d=/r- ft= 2/t
COS
8 V
(3.38)
and is thus proportional to the flow velocity V. Accordingly, flow velocity Vis represented by the following equation.
2ft cos
e1d
(3.39)
The flow rate is obtained by correcting for the difference between V

and the average flow velocity.
The Doppler method requires that fluid contain suspended particles, and thus cannot measure clean fluids as can the transit time
differential method.
(c) Characteristics of the ultrasonic flowmeters
(1) The detectors are attached to the outside walls of the pipe, requir-
ing no alterations of the existing piping. Furthermore, they allow
134
Chap.3
measurement of the flow rate without stopping the flow.

(2) The detectors do not come in contact with the fluid, so there is
no concern for corrosion or material adhering to them. Also, there
is nothing to obstruct the flow, so no pressure loss results from
their use.
(3) Disturbances in the flow velocity distribution affects measurement accuracy, so an appropriate length of straight pipe is required upstream and downstream of the place where the detectors
are installed.
(4) In process industries, flowmeters for use with liquids mainly measure water flows. Based on the particular characteristics of their
respective measurement principles, use of the two types of ultrasonic flowmeters is differentiated according to the characteristics
of the fluid measured in the following way.
Transit time differential flowmeters: Mainly clean water, water for
industrial use, raw water for water purifying plants, water for agricultural use, and other types of clean water.
Doppler flowmeters: Mainly drain water, factory effluent, and other
such dirty water which contains foreign particles.
3.4
Measurement of Pressure 33 J
In process industries, pressure itself is an object of measurement,

but in addition, temperature, flow rate, liquid level, and other measured variables are frequently measured indirectly by measuring pressure. Thus it can be said that pressure is the most fundamental measured variable.
The units of pressure standardized in the International System of
Units (SI) are the Pascal (Pa), and the bar, which can be used together
with the Pascal. Table 3.13 is a conversion table for the major units of
pressure now in use, including those of the International System of Units (SI).
When pressure is detected by a pressure transmitter, it can be displayed as absolute pressure, as a pressure differential, of as a gage
pressure. Absolute pressure is measured using vacuum as a reference
value, and is used in equations showing the relationships among pressure and other variables. Differential pressure represents the difference from some (non-zero) reference pressure. Gage pressure is pressure measured using atmospheric pressure as the reference pressure.
Thus, to show absolute pressure, the pressure indicated by gage pressure is added to the atmospheric pressure at that time. When it is necessary to distinguish absolute pressure from gage pressure, "abs"is appended to the pressure value to indicate absolute pressure, and"G" is
3.4 Measurement of Pressure
135
Table 3.13 Conversion table of pressure units

Pa
bar
atm
kgf/cm'
1 X 10- 5
mmH 20
mmHg
or Torr
1. 01972 X 10- 5 9.86923 X 10- 6 1.01972 X 10-1 7. 50062 X 10- 3
1 X 105
1.01972
9.80665 X 104 9.80665 X 10-1
9.86923 X 10- 1
1.01972 X 104
7.50062 X 102
9.67841
1.0000 X 104
7.35559 X 10 2
1. 03323 x 1o
7.60000 X 102
7. 35559 X 10-2
10- 1
1.01325 X 10 5
1.01325
1.03323
9.80665
9.80665 X 10-5
1. 0000 X 10-
9.67841
1.35951 X 10- 3
1.31579 X IQ- 3
1. 33322 X 102 1. 33322 x 1o-3
10- 5
1.35951 X 10
appended to indicate gage pressure.

3 .4.1 Pressure transmitters
A pressure transmitter is a device which converts the pressure of
a fluid that is the object of measurement into a standardized signal,
and then transmits the signal. Pressure transmitters are classified into
pneumatic types and electronic types.
(a) Pneumatic pressure transmitters
Pneumatic pressure transmitters are divided into absolute pressure types, differential pressure types, and gage pressure types (generally called pressure transmitters), according to the methods of pressure display described above. The operating principle of pneumatic
differential pressure transmitters is snown in Fig. 3. 72.
A : Twin diaphragm cap ule

B :Force bar
C : Diaphragm eal
D : Flexure
E: Range bar
F : Range wheel
G : Flapper
H:
ozzle
J : Pneumalic amplifier
K : Feed back bellows

L : Zero adjusling screw
Low pres ure
Fig. 3 72 Cutaway showing the principle of pneumatic differential pressure transmitter
136
Chap.3
The pressure of the fluid passing through the high pressure side
union is applied to the left diaphragm of the twin diaphragm capsule,
and the pressure of the fluid passing through the low pressure side union is applied to the right diaphragm of the capsule. If the pressure of
the high-pressure side fluid increases or the pressure of the low-pressure side fluid decreases, then the thrust on the twin diaphragm from
the left towards the right increases. This thrust moves the force bar,
which rotates slightly counterclockwise, with the diaphragm seal acting as a fulcrum, and as a result, the gap between the nozzle and the
flapper becomes smaller. Air normally flows out from the nozzle
through this gap, but if the gap narrows, the pressure inside the nozzle increases and the output of the pneumatic amplifier increases by a
larger amount. Part of the output is applied to the feedback bellows,
where it is converted to a force. This force is applied to the range bar,
which using the range wheel as a fulcrum, acts to pull back on the flapper, thus reducing the pressure inside the nozzle. The end result is
that, compared with before the pressure differential is changed, the
gap between the nozzle and the flapper has narrowed slightly, and the
output has increased and stabilized. In this way the two-forces are balanced; that is, negative feedback increases the linearity of the input-output characteristics of the pressure differential transmitter, and
improves the characteristics.
The zero adjusting screw is set so that the output corresponding
to the minimum pressure differential is 0.2 kgf/cm 2 The position of
the range wheel is adjusted so that the output corresponding to the
maximum pressure differential is 1.0 kgf/cm 2 If the range wheel is set
to a higher position, the amount of feedback is increased and the range
widens. Figure 3. 73 shows how the input-output characterstics change
with the position of the range wheel.
1.0,.----y-----,--------~
0. 8 1--~
;::..
.><
0.6
:l
is. 0.4
:;
Q.
8 0.2 ~----i-----i--------1
Rwuced span
O L---~---~~
ex3p~
a n~d~w~s~p~
a n~~
Differential pres ure
Fig. 3 73
3.4
Relationship between input-output characteristics and range wheel position
Measurement of Pressure
137
(b) Electric pressure transmitters

Electronic pressure transmitters are also grouped into the same
three types as are pneumatic transmitters. The operating principle of
electronic pressure transmitters is shown in Fig. 3. 74. Pressure is applied to the two diaphragms of the twin diaphragm capsule. This pressure is transmitted from the twin diaphragms to the sensing diaphragm through the sealing liquid. Two fixed electrodes are placed symmetrically on the left and right side of the insulator and an electrical
capacitance is formed between these electrodes and the sensing diaphragm. If there is a pressure differential between the high pressure side
and the low pressure side, the position of the sensing diaphragm will
change in proportion to that differential, and an electrical capacitance
differential will be produced. This capacitance differential is converted
an electrical current and output as a signal of 4 rnA to 20 rnA DC, corresponding to the measured pressure differential range. Zero setting
and span abjustment are done in the transmitter.
Low pressure
I
:-~..:.;.;..;~
L------Fig. 3 74
Diaphragm in contact with

process nuid
Structure for electronic differential pre sure sensor
3.4.2 Types of pressure detectors

To detect fluid pressure, the fluid is applied to a pressure sensing
element (detector), which converts the pressure to a force or a displacement. In pneumatic industrial instruments, a pneumatic input signal is
converted by the pressure detector to a displacement, and recorded,
displayed, or used perform a control action.
In industrial instruments, pressure detectors constructed of metal
whose degree of elastic deformation is proportional to the pressure applied to them are most often used.
138
Chap .3
(a) Diaphragms
A diaphragm is a thin plate fixed in place around its periphery.

The plate is corrugated to increase its effective area and give it good
elastic characteristics with respect to pressure. A diaphragm capsule
used in a pressure transmitter is constructed as shown in Fig. 3. 75.
Two diaphragms are attached with a metal backup plate placed between them. The backup plate has a small cavity filled with silicone
oil. This silicone oil transmits the pressure applied to the two diaphragms, and the oil in the cavity provides a damping action on the transmission of the pressure, thus increasing the stability of the entire
device.
Fig . 3 75
Twin -dia phragm ca psule cut away
The amout of displacement of the diaphragm is small, and normally it is used as a pressure-to-force converter. However, by combining the action of several diaphragms, thus increasing the displacement,
they can also be used to drive the indicator of a meter.
(b) Bellows
Bellows are made by stamping the outer surface of a thin-walled

brass, phosphorous bronze, or stainless steel pipe. The belows themselves have the characteristics of a spring, but when used as a pressure-displacement conversion device, they are used in combination
with a spring as shown in Fig. 3.76 to achieve good conversion linearity. The effective cross- sectional area of the bellows to which pressure
is applied given by the following equation.
3.4
Measurement of Pressure
139
Pressure
Fig. 3 76 Bellows paired with spring
Here, A is the effective cross- sectional area of the bellows, R 1 is the

outer radius, and R2 is the inner radius.
If two bellows are combined, and a vacuum is applied to one, and
the pressure to be measured is applied to the other, then absolute pressure can be measured.
(c) Bourdon tubes
A Bourdon tube is a seamless flat tube of bronze, phosphorous

bronze, brass, stainless steel, or beryllium bronze, monel, Ni-Span
or other such material, curved into an arc and fixed at one end. If an
internal pressure is applied to this tube, its cross section will change towards a circular shape, and the free end of the tube will move outwards. This displacement of the free end is proportional to the internal
pressure within the limits of elasticity. The displacement, usually from
about 1 mm to 8 mm, is amplified to drive an indicator. Bourdon tubes
are divided into C- shaped types, spiral types, and helical types (Fig.
3. 77). Of the three types, the C- shaped type is also used as a pressureforce converter.
(a) "C"Shaped type
( b ) Spiral type
( c ) Helical type
Fig. 3 77 Bourdon tubes

140
Chap.3
3.5 Measurement of Liquid Levef3 3 l
There are many methods of measuring liquid level, each having its
particular characteristics. A method is selected for use according its
suitability for the liquid being measured.
3.5.1 Float liquid-level meters
A float liquid-level meter, as shown in Fig. 3. 78, measures the
level of a liquid directly from the position of float resting on the liquid
surface. It is sutitable for measuring the level at a dam, clean water
tank, and so on. The up and down movements of the float are indicated
by the degree of rotation of a pulley connected to the float by a wire
cable.
Floal
Coomer weigh!
Fig. 3 78
F loal liq uid level meter
3.5.2 Pressure differentia/liquid-level meters

The principle of pressure differential liquid-level meters is that
the static pressure at any point in a liquid is equal to the product of
the distance from that point to the surface of the liquid, the density of
the liquid, and the acceleration due to gravity. Thus, if the liquid density and the acceleration due to gravity are known, the distance to the
surface (that is to say, the liquid level) can be determined from the
pressure. The method of tapping the pressure is different for open
tanks and closed tanks.
(a) Open tanks
A pressure differential transmitter or pressure transmitter is connected to the tank by a pressure tapping tube, as shown in Fig. 3. 79.
When a pressure differential transmitter is used, the liquid is tapped
for the high pressure side, and the open air is tapped for the low pressure side.
3 .5 Measurement of Liquid Level
141
Min imum
liquid
level
p ogh '
Fig. 3 79
Differenlial pressure
Liquid level mea urement of open ta nk
The following relationship exists among the pressure (gage pressure), the density of the liquid, and the acceleration due to gravity.
P= p1g(H + h1)
Here, P is the pressure (gage pressure, Pa), p1 is the density of the liquid (kg/m 3 ), g is the acceleration due to gravity (m/s 2 ), H is the plumb
distance between the minimum liquid level and the surface (m), and h1
is the plumb distance between the minimum liquid level and the pressure detector (m).
The relationship between the input and output of the transmitter
is linear, so the output of the transmitter changes in direct proportion
to changes in the liquid level.
(b) Closed tanks
When measuring the liquid level in a closed tank, the pressure

differential transmitter is used. The low pressure tap is the pressure
of the gas above the liquid in the upper part of the tank (Fig. 3.80).
The pressure of the gas is also applied to the high pressure tap at the
same time, so when taking the pressure differential, it cancels out and
so does not affect the output of the transmitter. Thus, the level of the
liquid can be known from the pressure differential.
High pressure tap pressure
Low pressure tap pressure
Pressure differential
PH= Pt9(H +h1) +Pc

PL=Pc
PH-PL= plg(H +ht)
Here, P c is the pressure of the gas in the upper part of the tank.
If condensation from the gas in the upper part of the tank collects
inside the tapping tube, the low pressure tap pressure in the tube will
change and the output of the pressure transmitter will be affacted. To
avoid this problem, the condensation is collected in a drain pot.
Another method involves using a relatively heavy liquid that does
not easily evaporate to fill the tube, as shown in Fig. 3.81. The pressure of the gas in the tank is then applied to the pressure detector
through this liquid. This method is called the wet leg method, as op142
Chap.3
Minimu m
liquid- level
lransmiuer
Drain pot
p,gh ,
D ifferentia l- pressure ( PH- Pt)
Fig. 3. 80
Minimum
liquid -level
Fig. 3 81
3.5 Measurement of Liquid Level
Liquid-level measurement
using a dry-leg system
D ifferentia l- pressure
transmitter
Liquid-level measurement
using a wet-leg system
143
posed to the dry leg method described above. In the wet leg method,
the pressure is calculated in the following way.
High pressure tap pressure
Low pressure tap pressure
Pressure differential
PH= Pig(H +hi)+ Pc

P L = Pz9hz + P G
PH- P L = Pig(H +hi)- Pz9hz
Here, p 2 is the density of the liquid in the wet leg (kg/m 3), and hz is the
height of the liquid in the wet leg (m).
When using pressure differential liquid-level meters, care must be
taken regarding the following points.
(1) The pressure differential transmitter must be positioned below
the minimum liquid level.
(2) Corrections must be made for changes in the density of the liquid.
(3) If there is a pulsating motion in the liquid, the output of the
transmitter will be unstable.
(4) The tapping tubes should be as straight as possible so as not to
trap air. Also, use of points that might leak should be minimal.
3.5 .3 Displacer liquid-level detectors

A displacer liquid- level detector is a round, tubular displacer combined with a force detector (Fig. 3.82). The buoyancy applied to the
displacer is equal to the weight of the liquid it displaces. Thus, if the
cross-sectional area of the displacer is constant along the radial axis of
the displacer, the notational force has a linear relationship with the liquid level. Accordingly, by measuring the notational force, the liquid
Pneumatic amplifier
Restriction
Nozzle
torage tank
Fig. 3 82
144
Liquid level measurement with displacer detector (Force balance type)
Chap.3
level can be known.

Referring to Fig. 3.82, the force applied to the displacer can be expressed by
F=pgA(H-h)- W
where F is the perpendicular upward force on the displacer (N), p is

the density of the liquid (kg/m3), g is the acceleration due to gravity
(m/s 2), A is the cross-sectional area of the displacer (m 2), H is liquid
level measured from the reference position (m), h is the position of the
bottom of the displacer measured from the reference position (m), and
W is the weight of the displacer (N).
The flotation force applied to the displacer is transferred to the
force bar, which, using the diaphragm seal as a fulcrum, transmits the
force to the force detector.
The displacer liquid level dectors have the follwing special features.
(1) They can be used to measure high-temperature, high-pressure
liquids.
(2) They have good sensitivity, good accuracy, fast response as well.
(3) They are suitable for short measurement spans.
(4) The measurement range can be changed by changing the length
of the displacer.
(5) If impurities or other objects adhere to the displacer, its volume
will change, generating measurement error.
(6) Corrections must be made for changes in the density of the liquid.
3.5.4 Purge-type liquid-level meters
Purge-type liquid-level meters are used for corrosive liquids, liquids that contain suspended objects, high viscosity liquids, and for
tanks that have already been constructed, underground tanks, and so
on, where installing tap tubes presents a problem.
A purge-type liquid-level measurement is illustrated in Fig. 3.83.
A bubble tube is inserted into the tank, and a fixed flow of air is
forced through the tube such that bubbles emerge from the end of the
tube. If the air flow rate is not extremely large, the air pressure in the
bubble tube can be considered to be the same as the pressure of the liquid at the end of the tube. Thus the liquid level can be measured by
measuring the air pressure.
Various materials, can be used for the bubble tube, such as stainless steel, mild steel and rigid PVC. The material used is chosen accoding to the nature of the liquid. Also, when air is not suitable as a purging substance, nitrogen or some other inert gas can be used.
145
Air-purge set
Fig. 3 3
Purge type liquid-level measurement y tem
To reduce the error associated with the air flow rate, the pressure
tap is placed near the bubble tube. Corrections are necessary when
there are changes in the density of the liquid.
3.5.5 Ultrasonic liquid- level meters

As shown in Fig. 3.84, an ultrasonic liquid-level meter measures
the time required for an ultrasonic wave emitted by a transducer and
reflected off the surface of the liquid to return to the transducer, thus
determining the liquid level. The relationship between the liquid level,
Fig. 3 84 Operating principle for ultrasonic liquid-level meter

146
Chap.3
speed of the ultrasonic wave, and time is given by the following equation.
2(L-H)=tu
Here, L is the distance from the minimum liquid level to the ultrasonic
transducer (m), H is the liquid level (m), t is time between the emission
of the signal and its subsequent reception at the transducer (s), and u
is the propagation velocity of the ultrasonic wave (m/s).
The velocity of the ultrasonic wave u varies with the type of gas it
travels through and the temperature, so the temperature of the gas
must be measured and corrections applied according to the results.
Ultrasonic liquid- level meters have the following special features.
(1) Measurement can be accomplished without touching the liquid.
(2) There are no moving parts, so inspection and maintenance are
easy.
(3) They are small and light, so installation and operation are easy.
3.5.6 Capacitance liquid- level meters
The electrostatic capacitance of two facing electrodes varies with
the dielectric constant of the material separating them. As shown in
Fig. 3.85, if the space between two concentric tubular electrodes contains a liquid, a change in the level of that liquid will change the capacitance between the electrodes as described by the following equation.
C- 2Jr{(e2-el)H+e,L}
ln (D/d)
-
Here, C is the electrostatic capacitance between the electrodes (F), c1

is the dieletric constant of gas (F/m), c2 is the dielectric constant of the
liquid (F/m), His the liquid level (m), L is the height of the electrodes
(m), D is the diameter of the outer tubular electrode (m), and d is the
Fig. 3 . 85 Operating principle for capacitance liquid-level meter
147
diameter of the inner tubular electrode (m).

The terms L, D, d, E 1 , and E2 are constants, so by measuring the
capacitance C, the liquid level H can be determined. The surfaces of
the electrodes are coated with an insulating material that must be selected considering the nature of the liquid being measured.
Capacitance liquid-level meters have the following special features.
(1) The detector has a simple construction.
(2) They have no moving parts, and thus a long service life.
(3) Changes in the dielectric constant that are caused by changes in
the temperature and density of the liquid being measured produce
errors in measurement.
3.6
Measurement of Displacement and Angle34l- 39l
Displacement and angular movement are objects of measurement

in themselves, but their measurement also is used in the displacementconversion parts of converters employed in the measurement of pressure ,flow rate and other industrial variables, and in the position-conversion part of recording device servo-mechanisms. Many methods
have been developed to measure displacement and angles; some representative methods are described here.
3. 6.1 Resistance potentiometer methods
In resistance potentiometer methods, displacement and angular
movement are converted to changes in the brush position of a linear or
circular slide resistor, thereby converting them to changes in electrical
resistance. These devices have a simple construction and have been
used from early times.
In wire coil-type meters, metal wire that has a low thermal coefficient of resistance, such as manganin wire, or constantan wire. However, although their linearity is good, a weak point is that noise is easily
generated by poor brush contact. CP types use conductive plastic resistance elements formed by heat-shaping a mixture of conductive particles, such as carbon, and plastic pellets. These devices have the merits of high resolution, and little wear due to smooth contact. Also, it is
possible to raise the resolution and stability of the wire coil-type devices by coating with this sort of conductive plastic.
3.6.2 Electromagnetic induction methods
The induction methods employ the change in inductance resulting
from displacement of a ferrite core or short-circuit ring. Although
there are many methods based on device structure, we will introduce
148
Chap.3
only a few of them here.

(a) Differential transformer method
As shown in Fig. 3.86, a movable core is placed between a primary
coil and two secondary coils positioned symmetrically on both sides of
the primary coil. If an exciting alternating current flows in the primary coil, a voltage is induced in the secondary coils. However, as this
voltage is differentially extracted, if the movable core is centrally positioned, the voltages in the two secondary coils cancel each other out
and the output voltage is zero. When the core moves, the induced voltage increases in one of the secondary coils, and decreases in the other. Accordingly, the differential output increases. The relationship between the displacement of the core and the transformer output can be
made linear by choice of coil winding size and distribution. The measurement range is from a few millimeters to several tens of millimeters,
and the linearity is approximately 0.1% to 1%.
Secondary coil ( a )
Oscillator
(Excitation)
Primary coil
Secondary coil ( b )
Fig. 3 86 Operating principle for differential transformer
Fig. 3 87 Structure for short-circuit
ring type displacement gage
(b) Short-circuit ring method
As shown in Fig. 3.87, a copper short-circuit ring lies in the middle of the magnetic path of two coils (Ll> L 2). Displacement of this
short-circuit ring differentially changes the inductance of the two
coils, and by measuring this change, the degree of displacement can be
determined. If the displacement is limited to a certain range by proper
3.6
Measurement of Displacement and Angle
149
choice of the two coils, L 1 and L 2 , the output will be proportional to

the displacement x, and the following approximation equation can be
applied.
X
a
L1- Lz
L1+Lz
iz- i1
i1+iz
Here, a is the effective length of movement of the short-circuit ring, i 1

is the current flowing in coil L~> and i 2 is the current flowing in coil L2
Thus, if (i1 + i 2 ) is held constant and (i2 - i 1) is measured, the displacement x will be obtained. Linearity is approximately 0.1% for a
measurement span of 1 mm.
(c) Variable coupling method
This method uses a coil formed on a printed circuit board, and detects displacement as a change in mutual inductance. The operating
principle is illustrated in Fig. 3.88. If an alternating current flows in
the exciting coil, and induced voltage is generated in the receiving coil
through the U-shaped ferrite core. If the core moves in the direction
of the X axis, the coupling area of the receiving coil increases, increasing the induced voltage. This voltage, when rectified and amplified,
yields an output that is proportional to the position of the core. The relationship between the displacement of the core x and the voltage generated in the receiving coil e2 is expressed by the following equation.
Here, m is a constant representing the relationship between the receiving area and the displacement (j(x) = mx), and a is the length of movement of the core along the X axis.
Figure 3.89 shows the construction and circuit diagram for the
device when used in an angle converter. The coil is a pattern formed
on a printed circuit board. For a 140 angle, the linearity is 0.1% or
Fig. 3 88 Operating principle for variable coupling type displacement gage
150
Chap.3
ore (Ferrite)
(a ) Structure for converting part
Q)
Q)
onstant amplitude oscillation

AC Amplification
H @~Output
@ Synchronous rectifocation
@ Smoothing amplification
( b ) Block d iagram
Fig. 3 89
Configuration of angle converting part
less. Angle converters and transmitter units for showing the degree of
opening of motor-drive valves are examples of the applications for this
device.
(d) Eddy current displacement gages
In this method, differences in eddy currents due to displacement
are detected as changes in the impedance of a coil. This principle is illustrated in Fig. 3.90. If an alternating current flows through the active coil used to measure displacement, an eddy current is generated in
the surface of a conductor that is drawn near to the coil. The magnetic
field from this eddy current produces a counter-electromotive force in
the coil, which changes the effective impedance of the coil. If the conductivity, magnetic permeability and thickness of the conductor are
fixed, then impedance becomes a function of the distance x between
the coil and the conductor. If the output voltage is detected by a coil
and condenser Wheatstone bridge circuit, that voltage V0 is expressed
by the following equation.
Vo=aexp ( -bx)+c
Here, a, b and c are constants.
If this output is made linear by a logarithmic amplifier, then
V= -bx+ln a,
which represents a proportional relationship between the displacement
3.6 Measurement of Displacement and Angle
151
target
Fig. 3 90 Operating principle for eddy current-type displacement gage
and the output voltage. In actual practice, a dummy coil placed near
the active coil compensates for temperature. By taking the excitation
frequency above a few hundred hertz, only the reactance portion is
measured. As there is no contact with the object of measurement, applications include measuring the axle displacement and vibration of rotating machinery. Linearity is 1% or less for a measurement span of
from 1 mm to 5 mm.
3.6.3 Magnetic balance method

In the magnetic balance method, the magnetic flux generated in a
magnetic path by a change in the angular position of a permanent magnet is detected electrically. This measured magnetic flux is balanced
with a magnetic flux from a feedback coil, from which the angle of the
permanent magnet can be known. Figure 3.91 illustrates this principle.
The stator is a saturable ferrite core of high magnetic permeability,
and when the permanent magnet rotates, the magnetic flux cPi changes
in proportion to the angular displacement of the magnet e.
c/J;=k,O
If output current / 0 flows through the feedback coil, a magnetic
flux >f is generated, as expressed by the following equation.
Jf=kzlo
Here, k1 and k 2 are constants.

Thus, a magnetic flux of 11>=>i -if>f remains in the leg of the
core, and if a current / 0 flows such that 11> becomes zero, the amount
of this current corresponds to the angular displacement. In detecting
the magnetic flux differential 11>, the direct current components of
152
Chap .3
Output
Eo
lo
F ig. 3 91
Magnetic balance- type angle converter
half-wave alternating currents flowing in exciting coils E 1 and E 2

change the magnetic flux in relation to 1 Specifically, the component
from E 1 adds to the flux, and the component from E 2 substracts from
it. The detecting coils 5 1 and 5 2 are differentially connected so that
only the voltage that is proportional to the angular displacement 1>t is
obtained. This voltage is amplified and the feedback current Io is obtained. This angle converter has a linearity of 0.2% for an angle of
40 .
3.6.4 Magnetic strain method

This method was developed for use as a servo-motor recording
device position converter. For details, refer to item 4.1.3 in Chap. 4,
Sec. 4.1.
3.7 Measurement of Rotation3>.o>
Tachometers are used to measure the rotational speed of moving
bodies, or in other words, the number of revolutions per unit of time.
The unit of measurement most often used is the number of revolutions
per minute (rpm). In industrial measurement, analog tachometers using generators as rotation detectors have been widely used in the past,
but for highly accurate measurement, digital counting type tachometers which employ frequency output detectors are used.
3. 7.1 Measurement using tachometer generators

Tachometer generators measure rotational speed by using alternating current generators as rotation detectors, applying the principle
that the voltage induced in the generator is proportional to the speed
3.7 Measurement of Rotation
153
of rotation . An example of a medium- speed (500 rpm to 5 000 rpm)

alternating current generator tachometer is shown in Fig. 3.92. It is a
single- phase, four- pole alternating current generator, having a permanent magnet attached to the rotor as shown in Fig. 3.93. When the permanent magnet rotates, the magnetic flux interlinking with the pickup
coil changes, and an alternating voltage is induced in the pickup coil.
This voltage is rectified and converted to a DC voltage output.
onnection cable
Fig. 3 92 Ac tachometer generater (Example)
'r==3~~n
- - l~..
t1__r+~ lnd~tor
!Pickup coil
Rotor
(Permanent magnet)
.J
~------- -- J
Readout circuit
Fig. 3 93 Ac tachometer schematic
By this method, low rotational speeds produce a low- frequency

output from the generator, and after rectification ripples in the DC output increase, causing the tachometer indicator to vibrate and making
measurement difficult. For measurement of low-speed rotation, a low
speed alternating current generator that uses a multiple-pole permanent magnet is employed. Such a tachometer can measure rotational
speeds of from 0 rpm to 20 rpm. The voltage induced in the generator
undergoes full-wave rectification by the built- in rectifying circuit, and
direct current output voltage is obtained. This output voltage is used
for remote measurement of the rotation speed.
Alternating current generators do not have brushes and commutators as do direct current generators, thus they are highly reliable and
easily made explosion-proof.
154
Chap.3
Pulse output sensors

These are sensors which count the number of pulses (frequency) of
an electrical signal that is proportional to the rotation speed of a rotor.
Typical sensing methods include the electromagnetic method and the
photoelectric method.
(a) Electromagnetic method
Figure 3.94 shows an example of this method. As the toothed
wheel of iron rotates, the magnetic resistance between the wheel and
the ferrite core varies periodically. Thus, a pulsed voltage signal that
is proportional to the number of teeth in the wheel is generated in the
pickup coil. At low speeds approaching zero motion, the output voltage
theoretically becomes small, so these devices cannot detect low rotational speeds. However, they have the advantages of excellent durability and not requiring a power source.
3. 7.2
(b) Photoelectric method
This method employs a light source and phototransistor or other

photoelectric converter to convert rotational speed into countable pulses. The light detecting methods include the incident light method and
the reflected light method.
The incident light method is illustrated in Fig. 3.95. A slotted disk
attached to the rotating part rotates between the light source and the
receiver photocell. This produces a pulsed signal that is proportional to
the number of slots in the disk. This system has the merit that if the
Receiver photocell
Output pulseg::::ID-
Lens
Light source
-~
Rotating disk with slits
Fig. 3 95 Photo-electric speed senor (Incident light type)
3.7 Measurement of Rotation
155
number of slots is increased, higher resolution measurement can be

made.
The reflected light method is shown in Fig. 3.96. The surface of
the rotor is patterned with marks of different degrees of light reflectivity, such as a black and white pattern or highly directional reflective tape. The receiver photocell outputs a pulsed signal generated by
the light reflected back to it from the marks in an on-and- off fashion.
The photoelectric sensors described above use either a light-emitting diode or an ordinary electric bulb as a light source, and a phototransistor or other such device can be used for the receiver photocell.
The photoelectric sensing method allows detection of rotational speed
without touching the rotor, so it is used for detecting high-speed rotation. The reflected light method places no load on the rotor, so it is
suitable for measuring low- torque rotations.
<D
Fig. 3 96
Receiver photocell
Rotating disc with dark and light areas
Photo-electric peed sensor ( Renexed light type)
3. 7.3 Digital counting tachometers

These are two representative methods of digital counting as used
in measuring the number of rotations. In both methods, the pulse train
signal output from the detectors is processed digitally, allowing highly
accurate measurement.
(a) Pulse (frequency) counting method
This method directly counts the number of pulse signals in a fixed
amount of time, yielding the average number of rotations for that period of time. A block diagram of this circuit is shown in Fig. 3.97.
The pulse signal picked up from the rotor is proportional to the
number of rotations. It undergoes waveform shaping and is transformed into a pulse- train signal. It is then fed to the gate. The gate is
opened for a fixed period of time by a control signal whose duration is
determined by a stable quartz oscillator. The number of input pulses
that pass through the gate while it is open are counted and displayed
as a numerical value.
(b) Rotation period counting method
The rotation period (time) of the rotor is measured, and the number of rotations is obtained from the inverse of the measured value. As
156
Chap.3
.....r~
I ,
lllillJ
jT-+J
J""L (T : G ate opening time)

CD
Rotating member
Speed detector ( Pul se output type)
D isplay
@ Waveform formation circui t
@ Main gate
Counter circuit
<1) Quartz (/Scillator (Reference signal)
Fig. 3 97
Control circuit for gate signal
Sohematic of pulce cou nting circui t
Selector witch
Signal waveform of each pan
CD
Analog outpu t
Pulse gene ra tor

Synchronous rec tifier
@ Wave form form ation

@ Period counting and inverse ca liculation
Liqu id crista! di splay (LCD )
Ripple filter
Fig. 3 98
Functional schematic of photo electric type tachometer
an example, the functional schematic and signal waveforms of a photoelectric tachometer are shown in Fig. 3. 98.
Special reflective tape (which is covered with many small spherical
lens such that incident light is reflected back to the source even if the
angle of incidence is not perpendicular to the tape surface) is attached
to the rotor. An easily focussed visible red light emitting diode (LED)
is used as the light source. The reflected light is sensed by a photo3.7 Measurement of Rotation
157
transistor, and converted to a pulse signal.

As shown in the signal waveforms in Fig. 3.98, the LED is fired in
pulses, and a pulse signal having a good signal-to-noise ratio is obtained. By synchronous rectification at the receiver side the rotation pulse
is obtained. Measuring the period of this rotation pulse and taking its
inverse yield the number of revolutions. This method has a short response time even for low-speed rotation, and makes high-resolution
measurement (0.1 rpm) possible.
3.8
Measurement of Composition
Gases, liquids and solids are all objects of composition measurement in industrial processes. The purpose of composition measurement
is to determine the actual constituents of a material. However, at
times it is also used as an indirect means to measure the properties of
a material when those properties cannot be measured directly. Thus,
there are many types of composition to measure, and many methods of
measuring them.
For length, mass, temperature, and so on, there are established
international standards, and thus there is little problem with traceability. However, in the measurement of composition there are many
different objects of measurement and measuring methods, so few standards exist. The current situation is that relative calibration, or sample
calibration, is done by comparison with what is considered to be the
most reliable reference instrument or standard. Thus composition
measurement is actually more like relative measurement than absolute
measurement.
This section describes the main instruments and methods used
measure the composition of gases and liquids used in industrial processes.
3.8.1 Gas chromatography 47 )- 49 )
In gas chromatography, the components of a sample are separated
and highly sensitive quantitative analysis is carried out for each component. This process can measure virtually any type of gas or any liquid
that can be vaporized. This ability to analyze multiple components at
one time is a special feature shared by no other analyzer, so the gas
chromatograph is used in many industries.
The basic structure of a process gas chromatograph is shown in
Fig. 3.99. It consists of an analyzing section and a computing control
section, and is used in conjunction with a sampling valve device. The
externally-attached sampling equipment produces the most suitable
measuring conditions for analysis of the gas or liquid by adjusting the
158
Chap.3
Car
ri~
Sampling
valve
~H
Column
Computing control
section
- ((.~rom l
J-D -
Analyzing section
processing and
sequence
output signal
r-
Recorder
Sampl ing system
Constant temperature oven
...__
Supervisory comp uter

Inlet Outlet
Gas vent
Sample (After passing through sampling equipment)
Fig. 3 99 Basic structure of a process gas chromatograph
sample pressure and temperature, removing water vapor, and filtering

out solid particles. It then conveys the sample to the analyzer section.
The sampling equipment also includes calibration equipment to facilitate calibration with a reference gas or liquid. These functions are essential for stable measurement.
The analysis section separates the components of the sample, and
coverts their concentrations to an electrical signal which is output to
the computing control section. The computing control section is made
up of several components. The waveform processor amplifies and converts the signal output by the analyzing section. The digital input/output circuit and processor controls the electromagnetic valve and the
constant temperature oven of the analyzing section, and other such
operations. The transmitter outputs the results of the analysis to the
supervisory compute or a recording device.
The basic structure of the analyzing section, as shown in Fig.
3.99, comprises the sampling valve, a column, and a detector. These
components are contained in a constant temperature oven controlled to
within 0.1 oc. A sample of fixed volume is taken in by the sampling
valve, and infected into the column under the pressure of a carrier gas.
The component separation process within the column is shown in Fig.
3.100. Although the sample consists of a gaseous mixture of various
components, the alternating adsorption and desorption that occurs between the filling in the column (fixed phase) and the carrier gas (moving bed) causes the various constituents to travel through the column
at different speeds due to differences in their distribution coefficiences
(the ratio of the amount in the fixed phase to the amount in the moving bed). Thus the various components appear at the column outlet at
different times, and the concentration signals for each component coming out of the detector over time comprise the chromatogram.
3.8 Measurement of Composition
159
Sa mple injection 0
Sa mple injection 0
Fig. 3 100 Gas sepa ra ting pr inci ple in col umn
The main structural parts of the analyzing section are described

below.
(a) Column and column system
Complete separation of the measured constituents in the shortest
possible time, and long-term stability are desirable features. Thus, selection of the column filler and ensuring the appropriateness of the
working conditions, and selection of the column system are important
matters. There are three types of column systems: pre- cut (backflush
and foreflush types), regrouping, and column bypass. The backflush
type is used for measuring minute quantities among the main constituents, shortening analysis time, and preventing contamination of the column. It is the most basic type of system, and is often used in combination with other types. Figure 3.101 illustrates a backflush column system.
The column by-pass system is used when measuring a mixture of
organic and inorganic gases. A separate column is used for separating
inorganic gases so that it does not to suffer damage from the organic
gases (Fig. 3.102).
(b) Detectors
Generally, thermal conductivity detectors (TCD) and flame ionization detectors (FID) are used, but electron capture detectors (ECD) are
Restrictor
Fig. 3 101
160
Chap .3
Back flu h col umn system
olumn I
M_A_ r
---------- -c::../
n !A Inorganic ga
o+ b+ d
Column :1
"
1\
Column
m Organic
Oeleclor-t
gas
Fig. 3 . 102 An example of ga separalion by column bypa
y lem
also used. Figure 3.103 illustrates the principle of a TCD. If a carrier

gas which contains the constituents is introduced to the measuring filaments, the difference in thermal conductivity of the two gases will
cause a change in the temperature of the measuring filaments. This in
turn changes the electrical resistance of the measuring filaments,
resulting in a change in output from the Wheatstone bridge that corresponds to the amounts of the constituents of the gas. For the carrier
gas, H2 , He, Ar, N2 , and so on are used, and most gases in the measurement range from 0% to 0.1% or more can be analyzed.
Oulpul
arrier gas
omponem gas/
Power supply " '
Conslanl currem L - ----'
Fig . 3 103 Principle of thermal conductivity detector
Figure 3.104 shows the principle of FID. Using FID, organic constituents can be measured over a wide range from extremely small
amounts (on the order of parts- per- million) up to 100%. Hydrogen gas,
with air added, is burned in a flame. If an organic material is introduced into the flame together with the carrier gas, the carbon in the
organic compound will undergo electrolytic dissociation. Further, if an
electric potential of several hundred volts is applied to the nozzle and
3 .8
161
~
+
--=-
Comversion
amplifier
Hydrogen fl ame -
Fig. 3 104
Princi ple of na me ionization detector
to the collector electrode above the flame, a direct current of about

I0 - 7 to I0- 12 amperes corresponding to the amount of organic carbon
will flow. By amplifying this current, a signal corresponding to the
amount of the constituent organic gases as will be obtained. It follows
that this method is not suitable for analysis of inorganic gases. Either
N2 or He are generally used as FID carrier gases.
Recently, gas chromatographs having two detectors built into the
analyzing section have become practical. An example of such an application is shown in Fig. 3.105. For the column by-pass system, having
two detectors has special merits such as shortening the analysis time
and allowing the degree of sensitivity to be selected through choice of
the proper detectors.
Fig. 3 105
Application example of double detectors gas chromatograph
As an example of analysis, Figure 3.106 show a chromatogram for

LPG. The fields of application cover a variety of industries including
petroleum refining, petrochemistry, organic and inorganic chemistry,
the steel industry, metal industry, ceramics and glass-making, the electric power industry, the food industry, and the pharmaceutical industry. It is widely used in industrial processes involving ethylene, polypropylene, blast furnaces, poly- vinyl chloride, BTX, LPG, polyethylene, and so on.
162
Chap.3
100
90
Detecting method
80 Carrier gas
Bridge current
0 Chart speed
r l min
:I
30
f----
CsHn l%
.C,H., l 0 o
c.H,.
20%
20
.c.H, .A.
J{1l
0
0
d I
et ectlo n
,lC,H,3%
_1.
Thermal co nducuvny
H2
300mA DC
.J
J
-
Sample ga injection_....
.I
I
Time(/)
Fig. 3 106 An example of LPG (Simulation gas) chromatogram
3.8.2 Infrared analyzers 50>

There are dispersion and non-dispersion types of infrared analyzers. Dispersion infrared analyzers are used in laboratories as spectrophotometers; non-dispersion analyzers are used for continuous measurement in industrial applications. In non-dispersion analyzers, the
wavelength used is fixed in advance, and the concentration of the target constituent is measured continuously from the transmittance of
that wavelength.
The absorption spectrum of infrared radiation absorbed by a gas,
is unique to the type of gas, and the Lambert- Beer law applies between the degree of infrared absorption and the concentration of the
gas. If / 0 is the strength of the infrared radiation before injection into
the measuring cell, I is the strength after exiting the cell, l is the cell
length, and C is the concentration of the gas, then the following equation holds.
I= / 0exp( -klC)
Here, k is the coefficient of light absorption. The values of k, l, and / 0

are constants determined by the structure of the cell and the type of
gas. The concentration C can thus be determined by measuring the value of/.
Figure 3.107 illustrates the principle of semiconductor and microphone-type infrared analyzers. To allow the tiny (about 2 mm 2) detectors to make the most use of the light emitted from the light source,
3 .8 Measurement of Composition
163
Mirror
.-----A--o_n_de...,nser microphone
r;=:;:::~t--tFixed electrode
--H"::;;_-r-t=:--- , I Fi lied ga
{Same ga a
o.q:::p~-====1?' mea uling component)
Window { aF2 )
Mirror face finished

gold plating internal
face
Window {CaF2 )
o+---11-+ifr... Sector and motor
~Magnetic
coupling
Light source
L...::::::;t"""Eiectric motor
(a) Semiconductor infrared ana lyzer
Fig. 3 107
(b)
ondenser microphone infrared analyzer
Semi onductor- and microphone- type infrared analyzer
that light (infrared) is split into two parallel beams by the upper mirror. One beam is used for the measurement and the other is used as
the reference. The reference light beam passes through the reference
cell, which is filled with air or N2 , and is then reflected off the bottom
mirror onto a semiconductor detector. The measurement light beam
passes through the measurement cell, and is reflected onto a semiconductor detector in the same fashion as the reference light. The component to be measured in the gas sample flowing through the measurement cell absorbs some of the measurement light, thus reducing the
strength of the light relative to the strength of the reference light.
The two light beams are alternately cut off from the detector by a
semi- circular rotary sector. This allows the detector to convert the
difference in the strengths of the measurement and reference light
beams into an alternating electrical signal that represents the concentration of the gas being measured.
The semiconductor detector output does not depend on the light
wavelength, thus it is used with a band- pass filter so as to eliminate
interference gas effects. This filter is formed by depositing alternating
layers of thin- film dielectric materials of high and low refractivity .
The interference effect of these layers permits only a certain range of
light wavelengths to pass through the filter, thus allowing selective
measurement of the target. Figure 3.108 shows the spectroscopic characteristics of a band- pass filter for use with CO, C0 2 , and CH 4
In the condenser microphone type of analyzer, the interior surfaces of the cells are first given a mirror finish and then gold-plated so
as to maximize the amount of light reaching the detector. If the meas164
Chap.3
Gas absorption
IY.
..
100
9or--ci-t, Filter.-7
g 80
-~
pectrum
0 : Filter ,........,
iller-
~ 6
c
r:
1-
3
Fig. 3 108
4
Wavelength (pm )
Spectroscopic charateri stics of a band-pass filter
urement light and the reference light are alternately injected into the
condenser microphone detector, infrared radiation is absorbed by the
filling gas (the same gas as the component being measured) , which
then expands due to heating. The movable membrane electrode is displaced by the pressure from this heat expansion. The movable electrode and the fixed electrode form a condenser whose capacitance varies according to the changes in the distance between the electrodes
resulting from the displacement of the membrane. This change in capacitance modulates the frequency of a square- wave output signal.
Although any gas that has absorption characteristics in the infrared region can be measured, this method is generally used to measure
the gases listed in Table 3.14. The concentration measurement range
can be increased by lengthening the cells, and the sensitivity can be increased by increasing the pressure of the filling gas. Infrared gas analyzers are widely used in all industries for such purposes as the analyTable 3.14
Measuring gas
Carbon monoxide
Carbon dioxide
Sulfur dioxide
Nitric oxide
Ammonia
Methane
Ethane
3 .8
Application examples of infrared gas anal yzer
Molecular formula
co
co,
so,
NO, NO,
NH,
CH,
C,H6
Minimum range
( ppm]
0 to 30
0 to 20
0 to 100
0 to 100
Oto1000
0 to 200
0 to 200
Maximum range
( %J
0 to 100
0 to 100
0 to 100
0 to 100
0 to 100
0 to 100
0 to 100
165
sis of furnace gases, process component analysis, pollution measurement, and atmospheric gas analysis. They are especially valuable for
measuring flue gases, which have temperatures in the range of 100oC
to 300C, contain large amounts of dust, and are saturated with water
vapor. The analyzer is used in conjunction with sampling equipment
that includes an electronic cooler/dehumidifier and triple filters.
3.8.3 Oxygen analyzers 51 )
The methods of measureing oxygen concentration are classified
into paramagnetic methods and electrochemical methods.
(1) Paramagnetic methods
Magnetic wind method
Magnetic force methods
Dumbbell type
Pressure detection type
(2) Electrochemical methods
Zirconia method
Electrode methods
Controlled potential electrolysis
Galvanic type
Polarograph type
The magnetic methods employ the force of attraction generated
when paramagnetic oxygen molecules are magnetized within a magnetic field. The electrochemical methods use oxidation-reduction reactions of oxygen, or oxygen concentration cells. At present, the zirconia method and magnetic wind method are mainly used in process oxygen analyzers. These methods are compared in Table 3.15.
(a) Zirconia oxygen analyzers
As shown in Fig. 3.109, a zirconia cell is constructed of a solid
zirconium oxide ceramic electrolyte covered on two sides by gas-permeable porous platinum electrodes. The zirconia is made conductive by
high temperature (600C or higher) oxygen ions (0 2 -), If gases with
different partial pressures of oxygen are introduced to the two sides of
the cell, an oxygen concentration cell is formed. The following two
reactions involving oxygen and oxygen ions occur. The oxygen reaction takes place at the negative electrode, and the oxygen ion reaction
takes place at the positive pole.
Negative pole:
Positive pole:
02+4e~20 2 202-~02+4e
An electromotive force expressed by the Nernst equation is generated between the two electrodes. If the temperature of the zirconia element is raised to 750C, the electromotive force E is given by the fol166
Chap.3
Table 3.15 Comparison between zirconia method and magnetic method

Method
Magnetic method
(Magnetic wind method)
Zirconia method
Item
Measurement range
Min. Oto 1%
Max. 0 to 100%
(Principally 1 to 100ppm
available)
Cell output
Logarithm
Nearly linear
Response time (90%)

(Case of gas introduced
from caribration gas
inlet)
1 to 5 seconds
5 to 30 seconds
Gas temperature
1 to 1400' C
0 to 50 ' C
Warm up time
20 minutes
1 to 2 hours
Interference of
coexistent gas
Combustible gas of CO,

H2 , CH 4
High thermal conductivity

gases such as C0 2 etc.
External sampling devices
Not required
required
Zirconia cell
Oz partial pressure
of reference gas
Min. 0 to 1%
Max. 0 to 100%
Polous electrode membrane ( Pl a tinum )
0 2 partial pres ure

..,,.,., ......,......,.,.,,. of measuring gas
Px
PA
In dry air,
( Oz partia l pres ,
is 20.95 %
Fig. 3 109
Principle of zirconia oxygen analyzer
lowing equation.
E= -50.74log
~:
-50.74log
z~g 5
(mV)
Here, P x is the partial pressure of oxygen in the gas being measured

(vol%), and P A is the partial pressure of oxygen in the reference gas
(vol%).
If dry air at 1 atm is used as the reference gas, then P A= 20.95
vol%. By measuring the electromotive force E, the partial pressure of
167
0 2 in the measured gas Px can be obtained by calculation.

The zirconia oxygen analyzer uses a direct-feed method, so sampling equipment is not required. As a result, the measurement response
is fast, and highly accurate, even for low concentrations. Thus these
analyzers are often used for control of the air-fuel ratio in combustion
furnaces and low combustion management.
(b) Magnetic wind oxygen analyzers
Figure 3.110 illustrates the principle of magnetic wind oxygen analyzers. The sample gas flows into the glass ring tube through the gas
inlet, then flows through the tube, and out the gas outlet. If oxygen is
present in the sample gas, the gas will be pulled from the left to the
right within the tube by the magnetic field of the magnet poles. If this
gas is heated by a heating coil, the magnetic characteristic of the oxygen will be weakened, and the gas will be pushed out to the right. In
this way, a gas flow (magnetic wind) that corresponds to the concentration of oxygen in the sample will be generated in the middle part of
the tube. The magnetic wind cools the upstream coil a more than it
cools coil b. Electrical resistance in the coils decreases with temperature, so an unbalanced voltage is generated in the Wheatstone bridge.
This voltage is amplified and converted to serve as an indication of the
oxygen concentration.
Although nitrous oxide (NO) and nitrogen dioxide (N0 2 ) can interfere with the measurement, normally they are present in very small
amounts, and because the magnetic susceptibility of oxygen is overwhelmingly large, oxygen is selectively measured.
Sample gas
Glass ring tube

Magnet
Fig. 3 110
168
Principle of magnetic oxygen analyzer
Chap.3
3.8.4 pH meters and ORP meters44 >6 >

(a) pH meters
The pH is a unit expressing the degree of acidity or alkalinity, and

is defined by the following equation.
pH= -log 10 (hydrogen ion activity)
Here, the hydrogen ion activity is the product of the hydrogen ion
concentration and the activity coefficient and represents the degree to
which ions in a solution participate in chemical reactions. When the
ions can move about freely, as in weak solutions, the activity is 1.0,
but as the ion concentration increases, the ions restrict each other and
the activity decreases from 1.0.
There are various methods of measuring pH, but the main methods are those using chemical indicators, hydrogen electrodes, quinhydrone, antimony electrodes, or glass electrodes. Here, the glass electrode type is described, as it is currently used widely in industry.
The glass electrode method is illustrated in Fig. 3.111. This method employs the electromotive force that is generated on both sides of a
glass membrane when the membrane separates two different solutions.
This electromotive force is proportional to the difference in the pH of
the two solutions. The actual configuration of this type of device is
shown in Fig. 3.112. A solution of known pH is put into electrode B,
formed from a thin glass membrane. If this electrode is then immersed
in solution A, the solution which is being measured, then an electromotive force is generated on the two sides of the membrane. The reference electrode is filled with a KCl solution, and an appropriate internal
electrode E 1 is immersed in this. The reference electrode is connected
Glass
membrane Glas membrane
1 internal liquid side
I
( e;)
Meas uring
liquid side
(e.)
I
I
H' ++H"
I
I
es
L ~
1 .. .
_1
-c---l
_t__
Hydrated glass layer
Fig. 3 111
3 .8
Principle of glass membrane
169
Fig. 3 112 Principle of pH measurement
with solution A through the liquid junction ]. The KCl solution inside
the reference electrode flows out through this porous ceramic liquid
junction and comes into contact with the liquid being measured, and a
fixed reference potential is obtained. This method is superior to others
in that it offers a wide measurement range, short measurement time,
excellent reproducibility, and easy operation. On the other hand, while
the strength and high resistance (several tens of M.il to several hundred M.il) of the glass membrane are points that require attention,
these factors are no problem at all for commercial products in actual
use.
From the Nernst equation, the surface potential generated on the
glass membrane can be expressed by
Eg
2.303RT
F
(pH;-pHs)+ EAs=(54.19+0.198 3t)
(3.40)
X (pH;- PHs)+ EAs
where R is the gas constant, T is the absolute temperature, F is Faraday's constant, tis the temperature("C), pHi is the normal pH 7 of the
buffer solution filling the glass electrode, pHs is the pH value outside
of the glass electrode, and EAs is the deviation from the theoretical value due to the asymmetry potential.
In actual practice, standard solutions are used to adjust for EAs in
the above equation, and for span calibration. The following ]IS standard solutions are normally used.
phthalic acid salt (pH 4.01)
neutral phosphoric acid salt (pH 6.86)
boric acid salt (pH 9.18 at 25 oC)
High input impedance and low bias current are desirable in the pH
meter circuitry, and recently MOS FETs and junction FETs have been
used.
Figure 3.113 shows an example of a recent pH meter designed for
industrial use. The wide application range (corrosion resistance, tern170
Chap.3
CD : Glass electrode
:Junction
: Liquid earth
@ : Conductive rubber
: 0-ring
:Cable
Fig. 3 113
A'
: Detector
body
: 0-ring
: Reference electrode
: Porous plastic
@ : Resistance temperature detector
@ : KCl solution
Internal construction of industrial use pH detector
perature, pressure, etc.), ease of maintenance (cleaning, calibration,

etc.), and high reliability are important considerations in the design of
instruments for industrial use.
Here we use equation 3.40 to calculate an actual numerical value
for the electromotive force. Taking EA s = 0,
Eg=0.198 4(7- X)T (mV)
where X is the pH value, and T is the absolute temperature.

(b) Oxidation- reduction potential meters (ORP)
If a reference electrode and an electrode made of platinum, gold or
some other chemically inert metal are inserted into a solution containing an oxidation- reduction material, an electric potential will be generated between the two electrodes. This potential is called the oxidation- reduction potential, and is determined by the strength of the acidity or alkalinity of the solution. Devices using this principle to measure the acidity or alkalinity are generally called ORP meters, deriving
their name from the oxidation-reductio n potential. Their main applications are in the control of cyanides in the treatment of waste water
from metal plating processes, control of the amount of sodium metabisulfite added in the treatment of chromic acid waste water, and so on.
171
E'
Q.
60
-~
c
-~
40
ec
-~
20
:>
c - 200
;;;
""'0
;;.: - 400
Fig. 3 . 114
8c
0
Oxidation reduction potential curve

a nd C concentratio n
Figure 3.114 shows an example of this. In addition to these applications, ORP is also used in bleach liquor production processes, and the
bleaching of textiles, paper, and pulp.
3.8.5 Moisture/humidity meters511

Moisture/humidity meters include those employing the non-dispersive infrared method in the near-infrared region (explained in Sec.
3.9), relative humidity meters which employ microwave absorption
characteristics, and for measuring moisture in gases there are the devices which detect the changes in frequency of a quartz oscillator.
Here, the zirconia humidity meter is explained. The zirconia humidity meter is for use in industrial processes to measure the moisture
content of high-temperature gases (max. 600 C) . The measurement
Zirconia cell
Sample
gas ~
/
Electrode
Vapor
10?%
Fig. 3 115
172
Chap.3
0 2 Concentration
indicator
Air
10:0 %
Principle of zirconia humidity meter
principle of this devices is shown in Fig. 3.115.

If a solid zirconia electrolytic cell (zirconia ceramic) is operated at
high temperature as described in Subsec. 3.8.3, the electromotive force
E is generated between the two electrodes. The strength of E is dependent on the partial pressures of oxygen in the sample gas and the
reference gas, and is defined by the Nernst equation.
E=- RT log JL (VJ
nF
a
Here, R is the gas constant (8.3143V C/Kmol), Tis the absolute temperature (K), n = 4, F is Faraday's constant (96 487 C/mol), y is the partial pressure of oxygen on the sample gas side of the zirconia cell
(vol%), and a is the partial pressure of oxygen on the reference gas
side of the zirconia cell (approx. 21 vol%).
If this humidity meter is used only for sample gases that are
mixtures of air and steam, and the humidity of this gas mixture is represented as x (vol%), then
y = (100-x) x 0.21.
From the above two equations (but with k = RT/2.303 nF), we obtain
E= -k log-!-= -k log (1-0.01x)
(mVJ
and the humidity x(vol%) can be obtained directly from E, the electromotive force generated between the electrodes of the zirconia cell.
The structure of the detector (Fig. 3.116) is the same as for the
zirconia oxygen detector. This device can measure sample gases in the
temperature range from ooc to 600C, so it can be inserted directly
into the environment of the gas being measured without sampling. Response time is fast (90% of the responses are within 5 seconds), and the
H2 0 measurement range is from 0% to 100%, so this device has a wide
Fig. 3 116 Detector construction
173
range of applications as a humidity meter for industrial processes.

Also, zirconia detectors that measure the humidity of ordinary
combustion gases rather than only air-steam mixtures have become
practical. In these detectors, the gas being measured is dehumidified
by an electronic cooling dehumidifier or some other such equipment,
and used as the reference gas. As a result, the application field of
zirconia humidity meters is expected to widen.
This device is used in all industries for the control of humidity
over a wide range of temperatures in drying and heating processes.
Through its use, quality and productivity are increased, and energy is
saved.
3.8.6 Turbidity meters41 >
In the treatment of clean water and waste water, the measurement of turbidity is becoming an important control indicator for manufacturing processes in every kind of manufacturing industry.
For continuous measurement, the methods available include the
transmission method, transmission-scattering method, scattering method, and the integration sphere method. The basic principles of these
methods are listed in Table 3.16. In the ]IS K 0101 standard for turbidity, one liter of water containing 1 mg of refined kaolin is defined as
having a turbidity of 1 ppm.
An example of the transmission-scattering method is as follows.
The measurement principle of the transmission-scattering method of
measuring turbidity is illustrated in Fig. 3.117. A long-life tungsten
bulb is used as the light source. The light from this source is converted to parallel light rays by a condensing lens, and then is passed
through a glass window into the measuring chamber. A measuring
chamber between 5 mm and 300 mm in diameter can be chosen according to the turbidity of the fluid. In the chamber, the light is reflected,
absorbed and dispersed by the material suspended in the fluid. The
transmitted light and part of the dispersed light pass through the chamber and arrive at the detector. The detector consists of two photocells.
The inner photocell (PC 2) detects the transmitted light, and the donut-shaped outer photocell (PC 1) captures the dispersed light. The ratio
of the output from these two photocells, which depends on the relative
amounts of transmitted and dispersed light coming from the measurement chamber, is determined by the calculation circuit. As a result of
these calculations, influences from fluctuation in the light source and
the color of the liquid being measured are eliminated. This is explained
in equation form below.
If the photoelectric current for PC 1 is / 1 and that for PC 2 is /2,
then the relationship of these values with turbidily within the measure174
Table
3 16
Principle a nd features of turbidimeter

Features
Principle
Tran mission method
measuri ng chamber
-_-_ ' )
I~
-:-:-
Photocell
Surface sca ttering method
Window dirts not influential

Liquid color influenti al
~
(~/'
Scattering a nd transmission
meth od
r-:=:=1 /,
- - / , )
~~
], Mirror
Chopper
Scattering and transmission

method
1 Photocell
W indow dirts. Liquid color

influential
0 to l , 0 to 5 000 ppm
Scattering method
Available for high turbidity
T rx /,
T <X (K - log/, )
0 to 2, 0 to 10 000 ppm
~0~
~
R ing type
T <X /,
0 to 0.2, 0 to 30ppm
Window dins, Liquid color,
influentia l
T rx l ,/l z
0 to 0 . 5, 0 to 20 ppm
Liquid color, not influential
W indow dins, influential
T rx !,/l z
0 to 0.5, 0 to 20 ppm
Liquid color, not influenti al,
Window dirts, influential
Integra tion sphere method

Integration
T <X / , f/,
0
to 5, 0 to 1000 ppm
To trap
~''~"
,,
Liquid color. not influential ,

te board Window dins, influential
T . Turbodoty, 11 Scattermg bea m , I 2 : Transmission beam, K : Constant
Tramsparent grass
Measuri ng chamber
Fig. 3 117
3 .8
Principle of measure m ent
175
ment range is given approximately by the following equations.

It= K exp (- [3/N)(Jt<o> + alNI2<o>)
!2= K exp (- {31N)I2<o>
Here, N is the turbidity, J1<o> is the current due to the scattered light
when the turbidity and color components are zero, 12<o> is the current
due to the transmitted light when the turbidity and color components
are zero, lis the effective length of the measuring chamber, a and {3
are constants determined by the shape of the measuring chamber, the
characteristics of the photocells, the properties of the material suspended in the fluid, and other such factors, and K is a constant determined
by the characteristics of the photocells, the length of the measuring
chamber, and the color component.
Taking the ratio of 11 and 12 , we obtain
The left side of this equation is proportional to the turbidity, and is

not influenced by color.
3.8.7 Conductivity meters42>

At a given temperature, electrolytic solutions exhibit a conductivity that is a characteristic of the substance and its concentration.
Thus, by measuring the conductivity, the concentration can be determined. Demineralized water, such as boiler water, has an extremely
small conductivity (theoretically 0.055,uSjcm at l8C), so by measuring
conductivity, the presence of impurities can be determined. In this
way conductivity meters can be used as water purity meters. A particular use for these meters is the detection of NaCl, so they are also someAC Power supply
Fig. 3 118
176
Chap.3
AC Current meter
Principle of liquid conductivity

measurement
times called salinometers.

To explain the principle of conductivity measurement, the method
generally used to measure conductivity in liquids is illustrated in Fig.
3.118. Two metal plates are inserted into the liquid. These plates serve
as electrodes. An alternating voltage of fixed amplitude is applied to
the two electrodes, and the current flowing through the liquid is measured.
The resistance between the electrodes Rc, is inversely proportional to the conductivity K, and can be expressed as Rc=f(l/K). In this
equation, f is a proportionality constant determined by the shape of
the electrodes, generally called the cell constant. It is expressed in units of cm- 1 If the impressed electromotive force between the electrodes is E, and because i = E/ Rc, the two equations yield i = (E/l)K. The
values of E and f are fixed, so by obtaining a value for i, conductivity
can be measured.
In actual practice, however, an electrical double layer forms on
the surface of the electrodes, and ions in the solution receive the electrons and electrical charge from the electrodes (electrode reaction).
Thus the impedance between the electrodes becomes extremely complicated. Figure 3.119 shows a typical equivalent circuit. Figure 3.120
shows the equivalent circuit in which the impedence between the
electrodes is represented as resistances and capacitances in series. To
RP
RL : Liquid impedance between electrodes
CL : Electro-static capacitance of liquid caused by
dielectric constant
Cn : Electro-static capacitance on the surface of
electrode caused by electric double layer
CF, RF : Electro-static capacitance and resistance on the
surface of electrode caused by electrode reaction
Cp, Rp : Parallel electro-static capacitance and resistance
composed by lead wires to electrode
Fig. 3 119
Equivalent distributed parameters between

electrodes
Fig. 3 120
Equivalent circuit of
impedance between
electrodes
177
measure RL, it is necessary to reduce the impedance due to electrode

reaction and other such factors, represented by Cs, and Rs, to a value
sufficiently small compared to RL. The values of Cs, and Rs change
greatly with the material, surface conditions and the structure of the
electrodes. Platinum electrodes that are covered with platinum black
and set facing each other in a parallel fashion is the optimum configuration. However, because of the strength of platinum black, cleaning,
and other problems, in actual practice platinum, graphite, stainless
steel and other such materials are used, particularly in process applications. A fixed alternating current voltage suited to the conductivity
measurement range is applied between the electrodes. The frequency
of this voltage is chosen so as to reduce impedance for the purpose of
avoiding polarization from electrode reaction and so on. As for the materials used in the parts other than the electrodes, ethylene tetrafluoride and hard vinyl chloride are used as insulating materials, and stainless steel is often used for the outer tube. As the conductivity of the
liquid being measured has a temperature coefficient, some devices also
have a temperature sensor built into the electrode, and perform a temperature compensation function. Figure 3.121 shows and example of an
actual electrode for process use.
Protecting
G lass tube
External SUS electrode

Thermistor
Platinum electrode
Internal SUS electrode
( a ) Cell constant 0.05 cm- 1
( b ) Cell constant 5 cm - 1
Fig. 3 121 Electrode of process conductivity meter
The conductivity corresponding to the temperature of the liquid can generally be
(1) Conductivity temperature coefficient for dilute solutions
178
expressed by the following equation, using a and {3. Although the values of a and {3 vary with the type and concentration of the solution,
for weak solutions (0.001 mol or less), the concentration factor becomes more or less constant.
Kt = Kts{1 +a( t- t.) + {J( t- t.) 2}
Here, ts is the reference temperature, Krs is the conductivity at the

reference temperature, and Kt is the conductivity at temperature t{"C).
For example, for NaCl at t 8 = 18 oc, a= 226 x 10-4, and {3 = 84 x 10- 6
Many highly concentrated
(2) Conductivity in concentrated solutions
with respect to convalue
conductivity
maximum
a
solutions have
in particcharacteristics
acid
Sulfuric
Fig.3.122.
in
shown
centration as
of 30%,
vicinity
the
in
maximum
a
having
complicated,
ular are rather
are
characteristics
These
92%.
near
and a secondary local maximum
water
pure
and
solute
single
a
comprising
for two-component systems
as the solvent. Adding even a minute amount of another component
may cause great changes in the conductivity.
0.8
0.7
uOo
~
Jl\.
/h 1"\.
1/
\
1\ 1\
0.6
(J)
.~ 0.3
0.
1/
\
H.so
.....!l o.4
.ii>
:::1
-g 0.2
0
u
'
HCI
0.5
lHNO,- I -
1\
\
_,
1\
[\ NaOH
' Na CI "\
0
0
I
I
20
40
\
60
80
100
Concentrat ion of liq uid [Weight %)
Fig. 3 122 Relation between concentration

and conductinity of liquid
3.8.8 Other composition measuring devices

(a) Sulfur-in-oil analyzers 52 )
The strength of x-rays passing through a chamber filled with
petroleum decreases according to the carbon, hydrogen and sulfur content of the petroleum. The mass absorption coefficient of sulfur is ten
179
times as large as that of carbon and hydrogen, so the strength of the

transmitted x-rays varies markedly with the sulfur concentration. Sulfur-in-oil analizers use this phenomenon to measure the sulfur concentration of petroleum.
The transmitted x-ray are detected by an ionization chamber. The
current generated in the ionization chamber, Is, is expressed by the following equation.
Is=Iso exp {- pt(J.t"C"+ t~cCc+ .usCs)}
Here, CH + Cc + Cs = 1; Is is the ionization current (A) when the measuring chamber is empty ; p is the density of the petroleum (g/cm 3); t is
the effective length of the measurement chamber (em); 1-lH. 1-lc, and 1-ls
are the mass absorption coefficients for hydrogen, carbon, and sulfur (cm 2/g); and CH, Cc, and Cs are the relative weights of hydrogen, carbon, and sulfur.
The values of the mass absorption coefficient of each element varies with the energy of the x-rays, but as can be seen from Fig. 3.123,
at 20 keV the coefficients of hydrogen and carbon are equal, and the
above equation becomes
Is=Iso exp (- pt{(J.ts- /-IH)Cs+ /-IH})
In this equation, Iso and t are constants associated with the detector,
and 1-ls and 1-lH are physical constants. Thus, by measuring the ionization current I s and the density p, the sulfur concentration Cs can be obtained.
IOO.----.--.------.-------,
'";;
.....
e 10
c...
' ()
!lu
c:
.>!
c..
5
1l
"
::!
"'
::E
X ray energy (keY )
Fig. 3 123 Energy-dependence of mass absorption coefficient
180
Chap.3
An example of the structure of a sulfur- in-oil analyzer for process

use is shown in Fig. 3.124. This device includes a transmitted radiation
sulfur detector, a density detector, and sulfur converter. It performes
the calculations described above, and outputs the value of the sulfur
concentrati on Cs. Refer to Subsec. 3.8.8 (d) for information regarding
the density detector.
An example of a sulfur-in-oi l detector is shown in Fig. 3.125.
Gamma radiation from a radioactive source (2 41 Am) strikes a silver target, producing 20 keV x-rays. These x- rays pass through the liquid
being measured and enter the ionization chamber. Teflon is used for
Sample inlet
Sulfur convener
Sulfur concentration
signal
(D Amplifier
Temperature/ Voltage convener

Frequency counter
Fig. 3 . 124 Sulfur-in-oil analyzer system configuration
Fig. 3 125 Construction of sulfur detector
181
the measurement chamber because it is not easily stained. A material

with a low mass absorption coefficient, such as beryllium, is used for
the ionization chamber window.
The transmitted radiation type sulfur-in-oil analyzer has the following special features.
(1) Relatively unaffected by the CH ratio ( 0.005 wt% sulfur/ 2C
/H)
(2) High pressure tolerance (1 to 20 kg/cm 2)

(3) Fast-response
(4) Long-term stability
In addition, this method allows a stronger measurement chamber construction than the analyzer which use low-energy fluorescent x-ray excitation.
The applications of the transmitted radiation sulfur-in-oil analyzer include crude oil and fuel oil inspection, monitoring and control of
fuel oil blenders, control of the operation of sulfur removal equipment,
and monitoring of sources of pollution. It is mainly used in the measurement range of 0 to 5 wt%. This analyzer uses a radioactive source,
so there are legal restrictions concerning its use, and safely precautions are necessary.
(b) Residual chlorine detectors 43>
In process water testing, residual chlorine is defined by the following equation.
Residual chlorine (ppm)
=free chlorine (ppm)+ combined chlorine (ppm)
Generally, residual chlorine is called total chlorine to distinguish it
from free chlorine.
(1) Free chlorine
If chlorine gas is absorbed into water, Cl 2 will
maintain the following equilibrium reaction.
Cb+HzO
HCIO+HCl
This equilibrium varies with pH. Generally, free chlorine exists in water as Cl 2 , HCLO, or CIO- as expressed in the above reaction.
(2) Combined chlorine
Pollution of rivers and so on results in the
presence of ammoniate nitrogen, such as ammonia, in the water. This
produces various chloramines such as monochloramine (NH 2Cl), dichloramine (NHC1 2), or trichloramine (NC1 3), depending on the pH of the liquid. The antibacterial power of these compounds is several tens of
times lower than that of free chlorine.
(3) Methods of measuring residual chlorine
The various residual chlorine measurement methods can be broadly classified into colorimetry
methods, and polarograph methods. Here, we describe the colorimetry
182
and polarograph methods.

(i) Colorimetry (the standard used in japanese water purification facilities) : This method, which is called the o-tolidine method, is based on a
reaction which exhibits a thin yellow to yellow-brown color corresponding to the free chlorine content when an o- tolidine solution is
added to the water being tested. It is used together with sodium arsenite. Accuracies of 0.50.05 ppm at 0.5 ppm and 0.1 ppm at 0.5 to 1.0
ppm have been recorded. However, repeatability is poor and there can
be large differences among the results produced by different operators. Furthermore, the results are sensitive to iron ions, high-valence
manganese, and nitrous acid ions.
(ii) Polarography: This method is used for almost all continuous industrial measurement. As shown in Fig. 3.126, two electrodes are placed
in an electrolytic solution. One of the electrodes has very small surface
area, and the other has a large surface area (about ten times as large
as the small one). If a variable direct electrical potential is applied between these two electrodes, as the voltage changes a graph like that
shown in Fig. 3.127 is obtained. (This graph is an example of an 12 plateau curve.) This kind of graph is called a polarogram, and the equipment is called a polarograph. If electrolysis begins, due to the direct
current voltage applied to the electrodes, concentration polarization occurs because the supply of ions from diffusion is disrupted near the
smaller electrode. As a result, a saturation current that is determined
by the ion concentration and diffusion rate of the liquid flows between
the electrodes. This current is represented by the flat part of the
graph, and is called a plateau.
7
6
< 5
..3
;:
4
~
"u
c
Ro1ary electrode
Reference electrode
-~
2
"
t'
Sppm
I
I
I
I
I
I
3 /
J
zv
'v
0
-"" 3ppm
I
I
2p~m
I
I
lpJm
I
I
0 - 0.1- 0.2 - 0.3 - 0.4 - 0.5 - 0.6
Applied vohage (v ]
Fig. 3 126 Principle of polarograph
(Example of polarogram of lz)
3 .8
Fig. 3 127 Plateau curve of lz
183
The platinum electrode method is one variation of the polarograph

method. Below, a residual chlorine analyzer that uses a rotating platinum electrode is described simply.
If potassium iodide is added to the liquid to be measured, iodine
atoms are liberated by reaction with the chlorine in the liquid.
Cl2+2I-~I 2 +2Cl-
By measuring the current that flows when these iodine atoms

become ions again as a result of electrolytic reduction, the total chlorine can be measured, In this case, platinum electrodes are used for
the working electrode and the counter electrode. In addition to the potassium iodide, the reagent includes a buffering solution composed of a
glacial acetic acid and glacial sodium acetate to stabilize the pH value.
Figure 3.128 shows the relationship between chlorine concentration and diffusion current. Although the process is theoretically the
same as for measuring free chlorine only, using potassium bromide allows selection of an appropriate pH value for the reaction liquid.
7
6
lr
1/
I
~ Electrode : 316 X I 3.516

Spot type
Temperature : 20 c
0o 1 2 3 4 5 6
Chlorine concentration [ppm)
Fig. 3 128 Calibration curve
(c) Alkalinity meters

Alkalinity is the alkaline portion consisting of bicarbonates,
carbonates or hydroxides contained in water converted by calculation
into an equivalent amount of calcium carbonate in milligrams per liter,
and expressed as parts per million (ppm). This measurement is made
by adding sulfuric acid. of a certain concentration until a fixed pH (titration point) is reached. The alkalinity is obtained from the amount of
acid added. Table 3.17 lists typical alkalinity conversion factors. From
this table we can see that the alkalinity of a liquid containing 2 ppm of
QH- and 3 ppm of C0 3 2 - is 2.94 x 2 + 1.67 x 3 = 10.89 ppm.
184
Table 3.17 Conversion factor to calium carbonate of components

Components
Item
Calium carbonate Hydroxide ion Carbonate ion Bi-carbonate ion

CaC0 3
OHco;HC0 3 -
Equivalent
Conversion factor
50
17
30
2.94
61
1.67
0.82
M-Alkalinity
(SO ppm)
3~----~----~-----L----~----~
10
20
30
Reagent volume [ml)
40
50
Fig. 3 129 Titration curve
(1) P alkarinity and M alkalinity

While P alkalinity is represented by
the amount of acid required to neutralize all of the hydroxides and one
half of the carbonates, M alkalinity is the amount of acid required to
neutralize all of the alkaline in the water. The titration curve for this
is shown in Fig. 3.129. The titration point pH values are different, as
shown in the figure. However, as the pH values for ordinary measurements of river water alkalinity are between pH 6 and pH 8, the alkalinity meters used in clean water treatment facilities are M alkalinity meters only, with no P alkalinity meters exsiting.
(2) Alkalinity and pH
When hydroxides and carbonates are dissolved
in water, hydroxide ions expressing alkalinity are generated, and the
pH value becomes higher. However, when bicarbonates are dissolved in
water, almost no hydroxide ions are formed, and the pH value does not
rise. Thus, depending on the composition of the alkali, there are times
when there is almost no relationship between pH and alkalinity.
(3) Measurement principle of alkalinity meters
Below, an example of
an actual intermittent titrator using the neutral titration method is
185
given.
A certain amount of liquid to be measured is drawn, and pH electrodes for detecting the end point of the reaction are inserted. The liquid is titrated until the end of reaction by adding an accurate concentration of titrating reagent (sulfuric acid) to the liquid. At this
point, the chemical reaction is expressed as
CaC03 + H2S04 - - + CaS04 + H2C03
The following relation holds between the amount of reagent consumed
and the alkalinity of the liquid being measured.
CaC03=C.J ~: Xl 000X50 (ppm)
Here, CR is the concentoration (N) of the titration reagent, WR is the
amount of the titration reagent consumed, f is a coefficient based on
the concentration of the titration reagent, and Ws is the amount of the
liquid being measured (g).
This relationship is not disturbed by other components that may
be in the water because the titration reagent reacts only with the alkaline components.
(d) Liquid density meters45 l
Density is a basic variable in the determination of the nature or
composition of a substance. The standard units for measuring density
are kg/m 3 or g/cm3. The methods of measuring the density of liquids include the ]IS hydrometer method, pycnometer method, and displacement method. For industrial use, there is the float-type density meters, which use the change in flotational force on a float; the
weight-type density meters, which use the change in weight of a liquid flowing through a U shape-tube of fixed volume; vibration-type
density meters, which use change in the vibrational frequency of a
pipe; and radiation density meters, which use change in the transmittance of radiation. Of theses, we will explain the widely used vibration
density meter and radiation density meter.
(1) Pipe vibration density meters
If fx is the frequency of free vibration in the horizontal direction of a pipe of uniform cross section filled
with liquid, the following equation is given.
fx
JI+K2Px
Here, Px is the density of the liquid being measured, K1 and Kz are constants determined by the inside and outside diameters of the pipe, the
pipe material, the modulus of longitudinal elasticity, and the length of
the vibrating part of the pipe.
186
CD Oscillation ampl ifier

Standard oscillator
Mixing detection
@ Power supply
Fig.
'@ Temperature compen ation
CV
Frequency/ Voltage converter

Voltage/ urrent conver ion
and linearizing circuit
3 130 Principle of pipe vibration type density meter
From this equation, the density of the liquid Px can be determined.

An example of this principle applied to a practical density meter is
shown in Fig. 3.130. Two pipes are used. These are symmetrical in the
direction of vibration, so forces applied to the upper and lower connectors negate each other so no energy escapes to the outside (the main
pipe system), and Q becomes large. The pipe walls are made 0.15 mm
thick, increasing sensitivity. The vibrator becomes a kind of tuning
fork. Combined with a positive feedback self- exciting oscillator circuit
as oscillating element, this vibrator vibrates continuously
(2) Ring mode density meters
Density meters which employ the circular vibration mode of pipes allow the detector to be small, and are applied in research. For these detectors, the natural frequency of vibration is given by the following equation.
fx=fo/ /I+~:
~ ~
Here, / 0 is the natural vibration frequency in a vacuum of a round tube

of radius R whose walls have thickness t, and which is made from a material of density p 0 The oscillation circuit for this device is shown in
Fig. 3.131.
187
detector __ ,
L__Phase
__________
j
Fig. 3. 131 Principle of self-oscillation circuit
of ring vibration type den ity meter
Radia tion source
Hard lead sh ield
Measuring liquid
Fig. 3 132 Principle of r -ray density meter
(3) Radiation density meters

For the continuous measurement of slurries and corrosive liquids, mainly gamma-ray density meters are used.
These meters are based on the principle that when gamma-rays from a
radioactive isotope (cesium 137 or cobalt 60) passing through a material
of fixed thickness, the radiation will be reduced as a result of absorption, diffusion, and so on. This principle is illustrated in Fig. 3.132.
3.9
B/M Systems'9)50 53)- ss)
In the paper industry, a .measurement control system called the

B/M system is used in papermaking and paper coating processes. This
system is comprised of a process computer system and sensors for
measuring the basic characteristics of paper, such as basis weight
(weight per unit area), water content, ash content and thickness. An
example of the configuration of a B/M system is shown in Fig. 3.133,
but see Sec. 10.5 for more details. The intelligent frame (Fig. 3.134),
which carries five kinds of sensors, including a beta-radiation basis
weight meter, an infrared moisture content meter, an ash content me188
Chap.3
Gateway unit
Field control
tation {BF S)
Fig. 3 133 System configuration
Fig. 3 . 134
External view of intelligent frame
ter, and a thickness meter, collects data on paper quality by shuttling

the sensors back and forth across the width of the paper.
The B/M system must have multiple functions in order to measure
the various paper parameters, and it must have high-perfor mance sensors which do not affect the raw materials or papermakin g conditions.
We will describe the operating principles of the detectors now in practical use. Table 3.18 lists the measureme nt principles and measureme nt
ranges of each type of sensor.
3.9.1 Basis weight sensor (B sensor)

The basis weight sensor employs a detector which detects the
3.9 B/M Systems
189
Table 3.18 Measuring range and detecting principle of B/M meter

Name of detector
Measuring range
Measuring principle
Source of 85 Kr: Basis weight

10 to 1200 g/ m2
Source of ' 45 Pm : Basis weight
5 to 100 g/ m2
displacement Paper thickness 0 to 400 t-~m
Basis weight meter
/3- ray transmission

tion method
Caliper
Eddy current
meter
(Floating target position detection)
Near infrared method
Microwave method
Moisture meter
Ash meter
Color sensor
absorp-
X-ray transmission method

(Ti target)
Spectrometer
Moisture 0 to 12% (In case of

basis weight 10 to 360 g/m2 )
Moisture 4 to 12% (In case of
basis weight 150 to 1 000 g/m 2)
Ash 1 to 35% (In case of basis
weight 30 to 200 g/ m2 )
Wavelength of 385 to 740nm
measurable with 5nm resolution
transmittance or absorption of beta radiation. The mass absorption

coefficient of beta radiation, fJ., does not change with the atomic number of the material being measured. Thus the absorption coefficient remains more or less constant for any kind of material, so this sensor is
universally suitable for measuring the area weight of whole sheets.
Also, the type of radiation source can be selected according to the
Air purge
Heater
Amplifier
Ionization chamber
Rad ia tion
ource
Fig. 3 135 Principle of ba is weight measurement
190
basis weight of the substance being measured.

Figure 3.135 illustrates the measurement principle. When the object of measurement is placed between the radiation source and the sensor, the amount of radiation transmitted I is expressed by the following equation.
I= Io exp(- J.tpt)
Here, t is the thickness of the material being measured (m), p is the

density (g/m 3), J1. is the mass absorption coefficient (m 2/g), and I 0 is the
amount of radiation transmitted when there is no material to be measured present in the system. Accordingly, if J1. is known, then by measuring the amount radiation passing through the material being measured, the basis weight tp (g/m 2) can be obtained.
In actual practice, the signal from the basis weight meter is converted to the basis weight from a calibration curve which approximates the absorption characteristics of the individual sensor.
3.9.2 Moisture sensors (M sensors)

The combined transmission/dispersion moisture sensor (N/T beam
moisture sensor) is a high-precision device whose output is almost unaffected by the basis weight or the color of the pulp. This is because it
applies a calculation formula derived from theoretical analysis to the
basis weight signal and four signals obtained from two light beams of
different optical path length which pass through the paper. The resulting precision is higher than either the transmittance moisture sensor
or the dispersion moisture sensor.
Figure 3.136 shows an example of the optical element construction. For the N optical path of the multiple dispersion optical element,
light from the source enters the light detector head after repeated reflections and being dispersed between the paper and the upper and lower disperison surfaces. On the other hand, for the T optical path of the
directly transmitted optical element, part of the light from the source
is reflected by a mirror, passes through the same filters, is projected
on the paper, and then enters the detector head. The filter wheel carries two monochromatic filters. One is the 1.94 JJ.m M filter, the wavelength absorbed by water molecules; the other is the 1.80 JJ.m R filter,
a wavelength not absorbed by water molecules. As the filter wheel rotates, a reference light (RN) and measuring light (MN) which follow the
N optical path, and a reference light (Rr) and measuring light (Mr)
which follow the T optical path are emitted successively. The optical
axis of the light detecting element is coaxial with the T optical path.
Each of the lights injected into the optical path are collected at the
PbS, converted into their four respective signals, and output. These
3.9 B/M Systems
191
Standard wheel
Filler wheel
Lamp
M Filler
R Filter
Mirror
<D
Heater for constant

temperature
Optical path
T Optical path
Output Temperature
signal control device
Fig. 3 136 Princi ple of moisture meter
four signals are processed numerically and displayed as a percentage

moisture content.
3.9.3 Calipers (paper thickness gauges)

Calipers are a direct measurement sensor. They can measure the
thickness of paper in micrometer units by a one-side, non-contact
method which applies the air-micrometer principle. The principle is illustrated in Fig. 3.137.
p , : Blow -down air pressure
Spring
P, : Push air pressure

Floating distance
ToT ;oating distance
l'oJ..,j=='f""=
--t
Paper thickness
G lass face
Fig. 3 137
192
Chap.3
Princ iple of caliper
Above the upper paper surface, a round metal target floats at a

certain distance as a result of air pressure. On the lower paper surface,
there is an eddy-current displacement meter, which measures the distance D to the target. The paper thickness is obtained by subtracting
the known fixed floating distance d from the measured value.
There are several holes on the bottom of the target, through
which air is forced out. When there is no paper present, these jets of
air hit the glass surface and create an air curtain. The target is supported by a thin, weak diaphragm. The target is pulled up by a spring,
and at the same time pushed down by a separate air pressure P 2 , and
maintains a fixed floating distance above the glass surface determined
by the balance of these two forces. If paper is inserted between the
glass surface and the target, the surface of the paper takes the role of
the glass surface and the target adjusts its position to the fixed floating distance above the paper surface.
3.9.4 Moisture sensor for thick paper
These devices use the fact that water absorbs microwaves, and so
are suitable for measurement of moisture in black paper and paper of
large basis weight, for which near-infrared moisture sensors cannot be
used.
Microwave absorption depends on moisture content MW, and
when absorption by the paper material is negligible, the electrical
transmittivity TP is given approximately by the following equation.
R,M Signal
Reference side
microwave
Rad iation
thermometer
Fig. 3 138 Configuration of moisture detector
3.9
B/M Systems
193
TP=Tpo exp (-,uxMW)

Here, TPo is the electrical transmittiv ity when the moisture content
MW is zero, and .u is the absorption coefficient.
The structure of the detector is shown in Fig. 3.138. Microwaves
(10.5 GHz) generated by the gun diode are time-divide d into a measurement signal (M) and a reference signal (S). Signal M is emitted from
horn A. After passing through the paper two times, it enters horn D
and is detected. Signal R is fed directly to the detector. The moisture
content is obtained after compensati ng the two signals for temperatur e
according to a radiation thermomete r, because the microwave absorption coefficient varies with temperature .
3.9.5 Color sensors
Color, opacity (which is the degree of non-transpa rency), and

brightness (which indicates color), are measured concurrentl y and without contacting the paper. The principle and structure of these devices
is shown in Fig. 3.139.
To supervisory
computer
CD Spectroscope
Preamplifier (Peak
AID Converter
hold)
@) Microprocessor unit
Fig. 3 139 Principle of color sensor
The paper is illuminated by two pulse-type xenon lamps at angles

of 45 with respect to the normal line. Light reflected from the paper
in the direction of the normal line passes through a slit leading to the
spectroscop e. The spectroscop e analyses the light into a spectrum
from 385 nm to 740 nm, with a resolution of 5 nm. The spectrum is
then projected onto a 1 024 element linear diode array. Below the object being measured (the paper), white and black plates are alternately
set. While the white plate is set, color is measured; while the black
plate is set, opacity is measured. The results of the measureme nts are
displayed as a wavelength spectrum. The units of color, brightness,
194
hue, saturation, hunter white (representing whiteness), and opacity are

also displayed.
3.9.6 Ash sensors
Using the same principle as the basis weight meter, the construction of the ash meter is about the same except for the use of the titanium target x-ray tube. The ash content of the paper (mainly oxides of
Mg, AI, Ti, etc.) have x-ray absorption characteristics four to five
times stronger than the pulp. Ash content is measured by measuring
the absorption and transmission of 4.5 ke V x-rays. Also, using fluorescent x-rays, the calcium in the CaS0 4 contained in intermediate paper
can be selectively and quantitatively analyzed.
To enable the detectors employed in the above B/M system to be
used under any environmental conditions of high temperature and humidity and paper dust flying about, internal pressure air purging is
used to protect the instruments against humidity and dust.
3.10
Signal Converters 56 ) 57 )
3.10.1 The purpose of signal converters

In process control systems, a variety of detectors are used to measure physical quantities, and these devices generate various signals.
These signals must be converted into a form that allows easy connection with other elements, such as recorders, controllers, and computers. Various types of converters for this purpose are described below.
(a) Signa/level converters
The analog signals generated by detectors have various voltage
levels, ranging from low level to high level voltages. These signals are
converted to a fixed signal level by passing through an amplifier. In
process control systems, signal voltages of 1 to 5V DC and 0 to 10
V DC are often used.
(b) Signal type conversion
There are also cases where the signal is converted to another type
of signal to facilitate processing. For example, converting a signal representing changes in resistance value into a voltage signal, makes the
signal easy to amplify or transmit. When the distance between the detector and the receiving instrument is great, it is possible to eliminate
signal attenuation during transmission by converting the signal into an
electrical current signal. An electric current signal having the span of
4 rnA to 20 rnA is generally used.
(c) Linearizing
Detector input and output characteristics are often nonlinear.
Nonlinear signals from thermocouples, resistance temperature detec3.10 Signal Converters
195
L - - -- -Fig. 3 140
_ j
Principle of linearization
tors, differential pressure flowmeters, and so on are made linear by a

signal converter before being transmitted to an indicator or recorder.
Figure 3.140 shows the principle of linearization. The symbols in
the figure have the following meanings.
P: The process variable being detected (pressure, for example)
Vx: The voltage signal generated by the detector
Vr: The voltage signal output from the converter
If the input/output characteristics of the detector, for example
when Vx oc P 2 and the amplifier feedback characteristic is defined by
/( Vr)= Vr 2
(3.41)
then, from the amplifier input current balance condition

Vx
Vr 2 _ 0
R-
(3.42)
and accordingly,
Vr=-ffxcxp
(3.43)
Which is to say that by passing through this converter, signal Vr,

which is proportional to the voltage, is obtained.
(d) Filtering
In process control systems, machinery that consumes large
amounts of electric power, like electric motors and electromagnet ic
valves are often used together with measuring instruments that handle
very small signals. Accordingly, noise that is synchronous with 50 Hz
or 60 Hz electric power supplies and pulse noise is often present. Signal converters can prevent the abnormal operation of signal- receiving
instruments due to these types of noise. Low-pass filters, made up of
sets of capacitors and resistors, eliminate noise components above 50
to 60Hz.
(e) Signal isolators
When processing multiple input signals or coupling signals into
different systems, the signals must be isolated from each other to pre196
vent interference. Transformers or photocouplers are used to insulate

the signals.
Examples of signal converters are shown below.
3.10 .2 Thermocouple signal converters

The following uses temperature sensing by a thermocouple as an
example of amplification of a low-level signal. Figure 3.141 shows the
electromotive force characteristic of a thermocouple. Although there
are differences according to the type of thermocouple, in all cases, the
electric potential signal is extremely small, ranging from an electromotive force of a few mV to several tens of mV. The impedence of the th>
50
5
i
40
"8
0 30
-5"
'o 20
;;
g.
0"
Fig. 3 141
Electromotive force charateristics of thermocouple
v,
v,
RJC : Reference junction

compeusation
PW / V: Pulse width/ Voltage
converter
SWR : Switching regulator

PC : Photo-coupler
Fig. 3 142
Principle of thermal-electric temperature connerter
3.10 Signal Converters
197
ermocouple is small, but it is necessary to perform the measurements

without current flowing in the thermocouple. Thus, high-input impedance amplifiers must be used.
Figure 3.142 shows an example of a converter circuit. Conventionally, the signal from the thermocouple was converted to an alternating
signal by a chopper before amplification, to avoid direct current drift
in the amplifier. After amplification, the signal is converted back to a
direct current. However, now low-drift monolithic operational amplifiers having high output impedence are available.
The thermoelectromotive force characteristic of a thermocouple is
determined by the temperature difference between the hot junction
and the cold junction (usually ooq, thus the accuracy of the cold junction temperature is important. Maintaining the cold junction at ooc by
using the thermal balance between ice and water is one method of insuring temperature accuracy. However, in process control systems, the
temperature of the cold junction is measured, and the thermoelectromotive force is compensated for the temperature. Resistors having a large
thermal coefficient or transistors are used to measure the cold junction
temperature. The device shown in Fig. 3.142 uses the latter method.
The relationship between the V8 E of the transistor and temperature is expressed as
7
AT
kT
VeE= Vuo--ln1-
(3.44)
where T is the absolute temperature (K), Vgo is the energy gap of silicon at T = 0 (K), k is the Boltz man constant, q is the electronic charge,
A is the geometric constant of the transistor, r is a parameter related
to the base region diffusion constant, and Ic is the collector current
(Vg 0 , k, q, A, and r are independent of temperature).
Thermocouples are exposed to high temperatures, so they are
quickly degraded by reactions between metal and surrounding gases. It
is very important to always keep in mind this thermocouple degradation and breaks in the wiring when designing measurement control systems. Thermocouple converters include a function to make their output either the maximum or minimum value when a break in the wiring
is detected. This is known as the burnout function.
Thermocouples are often used to measure temperature in electric
furnaces and so on. When the insultation is degraded, the heater voltage becomes common mode noise in the thermocouple signal. If the
output side of the converter is grounded, this noise current flows from
the input side to the output side, and there is danger of abnormal
operation of the converter and damage to the circuit. Accordingly, the
input and output of the converters are generally isolated from each oth198
Chap.3
er. In the example shown in Fig. 3.142, the amplified signal is converted into an optical pulse signal. On the secondary side, it is converted
back into an electrical signal, thus accomplishing the isolation.
3.10.3 Resistance signal converters

This is an example of signal type conversion. Figure 3.143 shows
an example of conversion from resistance change of a resistance temperature detector (RTD) into an electric voltage signal. The resistance
temperature detector is connected to one side of the Wheatstone
bridge, and the voltage imbalance created by resistance changes in the
detector is amplified.
Resistance temperature
detector
~---
---~
~+=Ae==l
Voltage
signal
I R, ~--(>-;-b
L___ _j
---/
R,
I
I
~esistance-v?ltage conv~_er___ j
Fig. 3 143 Principle of resistance-voltage converter
In resistance-voltage conversion, sufficient attention should be

paid to the effect of conducting wire resistance R1. For the circuit in
the figure, the bridge imbalance voltage is expressed as
(3.45)
L1e
(3.46)
Normally, when copper wire is used, R 1 is 10!2. or less. Thus if Ra

= 1(k!l), and a platinum resistance temperature detector (Pt 100!2.) is
to make a measurement in the range from ooc to 100C, then the effect on the conducting wire resistance when there is change in the environmental temperature is a very small 0.2%/10C. By using a
three-wire configuration in which the three wires have the same characteristics, the measurement error can be made small. Figure 3.144 is
an example of a resistance-temperature converter. Transistors Q4 and
Q5 comprise a burnout circuit to drive the output to the maximum value when any of the three input wires (A, B, or b) are broken.
Platinum resistance temperature detectors (Pt 100!2.) are used in industrial measurement. The resistance change characteristic of this
199
device is non-linear, as shown by the following equation.

(3.47)
Re = Ro(1 +at- /3t 2)
Here, R0 is the resistance at ooc (100 il), Rt is the resistance at temperature t CC), and a and {3 are constants.
For example, the maximum non-linearity error in the 0 to 500C
measurement range is approximately lOoC in the vicinity of 250C. In
the circuit shown in Fig. 3.144, the non-linearity error of the Wheatstone bridge circuit is added. This error can be compensated for by
proper selection of the resistance Rf> which determines the positive
feedback from the output of amplifier A1
I
I
I
I
I
I
:I
'
l _______ .JI
Span adjustment
Burnout protection circuit

Fig. 3 144 Example of resistance-temperature converter
3.10.4 Two-wire signal transmission

When the distance between the detector and the converter is
large, the two-wire electric current transmission method is used. Figure 3.145 shows this configuration. The two wires connect the detector
with the converter, supplying power from the converter and at the
same time carrying the signal component as variations in the current.
Resistance of lead wire
R./2
Vs
Ro
1 to 5VI
~---~
Fig. 3 145 Two-wire method current transmission system
200
Corresponding to changes in the physical quantity from 0% to 100%,

the signal is converted into a DC electric current signal of from 4 to 20
rnA and transmitted. When a voltage signal is required, a resistor is inserted into the path of the current, and the voltage signal is picked up
at the two sides of the resistor. The allowable lead wire resistance Rz
of the transmission line is given by the following equation.
. . .- Vs-/,Vo- Vr
R ~~
max
(,...)
u
(3.48)
Here, V 5 is the minimum voltage supplied from the converter (V), V0 is

the internal voltage consumption of the converter (V), Vr is the minimum operating voltage of the converter (V), and I max is the upper limit
of the signal current (A). The electric current signal used in process
control systems is being standardized at 4 to 20 rnA DC, so Imax is taken to be 20 rnA. The signal is converted to from 1 to 5 V according to
the resister R0 inside the converter, so V0 is taken to be 5 V. The minimum operating voltage of the converter is 12 V. If Vs is taken to be 25
V, then R 1 <400!1.
3.10.5 Pulse signal converters
In flowmeters, devices which output pulse signals are often used.
By counting the pulses in these signals, the flow rate can be determined. However, because the pulse signals from flowmeters are
high-level signals, and chattering occurs because relay contact points
are used, direct input to computers and so on is not possible.
Figure 3.146 shows an example of a pulse signal converter. After
waveform shaping, the pulse input undergoes input-output isolation by
a photocoupler. The isolated signal is then applied to the output circuit. The output is a transistor contact signal, thus compatible with
Photo coupler
v.
Output circuit
,.--.-----{(0 4
Derivative Wave shaping circuit
Load
Signal~~
Fig. 3 146
Pulse signal converter
201
Table 3.19 An example of input pulse convertor specifications

(1) Input signal:
Contact or voltage level pulse from two-wire transmitter system or
three-wire trsnsmitter system.
Input frequency:
0 to 6kHz
Minimum input pulse width:
ON time; 60 ps, OFF time; 60 ps
Input level:
Contact signal:
Contact type: relay contact or trausistor contact (open collector)
Contact resistance: closed, 200ll or less; open; 100 kll or greater
Contact rating: 15V DC, 15mA or greater
Filter: insertable by switch; time constant, lOms
Voltage level pulse signal:
Low level (EL): -1 to + 8V DC
High level (EH): + 3 to + 12V DC
EH-EL~3V
Swing:
Signal source impedance: 1kll or less
Load resistance on two-wire transmitter system writh power supply:
200ll, 510kll, lkll (selectable by switch)
(2) Output signal:
Transister contact (open collector)
Contact rating: 30V DC, 30mA
Output frequency: Same as input frequency
(3) Power supply for transmitter: 12V DC 10%, 30mA or less
computers. This device can accept as input relay contact, transistor

contact, or voltage pulse signals. For relay contact input, the filter
switch is turned on and chattering noise is eliminated. The purpose of
the load resistor is to convert the current pulse signal to a voltage
level when the two-wire current transmission method is used. An example of input pulse converter specifications is given in Table 3.19.
3.10.6 Computer input equipment

Process control systems which use computers accept multiple analog input signals. Figure 3.147 shows a computer input processing system, taking thermocouple input as an example. The multiple input signals are switched by a multiplexer, and then undergo A/D conversion.
For switching speeds of 1 to ten points per second, relay switching is
suitable, but for higher speeds, semiconductor switching is used.
Figure 3.148 is an example of a flying capacitor multiplexer. The
unselected channel switches are thrown to the detector side, and the input electric potentials are stored in capacitors C1 The switch for the
selected channel is thrown to the amplifier side, and the electric potential on the two sides of capacitor C1 undergoes A/D conversion. With
202
Chap.3
Terminal
block
Communication
interface section
Multiplexer
card
A/ D conversion card
gn! II
i;
i i! -~
+-lr:--!Amplifier
Thm="'' lft
convertor
~ L----------~
L-- - - --'
Reference junction temperature detector
Fig. 3 147
Input treatment system of computer
mV sigua l
No. I
Thermocouple
'
I. Sk
Burn-out
detector
I. Sk
Input
o. IS
L._.,
Thermocouple
I.Sk
-----,
I
I
o.l6
I
I
Control
data
upply
Multiplexer card
RJC : Reference jun~tion temperature detector
In case of mV input, connect input signal to
Terminal board
Fig. 3 . 148
0.
16
Flying capacitor type multiplexer
this method, the input signals are isolated from each other, and at the
same time analog filtering is accomplished during the conversion,
because the capacitors are charged by the input voltages over a fixed
time interval.
When multiple thermocouple signals are batch processed, equality
of the cold junction temperatures is important. Figure 3.149 shows an
example of a 15 input connection terminal board. The 15 input terminal
board consists of a copper printed circuit board designed for constant
temperature across the part where signal wires are fastened by
screws. The temperature of the printed circuit board is measured with
a resistance temperature detector, which sends a temperature com3 .10 Signal Converters
203
Fig. 3 149 Terminal board with integrated reference junction compensation circuit
- - - --
CD
A/ D Converter
Data register
A/ D Conversion control section
Fig. 3 . 150
_ _ _ _ _A! D Conversion . card
Power supply for burnout circuit

Signal selection control ection
onfiguration of A I D Conver ion
pensation signal for alliS inputs to the multiplex er.

Figure 3.150 shows the structure of the amplifier and the A/D converter. When thermoco uples having different measurem ent temperature ranges are mixed together, the A/D converter input range can be
adjusted and put on a common basis by using a variable gain amplifier.
When slow conversio n is acceptabl e, an integratio n A/D converter is
used; but when high speed conversio n is required, a successiv e approximation A/D converter is used. The resolution required in process control is 10 to 13 bits.
204
PRACTICE QUESTIONS
3.1
3.2
Describe the error factors influencing the measured values of industrial variables.
Why are differential pressure type flowmeters so often used? What are their drawbacks?
3.3 Describe the features of vortex flowmeters.
3.1
3.2
See Subsec. 3.1.2. in the text.

a. Wide application for liquids, gases and steam.
b. Many types of constriction devices are available; the structure is simple so
failures are rare.
c. Can generally used without calibration of actual flow, because constriction device
standards are well-established.
d. High-accuracy differential pressure transmitters are available.
The disadvantages are,
a. Pressure loss is high compared to other flow meters.
b. The differential pressure is proportional to the square of the flow rate.
Thus, measurement acccuracy falls in the low flow rate region (low range ability).
c. They cannot be used for some highly viscous fluids or slurries.
3.3 See Subsec. 3. 3. 6 in the text.
REFERENCES
1)
2)
JIS Z 8103-1984: Glossary of Terms Used in Instrumentation (in Japanese).

T. Senbon and M. Toyama: Industrial Measurements, pp.8-9, pp.240-242, Ohmsha
3)
T. Naito: Industrial Measurement Handbook, pp.23-26, Asakura Shoten (1976) (in
Japanese).
4)
JEMIS 022-1983: General Rule for Expression of Industrial Instruments Performance (in Japanese).
5)
M. Kanno: Electrical and Magnetic Measurements and Instrumentation, p.30, Corona Publishing (1982) (in Japanese).
6)
T. Ishii: Process Control Instrumentation Systems, pp.60-94, Denki Shoin (1973)
(in Japanese).
7)
The society of Instrument and Control Engineers, Temperature Measurement Devision: Temperature measurement (1981) (in Japanese).
8)
The Instrumentation Control Association Japan: Industrial Instruments Management Standard m, Management of Temperature Instruments, pp.68-69 (1978) (in
Japanese).
9)
Daido Steel Co. LTD.: Technical Information, Portable Digital Radiation Thermometer "STAR THERMO" (in Japanese).
10) JIS C 1602-1981: Thermocouples (in Japanese).
11) JIS C 1605-1982: Metal Sheathed Thermocouples (in Japanese).
Rejernces
205
12)
13)
14)
15)
16)
17)
18)
19)
20)
21)
22)
23)
24)
25)
26)
27)
28)
29)
30)
31)
32)
33)
34)
35)
36)
37)
38)
39)
206
JIS C 1604-1989: Resistance Thermometer Sensors (in Japanese).

]IS C 1606-1989: Metal Sheathed Resistance Thermometer Sensors (in Japanese).
]IS C 1610-1981: Compensating Lead Wjres. (in Japanese).
]IS C 1611-1975: Thermistor for Temperature Measurement (in Japanese).
JIS Z 8706-1980: Methods of Temperature Measurement by Optical Pyrometers (in
Japanese).
H. Kawada, et al.: Flow Measurement by Constriction Devices, Flow Measurement
Handbook, pp.55-65, Nikkan Kogyo Shinbun Sha (1979) (in Japanese).
The Instrumentation Control Association Japan: Industrial Instruments Management Standard N, Management of Flow Instruments, pp.18-28 (1978) (in Japanese).
JIS Z 8762-1988: Measurement of Fluid Flow by Means of Orifice Plates, Nozzles
and Venturi Tubes (in Japanese).
T. Kobayashi: "Field instruments," Yokogawa Tech. Rep., 28, 2 (1984) 77 (in Japanese).
]IS Z 8761-1977: Method of Flow Measurement by Float Type Area Flowmeters
(in Japanese).
M. Kanayama, et al.: "Fundament of Flow Measurement," Automation, 21, 7/12
Oval Engineering Co., LTD.: Training Text Book, No.A-001-1, p.35 (in Japanese).
F. Satori: Turbine Meter, Flow Measurement Handbook, pp.203-210, Nikkan
Kogyo Shibun Sha (1979) (in Japanese).
]IS Z 8765-1980: Method of Flow Measurement by Turbine Meters (in Japanese).
N. Kayama, et al.: "Recent progress in electromagnetic flowmeter,"]. Soc. Instrument & Control Eng., 23, 2 (1984) 20-27 (in Japanese).
]IS Z 8764-1980: Method of Flow Measurement by Electromagnetic Flowmeters (in
Japanese).
JEMIS 028-1985: Methods of Flow Measurement by Vortex Flowmeters, p.7,
pp.19-21 (in Japanese).
I. Ito: "Vortex flowmeters using stress detection method," Instrumentation, 34, 8
H. Ohwada: "Flow measurement using ultrasonic wave," Sensor Technology, 4, 2
Y. Yamamoto: Flow Measurement by Ultrasonic Flowmeter, Flow Measurement
Handbook, pp.261-262, Nikkan Kogyo Shibun Sha (1979) (in Japanese).
N. Kayama: "YEWSONIC ultrasonic flowmeter," Yokogawa Tech. Rep., 30, 2 (1986)
61-63 (in Japanese).
Yokogawa Electric Corporation: Industrial Instrument Handbook Pneumatic Instrument Volume, 2nd Edition, Tokyo Denki University Press (1974) (in Japanese).
B. Saito, et al.: "Air-to-current type 5551," Yokogawa Tech. Rep., 15, 2, (1971)
Hirano, et al.: "Con tactless transducer using a coil formed on a printed circuit
board," Text Books for No.23, Training course of Automatic Control (1979)
S. Shindo: "Miscellaneous measuring instruments," Yokogawa Tech. Rep., 25, 3
(1981) 27-29 (in Japanese).
I. Ohno, et al.: "Magnetic displacement transducer using a self-oscillating technique,"]. Soc. Instrument & Control Eng., 12, 1 (1973) 2-15 (in Japanese).
T. Ueda, et al.: "Displacement transducer using magnetostrictive delay line,"
Trans. Soc. Instrument & Control Eng., 17, 8 (1981) 858-864 (in Japanese).
H. Ohtu, et al.: "Programmable intelligent recorder t-tR100/t-tR180 series,"
40)
41)
42)
43)
44)
45)
46)
47)
48)
49)
50)
51)
52)
53)
54)
55)
56)
57)

S. Shindo: "Electronic measuring instruments," Yokogawa Tech. Rep., 25, 3 (1981)
H. Hirasawa: "Turbidity transmitter 'T/B cell'," Yokogawa Tech. Rep., 17, 1 (1973)
H.Ikegawa, et al.: "New electrolytic conductivity measuring equipment for process
use," Yokogawa Tech. Rep., 18, 2 (1974) 63-69 (in Japanese).
H. Kamata, et al.: "Non-reagent type free available chlorine analyer 'F/C Cell',"
M. Torno, et al.: "New series of liquid analyzers for process use," Yokogawa Tech.
Rep., 28, 1 (1984) 55-62 (in Japanese).
T. Minaki, et al.: "A digital density meter," Yokogawa Tech. Rep., 29, 2 (1985) 3-8
(in Japanese).
Yoshimura, et al.: Theory of pH and Its Measuring Method, Maruzen (1968)(in Japanese).
Y. Fukai, et al.: "New gas chromatograph readout GC6P," Yokogawa Tech. Rep.,
24, 2 (1980) 2-9 (in Japanese).
M. Maeda, et al.: "New thermal conductivity detector for process gas chromatograph," Yokogawa Tech. Rep., 27, 1 (1983) 27-32 (in Japanese).
Y. Hanaoka: "Process analyzers," Yokogawa Tech. Rep., 30, 2 (1986) 64-67 (in Japanese).
H. Takahara: "Type IRll, IR21 infrared gas analyzer," Yokogawa Tech. Rep., 24, 3
(1980) 16-20 (in Japanese).
H. Nakashima: " Process analyzers," Yokogawa Tech. Rep., 28, 2 (1984) 86-89 (in
Japanese).
S. Muramoto:"PC6 oil calorimeter," Yokogawa Tech. Rep., 24, 2(1980) 10-16 (in Japanese).
S. Kiyobe: Japanese Journal of Paper Technology (Mar. 1984) (in Japanese).
H. Ishikawa, et al.: "New sensors of B/M-CENTUM System in paper plant industries-ash, caliper and new moisture sensors," Yokogawa Tech. Rep., 24, 3 (1980)
3-15 (in Japanese).
K. Isozaki, et al.: "New 'WEB GAGE' -uses radiation to measure sheet thickness," Yokogawa Tech. Rep., 26, 4 (1982) 27-31 (in Japanese).
S. Miyazaki: Control for Chemical Pressure, Sangyo Tosho (1971) (in Japanese).
C. D. Johnson: Process Control Instrumentation Technology, 2nd Edition, John
Wiley & Sons (1982).
Refernces
207
Chapter 4
RECORDERS AND CONTROLLERS
In large-scale instrumentation systems, CRT display and printout

of data is often used, and accordingly the priority of recorder units is
decreasing. However, recorder technology is also rapidly developing
through application of electronic technology. This chapter describes
the non-contact type self-balancing mechanism, which has been problematic in conventional industrial recorders. New methods of printing
and so on are also described.
Controllers have developed from the early pneumatic controllers
through analog electronic controllers, digital controllers to the CRT
display control systems which have data communication functions. The
section on controllers in this chapter is limited to single-loop controllers, and while comparing analog electronic controllers with digital controllers, discusses controller functions. Control algorithms are described in detail in Chap. 2 and Chap. 8, so discussion in this chapter is
limited to the range of the basic PID form. The fine points of pneumatic controllers can be found in other works. I)
4.1 Recorders
Industrial recorders monitor the changes over time in process

measurement signals, and provide a permanent record of those changes.
4.1.1 Types of recorders
Industrial recorders are classified in the following way according
to the methods they employ.
. me th od {Self-balancing recorders
0 perat 1on
Raster scan recorders
.
{Pen recorders
Recordmg method Dot prm
. t"mg recor ders
4.1 Recorders
209
recorders
. .t me th o d {Analog
CIrCUI
..
Digital recorders
As most self-balancing recorders use the zero method, the principle of that method is shown in Fig. 4.1. The input signal is converted
by the preamplifier (signal conditioner) to a level suitable for transfer
to the servo-amplifier. In the servo amplifier, the differential voltage
signal between the output voltage from a potentiometer connected to
the pen and the preamplifier output is impressed on the input voltage
signal, and this differential voltage is amplified and used to drive the
motor. This is currently the most often used type of recorder. Raster
scan recorders will be discussed later. This section is devoted to description of recorders classified as pen recorders and dot printing recorders. These recorders are mainly manufactured in recording widths
of 100 mm, 180 mm, and 250 mm, and accuracies ranging between 0.1%
to 0.5% of full scale.
Preamplifier
E.
Fig. 4 1
Principle of self-ba la nci ng type recorder
4.1 .2 Recorder functions

As shown in Fig. 4.2, recorder functions can be classified into
measuring, processing and recording. In self-balancing recorders,
measuring and recording are mainly accomplished by servomechanisms. Chart feed and the printing mechanism are driven by a synchronous motor. Although these mechanisms are relatively simple, if other
functions such as character printing, display, computation, remote control, and so on is required, then the circuitry and mechanical components become more complex.
Analog- to-digital (A/D) conversion of the input signal allows efficient processing under control of a microprocessor. The following func110
Chap.4 Recorders and Controllers
Tempenture
Fig. 4 2
Function of Recorder
tions can be readily performed.

(a) Measuring
As shown in Table 4.1, there are many types of inputs that can be
measured, and the measurement range is also wide.
Table 4.1
Examples of input signal and range
Input type
Range
DC voltage
20mV to SOV
Thermocouple
TYPE R, S, B, K, E, J. T, N, W
Resistance temperature detector
Pt 100/SO
n.
Ni 1001120
(b) Recording
As shown in Fig. 4.3, the measured values, time, scale, and so on

are digitally printed, and have been conveniently used to grasp
long-term trends of input change in addition to the analog recording.
This is referred to as hybrid recording. Recording zones for each channel can be selected and important portions can be enlarged. Also, scheduled printing, in which information on conditions such as the range of
each channel, tag numbers, units, sensor alarms, the date, and chart
speed can be listed at fixed intervals on the chart.
(c) Display
In addition to the digital display of the measured values for each
: i~
;;
U~ l til l II' II
::: ~]
; ;~ I
I'
:
! ~ I' f;~
: :: ::! ~ ,
~ . -it~
II
.i......_., ....t._.,l
!I! llli !' ~I
Ill .I I
I ~
'1
"
-t'
'I
11
.
'
II
1',1
tr
~
I
ilt "::
!H
rf# tt ~
, ... ., . . . . .. ,
.,_,,,
: o< T,
IIIII 1i II
11 I !I
,l,.
!1:111
. Lci'
Ill '
I I
::::::
:~
.;:J
.II
! ...r,11 1~ .liJ.I
: ::::::7..:.' , II mir ' l1ITtlii'l'llr' mt-tHtil1cttl-fH-Mitt~~trl 411l ,:n,
II' ~::
..
~ ~~l
-trh+tli~H~--1-'-~ @
-4I1l
tl
f~
..,
l'l
l ~ll
~!
:::::
.-------
Fig. 4 3 Chart example (Analog + Digital record)
4.1 Recorders
211
channel, the year-month-date, and the time, measured values, alarm

settings are displayed, and at alarm times, flashing point bar graphs
are displayed.
(d) Computation
Inter-channel differential computations, industrial variable conversion of input voltage signals (scaling), linearizing, and square root extractions are possible. Also, taking advantage of memory storage, the
offset between pens can be compensated for. Multiple-point high speed
recording and printing using the raster scan method can be performed.
In addition to this, event recording, in which the data before and after
any event that occurs can be preserved in a special memory area and
the data can be printed on command.
(e) Alarms
From two to six alarms can be set for each channel. The types of
alarms include upper limit, lower limit, upper velocity limit, lower velocity limit, difference upper limit, and difference lower limit. These
alarms can be annunciated through relays as well as recorded.
(f) Data communication
These devices have GP-IB and RS-232 C communication capabilities, so communication with computers is possible. Thus they can be
used as computer output devices.
The operating principle and structural components of recorders designed for multiple functions through use of microprocessors are described in the following sections.
4.1 .3 Pen recorders
A typical pen recorder is shown in Fig. 4.4. Figure 4.5 is a block
diagram of this device. Commercial products are mainly 1- pen to 3- pen
models.
(a) Operating principle
The measurement signal, after amplification corresponding to a
Fig. 4 4 Pen recorder
212
nalog block
Operation control block
Signal conditioner
l n~put
<D
Reference i
junction . 1
compensation
tran i tor
:
(RJ )
:
:
L----..L-----1----- -
(Integration type)
circuit
@ 0 / A converter
Driver
One-chip P
<V Ultrasonic displacement
detector (Po ition
feedback element)
Ca lculation
Fig. 4 5
Motor control I
Functional b lock diagram of pen recorder (I pen)
pre- set measurement range, is converted to a digital signal by an integration type A/D converter. The converted signal undergoes linearizing and alarm processing in the operation control block, and is then
temporarily stored as display data in memory (RAM). It is subsequently converted to recording data compatible with the recording
equipment. The recording data is converted back to analog data by a
D/A converter and it is sent to the servo amplifier, where it is compared with the pen position signal obtained from the pen-position feedback element. The deviation (error) output is power-amplified by the
motor control IC and drives the servomotor. The position feedback element is an ultrasonic displacement detector, and has no mechanical
joints. The servomotor is a thin type brushless DC motor. These components make up a highly reliable contact- free servo system.
List printing and scheduled printing of logs is done by a small X- Y
plotter that operates independently of the servomechanism. The plotter is controlled by a dedicated one-chip microprocessor.
(b) Structural components
Figure 4.6 shows the structure of the A/D converter and the servo
unit.
4 .1 Recorders
213
Detecting coil
Fig. 4 6
Driving string
Transmission coil
Brushless DC motor
Servo amplifier
Configuration of servo- unit and A/ D converter
(1) Signal conditioner and A/D converter

Signal conditioners are available in two models: one for DC voltage thermocouple input and one for
resistance temperature detector input. The model for DC voltage thermocouple input consists of six ranges from 20 mV to 20 V, which depend on a potential divider and a variable gain amplifier. The output of
the amplifier goes to an integration type A/D converter which uses the
feedback-pulse-width modulation method. The obtained pulse width is
isolated by a photocoupler and is applied to the operation control
block. The operation control block has a high-precision reference voltage source and performs zero compensation for each range, zero compensation and full-scale compensation for the A/D converter, and other
types of automatic calibration.
In the resistance temperature detector model, the input signal is
converted to an electric potential by a resistance-voltage conversion
circuit consisting of a fixed current of about 1 rnA and a reference resistance of 300 n (accuracy of 0.01%). The resultant signal is fed to
the A/D converter. The control signal from the operation control block
is sent in serial form through the photocoupler to the analog block,
and undergoes serial-parallel conversion to control signals for the amplifier and all switches. Typically, the resolution of the A/D converter
is 14 bits, and its data output rate is 1/125 ms.
(2) Servo system
The deviation of the signal from the position feedback element and the recording position signal corresponding to the
measured value is amplified, and is sent to the DC motor circuit. In the
D/A conversion block, the recording position signal becomes a smooth
direct current signal by smoothing after pulse-width modulation by
300 Hz repetition. In the ultrasonic position detector, the pulse signal
flows through the drive coil, and the direct sound waves and reflected
214
sound waves received in the detection coil (which moves as a unit with
the pen) are converted into a position signal in the analog calculation
circuit.
(3) Brushless DC motor
For industrial recorders, tough, long-life
two-phase alternating current servomotors having no frictional parts
have often been used. Recently, however, small, lightweight brushless DC motors that generate little heat are coming into use. In these
motors, electronic circuitry takes the place of the brushes. Figure 4. 7
shows the structure of such a thin brushless motor mounted on a recorder. This is a rotating- field type DC motor, in which the armature
coils are fixed on the stator (printed circuit board), and the field magnet rotates. Hall elements are used for contactless detection of the relative positions of the armature coils and the rotating field. Motor control is performed by dedicated ICs which have built- in commutation
control circuits. By using this motor, a servo unit step response time
(90% response) of 1 second or less can be realized.
Fig. 4 7 Configuration of brushless DC motor
(4) Ultrasonic position detector

The principle of the ultrasonic position detector is shown in Fig. 4.8. If a pulse magnetic field is generated at the position of N0 by the current passing through the ultrasonic
drive coil N0 , the Joule effect will result in a mechanical distortion
within a ultrasonic waveguide made of magnetostricti ve material, and
the ultrasonic pulse will propagate. When these ultrasonic pulses (the
4.1
Recorders
215
/, = ( L - x ) l vo
t, = ( L + :c ) / vo
-Time
vo : Ultrasonic wave propagation speed
-L"' ( Dimensionless position signal )
= ...l.!.::...!!.
,, +,,
Fig. 4 8 Operating principle of ultrasonic pen position sensor
direct pulse from N0 to N~> and the reflected pulse from the reflecting
end) arrive at the detection coil N~> an electric potential pulse is generated in N1 by the Villari effect. As shown by the propagation waveform in Fig. 4.8, by measuring the propagation time of this pulse, the
position of the moving part can be obtained. Using the time width signals t 1 and t 2 for each sample cycle, the computation circuit carries out
the following computation.
l (a dimensionless position signal)

Because the computation for obtaining changes in the relative positions is done in this way, changes in environmental conditions such as
temperature, which change the speed of sound v0 do not affect the output. In this way infinitely high resolution and linearity of 0.1% or less
are achieved.
Other than the contactless position feedback element for use in industrial recorders described above, there are also magnetic detectors
and capacity type detectors. Also often used are contact type devices
such as coiled wire potentiometers, conductive plastic potentiometers
(CPP), and hybrid potentiometers in which the coiled wire slide surface
is coated with conductive plastic. However, contact type position feedback elements often require cleaning of the resistor and exchanging of
the contact mechanisms, so the movement will probably be toward con216
IV
...
""......
:::.::,
8
DC amplifier
Programmable AD convertor
DC amplifier
Rotary
switch
Relay
Digital
Stepping
motor
Yokogawa Tech. Rep., 2 8 , 2, p.34 (1984)
Wiredot
Raster
scan
Microprocessor
Digital
computation
ROM table
Ink hammer Synchronous

motor
(motor and
cam- driving)
Servo
Chart feed
Synchronous
Analog
Function
potentiometer comparator motor and gear
train
Internal sequence
control
Print
Alarm
Recording
Method
Linealize
Convertor
Signal
Scanner conditioner
Ana log
Function
Processing
Measurement
Table 4.2 Comparison between analog method and digital method for multipoint- reco rder
tactless devices in the future.

4 .1.4 Multipoint recorders
The configurations of analog and digital type multipoint recorders
are compared in Table 4.2. Digital recorders are also called hybrid recorders. They have a structure near to that of data loggers, and they
are difficult to distinguish by function. Figure 4. 9 shows the external
appearance of a hybrid recorder.
Fig. 4 9 Hybrid recorder
(a) Construction and operating principle

As shown in Fig. 4.10, the device is made up of an analog block,
an computational and control block, a recording block and a keyboard
block. The analog block consists of a multipoint measurement circuit
made up of a relay scanner and programmable amplifier, and a pulsewidth-modulation, integration type A/D converter circuit that has a
resolution of 4Yz bits. The computational and control block consists of
two microprocessors, ROM, RAM, and peripheral circuits. One microprocessor controls the analog block, the keyboard, and the communication interface; the other controls the recording block.
The recording block is made up of a wire dot hybrid printing head,
chart feeding mechanism, motor driving equipment, and so on. The
keyboard consists of various setting keys and an LED display.
Thirty direct current voltages and temperature input signals are
selecting in turn by the scanner, and undergo A/D conversion according to pre-set measurement ranges in the analog block. This data is
sent to the computational and control block, where, according to the
type of input, it undergoes linearization, scaling, alarm processing and
so on. It is then converted to display data and recording data, and
stored in RAM. The data for recording is converted into analog and
character printing data corresponding to their respective recording positions. Control up to this point is done by the first microproces218
ol>..
t...
fi}
'
......
-tl
Microprocessor ~P,
@ Interface driver
@ Microprocessor ~'
@ Key display
@ Relay output
Isolation
19
~
I
8 bits
8085
Block diagram of hybrid recorder
Computational and control block
Fig. 4 10
Analog t
block
Communication interface (Option)
Memory
<J) Alarm interface
Memory
Key/ Display interface
<!>
<D
Programmable amp lifier

A/ D converter(Pulse width modu lation)
@ Interface
<D
0
Input
:
DCV
'
( TC (Thermo couple)
RTD (Resistance
1
temperature detector)
T"m'':
booro. ( '' JO
I to 30CH
Scanner
arriage motor
Recording block
motor
Chart feed motor
Keyboard block
sor,pP1.
The second microprocessor, t.tP 2 , drives the printing head. As the
head moves across the chart from one side to the other in 0.1 mm
steps, t.tP 2 refers to the recording data. When relevant data is present,
t.tP 2 excites the solenoid of the printing head, causing a dot to be printed. This operation is repeated, and one scan is completed when the
head arrives at the end of the chart. Another scan is performed as the
head returns in the opposite direction to the first side. This recording
method is called raster scanning. As shown in Fig. 4.3, in the hybrid recording mode, 6-color analog recording is done on 250 mm wide chart.
Also, measured values and the date and time are printed on the left
side of the chart. While the alarm data are printed out on the right
side. The scan time is selectable, with a minimum of six seconds. In
logger mode, the measured values are printed across the entire width
of the chart. The scan interval can be set to one minute or any larger
value.
(b) Structural components
(1) Input scanner
In the same way as for pen recorders, there are
different types of scanners for DC voltage thermocouple input and resistance temperature detector input. The former uses low thermoelectromotive force relays for switching elements; the latter uses CMOS
semiconductor switches.
(2) Wire-dot printing heads
Figure 4.11 shows the printing head
structure. This is what is called the wire-dot printing head. It features
a two-layer construction, with solenoid drivers for analog recording
wires in the bottom layer and solenoid drivers for character printing
wires in the top layer. Figure 4.12 illustrates the recording method,
with the black dots in the head representing the wire array. As shown
in the figure, each wire corresponds to the six-color ink ribbon. The
device has wires specially for printing character, and can print one row
of 5 x 7 dot characters in one scan.
The wires are the free-flight type. As shown in Fig. 4.13, the wirRibbon
Ribbon cassette
Analog record wire
Platen
Fig. 4 11
220
Printing head mechanism
Bl ack
Blue - - - -----,
Green - - - - - ,
Purple- ------,
Brown
Red
---.,1--,r+t-h-1... Ink
1-'-.....___ _._.'--'--1 ribbon
: Printed dol
o : Unprinted dot
Fig. 4 12 Recording method

Wire free night time (Going)
Wire free night time (Returning)
Seuling time
8c
!lc
.~
"0
;;::;
.. .
,;:
~
c
:;
0
E
:!
u.
.. ..
~
3 .><
j 10
I <E!:
~ "'
6ms
-Time
Fig. 4 13
Wire movement
es continue free flight after the armature has hit the yoke, strikes the
ink ribbon and paper, and bounces back. This method can get a long
wire stroke, and adjustment and maintenance are easy.
The ink ribbon is a nylon cloth cartridge type. It has six separate
colors corresponding to the analog recording wires, and is moved forward by a motor with each stroke.
In addition to the color wire-dot recording method described
above, multipoint recorder methods also include the multicolor dot
printing pen head method, the heat-sensitive method, the color ink-jet
method, and the electric discharge method.
The external appearance of a recorder that employs the multicolor
dot printing pen head method is shown in Fig. 4.14. As shown in Fig.
4.15, this recorder has the special feature of recording the maximum
and minimum values input during the recording period in line segments.
4 .1 Recorders
221
Fig. 4 . 14 External view of process- use recorder
Recording period
Input ignal
TR
0 -,rl4++~+++~HH~,.---
Trace record of MaximumMin imum valu

o~~~ H ~~~~~~~
-
Fig. 4 15
Time
Relationship between input signal and

trace record of max ./ min. va lues
The dot printing recorders generally have from 6 to 30 dots, and a

recording speed of from 1 dot per 5 seconds to 30 dots per 5 or 6
seconds. However, in the future, with faster A/D converters and faster
processing through distribution of microprocessors to each block, even
higher speed multipoint recorders will become commercially available.
4.2 Controllers
A computerized multi-loop digital control system can conveniently

handle 8-loop, 16-loop or 32-loop multiple processes. However, here
we will describe the structure and functions of a single-loop digital
controller, which handles a single process loop. Multi-loop systems are
explained in Chap. 5.
4.2.1 Pneumatic and electronic controllers
As explained in detail in the history of industrial instrument development (Sec. 1.1), development of the field of electronics was accompanied by a move away from pneumatic controllers towards electronic
controllers, and as computers came to be used in process instrumenta222
ignal
Control valve
( a ) Pneumatic control
Fig. 4 16
Loop configuration of pneumatic and elctronic control
Table 4.3 Comparison between pneumatic and electronic control

Item
Pneumatic
Electronic
Transmission characteristics of signal
Slow
Fast
Usage in explosive gas atmosphere
Safe
Should be careful
Computer compatibility
Poor
Good
Actuator compatability
Good
Poor
tion, the trend became stronger. Figure 4.16 shows the pneumatic and
electronic type loop configurations for a flow process, and Table 4.3
lists the special features of each type. Suitability for use in an explosive gas environment is the greatest advantage of pneumatic controllers. However, compatibility with computers has become an important
point in recent years as central monitoring of processes using computers and digital control have become popular. Thus the trend is towards
electronic controllers. This chapter is mainly concerned with electronic
controllers.
4.2.2. Analog electronic controllers
Figure 4.17 is a block diagram showing the basic configuration of
a controller. Controller operation can be switched among three modes:
manual (M), automatic (A), and cascade (C). In the manual mode, the
output signal (manipulated variable) can be directly manipulated by
means of the manual control block. In the automatic mode, the controller itself compares the input signal (controlled variable) with the setpoint value, and the control computation block acts to make the deviation zero. The output signal (manipulated variable) is then sent to the
operation block. The only difference between the automatic mode and
4.2 Control/us
223
omrol computation block
@ Manual control block

@ Output block
Fig. 4 17
Basic configuration of controller
the cascade mode is that in the cascade mode the setpoint is obtained
from outside the controller.
The indication and operation part of this kind of controller is
shown in Fig. 4.18. The indicator arrangemen t is generally such that
the relationship between the setpoint value and the controlled variable
can be easily seen at a glance. Setting of the setpoint value and manual
operation is generally done from the front of the instrument, as shown
in the figure, and the setting functions for other control constants are
generally placed on the side.
f - Nameplate
FlC-3058
J=
.,. -~
60
Process variable pointer (Red)

Setpoim index (Blue)
~iO
..
1:.
t20
l:. o
Fai l lamp (Red)

Alarm lamp (Yellow)
'
'
!-- Setpoim up/ down switches
Operation mode transfer witches

miD (Q1) [Wo:J f - - with indicator lamps
I Ill "
(g-,__,--
1--- Output indicator
ml
1--- Manual operation lever
Fig. 4 18 Indication and operation section of controller
The methods for manipulatin g the setpoint value and output value
include position form setting methods, which typically use variable
resistors, and velocity form methods which change the setting relative
to the current value using push-button switches. In the example
shown in Fig. 4.18, as an example of any type of velocity form setting
224
switch, there is a push-button switch for increasing the value, and another push-button switch for decreasing the value. While either of
these buttons continues to be pressed, the value changes. However,
there are differences in the manipulating characteristics when used for
setting the setpoint value and when used for setting the output value.
For example, for the setpoint value, the ability to set the value accurately is the important consideration, and there is the single-speed, 40
seconds per full-scale change rate. For setting the output, on the other
hand, two setting change speeds are available. By pressing lightly on
the operating lever, the change rate is the same 40 seconds per full
scale; but by pressing hard, the change rate becomes a fast 4 seconds
per full scale.
(a) Control block configuration example
Figure 4.19 is an example of the circuitry of a PID controller
which uses an analog operation circuit. The deviation signal E undergoes derivative and proportional operations in operational amplifier A1
and integral operation in amplifier A 2 , and is then output. The proportional band, derivative time, and integral time are determined by the
constants of the resistors and capacitors, and the variable resistors can
be set to a wide range of values. Generally, however, resistance of at
least several megohms and capacitors having a capacitance of several
tens of microfarads are required. Thus in circuit design, care must be
taken to maintain high insulating resistance.
A switchover from automatic control (A) to manual operation (M)
or from manual to automatic is performed in front of amplifier A 2
Then, A2 acts as a hold amplifier in manual operation.
AIM
Fig. 4 19
Example of analog controller circuit
Manual operation block

The manual operation block has the important function of allowing
the operator to directly operate the valve at plant start-up or when abnormalities occur. The ability to switch from manual operation to automatic control or from automatic to manual without any sudden
changes in output, known as "bumpless switching," is required func-
(b)
4.2 Controllers
225
tion. In the circuit shown in Fig. 4.19 for example, when the A/M
switch is in the "M" position, the output signal is determined by the
electric charge on capacitor CM. The output signal can be manipulated
by the charging or discharging current flowing through the INC (increasing) or DEC (decreasing) switches in the manual operation block.
On the other hand, the entire output of amplifier A11 which operates on the deviation signal E, is stored in capacitor C1 so even if the
A/M switch is thrown to "A," the charge on CM does not change suddenly. Conversely, the charge on CM is not changed by switching from
"A" mode to "M" mode. In this way, bumpless switching between the
automatic and manual modes becomes possible.
(c) Control action
The simplest control action is the on-off operation. For example

when using a heater to control temperature, the heater is turned off
when the setpoint temperature is exceeded, and turned on when the
temperature drops below the setpoint. However, both overshooting
and undershooting from the desired temperature occurs, producing a
zig-zag control effect. To obtain a good control without constant cycling, PID control action is often used.
The basic formula of PID controllers was shown in Chap. 2
(Eq.(2. 7)), but an actual analog controller such as the one shown in Fig.
4.19 uses the following equation.
MV=l00(1 +_1){
PB
I;s
1+Tos
1+ ( To/m)s
}E
(4.1)
Here, MV is the manipulated variable, E is the deviation, PB is the proportional band, T1 is the reset (integral) time, To is the rate (derivative) time, m is the derivative gain, and s is an operator.
At plant start-up time, the values for PB, T1 and To can be set to
the most suitable values for the characteristics of the process using
the variable adjustors provided on the side of the controller. However,
as seen from Eq.(4.1), the proportional and integral actions and the derivative actions take the form of a product, and create an interaction
in the PID action.
In the third term, 1/ {1 + (To/m)s} is for incomplete differentiation,
so normally the value of m is chosen to be between 10 and 20.
4.2.3 Digital controllers
In direct digital control (DDC) using computers for process control, a large number (up to several hundreds) of loops are controlled at
essentially the same time. However, with the development of the
microcomputer, economical digital controllers which control several
loops or even a single loop have become practical, rivaling analog con226
trollers.
In digital controllers applying microprocessors, ratio control, cascade control, feed-forward control, nonlinear control and other such
control algorithms, which are difficult to achieve with conventional analog controllers, can be easily realized. Beyond that, digital controllers
can be equipped with preprocessing of the input signal, data transfer
to a supervisory system, self-diagnosis, and many other such functions.
(a) Configuration example
Figure 4.20 shows an example of a digital controller which uses a

microprocessor. The entire operational function of the controller is programmed in the system read- only memory (ROM). First, an analog signal is selected by the multiplexer. It then undergoes A/D conversion
and is read into the microprocessor as a digital variable. After digital
processing, the signal is converted back to an analog form by the D/A
converter and then output. This operation is repeated continuously in
intervals 0.1 or 0.2 seconds. During the intervals, the output signal is
maintained at a constant value until it is updated at the next cycle.
This method is a sampling control method, but as the sampling
time is ten times or more faster than the process time constant, its control ability is about the same as that of analog controllers. Also, every
type of computational and control function required of the controller
can be executed by the microprocessor, so these controllers have the
advantage of being suitable for diversified control methods.
It is desirable for the A/D converter used in digital controllers to
r--- ------- --------------,
Digital
Fail contact
-----~2.-.......J alarm
1/ F : Interface
I
I
L.----~~
_Jide.J!llnel _ _
Fig. 4 20 Construction of digital controller
4.2 Controllers
227
Comparator
Analog input 1
M icroprocesser
__...-r-i''>--- - ---1
Analog input 2 ___,....
0/ A
:~
SWJ
2
Analog output
~ Sampled
data holding amplifier
Fig. 4 21 Micro processor based A/ 0 - D/ A converter
have a resolution of at least 12 bits, and conversion speed of a few

microseconds. Generally, the successive approximation method is used.
The conversion procedure is controlled by the microprocessor, which
allows simplification of the circuitry. Figure 4.21 shows an example of
a single D/A converter being used for both D/A conversion and A/D
conversion. First switch SW 1 is turned on, and the comparator, D/A
converter, and microprocessor form an A/D converter by the successive approximation method, and the analog input signal is converted
into a digital signal. When converting the computation results of the
microprocessor into an analog signal, sw2 is closed, the signal passes
through the D/ A converter and is output to the sample and hold amplifier. The A/D conversion method is explained in detail in Subsec. 5.5.2.
(b) Controller action
For digital controllers, any type of control algorithm can be constructed. In the automatic mode, the following PID algorithm is generally used.
MV=~0~{Pv+
is
E+ 1+(tfm)sPv}
(4.2)
where PV is the controlled variable.

In this equation, the integral action term and the derivative action
term are independent additional operations, so the interaction between
reset time and derivative time seen in analog controllers does not occur. Furthermore, the proportional action is only with respect to
change in PV, so even if the setpoint value (SV) is changed, there is no
sudden change in the output, which is a desirable operating characteristic.
In the cascade mode, the following algorithm is largely used.
100{
1
Tos
}
MV= PB E+ 'n.s E+ l+(To/m)sPV
228
(4.3)
ChapA Recorders and Controllers
c Mode
e.t point value

Change point of setpoint variable
F ig. 4 22
PV Response to setpoint va lue chang
In this equation, the proportional action responds to the deviation

(setpoint minus measurement) signal of the controller. Thus fast response to changes in the setpoint value is achieved. The difference in
controlled variable response to stepwise change in the setpoint variable
for the two kinds of control represented by Eqs. (4.2) and (4.3) is
shown in Fig. 4.22. Digital controllers can in this way switch between
two or more control algorithms and execute the most suitable control
for the purpose at hand. The processing within the digital controller is
Operator console
with color CRT
Communication and monitoring system
Communication bus
Communication
function
Sequence control
function
Batch . Blending system
Fig. 4 23 Configuration of control system
4 .2 Controllers
229
a discrete in sampling interval T, and Equations. (4.2) and (4.3) should

be expressed in discrete form . However, for ease of comparison with
analog controllers, they have been expressed in a continuous form.
(c) Communication with a supervisory system
When considering the configuration of a process control system,
generally the principle of "distributed control and centralized information" is followed. Figure 4.23 shows an example of a system configuration. Digital controllers are arranged in a dispersed fashion , with a
CRT -equipped operator console as the heart of the system. Input data
coming through the communication bus from the controllers can be centrally monitored on the CRT, and at the same time even a change in
the setpoint value and output value for each controller can be manipulated from the supervisory system side. This type of configuration is
called a setpoint control system (SPC) or direct digital control (DDC)
system.
An example of the operating conditions of a controller displayed
on the CRT of the supervisory system based on the data sent from the
controller is shown in Fig. 4.24. This screen shows the input and output signals, and at the same time the high and low alarm settings of
the input signals (PH and PL). Thus abnormal process conditions can
be detected from the screen.
TUN IN(; PANf.L
F 400 3A GAS FLOW
PV
sv
MV
ov
PH
PL
DL
-.-
M
Ml
p
I
0
146 . 3
I 32 . 0
48 . 5
I 4. 3
190 . 0
1(\ . 0
so . o
100 . 0
0.0
150 . 0
0
2 . 571
as
0 . 450
cs
AUX I
6 . 9
AUX2
3 3. 5
79 . 1
AUX3
Fig. 4 24
I T 3
AU T
NM / H
NR
,.
f
0.0
"w
200.0
53.5
SEC
SEC
,.
,.
t
Mon itor and operation panel of controller
(d) Self-diagnosis function
One special feature of digital controllers using microprocessors is

self- diagnosis. The controller confirms its own operating conditions.
When it detects an abnormality, it promptly notifies the operator by
lighting a lamp or sounding a buzzer, thus improving the safety and
maintainability of the plant operation. The following are self-diagnosis
230
items.
(1) Microprocessor abnormalities
(2) A/D converter and D/A converter accuracy
(3) Range over of input signal
(4) Computational overflow
(5) Break in the electric current output
4.2.4 Programmable controllers
In analog controllers, there is a one-to-one correspondence between hardware and computation functions, thus the execution of complex control calculations is difficult. On the other hand, with controllers employing microprocessors, various additional control calculations
can be performed through software. Thus in addition to the basic PID
algorithm, complicated control algorithms can be performed easily by
assembling various software modules chosen according to the requirements.
Figure 4.25 shows a programmable controller set up for flow ratio
control. In controlling flow so as to maintain a fixed ratio between line
A flow rate and line B flow rate, the flow rate (pressure differential)
of line A is measured, the square root is taken, the ratio is found by division, the bias added, and the result serves as the setpoint value for
line B control. Accordingly, it is necessary to combine the square root
A Line
T~
x,
~[
,B Line
Fig. 4 25
4.2 Controllers
y,
A Line
Manipulated output
Application example of flow ratio control
231
operation, division operation, and addition operation.

Programmable controllers can be programmed by the either the fillin-the-form method (FIF), or the step writing method (procedure-oriented language, problem- oriented language in a narrow sense). In the
former, module specification items are written into a table to assemble
the program; in the latter, the functions to be executed are written as
instructions. We will describe the step writing method here.
(a) Computation principle
The operation of programmable controllers is built up of three
functions (input reading, computation, and output of the result), and
can be written as instructions. In Fig. 4.26, the program for a two-input adder and an arithmetic register operation are shown. The computation is done completely in a common arithmetic register S. Connection of the adder signal, that is to say the S register input, is performed by a LOAD instruction (written as LD). The S register consists
of stacks S 1 to S5 By the LD instruction, the data is pushed down from
S 1 to S2 with each input. The FUNCTION instruction is used to perform operations on the input data. The operation modules required by
control computation are prepared. They are written as symbols such as
+, -, -o-, x , ..;--, and so on. The data in the S register is used for
the number required by an operation, and the results are stored in
regsister S1 To read out the computation results, the STORE instruction (written as ST) is used.
XI
X2
tL1.Lt
LD XI} Input reading

LD X2
) amputation
ST Yll Output result
Yl
LD XI
LD X2
S T Yl
1/ 0 Register
. .
Anthmet1c
s,
. s
reg1ster s,
s.
s,
Lot
Fig. 4 26
Arithmetic tack regi_ster operation
(b) Structure of the input/output register

Figure 4.27 shows the structure of the input/output register.
Before execution of the user program, the analog input, digital input
and setting parameters are put into the XN, DIN, PN registers, respec232
1- SV
Input conversion
I
I
t
I DJii I
l .t. l 1~ 1
ob
In put regi ter

An alog input Digital input Variable parameters
II'UNt BSCI
ISIO_RE.Y
Output
conversion
'-
User program
ll.QAJ
0. 2 Sec.
Repeated
Analog output
Output
1
Xe ) I
reg1 ster ~
t
4- 20mA
Di gita l out put
~J
I
1- SV
Fig. 4 27 1/0 Register configuration
tively, as normalized values. In the user program, as described earlier,

the required input signal and parameters are loaded from their respective registers into the arithmetic register by the LD instruction. Then
the results of the calculation are connected to the output register by
the ST instruction. Next, the contents of the YNor DON registers are
converted to analog or digital output. The user program handles all signals as normalized digital data, so concern over process input/output
con version is unnecessary.
(c) Computational and control functions
Examples of the computational and control functions of a programmable controller are listed in Table 4.4. The basic PID control function
BSC, and the cascade function CSC, in which two controller units are
connected in series, are each defined as a different computational function. Thus, if the programmable controller is to be operated as a simple PID controller, then the program is simply written as follows.
1 LD Xl
2 BSC
3 ST Yl
4 END
The BSC functions are shown in Fig. 4.28. As a dedicated extended function of BSC, the A and FL registers can be used by the program. For example, when it is desired to set the setpoint value for cascade, the cascade setting input is connected to the Al register by the
ST instruction.
The program for the flow rate ratio control in Fig. 4.25 is written
as follows.
4.2 Controllers.
233
Read in X2 input
Calculate .JX2
Read in the ratio
Calculate P1 .JX2
Read in the bias
Calculate P1 .JX2 + P2
Connect to the cascade input terminal
Read in X1
Calculate .vXI
Basic control module
Connect the results to the operation output
1 LD X2
2
...;-
LD P1
*
*
LD P2
6 +
7 ST A1
8 LD X1
9 ...;10 BSC
11 ST Y1
12 END
PV Input
! Input Alarm
'V
A, Cascade
A:
Adaptive gain
Output
A, compensation 1I
A,
~- r+--
C IDevia tio n alarmf l
Output tracking
L__
High alarm
low alarm
Deviation alarm
Velocity alarm
FL,
FL,
FL,
FL,
I Nonlinear gain I
t
A,
set
Input
compensation
---,- ~
AI M ! +
PID
{ Sta ndard PI D
1 Sample>-and-hold PI
Batch PID
+
+
~ --
C/~(.'!.j
-- ------1 Tracking
M""''
FLo
.,,.,,,I
IOutput limiter I
Fig. 4 28 BSC Function
(d) Programmer and programming procedure
The program can be written either by using the programming panel on the side of the controller to write directly to the RAM, or by using a dedicated tool called a programmer to write to the ROM. The
functions of this programmer include:
(1) Program key-in,
(2) Display and set the computational constants,
(3) Test the finished program,
(4) Display the process input and output,
(5) Write to the ROM, and
(6) Print out the program listing.
234
Table4.4 Computational functions of programmable controller

Function name
Instruction code
Basic four function
+-*I
Signal conversion
Absolute
Square root extraction
Line segment
ABS
SQT
FX
(Absolue)
(Square root)
HSL
LSL
(High selector)
(Low selector)
HLM
LLM
VLM
(High limiter)
(Low limiter)
(Velocity limiter)
Dynamic functions
1st order lag
1st order lead
Dead time
Velocity
Timer
LAG
LED
DED
VEL
TIM
(Lag)
(Lead)
(Dead)
(Velocity)
(Timer)
Analog conditional decisions

Comparison
High alarm
Low alarm
CMP
HAL
LAL
(Compare)
(high alarm)
(Low alarm)
Logical functions
AND
OR
NOT
Branching
AND
OR
NOT
GIF
(AND)
(OR)
(NOT)
(Go if)
Control functions
Basic control
Cascade control
Selector control
esc
sse
BSC
(Basic control)
(Cascade control)
(Selector control)
Other
Analog I/0
Digital I/0
Branching
End
LD
ST
GO
END
(Load)
(Store)
(Go to)
(End)
Selector
High
Low
Limiter
High
Low
Velocity
lf(x))
Figure 4.29 shows an example of such a programmer connected to

a controller.
4.2.5 Batch controllers and blending controllers
Fluid blending and batch shipment systems use dedicated controllers. Figure 4.30 shows an example of the configuration of a batch ship-
4.2 Controllers
235
Fig. 4 29 Connection of programmer

Batch controller
(D Pulse now signal
Fig. 4 30 Batch
Temperature signal
hipment y tern
Batch set station
B Component
Fig. 4 31
236
Blending shipment ystem
ment system. Batch controlleres are PI controlleres that perform programmed control of the measured flow rate such that accmulated flow
produces a pre-set batch volume. The program is written to suit the
controlled object, and can be freely specified at the time.
Figure 4.31 shows an example configuration of a blending shipment system. The blending controller receives a set signal from the
bath set station, and performs PI control such that the integrated values of the set signal and the measured flow rate maintain a fixed ratio.
The control algorithm for blending controllers is expressed by the
following equation.
MV= 100(l__+ 1 2 + r,
PB s T.s
v
E=SV-PV
)E
(4.4)
(a) Display, setting, and operating functions
An example of the front display and operation panel of a batch controller is shown in Fig. 4.32. The batch sequence command switches
are located at the top, the batch set value and the totalized value can
be monitored at the same time on a two-stage digital display, and in
Self di agno tic lamps
ALM FAIL
LOAD PRE END
START
RESET
FLOWc::::lc::::)
_,c 5
i
STOP
C;;:c::::llc::::::
s 5 n un
(.1
Dml
_,::J
'-{
c 6
-Seq uence command switches

+- Instantaneous now display (Bargraph)
..._ Batch setpoim
..._ Data display
+- Data set switches
Data displ ay selector switches
+-Operat ion mode selector witches

+--Manipulated output indicator
-
Manual operation lever
1.1:;;;;;;;;;;;;::;;;;;;;;;::==~
Fig. 4 32 Front view of batch controller
4.2 Controllers
237
addition, all parameters can be set using this display.

(b) Process inputjoutput interface
The inputs to a batch controller include the flow rate input, compensation input, and contact input. Also, the form of the signal from
the pulse flow signal transmitter has not been standardized, so there
are differences in specifications such as the power supply voltage for
the transmitter, the signal amplitude, the frequency, and the load resistance according to the manufacturer. Thus it is necessary to provide
interfaces matched with these various specifications, and there is some
inconvenience concerning maintenance. Figure 4.33 shows an example
of a pulse input circuit constructed to support a wide range of transmitter specifications. The transmitter power supply and load resistance
can be switched to match the specifications of the transmitter. The
specifications of this pulse input circuit are listed in Table 4.5.
(c) Flow rate compensation
In batch blending systems, flow rate compensation is an important
Power supply isolation
Signal isolation
JUUl
Fig. 4 33 Pulse input circuit
Table 4.5 Pulse input specifications
238
Item
Description
Type of pulse signal

Frequency
Minimum pulse width
Voltage level pulse or contact

0 to 6.0kHz, zero elevation not available
50t.ts
Voltage level pulse signal

ON (EH)
OFF(&)
Pulse peak value (EH- &)
Input resistance
Power supply voltage
Current
3 to 24V DC
-1 to +8V DC
3V or more
lOk.Q or more
12V DC or 24V DC
8mA or less
Contact signal
ON
OFF
Signal source rating
Signal source resistance 200!2 or less

Signal source resistance lOO.Q or more
30V DC, 30mA or more
function for accurate trade. High accuracy is required, and the compensation method must be adaptable for various types of compensation
according to the type of fluid. The following three types of compensation computational formulas are used.
c=(l +a)f(p, t)
(4.5)
c=(l +a){ I+ 10-2 ,8(t- to)+ 10-6 r(t- to) 2}
(4.6)
c=(l+a){(Th)+0.5}, TI2=0 to 1.0
(4.7)
F=cFo
(4.8)
In these equations, c is the flow rate compensation coefficient, a, ,8,

and r are compensation coefficients that can be set freely, p is the density, t is the measurement temperature, t 0 is the compensation reference temperature, F 0 is the flow rate before compensation, and F is
the flow rate after compensation.
Equation (4.5) relates mainly to petroleum, and using the generally
used ASTM* (specific gravity-volume conversion table), automatically
calculates the compensation coefficient for the reference temperature.
The only specified data is the specific gravity of the fluid. The coefficient a is for specifying compensation for instrumental error of flow
transmitter, required depending on the properties of the fluid such as
viscosity. It can be used for various other purposes according to the
fluid being measured.
Equation (4.6) is a multiple-item compensation computation for
processes other than petroleum processes, and has a wide degree of
freedom. The temperature vs. density curve is approximated by a quadratic equation, and compensation coefficients ,8, and r are set. Equation (4. 7) is for special conditions for which the above Eqs. (4.5) and
(4.6) cannot calculate compensation, and is used by connecting with an
externally computed analog compensation input (TI 2). It is linearly converted over the range of 0.5 to 1.5 corresponding to a 0% to 100%
change.
(d) Flow rate addition and subtraction
In the shipment of fluid of volatile fluids such as LNG, several percent of the flow changes to a vapor and returns, causing error in the
shipment volume. The measured vaper flow is introduced to the flow
rate addition and subtraction function through analog or pulse input,
and compensates for the lost amount, thus making highly accurate
shipping control possible. A block diagram of the functions of a batch
*ASTM: American Society of Testing Materials.
4 .2 Controllers
239
Fl- 1
Fl- 3
TI-l
T l- 2
CD
Process variable input (Pulse signal)

Process variable input or added/subtracted
flow signal (I to 5 V)
Added/ Subtracted flow signal (Pulse signal)
(!) Compensation signal (PtIOO 0)
Compensation signal (I to 5 V)
Start, Stop, Reset signal
Cumulative flow totalizer

(With added / Subtracted flow signal and
compensation computation)
@Alarm
@ Instantaneous flow display

@PI Control
@ Manual operation
Batch end pre-alarm
Batch end alarm
@ Flow signal repeater (Pulse signal)
@ Alarm output
(j) Master pacing

@ I) ASTEM equation
2) Quadratic equation
3) General compensation equation
Flow setpoint program
Measured batch flow

@ Cumulative flow totalizer
(With added/subtracted flow signal)
@ Measured batch flow
@ Flow signal repeater (I to 5 V)

@ Communication
@ Manipulation signal (4 to 20mA)
(With compensation computation)
Fail signal
Reset signal
@ Batch flow totalizer (With added / Subtracted

flow signal and compensation computation)
Fig. 4 . 34
240
Block diagram of batch controller
controller are shown in Fig. 4.34.

(e) Batch sequence
The batch sequence is controlled by the start, stop, and reset buttons on the front panel, or by external contacts. At the same time, the
status of the controller can be known from lights or the output of external contacts. Figure 4.35 shows the relationship between batch sequence and flow rate settings.
Stop
Maximum flow set value
\I
E
3:
u:
Initial flow set value
I
I
I
I
stan
Reset
M a ier pacing
<7
Initial total ized-flow limit value
/O FF
'
'
\..J
Minimum set va lue
'
Time Prebatch quantity
Predicted leakage vo lume
t:l
Batch quantity
11
E
"
...J
~~
:$:
Stop status
:.~A :
PRE
! ~l
END
: ~:
LOAD
e
@
Prebatch Open
Open
Close :Ope~
Close
Open
Open
;"'"'
c
Batch
Open
Close
Close :Ope~
Close
C lose
Open
"'~
Reset
Close
Open
Open :Openi
Open
Open
Open
. Fig. 4 35
Batch sequence and flow rate sett ing
4.3 Computing Stations and Set Stations

4.3.1 Alarm set stations
An application example of an alarm set station is shown in Fig.
4.36. For safe system operation, control status is monitored by an
alarm set station, which is a system component that is independent of
the controller. This device monitors the absolute value of the measurement signal, the deviation between two signals, and so on. When an abnormality arises, its contact output sounds an alarm buzzer, may initiate a system shutdown sequence, or initiates some other such action.
An example of the circuit of an absolute high and low alarm set station
is shown in Fig. 4.37. Generally, the action mode is selectable accord4.3 Computing Stations and Set Stations
241
Fig. 4 36
Application example of a larm set station
v
Alarm action
---:v cgM
'----------<0
E,.
R,
~~
Ll
CD
Direct
action
Reverse
action
~Ligh t
emiu ing
' - - - - - --+-- + ----'D, diode
L2 GND
Fig. 4 37
Principle of ab olute high & low alarm set ta tion
ing to whether the relay is activated or deactivated when the alarm is

generated, with deactivation being the fail-safe direction.
After the input signal has reached the alarm setpoint, the relay
output is latched by a lockup function such that it does not switch
back and forth from the on and off positions with small changes in the
input signal. The relationship between the relay operation mode and
the lockup action is shown in Fig. 4.38.
4.3.2 Programmable computing units

In addition to square root extraction, temperature compensation,
and other such operations on the signal from the sensor, this device
also performs filtering for input to a recorder or controller. Conventional analog computers were also used for this purpose, but because
the computation was implemented in hardware, the accruracy of com242
For direct action
For reverse action
------r-,t
Energized
Energized
Deenergized
Low -
t
I
Alarm
setpoint
Low -
Alarm setpoint
High
()
Input
Fig. 4 38
1 Deenergiled
--'------- - High
Input
Lockup action
plicated calculations was poor and it was sometimes the calculations

were not possible.
In the same way as with the programmable controllers in Subsec.
4.2.4, the application of microprocessors allows complex computations
to be freely performed through the use of groupings of computational
software modules. Figure 4.39 shows an example of a programmable
computing unit. It dose not have a display or operating panel, because
it is not normally operated by an operator.
Parameter setting I
Fail lamp
Alarm lamp
Parameter setting 2
Nameplate
Fig. 4 39
External view of programmable computing station
Figure 4.40 shows the interior of the device. The program is written by the step writing method using a dedicated programmer, and
written into ROM. The specifications of this type of programmable
computing unit are listed in Table 4.6. This device can perform the
computations of ten conventional analog units, so it is possible to simplify the rack instruments, which have become overcrowded as a result
of current control requirements.
4.3.3 Manual set stations and manual operating stations
A function block diagram for the manual set station is shown in
Fig. 4.41. Figure 4.42 shows the function block diagram for the manual
operating station. The internal function configuration excludes the
4.3 Computing Stations and Set Stations
243
Fig. 4 . 40
Internal view of programmable comput ing ta tion
Table 4.6 Specifications of programmable computing station

Item
Description
Analog input
1 to 5V DC, 3 points (4 points available with no digital in

put)
Analog output
1 to 5V DC, 2 points
Digital input
ON-OFF signal dry contact or voltage, one point
Digital output
Transistor open collector, one point
Adjustable constant
One turn potentiometer, 2 points (0 to 100% scale)
Fail output
Transistor open collector, one point
Calculation function
Addition, subtraction, multiplication, division, and logical sum, etc., 26 functions
Program step number
Max. 99 steps
Cascade signal
(I to 5V DC)
Status input for C/ M tran fer
M
-
... - - - - - - - - - - - -
Setpoint output I Setpoint output 2

(I to 5V DC)
( I to SV DC)
Fig. 4 41
244
' Contact output
- ... - - - - J
C/ M Status
o utput
Manual set station (Setpoi nt output)
ascade signal
(I to 5V D )
latus input for

/ M transfer
Process variable
{I to 5V D )
'
I
I
I
PV
I
I
I
________ _____ .JI
-------------,
I
I
<D
ontrol output (4 to 20mA)

ontrol output (I to 5V)
r-----~~M~V~~
ontrol output ( I to 5V)

/ M Statu output
I
I
I
I
I
Fig. 4 42 Ma nual operating station (Manipulated variable output)
PID computation function from the controller.

The manual set station is sometimes used to provide a common setpoint value to a number of controllers. On the other hand, from the
system safety design point of view, the manual operating station is often used as a controller backup device. It is also used as the final control device in the application example in which the output of one controller unit drives a number of final control elements.
REFERENCES
1) Yokogawa Electric Corporation : Industrial Instrument Handbook Pneumatic Instrument Volume, 2nd Edition, Tokyo Denki University Press (1979) (in Japanese).
2) T. Utsumi : "Measuring instruments-Recorders ," Yokogawa Tech. Rep., 28, 2
3) H. Ohtsu, et al.: "Programmable intelligent recorders 11R 100 I tJR 180 series,"
4) I. Morishita : Process Control Digital Instrumentation Systems, The Society of Instrument and Control Engineers (1983) (in japanese).
References
245
5) S. Narita : Digital System Control, Shoko Do (1980) (in Japanese).

6) H. Tamura, et al.: "Electronic control system YEWSERIES 80," Yokogawa Tech.
Rep., 25, 2 (1981) (in Japanese).
7) Yokogawa Electric Corporation : Model SLPC, SLMC, SCMS, SPLR, SPRG Programmable Instruments Functions awd Applications, Yokogawa Technical Information, TI-1B4C2-02E.
8) Yokogawa Electric Corporation : Model SLCD (Style E) Indicating Controller Functions and Applications, Yokogawa Technical Information, TI-1B4C1-03E.
9) Yokogawa Electric Corporation : Batch-Blending Control System Overview, Yokogawa Technical Information, TI-1B4A3-01 (in Japanese).
246
Chapter 5
SYSTEM CONTROL EQUIPMENT
The computer was introduced to process control in the 1960's, and

broad advancement in digital control techniques was achieved. After
that, electronic devices such as microprocessors developed rapidly.
Hand-in-hand with the development of digital communications, distributed control systems first appeared in 1975. Use of this type of system
quickly spread, and today has achieved prominence among process control systems. This digitization has greatly transformed control systems, bringing broad improvements in productivity and product quality, and reducing manpower requirements. From a broad perspective
with respect to industrial processes, these automatic systems are
called process automation (PA).
In the area of discrete processes, where assembly work plays the
leading part, factory automation (FA) has recently been developing rapidly towards automation and labor reduction through combinations of
such automation machinery as computers, numerical control machines,
and robots.
In recent years, the control systems of entire factories have been
networked into unified systems which include production management
computers. Through the integrated system design, production-related
information is accurately transmitted among the system components.
Direct connection of this information and the control system shortens
production lead time, makes production more flexible, greatly reduces
production costs, and leads in the direction of a system matched to an
era of diversified manufacturing. This kind of unified production system centered around a computer is called computer-integrated manufacturing (CIM).
This chapter narrows the field of digital control, and explains the
most important basic points of distributed control systems for continuous processes, production line control systems for discrete processes,
and production management computer systems. The description uses
Chap.S System Control Equipment
247
actual examples as much as possible to facilitate understanding and

show how everything fits into the total system. The main features of
high-reliability technology, and the basic software such as real-time
operating systems and control algorithms, and other software which
form the basis of digital control are also introduced. These points help
clarify the trends in technology.
5.1
Overview of System Control Equipment 1 l
5 .1.1 Development
Instrumentation system equipment has progressed hand-in-hand
with the expanding requirements of the process industries. This progress has been supported by the development of electronics and control
technology. The system control equipment dealt with here is mainly
for digital control systems. Figure 5.1 shows an example of the historical development of such systems.
In Japan, the automation of process control (PA) for continuous
processes (fluid or continuous material flow) began around 1955, and after that achieved rapid progress. Now, the field of control systems has
'65
'60
~
CD
'75
'70
'85
.l
I
1962
ccs 2000
'ccs
~
CD
1972
YODIC 100
1966
'\ YODIC 500
EBS
,. First control
com p uters
First DOC
sy stems
CD Factory management system

@ Control system
Process control computer system
CD Centralized DOC system
Fig. 5 1
(j)
/!
1976
YODIC 1000
~,975 c
,..,
YEW MAC
CENTUM
ECS
1982
\YEWCOM 7000/8000
1970
YODIC 600
(Dual CPU)\
32 loops \
248
'80
.!SERIES
II
1979
YEW PACK
\
8 1 ~~~ps
l YF:W.SERIS 80
I loop
First distributed One-loop DOC

control system ' Factory management
computers
using microprocessor
Analog control system
Distributed DOC system
(j) Factory management computer system
@ Manufacturing .Jine control system
Minicomputers
History of industrial computer and control systems in Yokogawa Electric Corp.
Chap.5 System Control Equipment
firmly taken root, both technologically and conceptually. These systems are the backbone of plant operation, and are closely related to
productivity, product quality, safety and so on. Looking at the breadth
and depth of the increasing practicality of automatic control, it can be
seen that the work done by humans is being given over to machines
step-by-step as it becomes possible. Process control equipment is making the transition from pneumatic devices to electric or electronic
types, through the introduction of computers, and on to direct digital
control (DDC), distributed direct digital control, and single loop controllers. The digitization of control equipment is bringing about rapid development of all aspects of control systems.
One such aspect is controllability. There has been spectacular development in control theory, and this has had its effect on the field of
process control. However, the basis of control algorithms, even now, is
still the classical control theory, and the PID control method is the
mainstream of process control. Still, complicated calculations are possible through the use of digital computing equipment, and with appropriate switching of PID control parameters and nonlinear control, advanced control yielding high productivity while keeping in mind safety
limits is being achieved. Also, by combining sequence control and PID
control, the automation of plant start-up and the full automation of
batch processes has become possible.
Moreover, process control was performed by humans actively handling the control elements, thus managing the entire plant. However,
as plant construction became complex and highly developed, and the
pressure for reduction of labor costs became stronger, the dependence
of process control operation on computers became higher. Also, the
operational feature of process control changed from large instrument
panels to the completely new method of the CRT. The copious flexible
functions of the CRT allowed the revolutionary transition to graphic
display of infonnation, clear, simple and accurate alarm indication, and
simple operation based on one-touch operation that effectively prevents operation error. Furthermore, the meaning of operation itself
has been changing from the adjustment of control loop settings to management aspects including specification of product names, production
amount, product quality, and so on. Figure 5.2 shows the conceptual
trend in operation. The trend towards integrated control systems is expected to continue, and operation of these systems will undergo great
development making use of high performance features such as display
of data from other systems on CRT monitors, ultra-high density, and
artificial intelligence.
The cost-performance ratio of computers is decreasing by a factor
of 10 every ten years, with capacity and speed of computation increas5.1
Overview of System Control Equipment
249
In the 1930s
Mechanica l instruments
Field o peration
~~~:r~::=---~~
IC
In the 1940s
Large pneumatic instruments
ignal tran mission
In the 1950s
Sma ll pneumatic instruments
In the 1960s
Small electronic instrument
Dig ital computer
From 1975
Di tributed control system
C RT Operation
Fig. 5 2 History of man-machine interface

(Reprinted from "Prevention review" 140 p.I8
issued by Japan non -life insurance association)
ing. Powerful database management systems form the core of production management, such as production and process planning, and process line management such as optimizing calculations and process
simulations. The apportionment of these various areas of management
is becoming a goal of management by control, and the transformation
250
Chap.S
System Control Equipment
--_
r- -------- ----------------,
Management information system :Management information comrol:
t~~~~~t~~n-~~r~~---Production planning.
Operation planning
Product resullS management
..
Process control
Production scheduling
Optimiution Ana lysi

calculation
Data acqui itio n
etc.
c::;;
0~
;;;;
v-
"0
-gg
0
~
O.. v
raphic display
Comrol
DO , Sequence control
Monitoring Operation
Recordi ng ( Loggi ng)
etc.
Production
planning
control
level
Production pl anning
lnvemory comrol
o t ca lculation
Operation planning
Process
plan
control
level
Daily production planning

Production plan for each process
Progres control
Load distribution
Proces
line
control
level
Production data acquisition and reporting

Optimization. Ana lysis calculat ion
Recipe comrol Quality cont rol
Production line si mulation
l1;l
;;;-
ec
0
_!
1_
D i tr ibuted control sy tern
Production line cont rol

system
Process operation and

monitoring
PID control
Batch control
Advanced control
Data logging
Productio n line opera

tion a nd monitoring
Mac hine tool comrol
Transfer machine
Automatic inspection
Data loggi ng
Fig. 5 3 Trend of role in proces con trol com puters
of the control system structure into a hierarchical form is becoming

clear. (Fig. 5.3)
Taking the automotive industry as a representative case, discrete
processes like machining and assembly developed, quickly improving in
productivity compared to the original mass production or lot production. Regarding the control of these processes, numerical control (NC)
first appeared in 1948, followed by the machining center and DNC.
Then came the programmable controller (PC) in 1969, which came to
play an important role. Entering the era of diversity, low-volume production of a varied line of products followed the trend of the times. To
handle the business related to this, in those years the personal computer, which had become economical and spread very quickly, had begun
to be used in sales management. Although the concept of factory automation (FA) had begun to be known in the 1980's, in 1983 the special
features of the programmable controller and the personal computer
were integrated, and the FA computer was achieved. Thereafter, FA
made great and rapid progress.
Actually, however, FA has still not been fully developed. The automation of factories is not sufficient. Also, the combination of
5 .1 Overview a/ System Cantral Equipment
251
CAD/CAM with production management is still in the planning stage.

However, even though there are no large integrated systems, partially
automated lines consisting of combinations of numerical controlled machines and robots, and automated local material flows, as well as the
linking of assembly line and production management being steadily developed. These have reached a fairly advanced level.
5.1.2 Configuration of a total FA system

Recently, terms like "complete factory automation," and "total
FA" have begun to be widely used. However, these expressions mean
more than simply automating manufacturing processes. As shown in
Fig. 5.4, these concepts apply the latest in information processing, control and mechatronics technologies to such production activities in the
factory as design, production line, inspection, production management,
and orders and shipping. The goal is an integrated factory system designed for flexible, minimum-cost production. The important point
here is that the automated systems of all sections are connected together by unified communication network, managed by a central database
in such a way that the required data is sent to the section that needs it
at the time it is needed.
Order estimate
CAD
CAM
CAT
FA
:
:
:
:
Computer aided design

Computer aided manufacturing
Computer aided testing
Factory automation
~----------------------FA----------------------~
Fig. 5 4
Concep!Ual diagram of FA
Figure 5.5 illustrates the concept of the total FA network. The upstream PA (continuous batch process) and the downstream FA (assembly, inspection, and packaging) are subsystems with individual busses
in a distributed control configuration, but in total FA these subsystems are connected in the trunk line of a unified factory local area network (LAN). However, although we speak of them as connected, there
The terms "total FA," "PA" and "FA" are used. However, in this chapter, "total FA" has the broader meaning of a combined control system which includes PA.
"FA" refers to a control system in the narrow sense of a discrete process.
252
1--- - - - - - -- - -
Total FA
- - - - - - - - - --1
Administration management computer
f - - - L A- . J
~-----PA ----~~------ FA -----~
CD
Supervi ory computer

YEWPACK Package control system
CE TUM Distributed control y tern
@ Operator station
YEWMAC Production line control system
<V
Line computer
Line controller
Fig. 5 5
YS-80 Electronic control sy tern
Robot
Visua l in pection
@ Visual in pection tactile sensor

@ Contenuos Batch proces
@ Assembling
Inspection / Packing
Schematic diagram of total FA network (Example of Yokogawa Electric Corp.)
are few examples where the PA and FA subsystems exist together in

the same factory. Thus it is better to think of total FA as the connection of their respective subsystems.
When making a total system with a hierarchical structure such as
shown in Figs. 5.3 and 5.5, the roles of the subsystem functions for
each level must be determined. Table 5.1 lists rough standards related
to the characteristics and processing time of information for each
level. The upper levels require the capability to process large volumes
of data and long- term data storage;in the lower levels, real-time response is the important factor. Timing and data volume information of
a suitable level is exchanged among subsystems both horizontally and
vertically in the hierarchy by means of the communication network.
5.1
Overview of System Control Equipment
253
"'.....;;:
;;!
><::>
"~
.....
;;:
<;:)
(]
"'~
.....
"'
<,-,
{;.
(]
Production plan for

each process,
Progress control,
Load distribution,
etc.
Schedule control
management level
Production planning,
Inventory control,
Cost calculation,
Order control,
etc.
Production Weeks to
months
dept.
Per shift
or day
Hours
A few
seconds
1 month to Several tens M bytes

to several hundreds
3 months
M bytes
Several hundreds
M bytes or more
6 months
to 1 year
Several tens
seconds
Days
1 month to Weeks or
months
6 months
Factory
Production planning
management level
Several hundreds
M bytes or more
1 year or
more
Several tens
minutes
Data bank
capacity
Months
Months to
a few
years
Entire
company
Management
information
gystem
Months
Response time Data

Management Data
retention
acquisition (Transaction
Unit
time
processing)
cycle
Managing
period
Relevant
Dept.
Information peculiarity in each level
System
Table 5.1
~
v.
~
....
j"
~
~
;::
~
g
....::!
"'
~
~
;;;;
~
......
Relevant
Dept.
Managing
period
[Discrete process]
Control
Alarm
[Continuous process]
Control
Alarm
Operation
Monitoring,
Data acquisition
Production (Real time)

line process
Process line management Production For each

Monitoring,
shift and
section
Data acquisition,
day
Analysis calculation,
etc.
System
Minutes or
hours
2 sec. or less
Scan cycle lOO)ls to 10 ms

1 ms to 1 sec.
Scan cycle 100 ms or less

1 to 10 sec.
(Real time) 1 sec. or less
Minutes
A few ms
to 1 hour
Days to
weeks
Management Data
Response time Data
retention
acquisition (Transaction
Unit
cycle
processing)
time
Several tens K bytes

to several hundreds
K bytes
Several M bytes to
several tens M bytes I
Data bank
capacity
5.2 Distributed Control System 2 >-a>

5.2.1 Concept of the distributed control system
The basic architecture of the distributed control system, as shown
in Fig. 5.6, consists of an operator station, a control station, and a communication system. The operator station collects data relating to the
process operation and can display and manipulate those data. The control station contains control functions such as for the DDC. According
to the scale and type of object process, multiple stations are used. However, the degree of dispersion differs according to the number of control loops to be processed. There are 1-loop, 8-loop, and 40-loop types.
The communication system exchanges data at high speed between the
operator station, the control station and other stations.
The distributed control system consists of the four interfaces depicted in Fig. 5.6. The process interface is the interface between the
distributed control system and the plant (i.e., measurement sensors
Man-machine interface
Display and operation functions
Man-machine data-base
Communication
..c
0
B ] systems
<IS
!i
.s
way
c5
.Control
Control
data-base
Process interface
Fig. 5 6 Architecture and four interfaces of distributed control system
256
and final control elements), The control stations receive measurement

signals from sensors such as for temperature, pressure, and flow rate,
and perform control calculations in accordance with the deviations
from the setpoint values. Output signals are then sent to the final control elements to perform compensatory actions.
The man-machine interface is the interface between the distributed control system and the operator. It does central monitoring of the
plant and permits the operator to perform operations. The operator
console is composed of a powerful microprocessor, CRT, and keyboard.
Many different operations can be carried out by switching displays on
the CRT. Operating each plant using the CRT as above is called "CRT
operation." System operation using electronic instruments arranged together on the control panel for control loop unit, is called "board operation."
Other interfaces consist of a supervisory computer interface and
control sub-system interfaces. The supervisory computer interface connects the distributed control system to a supervisory computer. It
transmits control data and receives supervisory operation commands
and optimal settings.
The control sub-system interface connects the distributed control
system to other types of instruments such as the PLC (programmable
logic controller) or a composition analyzer to integrate plant operation.
The engineering interface permits communication between the distributed control system and the engineer. It permits system build-up
and software maintenance in the distributed control system.
The main system names for some representative products available at present for this type of distributed control system are listed in
Table 5.2. Figure 5.7 shows a typical configuration with Yokogawa
Electrics Corporation's total process control system.
CENTUM is a total distributed control system comprised of a
Table 5. 2 Main distributed control systems
Manufacturers
System name
Yokogawa Electric Corp.
CENTUM (40 loops type or 8 loops type)

YEWPACK (8 loops type)
YEWSERIES 80 (1 loop type)
SPECTRUM
MICREX-P
Unitrol EX Series
MACTUS
NETWORK-90
TOSDIC
TDCS-3000
The Foxboro Company

Fuji Electric Co. , Ltd.
Hitachi, Ltd.
Mitsubishi Electric Corp.
Bailey Japan Co., Ltd.
Toshiba Corp.
Yamatake Honeywell Co., Ltd.
5.2 Distributed Control System
257
.....
.;;.
~::1
::1
.....
~
;!!
~
"';;:-
..
-~
<D
~CENTUM '~=: :-,r. .=: =r:.,.=, ~.
Ma nagemem computer
YEWCOM
~ ~ ~
Sub -comro l systems e.g. PLC
Fig. 5 7 Tota l process control system of v okogawa Electric Corp .
<D
.~:~
n~
-~:~ e ~:-9CFCS2 (AJ CFCD2 (D) CFBS2 !BJ CFMS2
Mul iple- processor control sta tion
@ Field monitoring station
unll
----
M~nllori ng
0-9UFMH
HL Bus
g uoPs
O perator
comso le
Field gate
way unit
lcrrw 1
Bu~---~,',,, ',~--------~H:!_!~~B~us~-~-L~---1~--=I:='p;;~;~
Wide-a rea- network
e 3.:f-9UFCH
Basic
.
control umt
YE WPACK
HF
GP- !B
Basic contro l statio n

Duplex co mrol statio n
YE WSERI ESSOJ
~--
PHONE
RS- 232C
powerful CRT operator station, a control station which performs with

various process instrumentation, and a gateway unit. There are four
types of basic control stations: a 40-loop basic control station, a duplex
control station which duplicates the basic control station, an 8-loop
N: 1 backup-type highly distributed control system, and a field monitoring station which monitors processes.
There are two types of gateway units: a communication gateway
unit for the supervisory computer and a field gateway unit for a controller sub-system. CENTUM has a wide range of uses in large scale
continuous processes as well as batch processes.
YEWPACK is a distributed control system on a relatively small
scale. It has 8 loops per basic control unit, and is ideal for the control
of batch processes. The YEWSERIES 80 is a 1-loop board instrumentation system. It is suited to small to large scale instrumentation of
continuous processes. These systems can perform monitoring and
operation by communicating with the CENTUM operator station. Figure 5.8 shows a general view of several CENTUM operator stations
and a control station.
Fig. 5 8 External view of CENTUM
5.2.2 Configuration of the distributed control system and its functions

(a) Configuration of the control station
As shown in Fig. 5.9, the control station consists mainly of a station control nest (SCN) and 1/0 nests. The SCN has a microprocessor
and performs process control, communication control, and duplex control. The 1/0 nests contain interface cards for input and output. Signal
conditioner nests and terminal boards can be installed in the rear of
the cabinet, as necessary. Signal conditioner cards provide signal isolation and standardize signals from the process to 1 to 5V DC. Output
conditioners provide isolation for 4 to 20 rnA DC. As shown in Fig.
5.10, analog signals such as 4 to 20 rnA DC outputs.
As shown in Fig. 5.10, analog signals such as 4 to 20 rnA DC from
259
Roof fan unit
est fan unit

1/ 0
ests (Front)
Signal conditioner or
terminater boards
(Rear)
Power distribution
unit
Fig. 5 . 9 Configuration of control station
the field transmitters and thermocouple inputs are linearized and/or

normalized to 1 to 5 V DC by signal conditioner cards. The inputs are
cabled into a group of 8 or 16 points to the 1/0 card. These signals are
converted to digital values by the 1/0 cards, transmitted to the processor via the Sl bus to become input signals for control algorithms,
alarm monitoring, etc. The output of the control computations is returned to the 1/0 cards via the Sl bus.
The digital control output is converted to 4 to 20 rnA DC analog
output signals and sent to valves and other such devices via signal conditioners.
The MAC 2 card effectively combines 8 point analog inputs and 8
analog 4 to 20 rnA DC control outputs. However, failure of the MAC 2
card means failure of the 8 control outputs, creating a problem in process operation. A duplex configuration (two cards) is usually implemented to avoid this. The SI bus is an internal station bus which connects the SCN and the 1/0 nests. Because control of all input and out260
HF Bus
Station control nest
(SCN)
)..,:::reo::
N
2c
0-
u -t:
il
~"
Q.
0..
0..
"
Cl
1/ 0 est
(CNEST - DI)
I~
Signal cond itioner nest for

: mu lt i-point control (CNM)
4 to20mA
TC
RTD etc.
D)
~~ ~~ ~ ~
!
I
t it
{I
1/ 0 Nest
(CNEST- DI)
CIP
CIP
I ~~
3 3
4to20
4 to20mA DC
Signal .
conditioner
nest for
1/ 0 signal
(CNC)
It
Sl Bus
I
o.l--l
Sl Bu
o.2~
'
q=l
='U niversal terminal board (U2 B..l
Fig. 5 10 Signal fl ow diagram or control tation
put signals becomes impossible if the SI bus fails, reliability is increased by making it duplex. Card types NC 3 and NCU 4 are interface
cards for sending and receiving signals between 1/0 cards and the SI
bus, and between the SI bus and the processor. The duplex control section selects a standby processor if the processor or HF bus interface
card fails.
Contact input signals connected to terminal boards are cabled to
the 1/0 card in 16- to 32-point groups. These signals are combined into
261
digital value at the I/0 card and then sent to the processor via the SI
bus. They typically become inputs to sequence control tables. The action output of the sequence control tables is again transferred to the
I/0 card via the SI bus. The output signals are cabled to terminal
boards for connection to field circuits, in order to perform motor
start-stop and lamp on-off.
Table 5.3 shows the different types of signal conditioner cards and
Table 5.4 shows the different types of I/0 cards. The signal conditioner cards isolate the system from noise from the field. On-line maintenance is possible because failed signal conditioner cards and I/0
cards can be replaced without affecting other cards in the same nest.
The processor uses a 16-bit microprocessor and memory consists of
both ROM and RAM. The ROM contains programs such as for the
DDC, and the RAM holds system configuration data, set and measured
values, etc. The contents of RAM not lost as a result of power failure
because of a battery backup system. The functions of the control station start automatically during normal power-up. Also, this processor
is connected to a stand-by processor via a duplex control section. The
database in the memory of the stand-by processor is continually
up-dated to equal the database in the control processor.
Table 5. 3 I/0 card list for signal conditioner nest
For input
signal
Card type
Card name
CM1
CT5
CR5
CS1
CH1
CH2
CH5
mV input card
Thermocouple input card
Resistance temperature detector input card
Potentiometer input card
Input isolation card (1 to 5 V input)
Input isolation card (0. 2 to 1 V input)
Input isolation card (1 to 5 V input with square
root extraction)
2-wire transmitter input card
2-wire transmitter input card (with square root
extraction)
Pulse train input card
CAl
CA5
CPl
For output
signal
ceo
CAD
CHO
CXl
Control output isolator card

Output isolator card (4 to 20 rnA output)
Output isolator card (1 to 5 V output)
I/0 through card
(b) Process data highway
The data highway is a local network in which data is transmitted

bit-serially along a single coaxial cable. The following features are included in the system, according to typical industrial requirements for
262
CJ1
"'....
"'~
"'
"'>:>..
<;.:
"'
~
....
i\:,
MAC2
VM1
VM2
PM1
ST2
ST3
ST4
PB5
MX2
MX3
MX4
MX5
LCS
LCU
LDI
Model
Multipoint analog control I/0 card

Multipoint analog input card
Multipoint analog I/0 card
Multipoint pulse train input card
Multipoint status I/0 card
Multipoint status input card
Multipoint status output card
Multipoint push-button input card
m V input multiplexer card
m V input multiplexer card
High level input multiplexer card
High level input multiplexer card
Loop communications card
Loop communications card
Loop display interface card
Name
S-loop control I/0 (1 to 5 V input; 8 points, 4 to 20 mA output ; 8 points)

1 to 5 V DC input; 16 points
1 to 5 V DC input; 8 points, 1 to 5 V DC output: 8 points
pulse input; 16 points
Input (contact or voltage) ; 16 points, output (Tr contact) ; 16 points
Input(contact or voltage); 32 points
Output (Tr contact) ; 32 points
Contact or voltage status change signal; 16 points
mV, thermocouple non-isolated inputs: 32 points
mV, thermocouple isolated inputs: 16 points
Voltage non-isolated inputs: 32 points
V oitage isolated inputs; 16 points
Interface for YEWSERIES 80; 8 units
Interface for loop display unit; 4 units
Interface for MAC 2; 8 units
Functions and specifications
Table 5. 4 I/0 card list
control use.
(1) Transmission distance: The transmission distance is several kilometers, and can be adjusted to suit the scope of the plant.
(2) High reliability and safety: This is the only centralized part of the
distributed control system. In addition to using duplex redundancy
for the transmission path and communication control interface, a
system is required to prevent the propagation of the effects of errors among connected stations.
(3) Real time: In order to display data dynamically on the operator
console, event-driven data such as alarm and fixed-period scan
data are transmitted along the data highway. For example, at each
station, 4 to 20 bytes of data are transmitted in less than 20 ms, as
frequently as several hundred times per second.
(4) Resistance to noise: Transmission lines are often installed in areas
of plants having high levels of electrical and RF noise. Error control for countering this noise and with error detection and recovery techniques is required.
(5) Expandability: The system can be easily reconfigured when expanding the plant.
(6) High availability and maintainability: Breakdowns are easily spotted and repairs are quick. The failed equipment can be removed
and repaired while other lines carry on normally.
(7) n:n transmission: Each station is able to enjoy mutually effective
communication with other stations. There are no master-slave relationships.
(c) High system reliability
The technology for high reliability is based on three ideas: 1) To
be difficult to damage (fault avoidance), 2) Not to be unduly influenced
by failure (fault tolerance), and 3) To be able to be repaired without
stopping the system (on-line maintenance). Important parts which are
shared by the system (parts which greatly influence the entire system
when they fail) are all of duplex configuration to increase fault tolerance. System redundancy and maintainability are directly related. The
more reliable a system is, the more the simpler the maintenance requirements are.
Errors detected by self-diagnosis of system components appear as
maintenance information on the operator station CRT or on the printer. This makes quick diagnosis and accurate repair possible. Table 5.5
gives examples of CENTUM system redundancies.
With this highly reliable design, for example on a control station
(CFCD 2), an availability factor of over 99.99% can be expected. This
has been proven from field data of the CFCD 2. Figure 5.11 shows the
calculation of the availability factor of the SCN section. The processor
264
Table 5. 5 CENTUM system redundancies

System components
Applied technology
COPS
Duplexed processors
HF bus, coupler
Dual-redundant HF buses
CFCD2
CFBS2
SCN (processor and memory unit) Duplexed processors

Duplexed internal buses and conInternal bus, N C 3 card
trol cards
Duplexed cards
CNEST Power supply card
Duplexed analog I/0 cards
MAC2
CIU
Internal bus, NC 1 card
CFCU 2 Power supply card

Processor and memory
card
MAC2
Duplexing control
Back up by BCU
Duplexed internal buses and control cards
Duplexed cards
n : 1 back up by BCU
Duplexed I/0 cards
Standby side
Control side processor
( a ) Reliability model of duplex SCN configuration
--------------~~~------( b ) Reliability model of single SCN configuration
Availability of duplex type

A= AscNCM !1- (1-Ar) (1- ArAoxs) l
Failure rate
[fit]
Availability
SCN common section
60
The availability of duplex type is Processor

3 digits better than that of single Duplex control section
type (0.9998).
24 300
A,
3 200
Aoxs
= 0.999 999 459
AscNCM
0.999 999 5
0.999 81
0.999 974
Fig. 5 ~ 11 Availability of duplex type SCN
section includes the processor, memory, HF bus interface card, and SI

bus interface card. The SCN common section has several logic circuits
which make up the control enable logic of the duplexing control section. The duplexing control section can detect failure through self-diagnosis. To enable on-line maintenance, control right can be fixed with a
5.2
Distributed Control System
265
Supervisory computer
~
!!
-"
1::. Q.
t
c:
.!2
~
.:""
a
0
E
"C
c:
"'c:
.2
&_E
1
"
Serial primer
Summary printing
Alarm sammary
printing
Histor ical mes age
priming
Logging
.!2
;;
c:
-~
""
omrol station function
.:
9
a0
E
c:
o-~
c:
0
~
~
c0
u
.,._
"C
1
"ii
~
(D Operator station ,Display / Seuing

alarm , monitoring, functions
Builder maintenance functions
@ Communication functions
@ Annunciator function
Fig. 5 12
266
Computational functions
Feedback control function
Sequence control functions
Analog input / output
Digital input / output
Distributed contro l y tem function configuration
Chap.5
selector switch. It is considered that the pass of a standby section includes the duplex control section. This makes the duplex system even
more effective.
5.2.3 Feedback control
Figure 5.12 shows the functional configuration of the distributed
control system. Feedback control function and sequence control function processing is done at the intelligence section of the control station.
(a) Feedback control function
The feedback control function consists of a combination of loops,
as shown in Fig. 5.13. Examples are the single loop and the cascade
loop. The single loop consists of a software instrument (a function unit
derived from a control algorithm and function parameter, set beforehand) which processes one input and one output. The cascade loop consists of two instruments combined in a cascade. The instruments are indicated by circles in the figure. Each instrument is assigned a tag number such as FIC 100 (meaning flow indicating controller No.100) for
each loop. Tags are used for loop identification. Table 5.6 is a list of instrument types.
( a ) Single loop
Fig. 5 13
( b ) Cascade control loop

Loop configuration
Figure 5.14 is a functional block diagram of a digital control or

algorithm. The input signal conversion process linearizes a non-linear
input signal of a flowmeter and other such device, and then puts it
through a digital filter. A first order lag filter is applied as a digital filter. The compensation/totalization process performs flow temperature
and pressure compensation, etc. In addition, flow rate can be totalized.
The alarm check function detects input signal errors and alerts
the operator by sending a warning which is displayed on the monitor
of the operator station, and may be output to the printer. Alarm status
can also used as a condition for testing shifts in process for sequence
control action . For example, it is can be used to test if a temperature
or pressure has reached a certain value.
There are several types of alarm checks. The input open check
267
Table 5. 6 List of instrument types
Instrument type
Instrument function
Indicators
Input indicator
Controllers
PID controller
PID controller with dead band
En-type PID control
En-type PID controller with dead band
PID controller with batch switch
Two-position ON -OFF controller
Three-position ON -OFF controller
Pulse duration ON-OFF controller
PD controller with manual reset
Blending PI controller
Manual loaders
Manual loader
Manual loader with input indicator
Auto/manual station with output transfer switch
Ratio set units
Ratio set unit
Signal selectors
Auto selector
Singal selector
Selector switches
Three-pole three-position selector switch

Data set switch
Computing units
First order lag unit

First order lead unit
Union filter unit (ramp unit)
Lead lag unit
Dead time unit
Dead time compensation unit
Moving average unit
Cumulative average unit
Line-segment function unit
Special line-segment function unit
Calculating unit
Data set units
Data set unit
Program set units
Program set unit
Batch set units
Batch set unit
Batch data set units
Batch data set unit
Data acquisition units
Data acquisition unit
Station data link
Station data link
268
Chap.5
1 to 5V Dc
,...-
-E
~
;:;
---=
-~
""~c r-
0
0
---.:
- "" - ""~
0
4 to 20mA DC ......
;:;
.5
-~
0
0
c
c
Input signal
conversion
Square root
Non-linear
transfer
function
rPulse train
input
Alarm check
Compensation Input open check
High/Low alarm check
Totalization
Velocity alarm check
~rithmetic
operations
Totalization
Output open check

Output high/low limit
Output velocity limit
Pulse width output
ON-OFF output
Output signal handling
PID Control
etc.
~
ettm~
Indication Control
f----c
0
g""~~
g Caseade
....!
set
Fig. 5 14 Digital (Control unit configuration)?
tests for an open input signal. There is the high/low alarm check for input of measured values, and the deviation alarm check which checks
for deviations from the set values. The velocity alarm check tests
whether the change rates of an input exceed a limit value or not.
For indication and control functions, there are the PID control
algorithm and the PID control algorithm with a dead band. The output
signal handling function checks for errors in the output wiring (output
open check), can apply a variable output limit to the controller output,
and finally outputs the manipulated variable to the valve. Output limit
is a function to perform the process operation safely. It consists of an
output velocity limiter and output/high/low limiter. The output velocity limiter limits the incremental change in output during an output processing cycle so that changes in output are smooth. In the output/high/low limiter, if the output reaches the high or low limit value,
the output value is limited to the high or low limit setting. In addition
to the 4 to 20 rnA analog output, there is a pulse width output operate
motor-operated valves. Also, an on-off output which can be used for
controlling electrical heaters, etc.
The loop connection function corresponds to the wiring of analog
instruments. It specifies the instrument signal input and output destinations, and the setpoint connections of the cascade loop. Also, the
loop connection function transmits the loop status at the lower loop of
the cascade loop and clarifies the cascade loop open/close status for the
upper loop. In this way, operability is enhanced.
(b) PID control algorithm
The PID control algorithm of the analog controller is expressed in
the following formula.
269
100( e+r.
1 edt+Todt
de)
M=-p
(5.1)
Here, M is the output value, P is the proportional band (%), e is the

deviation (desired value - measured value), T, is the integral time,
and Tn is the derivative time.
The PID control algorithm for digital control is expressed by the
following difference equation.
(5.2)
Here, Mn is the output value, en is the control deviation ( =mn -sn), Sn
is the setpoint value, mn is the measured value, Aen= en- en-1> K is the
proportional gain ( = 100/P), T, is the integral time, Tn is the derivative time, and r is the sampling period. The variable n expresses the
number of sampling times.
Equation (5.2) changes to the following.
(5.3)
Here, L1Mn = Mn- Mn-1 and L1 2en = Llen- Llen-1 Usually, Equation (5.2)
refers to the positional and Equation (5.3) refers to the velocity or
incremental change. Incremental output L1Mn has to be accumulated and
converted to positional output Mn. However, when compared with Eqs.
(5.2) and (5.3) allows bumpless switching between manual and automatic
operation by putting the manual operation value to the initial value at
that point.
Also, with the positional type, even if the output value is saturated at 0% or 100%, L:en is calculated. Even if the sign of control deviation en is changed, the output recovery is delayed by L:en which accumulated during the preceding interval, and the tendency to overshoot
arises. This is called reset windup. With the incremental type, even if
the output gets caught in the output limit, the output begins to change
rapidly along with the change in en, because the accumulated value is
not in the output value itself. Reset windup and over shooting can be
prevented. This integration term is only in the output value itself, and
the inherent property, in which various changes in the control status
appear only in the changed portion of the output value. It is favorable
for advance control, which uses a variety of changes the control parameters and control modes.
When constant-value control is performed, .Jsn=O, and since
L1en=L1mn-L1Sn in the incremental type Equation (5.3), this Equation (5.3)
changes as follows.
270
L1Mn=K(L1mn+ ;. en+
T; L1 mn)
(5.4)
Even if the setpoint value is changed in a step-wise fashion, that

change does not drive the value excessively from the proportional and
derivative terms. Because only the integral action responds to a setpoint step change, Equation (5.4) has the advantage of not over-driving
the process. However, because there is a delay tendency of response to
a setpoint value change at the secondary in a cascade loop, Equation
(5.3) is often better where Sn changes moment to moment. The use of
Eqs. (5.3) and (5.4) should be selected according to the situation. The
correct algorithm is automatically selected in the Yokogawa CENTUM
system, unless otherwise specified.
5.2.4 Sequential control
The sequential control function is used for sequential processes
such as process startup and shutdown, and batch process switchovers.
It is complementary with DDC and is particularly effective for automatic plant operation.
(a) Sequence description technique
There are many techniques for describing sequence operation including Boolean logic and relay ladder diagrams, however, in the
design stage, the following techniques are often used (Fig. 5.15).
(1) Logic circuit
Logic circuits are based on the symbols AND and
OR. They are commonly used for interlocking processes. The symbols
In~
Input ~ Output
( I ) Logic circuit
( 4 ) State transition diagram

Rule I: Rule21 Rule3.
Conditions
Operations
( 2 ) Flow chart
( 5 ) Decision table
ProcessEquipment I
--Equipment 2
~ipment3
( 3 ) Time chart
Fig. 5 15 Sequence description technique
271
used in Fig. 5.15 are prescribed by ]IS and U.S. MIL 806B (Military
Specifications and Standards). For use in processes, there is the ISA
(Instrument Society of America) standard.
(2) Flowchart
Flowcharts are used in computer programming and
are excellent for describing operations which follow a fixed sequential
procedure. For example, flowcharts are often used to describe the startup operations of processes.
(3) Time chart
Time charts are easy to read because they define the
device operations according to time flow. Usually, time charts are used
to describe the time-relevant operations of batch processes in which
products are finished while moving, in order, between many small machines.
(4) State transition diagram
Not related to time, state transition diagrams are effective for expressing what conditions several processes
require for transition, or in what way a command, such as manual, automatic, stop and re-open, operates in the flow of a sequence operation.
(5) Decision table
In the decision table, conditions and operations are
lined up one by one to describe, in chart form, which operation should
be performed under a combination of conditions. The decision table is
suited for describing not only parallel operations and interlocking, but
also sequences in general.
(b) Example of sequence control junction
The following is an explanation of the CENTUM decision table,
shown in Fig. 5.16, as an example of a sequence control function.
The condition signal is written into each vertical row of column B
of the sequence table, and the condition signal combination, as the
operation condition, is written into each vertical row of column C. To
operate, enter Y (Yes) for the condition signals which as a condition
must be activated, N (No) for signals which must not be activated, and
leave blank for a signal which has no relation to operation. Write the
action signal into section D. Use Y ("on") or N ("off") in section E, for
the actions to take when the above conditions are satisfied, and leave a
blank for those which will not be operated (have no relation). The columns of sections C and E correspond one-to-one. Rows F and G are
used when doing stepped sequences at row A. Write either the number
of the next step to proceed to in row F when the conditions are satisfied, or enter a step in row G when the conditions are not satisfied.
Leave blank when the conditions have no relation.
Condition signals can use only contact input/outputs, internal
switches, operation and stop statuses of timers and counters. The
switch position of the switch unit, which switches the path of the signals, and alarm status of the feedback control function can be de272
Chap.S
tep o. description
ondition signal description
o ndition rule emry
Action ignal description
Action rule entry
~ ~"CdfD.-l
0 00
-
t.:P N I
I
I
- 1-- ..
~-- ~
;:;;
r--
-~ .2B
Rule
: - ~-- .:::
-g
: -~8
:-~"I
-r----
r-- .. ~--~
;;;
r--
- 1-- c
1-- .9
1--ti
_ !-- <(
INIJ
J'>lfN
(Yf~TifLJ.)
I
I
-J
..J
Decision table
Manual loader unit
4 to 20 mA
Manua l loader unit
feedback comrol function
Fig. 5 16
Example of CE TUM decision table
scribed. For action signals, in addition to contact output and internal

switch setting and resetting, and timer and counter starting and stopping, change of loop status of the feedback control function (for example MAN to AUTO) and change of the switch position no. on the
switch unit are possible. In this way, by combining the sequential control and feedback control effectively, process startup and shutdown
can be automated, and shifting of process in batch process control is
273
tv
....
..,;;,::!!!
.;;
~
;:
~
.....
;;,:
....
:!!!
~
"';:-
(;-.
Additive
Material
H,O
I
I
Fig. 5 17
frlQD~t~~~p~essure
N,
I
Tl02
Air 1
--
0
1
'
Start push-button
l [tPB
I
.2:
>"'
"
g_
00
complete
~
a"""
Dl03
Vw
Output [%]
100
Steam
Cooling water
-~
[>I(]
~102
62iQD
Polymerization Polymerization!
[>I(]
in progress
complete
Depressurization Vs
~)
IX102
,-----
1
~
2
3
XIOI
IJ-;;ckette-;;;~~;--.e-
-\.__)
Polymerization process heating sequence
I
I
I
I
I~:."~a!._t:_m_!'~~u~~ ___ ~1 ________
Internal temperature
set value
I; I~
ir:t:r----~~~~:_:_ _~:_ _ :::\

~
~t
-~
Fig. 5 18 Time chart of polymerization process heating sequence
simplified.
A descriptive example of a sequence table application is shown for
the polymerization process heating sequence in Fig. 5.17. Figure 5.18
is a time chart for the process.
(1) After completion of the charging process, push the start push button to begin the polymerization process. The polymerization in
progress lamp lights and the depressurization complete lamp goes
out. Turn the switch (X102) to the manual station device (D103)
side. The output of the D103, via the manual station with switch
H104, causes the cooling water or steam valve to move. At plant
side, the steam valve (Vs) opens completely and the jacket heats
up quickly.
(2) When the measured value (PV) of an internal temperature controller (T101) reaches non-alarm status (NR) (PV>T1), turn switch
(X102) to the manual station (D102) side as a previous step for
fixed value control of the jacket temperature. Also, start timer
(U700). At plant side, the steam valve (Vs) closes completely and a
cooling water valve (Vw) opens a fixed amount.
(3) After time is up on the timer (U700), turn switch (X102) to the
jacket temperature controller (T102) side and turn a switch (X101)
to the manual station (D101) side. Put the jacket temperature controller (T102) to AUTO and cascade connect the jacket temperature controller (T102) and the manual station (D101).
(4) When the measured value of the internal temperature controller
275
(TlOl) reaches the high alarm setting (PH) (PV>T2), turn the
switch (XlOl) to the internal temperature controller (TlOl) side
and cascade connect the internal temperature controller (TlOl)
and the jacket temperature controller (T102). Put the internal temperature controller (T102) to AUTO and start a timer (U701).
(5) After time is up on the timer (U701), when the measured value of
the pressure indicator (PlOl) reaches the non-alarm status (NR)
(PV> P 1), the polymerization in progress lamp goes out, "polymerization complete" is printed and the polymerization complete lamp
lights up.
(6) When the measured value of the pressure indicator (PlOl) drops
below the low alarm setting (PL) (PV> P2), the polymerization
complete lamp goes out and the depressurization complete lamp
lights up.
Figure 5.19 shows this example in the form of a sequence table. Corresponding inputs and outputs are allotted as follows.
Status input:
Start push button PB:
N300
Status output:
Polymerization in progress:
N700
Polymerization complete lamp:
N701
Depressurization complete lamp:
N702
Internal switch:
Charging process end flag:
NlOO
Polymerization in progress flag:
NlOl
Timer, printer:
Internal timer:
U700
Internal timer:
U701
"Polymerization complete" print-out: U400
The "polymerization complete" print-out message is output to the operator console.
5.2.5 Man-machine interface

Plant operation by a distributed control system is as shown in Fig.
5.20. In the even of a buzzer sound which indicates an alarm, the operator gets information regarding the cause of the alarm at the operator
console. He judges whether to leave the situation as is or to deal with
the problem. When left as is, the system maintains the original monitoring status. To deal with a problem, he may drive various valves
from the operator console or take other actions, and return to the original monitoring status. Event-driven information cautions the operator
at the operator console, and it is used for notification of process
alarms and operator guide messages. Buzzer sounds are used. Plant status information is provided by the operator console in response to operator requests. The information is provided visually on the CRT or a
logging printer. Operator console operation refers to operation whereby the operator obtains desired status information from the operator
276
. ,u
Hll
II oo- 1 v soJ
101-
101-II
lt/ 100
II
/(..100
1).....
..
..
.
.
..
..
..
Tt#l , II~
" ,.,,,
Till/. HI
I~
WR
"- UTOO . HI
PoOI . LO
"1'--
V100 . /0P
"
I L V10I
lO...... V70/ lOP
II ..... "101
llll.....
K..
K::
&:..
~ ~....:
l -
Ill-
102
"'"~ - r
... _
116 -
)t. I
102. :1
~ 10;1 .
IIJ_ IT /OZ.AUT
IT.IOZ.CAS
lit- TIO . AUT
110- U700
Ill- v 701
Ill V 701.
I l l - U.fOO .
llC- 11 ?o> .
liS- H 101 .
H
I
H
H
116 -
01 10
5
6
7
8
9
'
~~~
I
II lll U IS I
""
Ill
y
y
rr
y
NYYYYY
.
..
..
..
.
..
111011 . II
'"'!I
.. ., .... '
lfl\t\U~\o
~Htl"ll .
"'
Ill"< Ill'
y
y
y
y
'fY
II y
N
N
"
'(
N
"1-
~
It/_
N!:XT
~~m~o9
l1n-l
CD
Charging process end flag

Stan push-button (Panel)
(!) Internal temperature <i:TI
Internal temperature <i:T2
Internal pressure :a;PI
(j) Internal pressure :a; P2
Timer, time-up
El.St
Polymerization temperature
rising sequence
TlU.N
~"UP
Timer, time-up
Timer reset state
@ Polymerization in progress flag
@ Jacket temp. controller auto
Internal temp. controller auto

@Timer
Polymerization in progress lamp @Timer
(ij) Tl02 Setting transfer (0101 side) Polymerization complete lamp
TI02 Setting transfer (TIOI side) @ Polymerization print-out
Polymerization print-out
Output transfer (0103 side)
@ Output transfer (0 102 side)
Depressurizing complete lamp
On polymerization flag
@ Output transfer (TlOI side)
Timer, reset state
Fig. 5 19 Entry example of polymerization temperature rising sequence
277
Distr ibuted control system
<D
tat ion
Moni toring, Recognition

Decision
Contro l station
Operation
@) Event driven information
Operator console operation
Plant operation
Status information
Operater console operation
Pla nt
Fig. 5 20
Plant operation by distributed control system
console. This does not affect plant operation. The less the operator console operation done the better. Plant operation refers to actions which
influence the plant such as changing setpoint values and operating
valves. Although a configuration which prevents incorrect operation is
necessary , it is also necessary to be able to perform an operation
quickly during an emergency. The operator console is designed considering the factors described above. The CENTUM operator console is
explained in the following paragraphs.
The operator console, as shown in Fig. 5.21, consists of the display
section and the keyboard section. The display section consists of the
system messages section, the main display panel, the sketch pad section, and the soft key labels section. The system messages section displays messages which notify the operator of information from the operator console. It is used mainly for notifying about process alarms and
operator guide messages, and for messages operation error messages.
On the main display panel, a function panel (explained below), which is
selected through key operation, is displayed.
The sketch pad section displays data entries done by the operator,
measured values, and the alarm status and the loop status of one designated loop. The soft key labels section displays the variable labels
which correspond to each function panel. They are used to call up
data, and allow the display of settings and other function panels, with
one-touch operation. The keyboard section consists of the tuning keyboard and the operation keyboard. The tuning keyboard is mainly used
278
Chap.5
olumn
r----- 80 Column
40
System (2 rows)
messages
/
(Large characters)
(Small characters) -----o-1
TA K L V L
TROL
84.1 2.28
U :43
F91% X
f'V
a uu"
'II~'
?t IZ
- sv
Main display panel

(25 rows)
PU)Q Jt
CHARGE
( CAL.A R~t STA11JS]

f"912 X
L911 X
~r
SCketch pad \...
(3 rows)
LO
~r
1..0
L-----~
~fliO 'lanel
ooJOoooo
OOOJJOOO Syo DeJOY
0000000 00000000
00000 aoornooo 8 88~ !

ooooo
oorn oo o~ 09~
~
ooooo
oooo
~~~
~
00000000 oooorn. . ~n ~key
ke~
EJ
Alarm ack nowledge

key
A. Panel select
DQt?key
oooooooo m~GJt;JbEJ~[;) C2c~rJ~eys
88888888
@JOOOO~~g
00000 000
r::Jr::JQQ(;::;)Cl
r.::1
ooogooo
~~Go EJt.~
Tag number keys Auxi liary panel keys
Fig. 5 21
Touch pad
onfigration example of operater con ole
for switching loop statuses and changing controller setpoints and output values. The soft keys are labeled by the screen. The speaker is
used to alert the operator by various tones of an alarm. Optionally, it
can provide voice communication. The operation keyboard consists of
function keys, panel select keys, auxiliary panel keys, a touch pad and
cursor keys, an alarm acknowledge key, tag number keys, data keys,
and system utility keys. The function keys, allow one-touch call up of
279
user-defined panels. They also call up the printing of logs. Each key
has an LED which flashes or lights upon the occurrence of specified
process alarms to urge operator action. The panel select keys call up
the operation monitoring panels. The auxiliary panel keys are used to
scroll pages in the function panels. The touch pad and cursor keys are
used to move the cursor for selecting items in the function panels. The
alarm acknowledge key is used to acknowledge the occurrence of an
alarm. The tag number keys are used to designate tag numbers. The
data keys are used to input various data. The system utility keys are
used to call up the system test, system maintenance, the system builder, and other engineering functions.
Function panels follow the format shown in Fig. 5.22. The operation monitoring panel is used for process operations and is hierarchical. The operator guide message panel outputs user pre-defined guide
messages, as shown in Fig. 5.23, according to instructions from the sequence control function. It is able to guide the operator during operation. The alarm summary panel, as shown in Fig. 5.24, displays process
alarms in order of occurrence. The display for an alarm disappears
when normality is restored, unless an alarm lock is specified. The display is designed to make the occurrence of an alarm readily apparent.
The overview panel, as shown in Fig. 5.25, displays in color on one
screen the alarm statuses of up to 800 instruments. It features the ability to recognize operations as patterns. The control panel, as shown in
Fig. 5.26, is made up of eight controller faceplates. In this example,
eight controller faceplates are shown. The bar displays in the middle indicate the measured value, the arrows on the right side indicate the setpoint value, and the arrows on the left side indicate the controller outOperator panels
Standard operator panels
::===;-;:==:=:;---;:.= =:;
Process
status
panel
Utility
panel
User defined Trend

operator
recorder o
_p~ _P.anels :
Fig. 5 22
280
File display /
: Update pane
Function pa nel configuration
Fig. 5 23 Operator guide message panel
Fig. 5 24 Alarm summary panel
! ~~~-~ ~'~~~w.
: 1..~.::.::~.:~...
:t~~~.
... ,J;;r
! :t::::::::
;;::::::: :~.==~~=
;:: F:::::::: ;;t::::: :~.::::~:
.~~==== ~t~~::::
;:::::
!~:1::::::
I
I I l l I I .,~II
I
l
l
I
.111111111 .t~L...........
.lllllllll-+:'111... I l l
:m:: :
ill
~~
,111111111
I,
IIIII
i
iiiiiL
.i
1~11
Il
ll lIl
hllllllll
!'11111111
'0:
111111
1; ~. 'li
(11111111
~:
;~::::::::
,_,. _. l.
.:
I: .
ttl""
:111111111
...........
'
\;
'L.r...
IIIII
o,IIIIIIIL.
\~
t'L.
!l
t;
.,
'1411111111
!\~r~~~;;: ::;~~r~,;== o;~m.:;: :;;,~;.r?:~;,:
,ti
j
Wi
:~
~I
:!
""~'~J~ ~..~: i;;~ ~~ . ~

II
't4.E
Fig. 5 25 Overview panel
281
Fig. 5 26 Control panel
Fig. 5 27 Tuning panel
put value. PV, SV and MV are the corresponding digital displays.

Setpoint and output values can be increased and decreased in a
gradual manner with the soft keys of the tuning keyboard, in accordance with AUTO and MAN status. Operation error from incorrect
entry of data values is prevented. L901XX and 7DC-D5, etc., are displays of tag numbers and tag comments. The frame surrounding
F12100 is called a tag marker. Tags signifying a test operation, etc.,
can be attached to the faceplate and operation can be prohibited when
specified. The tuning panel, shown in Fig. 5.27, displays all loop parameters. It is used for tuning loops, changing alarm settings, and other
control parameters. The parameters PH, PL, VL and DL are the alarm
settings (measurement high, measurement low, velocity alarm, and
deviation alarm, respectively). P, I and D are, proportional band, inte282
Chap.S
Fig. 5 28 Graphic panel
Fig. 5 29 Trend group panel
gral time, and derivative time. A trend display is provided for tuning.
The graphic panel is a display in which the user can follow and direct operation. It is the center of operation. An example is shown in
Fig. 5.28. Colors and graphics change in accordance with alarm
changes, and operation status can be made intuitive. The trend overview panel and trend group panel replace pen-and-ink recorders.
Trending on a CRT brings a high 'degree of freedom and ease of use in
the display of large amounts of record points, permits enlargements
and reductions, and operation is possible while watching process
changes over time. Also, data can now be easily stored on auxiliary media such as floppy diskettes. Use can also be made for trend data for
quality control. Figure 5.29 shows a trend group panel. The file display
change panel can be used to service simple data bases for logging printing data and printing format files, storing brand name data for brand
283
control, and storing recipe data files. The process status report panel
allows the operator to retrieve process statuses under some conditions.
The utility panel displays help messages for settings of each type of
system constant such as time settings, and for when incorrect operation occurs.
5.2.6 Communication with other systems

With the advance of computer communications technology, total
FA systems have become a reality. Distributed control systems, in
meeting the demand, have also acquired standard functions for communicating with other systems. As an example of this, the CENTUM system will be explained below.
For interfaces between computers used for control applications, as
shown in Fig. 5.30, the GP-IB or RS-232C are widely used. Beyond
these, standard protocols are used which use data with tag numbers
and industrial units which have ASCII character strings as their base.
These are unified as presentation interfaces so that communication
gateway units (CGWUs) can interface with supervisory computers in
exactly the same way that the operator station sends and receives character string information to the operator. Table 5. 7 shows the message
formats used in a controller. Figure 5.31 shows the information flow
for process data acquisition by a supervisory computer, YECOM.
(1) Configuration information of the control system, tag lists and
data conversion tables are downloaded from the operator console,
and the communication gateway unit begins to function.
(2) The message "message omitted here" is sent to the CGWU via
GP-IB from the user program of the YEWCOM.
(3) The message received at the CGWU is transferred to the command process routine via the GP-IB driver.
(4) The command process routine can call up the DDC process routine and sequence data process routine separately by message format or tag number. For this example, the DDC data process routine is called.
(5) The DDC data process routine changes the tag numbers to station
numbers and loop numbers (the tag order number), the data types
are changed into data numbers, and a request for data extraction
is put to the HF bus driver.
(6) The HF bus driver extracts controller data via the HF bus and
passes it to the request origin (DDC data process routine).
(7) The DDC data process routine converts the data format inside the
controller to character-format industrial-unit data from the type
of instruments and the data type.
(8) The command process routine arranges the message format and
284
Operation
Management
Management
computer
Operator
;:.
--------;~-_!~~.;_~~~'!"'..___---Operator
station
Tag list
TAGl\1_0.
SPAN
BIAS
UN I
COPSV
CGWU
CD
Tag list
AGNC.
SPAN
BIAS
UNJ
(RS- 232C )
GP - IB
-----------Communication
un it
7 gateway
I
MSGI
HF Bus
CFGW
Control
data - base r
Control
station
CD
Display charactor string

Display, Operation
ormalized data
@ Charator string MSG
<V
Interface conversion
ormalized data
Communications
Equivalent
Field gateway unit
) Control
.1
II
II
1data - base
------------Control subsystems
<V
i'-
( RS - 232C)
GP- IB
----------
PLC
Gas chromatographs
Tank guages
etc.
Fig. 5 30 CE TUM Communication gateway
passes it to the GP-IB driver.

(9) The GP-IB driver sends the message "message omitted here" to
the YEWCOM via the GP-IB.
(10) The GP-IB driver of the YEWCOM passes the received message
to the user program.
5.2.7 Engineering
The distributed control system uses computer technology to the
fullest, however, it greatly differs from computer systems in the way
of the program is written. Computer systems are programmed and
tested using assembly language or a general-use programming Ian5.2 Distributed Control System
185
~
....
.;;
g
....:ll:
;:"'
t;.,
n1; Data number (1 to 32)

nz ; Data number (1 to 8)
Tname; Tag No.
Dtype; Data type
Message from HF bus side
DDC/SEQ data set by loop No.
uDcduu Dcd CRLF

uDatau uData CRLF
uDcduData u
uDcduData CRLF
TPunz
DGuSTun1
DGuSTun1
DPuSTunz
Return
Send
Return
Send
CRLF
uFMumnou PAmes CRLF
MS
MS
Send
Return
uDatau uData CRLF
uDatau' uData CRLF
TPunz
Send
DPuST unz
uTname, DtypeuData u
uTname, DtypeuData CRLF
TGun1
Return
Return
uDatau uData CRLF
T-cun1
Message type
uTname, Dtypeuu
Tname, Dtype CRLF
Send
Data; Data
Dcd; Data identifier
ST; Station number
PArnes: Alarm message
DDC/SEQ data acquisition by loop No.
DDC data set by Tag No.
Function
DDC data a-cquisition by Tag-No.
Table 5. 7 Message type used in a controller
~
.._.
"';;:-
~
.....
~
.....
;::
c;;
.....
't
;:;.:
01
(FORTRAN)
User
program
'---
HP-IB
Driver!-
r-
1....--
~ ~-
oo:s
u o..e
~;;
==
-ooo
1-
- ..J... j..
Data type
$(conversion)
Address in
instrument list
0 to I Data
t (conveJNion)
Engineering data
Charncter string
code
Sequence data processing
Loop No.
Tag No.
$ (conveJNion)
Station No.
1-
.-
Instrument Data
configuratior conversion
information
list
table
DOC data processing
rf...
Tag list
HF Bus
driver
.J>
II
MProcess , perator
HF Bus
program
Fig. 5 31 Information flow in process data acquisition
II
I
s_equencel
hst
Control station
DOC
List
(0 to 1
Data)
DOC Data, Sequence data

HF
Bus
HF Bus
co~munication
pro ram
c:::Jt::lc:::Jc:::Jc:::JOc::J
~GWU mformat10nd' station
Down Io.ad of
"Control
station
information
The tables of CGWU are

subset of COPS tables
I
,..'ou\Ge a\\\te\\a\\ce
'tJ
'\:a'o\e
COPS C2 G'~<\J \ a
Information
LISTS
~o~O 0
for control
0
tt\
station
\'tolbta
configuration ....--- ~...-!)
Example
Transmission(YEWCOM-+CGWU)TG._.2._.T001,PV.... TOOl ,AS{QD(QJ )~
Reception
(YEWCOM<-CGWU)TG .... 2 .................... 150.9 .... NR ............................... (QD
...,
2P-IB~
Tag No.
Engineering
unit data
.Character
string code
-c:
-........
1...--
HP-IB
Driver
.--
YEW COM
CPU
.....
Auxiliary table
CGWU Communication gateway unit

CGWU Lists
[fNTUM
INSTRUMENT SPECIF ICATIONS

STATIONNO. r n
LoopNu m~r
T19 N1.1mb~r
I I I I I I
lnururnent Mod"
Conuol Proc.euing
Control PerKKj {Sec. )
Input Pt<>Uing
Gam m Dead Band
Seale High Limit
MV Tracking
State Low Limit
110 Compenutlon
EnginHring Unit
Ro~tio
Set Unil
Output Procen ing
Ro~tio
I I I I I
Scale
Oireet/R~er~
Ouu~Jt
Full Stroke
Alrm Output Point
Min . O u t put
Time PJopononlng 0n1011
Period
Period {Se-c.] 101 Velocity
Compen\alion Output Detf
Al1rm Oucpt~ t Point
Computing Moduli
Positioml/lncrememal
Program Set Uni1
I-
Al1rm Output Point
~LII'i'l
1~
l"'
7S W
Sp,n
Tv""
CMP-OOC
I 7 8 D JC A LC I I NL I
r.~:.ue::::J
~:.;:..
..
on Sklc
on - $adt~~
7PG
O.at.a <J.
Dt.ta <J.
UI - IU.
71'1 .
""
T1me
IJI
I
NL
~
,;;;
Vtlocitv limit
8 addash
V e!octtv AJ,Jfm +/-
Oevlilton Alarm
7SL
(rd. potn. ln 11!1)
Unked ..en
' c=:J c=:J ' c=:J [=:J

c=:Jc=:J l'~ Linlc!:dll
c:::::J [=:J
c:::::J
c=:J~
c=:J~
..
c=:J [=::::J ~
[=::::J [=::::J
c:=:J
[=::::J [=::::J
' [=::::J [=::::J
c=:J [=::::J
'"[=::::J [=::::J
'' [=::::J [=::::J
otc=J [=::::J
u c:::::J [=::::J
,, [=::::J [=::::J
'~
~
c:=:J
"'c:=:J
c:=:J
nr:ott) ErtC~ftutiftl Un.c Data
..Output Poi111 ., C'olumn un
""
II\
~llolhlo~S
YOKOGAWA
Fig. 5 32
288
7S L
..
nOt~
tfX'Cirff'd ror NL.
ENGME,;J
r<
Pulse Wid th Output
Alarm Proceuing
Discrim inative High/Low Limit
Alarm
;;
Acton
Ouuxn Tracking
Pulse R.ue
Totall ztf/ No. of Comi)Utational
Lin
OAT(
~v
8Y
,LANT N'AMIE CJOI!I NAJo!EI
WS 34B2X2-2JE
:Jr d Ed l iDOI Juf\.. 191!17

"on1..:11"J.e~ IT A )
Example of work sheet (Instrument specifications)
Chap.S
guage such as FORTRAN after a system engineer consults with the

user and writes the functional specifications.
In contrast to this, the distributed control system is completed
with the user or sales engineer filling in worksheets which includes
everything from system and hardware configuration to configurations
of feedback control functions and sequential control functions. This is
how basic specifications are decided. These worksheets are called a
"fill-in-the-form" (FIF) style worksheets. Their sole purpose is to permit communication of system configuration in the language of an instrumentation engineer. The contents of the worksheets are input into
the system using the builder/maintenance function (system creation
and alteration) which follows an interactive format, and the system is
created. The created system is then tested using the test functions
Fig. 5 33 Instrument specification maintenance panel
Fig. 5 34 Sequence table maintenance panel
5 .2 Distributed Control System
289
such as the status display function and the integrated retrieval function. Corrections are done using the builder/maintenance function. After the tests are completed, the distributed control system is finished.
Using the self-documentation function , complete documentation is
printed out for the system, to simplify maintenance. Figure 5.32 shows
the worksheet, Figure 5.33 shows the builder/maintenance function ;
Figure 5.34 shows the status display; and Figure 5.35 shows an example of self- documentation.
!N 5TRUt1U41 SPtC
CONT~Ul
TAG
"
[XAI1PLt
l!tl. ~ . d F. SUI
'IS J$211
PAMAMTER
F~UZXX
70COS
IN~T TYPE
70COS CONTROL ALG
CONT PERIO[I
INPUT CONV
GAP GAIN
SCALE HI
I UU. 0 MEA $ TRACI-: ?
SCALE lO
U. 0 1/0 CON PEN ~<
ENG UNIT I
OlTPUT SFEC.S
NliM~~~
R SC ALE HI
F"1..1L $ E ~ATE
SLIMMA T I ON
[/R ACT I ON - ~
CUT TRAC I; ?
VELOC LIMIT
Pll OUT SPEC$
ALARM SPEC
H/L GROUP
STROI ~ E
BACKLASH
ANN P(IINT
VEL +/
PERIOD
ANN POINT
r, p,~
MIN WITDH
ON/OFF PER I
COMPE NS OLIT
.,,.
CO MP OLIT PIV
I-
FROG SE T LIN I T
ANN POINT-
TIME SPAN
TYPE
SPEC IFY ALSO
CMP D OC
LOOP CONNECT
TUNING
PV
~~lllltltlltlt
I
I
I
I
X901XX
I
I
7SW-33
SV
MV
PH Su. O
PL 2!J.ll
Vl IOU.U
#IIIIIIII SWIO I <- -l

'
MH 1\JU.U
111 1 1111111111
1--> I SET
'
37.9
37.9
37.9
F~02XX
70COS
SUB
I
I
I
I
1->I IH IIIIIIIOUTI I
I
111111111 1 Ill
I
PV
I
I X992XX
I
7CMEX
I
I
111111111111111
11L
OV
0 . ()
O.u
P 20U.U
I
20
D
0
BS
O.u
CS I.UUU
DL IOU . O
Kl
I
SH
K2
OOC PA~AHtHM
FILTER NO
I
SACKUP
0
PIIT OFF
'"f.Afl PHS
SCAfl PHU
Fig. 5 35
290
Example of self-documema1ion
5.3 Production Line Control System 7 ) -to )

The implementation of mechanization and automation of production lines, where products are worked upon and produced, are very important factors concerning plant automation. Many difficulties lie, however, in the fact that both materials and information must be handled
in such systems. In this section, an explanation of the different types
and the overall position of the systems for controlling production lines,
and a summary of FA computers for production line control which
have come into wide use lately are explained.
5.3.1 Summary of production line control systems

FA production systems, as shown generally in Fig. 5.36 form a
hierarchical structure reaching from the production line level and spanning the entire company. Computer systems, which form the core, are
utilized at each level, and the type of computer system used depends
on what goes on at that level. At the company level, general purpose
computers are used, while at the factory and production department
levels, computers for production control are used. Many types of computers are used at the production line level depending on the type and
scope of the production line, and the type of control system. These
computers are generally referred to as "production line control sysLevel
Man agement a nd control items
~
Company level
Management information system
Factory level
Factory management Production

planning/ Order controi/ Costcontrol
Production
depanment level
Production depanment management

Proces scheduling/ Progress control
Product ion management

computer system
Production
line level
Fig. 5. 36 The hierarchical tructure of FA manufacturing sy tem
5.3
Production Line Control System
291
terns." A factory automation system at the production line level, which

uses a production line control system, will reflect the diversity of the
production line. For this reason, such systems are difficult to classify
into a single group. Also, proceeding with factory automation in sequence from the actual site, a system configuration becomes desirable
which starts from the factory floor, bottom up. This is the layer where
physical and mechanical information from the production line are converted to logical information which can be processed by a computer.
This layer, which is the basis of total factory automation, is one of the
most important elements of the system and is the key to its success.
5.3.2 Types of production line control systems
Production line control systems come in many types. The span extends from computers which would be better considered controllers
than computers, such as those which are embedded in every machine
in the plant and those which use a PC (Programmable Controller) for
main sequential control, to computers which classify more closely as
production control computers which center around data processing.
A summary of a typical production line control systems in actual
use as follows.
(a) Specialized controllers for FA equipment
Controllers for robots, NC machinery, automatic warehouse stacking cranes, and other such FA equipment are embedded in the body of
the equipment. Most of these controllers use 8- or 16-bit microprocessors, and both the hardware and software are specially designed to optimize the performance and function of each piece of equipment. The
program is generally stored in ROM. Although in some cases, the operating instruction set is fixed, normally instructions are received from a
programming device, paper tape, floppy disk, or from a supervisory
computer by communication. These operating instructions are stored
in RAM. Usually a battery back-up power supply is used to insure that
the memory contents are not lost in case of power failure.
(b) Programmable controllers
Programmable controllers (PC) are control computers that have in
recent years been replacing the formerly-used relay boards in sequential control. Programmable controllers are more easily changed and expanded than relay boards. They offer great flexibility, and their high
reliability facilitates maintenance.
An example of PC hardware configuration is shown in Fig. 5.37.
The sequential control procedure is input in advance from any kind of
terminal, and stored in the user subprogram memory. The control procedure is as follows. First, the input signal is read in from the 1/0
device and stored in the data memory (input processing). A microcom292
Progra mming
device
Personal computer
ii
C PU Ma in part
Micro
computer
System
program
1/ 0
in terface
Fig. 5 37
Co nfiguratio n of programmable controller
puter reads the user program, in order, and performs logical operations based on the status of input signals stored in the data memory.
The results are then stored in the data memory (operation processing).
When processing of the user program is finished, the operation result
stored in data memory is output by the input/output device (output processing). This one period: input, operation, output, is called "one
scan." This is executed cyclically as the sequence control program is
advanced. The scan period differs according to the type of PC and number of steps in the user program. Typically, a speed of 10 to 50 ms is
required, but for high- speed machine control, a speed of 1 to 10 ms is
required. Ladder diagrams, which resemble operational sequence diagrams for relay boards, are often used to express user programs (Fig.
5.38). Ladder diagrams are easy for technicians experienced with relay
boards to understand, but are not always the appropriate method for
expressing the overall progress of sequence control. Flowcharts, Boolean algebra, and decision table methods are also used.
In addition, with FA becoming more and more widespread recently, programmable controllers often incorporate high level features
such as analog control, data processing, and external communications
functions, but FA computers (described later) are usually used in applications requiring such functions, and in general programmable controllers are used as small, relatively inexpensive on-site controllers.
5.3
Production Line Control System
293
Fig. 5 38
Exa mple of ladder d iagra'm
(c) Computer system
Programmable controllers are widely used as inexpensive FA controllers, but as FA advances there is a demand for more general purpose data processing functions, various man-machine interfaces, communications with host systems, and management functions for all
kinds of production lines. Production line control systems that use computer systems allowing the construction of more general purpose systems than programmable controllers are increasingly being applied.
Computers based on existing minicomputer and personal computer systems are common. In addition to these, special purpose production line
control computers, i.e. FA computers are also available on the market
and are becoming widespread.
The features of these computers as production line control systems are outlined below.
(1) Minicomputer based systems
Minicomputers such as the YODIC
and YEWCOM from Yokogawa Electric Corporation have long been
used in the field of process control (PA). They have also been used in
the field of FA, mainly for production management. They are commonly used in production line control systems with a centralized type
system construction for controlling several processes. They are also
used as front end processors of large scale production management systems.
(2) Personal computer based systems
When 16-bit personal computers
294
came onto the market they became popular in the field of office automation, and their prices dropped quickly. This contributed to their
gradual appearance in the field of FA. Factors such as reliability, process input/output functions and expandability are of concern regarding
the use of personal computers for control on production lines, but they
are widely used for independent management in relatively small scale
systems.
(3) FA computers
These are specialized computers for application in
production line control systems and have the following features.
(1) Powerful real-time functions and high reliability.
(2) A wide range of 1/0 interfaces.
(3) Sequential control functions.
(4) Powerful communications functions for communicating with host
systems and FA equipment.
(5) Flexible system structure and easy system construction.
Following the release of the YEWMAC FA computer, by Yokogawa Electric Corporation in 1983, FA computers were released by
many other companies and this new computer market quickly expanded.
5.3.3 FA computer systems
The system structure an actual FA computer developed for production line control is explained below.
(a) System structure
For total control of a production line, it is important that in addition to the transfer of production command data and accumulation of
results data from each process in the vertical direction, tracking data
and quality data flow between each process in the horizontal direction.
The system structure of a total production line control system using personal computers is shown in Fig. 5.39. The personal computers
that control each process are connected to a host process management
<D
Per onal computer

Tracki ng information
Qua lity information
Production Iine
Fig. 5 39
5.3 Production Line Control System
295
computer, and normally, transmission of data between processes is performed via this management computer using some common communications protocol. In this type of system the performance of the communications link between the computers and ease of data transfer procedures are important points.
FA computers are designed with this kind of total production line
control system in mind, and in general have the distributed type structure shown in Fig. 5.40. The control units controlling the process, and
the management unit that controls the total system all have their own
processors, and are connected together via a high speed data bus and
function as one system. Each unit can directly and simply communicate at high speed, and users can easily access production data JIT
(Just In Time). This is necessary to construct high efficiency production control systems.
* Tracking
information
Quality information
Production li ne
Fig. 5 40
Production li ne control system by FA computer
(b) System example
To illustrate the general concept of FA computers, Figure 5.41

shows the system structure of an inspection process employing Yokogawa Electric Corporation YEWMAC FA computer. The YEWMAC
system consists of line computers which handle management functions
(MLCC; Manufacturing Line Control Computer), networked via the
ML bus to line controllers which handle the control functions (MLCU;
Manufacturing Line Control Unit), thus forming a hierarchical distributed system. The MLCU (equipped with a 16-bit microprocessor) and
the MLCC (with 16- or 32-bit microprocessor) run user programs written in distributed real time BASIC which has inhanced functions for
FA use. The MLCC and MLCU are shown in Figs. 5.42 and 5.43. In the
system shown in Fig. 5.41, an MLCU reads the product type of each
product via a bar code reader, and controls the various measuring devices under control of the inspection program corresponding to that
product type. Based on the results of the measurements, the DI/DO signals are directly controlled to control the product flow. The inspection
results are also displayed on terminals in the factory to inform operators. The MLCC simultaneously stores and manages the inspection pro296
Line computer (MLCC)

YEWMAC300 Super
YEWMAC300
32-bit Microprocessor
16- bit Microprocessor
2-MB Memory
I- MB/ 2- MB Memory
20- MB / 40- MB Disk
20- MB Disk
ML Bus. ML2 Bus
ML Bu
~===================~
ML2 Bus
Supervisory computer
YEW COM
FA Computer system
Line computer
YEWMAC300
.~
Line controler~
A Type
Inspection guide
Fig. 5 41
Line controller(M L U)
16-bit M icroprocessor
5 12- KB Memory
Sequenoer runction
Combined type or A
a nd S types
Example of F A computer system (YEWMA )
Fig. 5 42 External view of YEWMAC 300 MLCC
297
Fig. 5. 43 External view of YEWMAC 300 MLCU(C- M Type)
gram master file and accumulates the quality management data, inspection results table data, and operation results data. It displays the data
on the management terminal, outputs to the printer, and transmits
data to the host computer as necessary. From this example it can be
seen that the important points for an FA computer are, in addition to
the status of the system structure, the production line interface (process interface), man/machine interface, communication interface, and
the software to control them.
5.3.4 FA computer system hardware
(a) Process interface
A production line control system must connect to the many types
of sensors and actuators used in production lines, and control the FA
equipment. Table 5.8 shows the types of 1/0 cards and the sensors and
actuators they are used with.
Features of the 1/0 cards for FA compared to those for PA.
(1) A diverse variety of 1/0 signals
Unlike PA, in the case of FA, the
1/0 signals to the system are not assumed to be uniform. It is necessary to install an assortment of 1/0 cards to handle the variety of signals from the sensors and to the actuators.
(2) Multi-function card
For flexibility it is desirable to have a
multi- function card that includes a variety on 1/0 specifications on a
single card.
For example, a multi range analog 1/0 card, a communications
card compatible with a variety of communications standards such as
RS- 232 C, RS-422, and current loop, and a card for use with the factory terminal to handle the CRT output and keyboard input.
198
Chap.S
Table 5. 8 Example of I/0 card, sensor and actuator

Type of I/0 cards
Example of combined sensors and actuators
Analog input
Potentiometer, Differential transformer, Straingauge, Thermister, Thermocouple
Analog output
Electric motor, Recorder
Digital input
Interrupt signal
Pulse input
Microswitch, Limitswitch, Electrophoto switch,

Push-button switch, Rotary encoder
Digital output
Pulse output
Lamp, Solenoid valves, Electromagnetic contactor,

Servo-motor, Stepping-motor
(3) High speed

The demand for high response speed is greater for
discreet processes than for continuous processes, particularly for mechanical system control. Scan periods are normally 10 to 100 ms, and
the required control periods may be 1 to 10 ms. Accordingly, to comply
with the high speed, I/0 cards for the production line control system
should be capable of high speed. Many I/0 cards have their own processors for high-speed I/0 control. Process interface cards are listed in
Table 5.9. General purpose I/0 cards, communications cards, and
man/machine interface cards are also listed in the same table.
(b) Communication interface
The FA computer system communications interface is classified
into two types; host and field communication.
(1) Host communication
FA computers receive production planning
and manufacturing data from the host computer, and send back the actual production results and the production line operating status. This
communication with the host computer is performed at every batch,
such as a day or a shift. It is important that the production line control
system be able to operate as autonomously as possible. Figure 5.44
shows an example of a control system for an NC machine. Here, the
FA computer receives the required NC data for the day or shift and
performs the processing locally in its own system while the machine is
operating. This improves host system performance, and enables more
responsive downloading of the NC data for the variety of products to
be manufactured.
The communication procedure with the host is not fixed but the
asynchronous procedure, BSC (Binary Synchronous Communication),
and SDLC (Synchronous Data Link Control) are often used. For connection to general purpose host computers, terminal emulators, and RJE
299
"'~
{;_
~
;:
::!
("')
-"'
"'~
(,
.....
("')
.....
(Note)
AC 1
DC 1
SCI
PU 1
SRO
TRO
RL 0
3568-05
3568-04
3568- 02
3568- 01
0: Incorporate microprocessor
Sequence control use

l/ 0 cards
Inspection use 1/0 cards
PM 1
ST2
ST3
ST4
PB6
GB1
TV3
LCS
RS2
RS3
VMI
VM2
AN 3
AN4
HAD!
General use 1/0 cards
16 points, AC 100 V ON -OFF input

16 points, DC 24 V ON -OFF input
16 points, contact input
2 points, up-and -down pulse counter
16 points, solid state relay switch
16 points, transistor output
16 points, relay output
10 points 20/ 50/ 200 mV input

Thermocouple, Type R/ K/ 1IT IE
10 points. 1/ 10/ 50 / 1 to 5 V input
Universal counter
DC- 1 MHz, pu lse totalization
Frequency/ period/ measurement of frequency ratio
2 points, 1 to 10 V programmable voltage output.
with sweep function
16 points, 1 to 5 V input
8 points, I to 5 V input/ output each
16 points, 10 V input
8 points, 10 V input/output each
16 points. 1. 25/ 2. 5/5.0 V input
Resolution 14 bits
50 K samples/ second
16 points, pulse train
16 points, digital input/ output each
32 points, digita l input
32 points, digital output
16 points, push - button input
GP-18 interface
Color TV / keyboard interface
YEWSER IES 80 intereface
Multiboard (4-ch) RS-233 C
RS-232 C, RS- 422, 2 ports/ current loop
Function
Table 5. 9 Input/ output card used with MLCU

Type
Use
0
0
0
0
0
Microprocessor
Productio n management computer
FA Computer
--- -- - - --- ---- -,

Line computer
I
1
ML Bus
ltltD
C Machine Bar-code
reader
Fig. 5 44
Data flow in N C machine control system

(YEWMAC' example)
(Remote Job Entry) are required. With these, communication is possible without any particular need for awareness of the communications
procedure.
Machines with intelligence such as robots,
(2) Field communication
Programmable controllers, inserters, and NC machines are widely used
on production lines. From now, connection to FA equipment will in
general be via RS- 232 C, RS- 422, current loop, and GP- IB. There will
be a variety of communication procedures such as the asynchronous
procedure, or basic procedure, depending on the FA equipment, and
the production line control system will have to be able to flexibly respond to these.
(c) Man/machine interface
At present because of technical and economic considerations, there
are few examples of totally automated factories with no human operators, and in fact systems that harmonize the high level decision making ability of human operators with machines are more practicaL Accordingly, the man/machine interface that the operators and technicians in the factory use to input and receive necessary data is a very
important factor in the system.
It is not a requirement that the display on the
(1) Factory terminal
operator's terminal have great resolution for displaying large amounts
of information. Rather, it is desirable that the information be displayed
in large size characters that can be seen and confirmed from a distance. CRT and LED displays are often used, and recently demand for
Chinese Character displays and color displays is increasing because
301
they are easy to read and understand for Japanese operaters. Keyboards with numerical keys and just a few alphabetic keys are commonly used. They are installed throughout factories, so low cost is a
strong requirement. Figure 5.45 shows an example of this type of terminal.
Fig. 5 45
Example of FA terminal
There are a wide variety of man-machine interface devices used to

meet the needs of factories. These include bar code readers for reading
the manufacturing numbers off parts and manufactured goods, sign
board readers for single product flow type production systems, and
voice recognition systems for use in sorting systems.
(2) Management terminal
The management terminal informs production line supervisors and operators of a variety of management data relating to the production line and detailed information about the status
of the production line. Due to the increasing complexity of the data to
be displayed, and for human engineering reasons, displays are generally CRTs that can display in Japanese, with color and full graphics capability. Display resolution of about 640 x 400 dots is required and
14- inch screens are usual, although for process control use, 20-inch
screens are also used. Normally, these terminals are also used for program development, and general purpose keyboards (e.g., ]IS specification) are generally used. Flat keyboards covered with silicon rubber
are also used because of their closed nature.
(d) Sequential control functions
Sequential control is often used in production line control, and fast
cycles of 1 to 10 ms are often required. To avoid placing to much load
on the CPU, it is desirable to have a dedicated sequence control processor to handle these high speed sequence control tasks. Unity and coordination between the processing done on the CPU and the processing
done on the sequence control processor is important. It is also necessary for the program running on the CPU to be able to quickly access
the sequence control processor data, and upload and download the se302
quence programs. The YEWMAC can be equipped with MLCU dedicated sequence control processor cards, and access from BASIC on the
CPU is easy.
5.3.5 FA computer software
In recent years the percentage of software development cost in
computer system development has risen sharply. FA computer systems
are not just hardware, and the software must also be carefully considered.
The software consists of the system programs supplied as standard with the system, and the programs created for the applications of
individual users. The program development environment and operating
environment are heavily influenced by the content of the system programs. The important operating systems, programming languages, and
software packages are discussed below.
(a) Operating systems
The operating system (OS) is the nucleus program that controls
the operations of the software system.
The features required in operating systems for FA computers are
as follows.
(1) Real time functions: Fast response to external events, priority
interrupt processing, efficient allocation of resources, and high
speed task scheduling.
(2) Multitasking functions: Parallel processing of many tasks such as;
data acquisition and screen display, and external communications.
(3) Efficient 1/0 management system: Efficient and consistent management of process I/0, the man/machine interface, and communication interface are required.
The operating systems generally used in FA computers can be classified as shown in Fig. 5.46.
Operating systems from computer manufacturers, which are written exclusively for their machines, have the advantage that they are designed to optimally utilize the special features of these machines, and
FA ~p~
"""'""
OST.OS from--~"'~'~
Marketed OS[ Real time OS ex. RMX
RMS
Fig. 5 46
General purpose OSex. CP/M

MS-DOS
UNIX
Classification of FA computer oriented OS
303
are efficient in terms of functions and performance. On the other

hand, commercially available operating systems, which have started to
come into use because of the popularity of personal computers, have
the advantage that the various types of software marketed for the
same OS can be used relatively easily. Marketed operating systems can
be classified into real-time operating systems and general purpose operating systems. However, as stated previously, FA computers require
real-time operating systems. There are more software tools and computer languages for software development for the general purpose
operating systems, so some systems are switched between the two
types of operating system depending on the application.
(b) Languages
There are a wide variety of production lines that FA computers
control, and compared to the case of process control, they are difficult
to typify. For this reason, procedure-oriented languages such as assembler, BASIC, FORTRAN and PASCAL are normally used rather than
problem oriented languages. It is necessary to consider the following
points when adopting these languages.
(1) Ease of learning for factory technicians who are not computer experts.
(2) Efficient program development and simple debugging.
(3) Execution speed of the developed programs.
(4) Memory space required by the developed programs. (High coding
efficiency)
(5) Ease of transferring the programs to other systems.
The features of procedural orientated languages for FA.
(1) Assembler
Programs written in assembler have the advantage
that they can be fast and compact. Disadvantages are that assembler is
difficult to learn, program development is inefficient, and that programs are difficult to transfer to other systems.
(2) FORTRAN, PASCAL and C
FORTRAN is widely used for scientific calculations in process control. PASCAL features rich data structures and is suitable for writing structured programs. The C language
used to write UNIX* closely resembles assembler language and features a large number of operators, and the ability to perform the bit
control necessary in control and measurement programming.
Program development using these compiled languages is fast compared to assembler, but more memory space is required and execution
speed is slower. They are more oriented to specialists than BASIC
which is described below.
(3) BASIC
This language came into wide use as personal computers became popular and it is easy for beginners to learn. With the nor-
Unix was developed by AT&T and is a registered trade name.
304
mally used interpreter type there is no need for compiling or linking,

so debugging is relatively easy. Slow program execution and poor real
time functions can be raised as disadvantages, but in recent years,
powerful BASICs that solve these problems have appeared.
One example of such a BASIC is the distributed type real time
BASIC used in the Yokogawa Electric Corporation FA computer YEWMAC. (Refer to Subsec. 5.2.3 (b) regarding YEWMAC).
The BASIC that supports YEWMAC has, in addition to the functions of normal personal computer BASIC, the following functions for
production line control.
(1) Using time slicing, multiple programs can be run on the MLCC.
(2) Distributed processing functions for sending and receiving data
between the MLCC and MLCU.
(3) Real time processing for timed and unpredictable events. (Refer
to Table 5.10)
Table 5.10 List of real time statements
Real time statements
Function
WAIT
ON TIME
OFF TIME
ON TIMER
OFF TIMER
SIGNAL
ON EVENT
OFF EVENT
DISABLE
ENABLE
CHAIN
START
SETTIME
SETDAY
RESERVE
RELEASE
Waiting for execution of program

Branching by time
Cancellation of branching by time
Start timer
Stop timer
Notice of event occurrence
Waiting for event occurrence
Cancellation of waiting for event occurrence
Inhibits interrupt
Cancellation of interrupt inhibit
Chainning of programs
Start program
Set time
Set data
Declaration for use of common resources
Cancellation of declaration for use of common resources
(4) Dedicated statements for I/0 cards and sequencer cards (32
statements).
(5) High level graphics such as closed-region filling color, and Window/View (35 statements).
(6) MLCU debugging can be performed from the MLCC.
(c) Software packages
Using software packages is one way to improve the efficiency of
software development. In the world of office automation (OA), a large
number of packages have already been developed, and this so called distributed software is widely used. The FA field is difficult to typify and
305
Man-machine package
Interactive panel generati on package

Trend d i play package
Operator co nso le input package
ommunica tion package
3270 Emula to r
YE W OM Terminal emulator
Operation package
Statistics operat ion package

umerical analysis package
Matrix operation li brary

System package
Data base
Report generator
IBM fo rmat F D
conversion package
Undefi ned character utility package

Volume/ File operatio n utility package
Sequence contro l bu ilder/ maintenance/ monitor package
Application package
Graph package
NC Dat a down - load library
Fig. 5 47
Software package (In ca e of Y E WMA C)
so there are not many software packages available compared to the

number of OA packages available, but from now their use will gradually increase. Examples of package software that support FA computers
are shown in Fig. 5.47.
5.4
Computer System Equipment for Production Management 11 l -I 3 l
5.4 .1 Computer components and configuration

Production management computers must generally meet the following requirements.
(1) Have a real-time operating system capable of fast response to the
control system
(2) Be capable of processing and storing large volumes of data
(3) Have a redundant architecture for high reliability under roundthe- clock operating conditions
(4) Be capable of highly reliable long-term storage and frequent updating of production information
(5) Possess a human interface that allows easy, highly responsive
operation
(6) Have software tools such as data base management packages, pro306
Chap.S
cess data acquisition packages, and human interface packages, that

allow easy system construction
(7) Be capable of forming a computer network or factory network
for system integration
(a) System configuration
A production management computer system generally consists of
a high-performance central processing unit together with large-capacity auxiliary memory devices, operator console and other peripherals,
process input/output equipment, a distributed control system interface,
a factory LAN interface, an interface for communication with other
computers, and other such components. The YEWCOM system produced by Yokogawa Electric Corporation shown in Fig. 5.48 is an example of such a production management computer system. The detailed
hardware configuration of this system is shown in Fig. 5.49.
Fig. 5 48 Production management computer YEWCOM
(b) Central processing unit (CPU)

A processor that can be classified as a minicomputer or super minicomputer is used for the central processing unit. Formerly, 16-bit
word length processors were mostly used. Recently however, 32-bit processors have come to be used as well, due to increases in the volume
of information processing. Also, recent production management computers adopt the latest high-performance technologies from high-end
general purpose computer systems, in addition to the real-time capabilities that have been developed on the process control computer. Previously processing performance of about 0.5 MIPS* was common, but recently performance has been raised to the 3 MIPS to 4 MIPS level. In
addition, main memory capacity has expanded from a few hundred kilobytes (K) to several megabytes (M).
The following are techniques that are contributing to the increase
in CPU performance.
MIPS (million instructions per second) is one measure of processing power. It

represents the number of instructions the processor can execute in one second.
5.4 Computer System Equipment for Production Management
307
I HP-18 intc:rface card

HP IB ex tender card (Optical / coaxia l cable)
2 Asynchronou seria l interface: card
(RS -232 C . 20mA cum mt loop)
Asynchronous seria l interface muluplcxer card (R - 423,1o,)
OS/ 100 1 interface card (Distributed network)
Serial in terface card
Floa ting point procc:ssor ca rd (optional)
r- - -
__ ,
I
'
1 I YEWMAC) I
'
I
L. -- -- -- ~
Produclion line control syslem
F ig . 5 49
Configuration of production management computer YEWCOM
(1) Application of very-large-scale integration (VLSI)

(2) Use of cache memory
(3) Introduction of pipeline processing
(4) Use of floating-point co-processors
(5) Use of firmware for the operating system (OS)
(6) Instruction sets corresponding to high-level programming language
(c) Memory devices
There is an inverse relationship between the operating speed of

memory elements on one hand, and memory capacity and cost per unit
308
Chap.S
High Severa l tens

speed
to 200n. L-....,....-...........1
300 to 500ns
Several tens ms
evera l hundreds K
everal ten
(j)
ache- memory
Ma in memory unit
Auxiliary
torage device
M bytes Large ca paci ty

( Inexpensive)
Fig. 5 50 Hierarchical configuration of storage device
of memory capacity on the other. Thus a hierarchical configuration of

different types of memory devices such as shown in Fig. 5.50 is generally used.
Semiconductor memory devices have become reliable and inexpensive, so dynamic random access memory (DRAM) units are used for
the main memory device. Semiconductor memory requires little electric power, so battery back-up is a possible measure against the volatility of the DRAM. Also, using error correcting code (ECC), one-bit errors can be corrected and two-bit errors can be detected. Thus high reliability is achieved. Cache memory is an ultra- high speed buffer memory placed between the main memory device and the CPU in order to increase the effective speed of the main memory. It uses static random
access memory (SRAM) or memory elements using the same circuit
technology as the CPU, and has an access time that is equal to the cycle time of the CPU. The copies from the most recently accessed memory remain in the cache memory, and speed is increased because memory access is localized within small memory address range during a program executing period.
Fixed- head magnetic disk (hard disk) equipment is mainly used for
auxiliary memory devices. The hard disks in general use have capacities ranging from several tens of megabytes to several hundreds of
megabytes, although progress is being made towards smaller disks of
larger capacity. Magnetic tape devices and tape streamer devices are
used for data back-up and long-term storage.
(d) Input/output control and intelligent input/output equipment
The most common method of input/output (1/0) control uses a common 1/0 bus interface. The date transfer rate of the 1/0 bus ranges
from 2 Mbyte/s to 10 Mbyte/s, and the trend is towards higher speeds.
One method for control of data transfer is program control, in which
the transmission of each word of data is controlled by a program. Another method is direct memory access (DMA), in which data is exchanged by main memory and 1/0 devices directly and at high speed,
5.4
Computer System Equipment for Production Management
309
without depending on the CPU. In large computer systems, there is

also a dedicated channel interface for high-speed 1/0. As a means of increasing the 1/0 throughput, intelligence is being added by equipping
each 1/0 interface with an 1/0 processor to perform 1/0 control independently of the CPU.
Figure 5.51 outlines the intelligent functions of the 1/0 interface
and I/0 devices of the YEWCOM system Since each I/0 device handles
the complicated processing required for its own operation, the application programs are considerably simplified. The I/0 processor controls
DMA transfer, interrupt processing, 1/0 wait, and so on. It also controls protocol among I/0 devices, and processes command strings. Intelligent I/0 interfaces include the byte multiplexer for asynchronous
communication, HP-IB interface, the binary synchronous communication interface (BSC), and the high-level data link control (HDLC).
Along with bringing intelligence to the 1/0 interface, the various
types of I/0 devices themselves are being given intelligence by the
addition of microprocessors. This allows 1/0 processing to be done in
parallel with CPU operation. Also, because CPU overhead for 1/0 response no longer depends on the number of I/0 devices, high overall
system throughput and speed of response are achieved.
(1) Intelligent process input/output device (SPACE 1/0)
The SPACE
I/0 intelligent process I/0 device shares the burden of 1/0 processing
by downloading a program from YEWCOM. The program is written in
MML (measuring and monitoring language), a specialized measurement
programming language that is similar to BASIC. Figure 5.52 shows the
apportioning of processing for the case of analog input. SPACE I/0
gathers process inputs at fixed intervals. YEWCOM receives data that
has been filtered, linearized, and converted to industrial units in
SPACE 1/0, thus it can devote itself to the processing of the data.
(2) Communication gateway unit (CGWU)
The CGWU is an intelligent
YEW COM
]
se
Status,
Industrial
un it data
~
'\r--'
(j) Industri al un it data, Status
Linearizing and industri al uni t conver ion
Fig. 5 52 Functionnal configuration of intelligent process 1/ 0 device
310
l.o,o
........
::!
"'
.....
~
i:f
1:.
::!
::!
~
~
<:;
ci
'"'o
::!
"'
.....
i:f
~
;::
i:f
"'~
~~
iJ)
Process control and alarm

monitoring
Sequence control
Process operator console
Logging such as daily repon
System generation by tilling in
Builder and maintenance
e
e
e
e
Japanese la nguage terminal function

Color graphic fu nction
Background panel storage and display
function
Panel control function
Interactive panel generation function
Intelligent terminal
e
e
e
e
e
e
(This unit simplifies communication with a distributed

control system)
Character stream
Process data name and data type
Industrial unit data conversion
Character code data
Communication control between multistations
Da ta buffering
Communication geteway
interface card
Japanese character display

Offiine sentence edition
Alphabet-ideogra m conversion
Panel copy and print-out
Filing function by nexible d isk
Japanese character terminal
e
e
e
e
e
Fig. 5 . 51 In telligent functions of 1/ 0 interfaces and 1/ 0 devices
Distributed control system (CENTUM)
e
e
e
e
e
e
MML Language
Data acquisition
command
Linearize command
Data buffering
Operational function
Multitasking
Self-diagnosis
SPACE 1/0
e
e
CS-80
Transaction oriented command
Device control function a nd
logical device function
eJnput-output format check and
preparation
Self-diagnosis
Automatic error logging
Off- line backup
Magnetic disk device
Other computer
(Intelligent in interface card)
Communication protocol such as BSC,
HDLC etc.
communication control device that connects the various kinds of stations on the HF bus of the CENTUM distributed control system to
YEWCOM through the HP-IB. CENTUM data can be accessed by tag
number and data type in character form, and the industrial unit data
can be collected and set. As shown in Fig. 5.53, conversion between tag
number and station loop number, and industrial unit conversion is
handled by CGWU intelligence. Because the communication procedure
is performed with completely logical data, communication can be done
with no awareness at all of the arrangement or format of the data on
the OENTUM side.
cc wu
YE WCOM
CG WU
YEWCOM
LL L t CD
TG
32 TOO tA.
Pv ... .
(D
CG WU
TG 32
CD
YEWCOM
H F Bus commun ica tion
I
@
- z . _ CENTUM
tSo_g .. .
Data type
@ Tag number
CV
T ag No.
Data quantity
(D Command
Communication
Station
o . Loop
@ Industrial unit binary data
@ Industrial unit binary data

@ 0 to 1 data
o.
Data type
Fig. 5 53
Address in instrument list

Ind ustria l un it character string data
Da ta co nversio n in communica tio n geteway un it
(3) Intelligent terminal and operator console

In Fig. 5.54, the Y1720A
is shown as an example of an intelligent terminal. An example of its
display screen is also presented. In this terminal, as shown in Fig.
5.55, the required number of fixed background screens, such as process flow diagrams, are prepared in advance using the terminal's interactive screen construction program and stored in the terminal file.
When the desired screen is called up from the keyboard, first the background screen is displayed. In parallel with this, the process data is received from YEWCOM, and the foreground screen is superimposed on
the background screen, greatly improving response. The definition of
the information required for the foreground display can also be handled by the interactive screen construction program. Moreover, terminal intelligence can realize the features of a dedicated operator console. Most of the functions required for good operability, such as input
and output data editing and checking, and display of system messages
can be performed by calling up the appropriate screen by means of a
menu. Figure 5.56 compares the response time for an ordinary terminal and an intelligent terminal. For intelligent terminals, the response
312
Fig. 5 54
Intelligent term ina I

Yl720A/YI725A
Fig. 5 55
5.4
Intelligent terminal (Y 1720 A)
Panel ID
/ Data ID
Displ ay data/ Entry data
Functional configuration of intelligent terminal (Y 1720 A)

and operator console (Y 1725 A)
(Shaded portion shows background processing)
313
201----------Intelligent terminal
----- Terminal
Terminal number
Fig. 5 56 Response time of intelligent termrnal
time increases little as more terminals are added. Display within two
to three seconds in any system configuration is a great advantage of intelligent terminals.
(e) Duplex system configuration
The three types of redundant system configurations shown in Fig.
5.57 are representative of high-reliability systems designed to improve
reliability.
Coincident
check
(a ) Dual configuration
( b ) Duplex configuration
Fig. 5 57
( c ) Loadsharing configuration
Duplex system configuration
In this configuration, two CPUs with syn(1) Dual configuration

chronized clocks perform the same processing. When an inconsistency
occurs, each of the CPUs performs self-diagnosis. The normal processor that passes the diagnosis continues processing and the other is cut
off. When one of the CPUs goes down, complete continuity of processing is attained. However, this method requires special hardware
design, and thus has less flexibility than other methods.
In the duplex system, one CPU does the
(2) Duplex configuration
processing, and the other CPU stands by. If the active processor goes
down, the processing is automatically switched to the stand-by unit.
Although with this method processing stops for just an instant, the
availability factor can be greatly improved. This method is implemented by adding a simple duplex control device to two CPUs, and can thus
314
Chap.S
be widely used.
(3) Load-sharing configuration
In this method, the processing load
is apportioned between two CPUs. If one of the processors goes down,
its share of the processing is taken up by the normal processor. During
normal operation, the system has about the same performance as two
CPUs. However, during failure, engineering for degeneration of function becomes important.
Figure 5.58 shows the duplex system configuration used in YEWCOM. This configuration can apply either the duplex method or the
load-sharing method. The duplex control section monitors both CPUs.
When it detects failure in the executing CPU, it switches the I/0
device group over to the stand-by processor. The dual-port disk holds
common data. The dual-port disk controller also provides a means of
communication between CPUs, and has a buffer and a resource management table for that purpose. Thus when there is no other means of
Duplex control unit (DCU)
Serial communication
line (Data, status)
System console
System disk
Serial communication
line (Data, Status)
System disk
Fig. 5 58 Duplex system configuration in YEMCOM
5.4
315
communication between processors, the controller can perform effective common resource management.
In production management computer systems, high data reliability
is required to cope with long-term storage and frequent updating.
Thus data back-up capability is very important. This system uses a
mirrored configuration for the dual-port disk. In this mirrored disk,
data is written to two disk devices at the same time. If either disk
fails, then processing continues using the other disk, and there is no
loss of data. The system also supports recovery by making the data in
the two disks equivalent after repairing the failed disk without disturbing system operation.
5.4.2 Software for production management computer systems
(a) Software configuration
The software configuration is shown in Fig. 5.59. The core of the
configuration is an operating system that supports both real-time processing and large-volume batch processing. A development support
package consisting of a language processor, interactive screen editor
and various utility programs for program development provides an effi-
Fig. 5 59 Software configuration of production management computer
316
cient environment for programming and system generation. The basic

package contains the standard tools used for any type of application.
For process management computers, the real-time data base and network communication packages are particularly important. The production management package is a set of packages for facilitating development of application software for production management systems. It
consists of a human interface package, a process data acquisition package, and so on.
(b) Operating system
The operating system (OS) is a program that manages and controls
all of the various system resources, including the hardware components, software, and data for the purpose of efficient operation. The
following are the basic requirements for a real-time operating system.
(1) High throughput
(2) Fast response
(3) Ease of use
(4) Reliability, availability, and serviceability (RAS)
(5) Integrity
(6) Security
(1) Task management and multiprogramming
One of the most important functions of real-time operating system (OS) is the parallel execution of multiple programs. This is called multiprogramming. In
real-time OS, the object of each executing program is called a task, or
a process, and system resources such as CPUs, I/0 devices, programs,
and data, are allocated using the task as the basic unit of management.
(i) Interrupt control
Interrupt control refers to a system design in
which the CPU is notified when certain events occur, causing the currently executing task to be interrupted to allow execution of a higher
priority task. The following are some of the different types of interrupt.
(1) 1/0 interrupt: notifies of the completion of an input/output operation.
(2) Timer interrupt: causes an interrupt at a specific time or a specific time interval.
(3) Power stop/restore and error interrupt: notifies when power stops
or is restored, and when a CPU error occurs.
(4) Program interrupt: Notifies of an error in the machine language
code.
(5) SVC interrupt: generated when a user program executes a SVC
(supervisor call) instruction call to the OS.
(ii) Resource management
In multiprogramming, multiple tasks are
performed at the same time, and each task requires system resources.
Simultaneous requests from various tasks for resources are received,
317
and the tasks are arranged into queues to await the availability of each
required resource. When a task finishes using a resource, the resource
is made available to another task from the queue. The task dispatcher
allocates use of the CPU, the most important resource in the system.
This task switching is called dispatching or scheduling. Tasks are generally managed by transitions among the following four states.
(1) Run : being executed (use of the CPU secured).
(2) Ready: can be executed (resources other than the CPU are secured), but is waiting while a task of equal or higher priority is using the CPU.
(3) Wait: unable to execute while waiting for completion of an 1/0
operation , a signal from another task, etc.
(4) Idle : no activity.
Figure 5.60 illustrates the concepts of priority processing for foreground tasks which require real-time operation, and parallel processing for 1/0 wait. Also, when several users are operating different terminals on the same CPU , such as for program development, and editors, compilers and so on are executed as background tasks, it is necessary to allocate CPU time fairly among them. Figure 5.61 illustrates
the concept of time slice processing for fair execution of multiple background tasks.
~[@o0
(Priority A > B> B>C> D)
B A
Time
( a ) Pri ority processi ng
Execution
Wa it
0 CPC program execution

;
End of a program
Execution
Program s. jj~~C~~~
Program A 41-.,.'----I.LLIO~_;,.,...-L
1/0Wai t : Waits for the end of

inputoutput processing
Exection
(b) Parallel processing
Fig . 5 60
Priority processing and parallel processing
(2) Memory management

When multiple tasks are executed in parallel, it is desirable to have many programs in main memory. However,
there are physical limits to main memory size. The OS, subroutines
used in common by tasks, and common data are maintained in main
memory. Highly urgent tasks and tasks that are very frequently used
are also placed in residence in main memory. Other programs are
stored on disk, and loaded into main memory as needed for execution
318
Chap .5
Time slici ng interval

Task sta rt
A,B,C a nd D are
backgro und ta ks
with th e a rne pr iority
is end of a program
Fig. 5 61 T ime slice processing
as non-resident tasks. In such cases, it is normally necessary to divide

the program up into a number of segments according to its size, and
load the segments into main memory as the processing progresses.
This is called overlay processing. Virtual storage allows the programmer to regard the main memory and auxiliary memory as one large
main memory unit, without concern at all for the size of the program.
(3) File management
Files are collections of logically related records. File management is a function that provides unified access to
logical files for data input/output and storage without concern for the
physical configuration of memory or 1/0 devices or differences in physical interfaces. File access is accomplished by means of the file name.
Using the file name, the file managemant finds the physical memory
address and file attributes in a look-up table (directory) in memory,
File rna nagement

uti lity p rogram
Disk type file
"E
' - - -- - '
Ma in
storage file
"'
"'"
E
User
progra ms
Fig. 5 62
Non d isk type file
Fi le ma nagement con figuration
319
and performs the I/0 operation on the appropriate memory device.

Figure 5.62 shows the configuration of the YEWCOM file management system. All magnetic disk space is managed as files. The user
program specifies the disk space it needs, and can use this space exclusively or in common with other programs. Also, when high-speed file
access is needed, blocks of main memory can be used as a virtual disk
device, and accessed in the same manner as a disk. File management involves sequential devices such as magnetic tape and printers as well as
random access devices like hard magnetic disks and floppy disks. Each
device is given a logical device number (LU). By specifying a LU, terminal input and output can also be accessed as a file. A file management package is a collection of subroutines that can be called by a user
program. It consists of a group of programs for creating, opening, closing, and deleting files, and for disk file accessing. File management
utility programs allow file creation, deletion, storage, copy, packing,
and so on from the operator command level.
(c) Development support software
Although software development is also done on large, general-purpose computers, in recent systems it has become possible to do program development on most target systems in parallel with real-time
processing. FORTRAN is the most widely used programming language. FORTRAN 77, an extended function structured language that
can be used for real-time system, has become the ISO standard. In
1982 it appeared in revised form as JIS FORTRAN (X 3001), although
it remained substantially the same, and is widely used. To address the
serious problem of how to increase productivity in software production
and insure high quality, software development support has been studied in great earnest. Various kinds of interactive utilities such as interactive screen editors, on-line relocatable loaders, on-line system generators, and symbolic debuggers, are being perfected, and utility programs such as static analysis tools, coverage measurement tools, and
operability analysis tools are coming into use.
Computer workstations have become cheaper and more powerful.
With advances in development and standardization of data communication systems centered around local area networks (LAN), the trend is
toward a software development environment in which workstations are
integrated with host development systems. The UNIX operating system and the C programming language are widely used, and are becoming the industry standard.
(d) Basic software for production management systems
Figure 5.63 shows an example of a YEWCOM production management system configuration and data flow model. The IMAGE/1000 J
data base management system is the core of the system. The basic
320
Acctuisition of
PV,AS,
LS,S M .--------,
CGWU
SV Settin!\
CRT display
and maintenance
t:f>
.-
----------"-'
'1U:-:-s-er___,pr'-o-gr-am-- l
CENTUM
Printing
L..----~_ ___:U:__., User sub-routine
~:
User program
Mean value,
Cost calculation
statistical data
PV : Process variable
Statistical data
calculations
:
AS : Alarm status
calculation
Calculation of SV :
LS : Loop status
Load distribution
etc.
L___________ _
SM : Summation
Required
SV : Setpoint value
estimation
etc.
maintenance
Fig. 5 63 Configuration of production management system
package comprises the human interface (QUERY and REPOCALC), the

process data acquisition packages (PDHS), and a package for communication with various kinds of computer move over many types of software packages and user's programs are added to met the target system
requirements.
(1) Data-base management system
If, for example, a part number appears in multiple files, such as the product file, the quality control file,
the price file, the production performance file, and so on, then part
number data will accumulate job-by-job. This results in redundancy
and is a source of possible data conflict. A data base is a collection of
data structured by logical relations. The purpose of the data base is to
allow the data to be shared.
The IMAGE/1000 J data-base management system (DBMS) for the
YEWCOM system is a network DBMS that can handle Japanese text.
This system offers data independence, data consistency, data security,
and high operability. Data independence means the accessing program
is not affected even if the data base is altered. Data consistency is
brought about by the deletion of redundant data. Data security results
from many kinds of read/write protect functions. Also, a high-level
data manipulation language allows a high degree of operability. These
capabilities provide the operator with large amounts of production data
in a useful form. On the other hand, if data addition and deletion functions are not required after the system has been constructed, such as
for a process data acquisition system, then the data structure can be
simplified and a simple DBMS designed for high performance can be
used. The YEWCOM data cell is an example of this. As shown in Fig.
5.64, the data cells have a two-dimensional structure, and it's records
consist of several items which differ in word length and data format.
321
Header
Data table
]ilbl l!;t B
hem I
1_
-hem
hem 2
h em m
1 1_ _
Fig. 5 64
Data construction of data cell
A group of these records comprises a data cell. For example, if item 1

is allocated to tag name, item 2 to PV, item 3 to SV and so on, this collection of process data would make up one data cell. The data cell access statement can be written using the item name in correspondence
with the FORTRAN READ/WRITE statement. Ease of programming
and high performance at execution time are achieved by using a cell
preprocessor, which determines the physical offset of the item name
from the beginning of the record. There are many other utilities, such
as the date cell generator, which constructs data cells from a data cell
definition table, and the cell scope, which allows interactive inspection
and changing of cell contents.
(2) Human interface package
In production management systems,
the screen construction support package is particularly useful because
programs for CRT display screens, and forms and tables have to be
written very frequently. The report calculation package (REPOCALC)
is a typical example of software for display of listings on the CRT or
printing of forms on the printer. It has the following functions.
(1) Creation of screen specifications for any kind of listing and form
printout using a high- level dedicated language
(2) Automatic continual updating of process data display by means of
a recursive screen display function
(3) Hourly, shift time and daily print outs at fixed times are also possible
As shown in Fig. 5.65, an output screen can be constructed easily
by creating a FORMS file using the editor. The input/output data definition section can be used to easily define search conditions, I/0 editing, data checks, simple calculations, and so on.
(3) Process data acquisition package
The process data acquisition
package (PDHS) is a program which collects data from the distributed
control systems CENTUM and YEWPACK, the production line control
system YEWMAC, and the intelligent process input and output device
322
FORMS File
010
1
2
Oota base name

User level
, _
, 125 ,:10
' 'lEWIS
, 1
' .,
0 .8- Access mode

LPLU
c:.t;
Oc
.;:
c:
Max. Character numMr: eo
Max. line number

Refr<Sh lime (see)
o::::
"'O
: .SO
' S
:
8 Hard copy unit no.

I(""
lg
1011
Paru ma..ster
Pans No.
'I
Ol(tQ)(
I I
Quantity
:(
Pan name :(
Rate per hour:(
Unit price
:1
tiT
find dttlnHiUon :
endf
tln4 Plfli .IU'INt
L-Find value
t1 , PARTNO :
12 PARTNA :
U AKCODE:
14 QUNT
IS RATE
16 PRICE
O!Jid
: ( ZlZl.99:
CRT P~a~n~e~
~ --------------------Parts masler
Parts o. : c::::::::::::J Pans name : R 65E250A
AKCODE : T
Quanlity : 0500
Rate per hour :
Unite price : 240.00
Fig. 5 65 Output panel generation by REPOCALC
SPACE 1/0. PDHS stores the data in a centralized data base. The data
acquisition specifications are defined by filling in blank items in a
work sheet. An example of such a work sheet for defining the process
acquisition method and so on is shown in Fig. 5.66. The work sheet is
used to specify the acquisition interval, what process data to obtain,
and where in the PDHS data base to store the data. In this example,
323
SCAN & USER'S PROGRAM DEFINITION PDHS FORM SUI

TAG NAME
FlO!
-----------
Fl02 ,
-----
END ,
1st blank space:Tag name
.................
2nd blank space:Tag No of CENTUM, YEWPACK
3rd blank space: Data acquisition destination
----- ------
--------
-------------
SCAN & USER'S PROGRAM DEFINITION PDHS FORM SU2

PERIOD & PHASE ...... ~~--- ......... ~.......
1st blank space:Period, 2nd blank space:phase
SCAN DATA
PV
ASTS ,
----------- ----
SV
PV
AS
SV
-------------- --
---
-----------------
--------- ---- -- -
........~.s..... , ............
LSTS ,
END ,
.................
-----------------
; 1st blank space: Item name

; 2nd blank space: Tag No of CENTUM, YEWPACK
; 3rd blank space: Data type name
; 4th
blank space: Data acquisition destination
Fig. 5 66 Work sheet of process data acquisition package (ADHS)
PV, AS, SV, and LS for tag numbers FlOl and F102 in the CENTUM
system are coliected at 60-second intervals. The acquired data is
stored in the PDHS of the YEWCOM system as items named PV,
ASTS, SV, LSTS and having the tag names FlOl and F102. In addition,
by using the work sheet to define the PDHS data base and the conditions for executing a user program for average value and accumulative
value calculations, a flexible process data acquisition system can be constructed easily. The data stored in the PDHS data base can be displayed on the screen and printed out by using the above-mentioned
REPOCALC human interface package.
(4) Inter-computer communication package
Communication between
computers is becoming increasingly important considering progress in
distributed configurations and total system integration. Figure 5.67
shows overall configuration of communication in the YEWCOM system. The communication techniques for YEWCOM are classified according to which type of computer system is involved in the communication.
(1) Communication between YEWCOM systems (DS 1000-/V): DS
324
Another type
computer
General
purpose
main flame
,...-------, BSC: Binary synchronous

communication
Another type RJE : Remote job entry
computer
MRJE : Multileaving remote job
entry
X.25 : Packet switching network
interface standard of CCITT
--------......
_HP 3000
Another type
computer
computer
produced
by HP
YEWCOM
DS/lOOO
YEWCOM
YEW COM
-N...__ _ _.....J
Fig. 5 67
Computer network in YEWKOM system
1000-IV is network software which realizes a distributed processing system by interconnecting a number of YEWCOM systems.
This network has a topology-independent architecture, and has no
preferences among line, star, or ring configurations .. It features
communication between programs, remote file access, remote command processing, remote data base access, and other such functions. It also possesses many functions for reliability, operability,
and maintainability, such as dynamic message re-routing. In dynamic re-routing, the communication route is automatically
changed when a failure occurs.
(2) Communication between YEWCOM and other computer systems: The
communication method can be selected from among following, according to the amount of transaction, amount of data involved,
and the communication function possessed by the other computer.
1) Binary Synchronous Communication (BSC)
2) Remote Job Entry/Multileaving Remote Job Entry (RJE/MRJE)
3) X.25 (CCITT packet exchange network)
4) IBM 3270 Emulator communication
The recent trend is towards total factory management, and for
this reason standardization of a horizontal type of communication line
among multivendor systems (LAN) is rapidly progressing. The importance of this type of network will continue to increase.
5.5
Data Communication and Equipment
5.5.1 Data communication and standards

In the application of computers to the fields of measurement and
data processing and process control, the system must quickly convert
5.5 Date Communication and Equipment
325
from the user data input to the output processing results. In other
words, the need for real-time processing is exceptionally great. Recently, progress in microprocessors and the technologies applied to
them have advanced the intelligence and decentralization of measurement instruments and control devices. The result of these advances is
that data communication technologies that allow accurate and
real-time communication of the required data among these types of
equipment, and also between these devices and computers, is becoming
extraordinarily important.
Date communication can be described as the linking of intelligent
equipment such as computers and control devices by means of communication lines through which digital data is mutually transferred with
good efficiency according to certain procedures that are determined in
advance. To send correct data to the destination system, several methods and types of specifications must be the same for both communicating systems. These include the types of communication cables and connectors, the electrical characteristics of the signal, the signal delivery
procedure, and the method of expressing the data. It is advisable to follow as much as possible the standardized methods and specifications,
which are known as communication protocols.
One of the most frequently observed items in standardization is
the set of standards known as the open systems interconnection (OSI)
Table 5. 11 OSI reference model
326
Layers
Function
Layer 7
Application
Provides all services directly comprehensible

to application programs
Layer6
Presentation
Restructures data to/from standardized format

used within the network
Layer5
Session
Synchronize & manage data
Layer 4
Transport
Provides transparent, reliable data transfer

from end node to end node
Layer 3
Network
Performs packet routing for data transfer between

nodes
Layer 2
Data link
Improves error rate for frames moved between

adjacent nodes
Layer 1
Physical
Physical
medium
Encodes and physically transfers bits between

adjacent nodes
Transmission line
Chap.5
basic reference model created by the International Organization for

Standardization (ISO). As shown in Table 5.11, the OSI model classifies
all the functions which are indispensable for data communications into
seven layers, ranging from the physical layer implemented in hardware
(including the function of converting data to an electrical signal for the
transmission line) to the application layer implemented in software (application-oriented functions such as methods of expressing messages
and file transfer procedures). In addition, the functions of each layer
are defined and serve as guidelines for standardization.
Recently, the demand for total network integration of production
information systems on the factory and enterprise level has become
high. For this reason, the factory communication network MAP (Manufacturing Automation Protocol), whose standardization and practical
dissemination has been proceeding gradually with the cooperation of
standardization organizations and many enterprises, centering in the
United States and including Europe and Japan. Under the primary leadership of the General Motors Corporation, MAP is also conforming to
OSI.
On the ot}J.er hand, there are also many widely-used communication networks and protocols which do not entirely conform with international standards. One such example is the Systems Network Architecture (SNA), which has been the IBM Corporation standard network
since 1974. It is also used worldwide.
Also, in Japan there are many examples of practical optical communications systems. This is another area in which standardization is proceeding slowly. In any case, when the use of non-standard protocols is
unavoidable, it is necessary to make detailed preparations and carefully design for stability and reliability.
5.5.2 Methods of data communications
There are many data communication methods, but if we represent
them in terms of communication distance versus communication speed,
we obtain something like Fig. 5.68. The classification is not entirely
clear, but the following rough categories can be distinguished.
(1) Communication between devices in close proximity
(2) Communication at intermediate distances
(3) Wide area networks
(a) Communication at close proximity
Within testing laboratories and so on, automatic measurement and
data processing is often carried out with various measuring instruments connected to a computer. The IEEE-488 instrument bus, specified by the Institute of Electrical and Electronics Engineers (USA) is
widely used for such purposes because it allows easy reconfiguration
327
10M
~ lOOk
"
0
-~
WAN (Wide area network)
lOk
-~
"
lk
100
10
100
lk
lOk
lOOk 1 OOOk
Communication length [ m)
Fig. 5 68
Communication methods
of the system for various experiments and instruments can be easily

mounted and dismounted. In the atomic power field, the Computer Automated Measurement and Control Equipment (CAMAC) standard has
been established and is used for about the same purposes. Also, for connection of computers to terminal equipment such as CRT terminals
and printers, connection of various types of measurement instruments
and automation machinery (robots, NC machines, etc.), the RS-232 C
interface standard set by the American organization, the Electronic Industries Association (EIA), is being widely used. Recently the RS-422
and RS-423 standards have been set as an extension of the RS-232 C to
longer distance communication. These are replacing the RS-232 C, albeit slowly.
(b) Intermediate distance communication
In the field of measurement and control, PROWA Y (Process Dataway) specified by the International Electrotechnical Commission (IEC)
is the most important method for exchange of data among dispersed
computers, control devices, terminals, and so on at distances of up to
ten kilometers. The PROW AY standard is a means of highly reliable,
high-speed data communication between the operator console and multiple filed control units in a distributed control system. It served as a
model for the F bus, the predecessor to the HF bus of the CENTUM
system. The major feature is the baton-passing method used in the
CENTUM system. MAP and PROW A Y are examples of local area networks (LAN), which have recently been receiving much attention as a
means of integrating dispersed equipment in various environments
such as offices, research centers, and factories.
The modem is often used as a simple and convenient way to make
a fixed connection between two devices. By introducing a modem be328
Chap.S
tween devices equipped with the RS-232 C interface described in Subsec. (a) above, the transmission distance can be extended. There are
different types of modems. Specialized modems for on-site communication lines are used to extend the distance between connected devices
within a facility. Public communication line modems and private communication line modems can connect devices that are far apart on a
one-to-one basis through lines provided by commercial communications enterprises.
(c) Wide-area networks
Wide-area networks are a method of connecting devices across
large distances through commercial communication systems. Data is
transmitted over the communication lines in the form of data blocks,
called packets. Information on the destination and packet sequence
number are included in the packet, thus the communication circuit utilization efficiency is very high because many users can share a circuit.
This method can be used for on-line transmission of orders, production
plans, and so on from the head office to distant factories.
Below, the communication methods which are particularly important in measurement and control are explained in concrete terms.
5.5.3 The IEEE-488 instrument bus

The IEEE-488 instrument bus is most often used to connect computer and personal computers to measurement instruments, plotters,
Equipments
Function
Computer
Plotter
Measuring
instrument
Fig. 5 69
5.5
Connection example of IEEE- 488 instrument bus
Date Communication and Equipment
329
and other such peripheral devices in testing laboratories. In this way,

automatic measurement, analysis, data processing and recording can be
accomplished. The functions of the connected devices can be classified
into three types: controller, talker, and listener. Each device must be
equipped with at least one of these three functions (Fig. 5.69).
The controller performs such management functions as specifying
the talker and listener, and releasing a connection. The talker loads
data onto the bus, and the listener downloads data from the bus. Generally, computers are equipped with all three functions, while measurement instruments are talkers, and plotters and other such devices are
listeners. As an example of a simple system, the computer performs
the controller function as required and specifies the instrument as the
talker, and itself as the listner, and then receives the measurment
data. After completion of data transmission, the disgnations for both
talkers and listener are released. When data analysis is complete, the
Table 5.12 IEEE-488 instrument bus signal
Function
Classification
Signal line
Handshake
signal
DAV
NRFD
NDAC
Data valid
Not ready for data
Not data accepted
Data is valid
Not ready for data
Data is not accepted
Control signal
ATN
REN
Attention
Remote enable
IFC
SRQ
Interface clear
Service request
EOI
End or identify
Command designation
Remote control
designation
Reset of interface
Interrupt notice to
controller
End-of-data
Data
DI 01 to 08 Data input output
Data
Next cycle
H
DI 01 to 08
Talker
DAV
L
H----.
L
Listener! NRFD:
NDACH---~--~~
L
Fig. 5 70 Data transfer cycle timing
330
Chap.5
computer specifies itself as the talker and the plotter as the listener. It
then sends the analysis data to the plotter for recording.
The bus has 24 signal lines, including 5 internal control signal
lines, 3 handshake signal lines for data transfer timing, 8 data lines for
8-bit parallel data, and 8 common bus lines. The bus line names and
functions are shown in Table 5.12. Figure 5.70 is the data transfer timing chart. The data transfer is asynchronous, and is matched to the slowest among the talker or multiple listener devices.
5.5.4 The RS-232 C interface and modems
The word modem is an abbreviation for modulator-demodulator.
These devices are used to extend the communication route. They perform different functions at the sending and receiving ends. At the
sending end, a serial digital signal is used to modulate an analog signal
(using, for example, frequency modulation), which is sent out on the
communication circuit; on the receiving end, the analog signal is detected and demodulated to reconstruct the original serial digital signal.
Modems for connection to private or public communication lines must
be authorized and conform to standard frequency band and transmission power specifications. They must also be designed to transmit signals while avoiding external interference from adjacent communication
lines and be approved.
The interface between the modem and the devices it is connected
to (computer, terminal, etc.) is standardized in terms of connectors, sigTable 5.13 RS-232C signal
Pin
arrengement
of
connector
140
150
160
170
180
190
200
210
220
230
01
02
03
04
05
06
07
08
09
010
011
240012
250 013
Code
FG
SG
SD
RD
RS
cs
Pin
No.
1
7
2
3
4
5
DR
ER
CI
CD
ST2
20
22
8
15
RT
17
ST1
24
Circuits marne
Frame ground
Signal ground
Send data
Received data
Request to send
Clear to send
Data set ready
Equipment ready
Call indicator
Carrier detector
Send signal
element timing
Received signal
element timing
Send signal
element timing
Signal direction
From MODEM
0
0
0
To MODEM
0
0
0
0
0
0
0
331
t:::
"";:;:
....
~
~
:::!
;:;:
....
"'~
~
c.r,
tv
Transmission distance
15 m
3 V or more
1.2km
(100 kbit/sec. or less)
(0. 2 mm' twisted pair cable)
200 mV or more
1.2km
(6 kbit/sec. or less)
(0. 2 mm' twisted pair cable)
200 mV or more
400mV or more
1, 5 to 6 V (plus, minus)
4 to 6 V (plus, minus)
2 to 6 V (plus, minus)
5to15V(plus, minus)
Transmitted signal voltage
Received signal threshold voltage
Differential type
Differential type
Single end type

Differential type
Differential type
Differential type
Single end type

Single end type
Transmitted-received circuits
(Trans.)
(Receiv.)
n:n
1:1
1: 1
1:1
RS-485
RS-423
RS-422
Connection mode
RS-232 C
Table 5. 14 Typical electric specifications of RS-232 C, RS-422, RS-423, RS-485
nal lines, electrical specifications, data exchange method, and timing.

One such standard is the RS-232 C interface. Table 5.13 shows how the
signal lines are used in this standard. The typical electrical specifications are listed in Table 5.14.
Serial data is passed between devices by either the asynchronous
method or the synchronous method. In the asynchronous method, data
is handled in units of one character (6, 7, or 8 bits). As shown in Fig.
5. 71, data is delimited by appending a start bit before each character,
and a stop bit after each character. The receiving device recognizes
the beginning of the character by detecting the start bit, and from that
time reads each bit making up the character at fixed time intervals (determined by the transmission speed).
Start bit
1
.
Data btt
Parity bit
, l
1 =r.Ji--T--T--T--...---r--T--T--~rF
LSB :
:
:
:
:
:MSB
Stop bit
0 ----
--L-.l--.&.--L--L--.J--.1.--.&..-
Fig. 5 71 Asynchronous data
Synchronous transfer, on the other hand, sends data in units of

character strings, which are made up of a .number of characters. As
shown in Fig. 5. 72, no start bits or stop bits are added as is done when
the transfer unit is one character. Bit position is detected by synchronization with a special clock signal, and characters are distinguished by
detecting a special synchronization character at the head or in the middle of the character string and then counting bits as the data is received.
Modem clock
Data bit
i0
I
0 0 1 1 0 0,1 1 0 0 0 1
O!
'--Synchronization charactor3218,_....,_____ Charactor data 63,.--:
Fig. 5 72 Synchronous data
Recently the synchronous transfer method called the High Level

Data Link Control Protocol (HDLC) has been standardized. This method uses frames of variable bit length as the data transmission unit rather than single characters or character strings. It is being used in wide
area networks for connecting computers. In this method, as shown in
Fig.5. 73, a special bit string (01111110), called a flag sequence, is
places at the beginning and end of the frame. These flags are used to
333
1 - - - - - - Frame --------+1
Data
Check
code
F; Flag sequence
C; Control information
A; Secondary station address
Fig. 5 73 Data frame of HDLC
detect the frame and the bit positions. In the figure, the address represents the address of the secondary station corresponding to the main
station, and the control information specifies the data frame type.
When the transmission line is long or in factory environments
where there is much electrical noise, noise errors may be generated in
the transmitted data. Thus, a means of detecting and correcting these
errors is needed to allow transmission of correct data. In asynchronous
transfer, a parity bit is added to each character unit of data; in the synchronous method, a check code that has a high error-detection power
is appended to the end of each data unit (either character string or
frame). Using these features, the receiving side can detect whether or
not there is an error in the data. If an error is detected, the receiving
side issues a request to the transmitting side for re-transmission.
The RS-232 C interface can be used for direct connection of nearby devices (within 15 meters) as well as for the widely used standard
modem connection. Thus it is standard equipment on many computers
and control devices. As a substitute for the RS-323 C interface for
transmission without using a modem at relatively longer distances
(several hundred meters or so, depending on the cable diameter), the
RS-422 (differential type), RS-423 (single-end type), and RS-485
(differential multi-drop type) interfaces have been standardized, and
their use is spreading. The electrical specifications of these interfaces
are listed in Table 5.14.
5.5.5 Local area networks

Local area networks (LAN) were developed as a means of compact,
high-speed data exchange through a single medium among a number of
computers, workstations, controllers and so on located within a limited
area such as in the same building, or on the same floor. Many types of
LAN have been developed for various purposes, but they can be simply
classified according to their physical configurations (topology). Figure
5. 74 illustrates three topologies: the bus, the ring, and the star configurations. The methods for transferring data to the transmission medium (access method) include the token passing (or baton passing)
scheme and the CSMA/CD method.
The bus topology has high reliability because the transmission
334
Connected equipment
Bus
Ring
Fig. 5 74
Star
Topology
route contains no active elements (transistors, LSI circuits, etc.) in series, unlike the ring and star configurations. Thus the bus topology is
suitable for factory on-line networks. The PROW AY and CENTUM
HF bus use the bus topology. In the ring topology, the signal is intercepted and retransmitted at the connecting point for each device
(node). Optical fibers are often used as the transmission medium for
LAN. The star topology is used in telephone exchanges, and seldom
used in factories.
In the token pass method, a frame (called a token) makes rounds
to all the devices on the LAN. Only the device that is to receive the token can send a signal to the medium. In the CSMA/CD method on the
other hand, any device can send a signal at any time if there is no signal on the medium, thus signal collisions may occur. At such times,
each device is made to wait for a set time, which is different for each
device. The CSMA/CD method is efficient when the traffic load is
light. However, when the load increases, multiple conflicts develop
regardless of the waiting. As a result, efficiency is lost and response
time is lengthened. Degradation of response time beyond the limit is
unacceptable in a network for factory control and management, thus
the token passing method is most often used for factory networks.
MAP and PROWA Y are examples. ETHERNET (bus-type) is a typical
example of CSMA/CD. It is used extensively for light-load situations
such as in offices, and in research and design facilities.
Standardizatio n is currently proceeding in IEEE. Here, three standards have been discussed. Of these, brief specifications for the
IEEE-802.4 method, which is intended for factory use and is also employed in MAP are listed roughly in Table 5.15, together with the specifications for PROW A Y.
5.5.6 Optical communications
As a transmission medium, optical fibers have such advantages as
insensitivity to electrical noise, the capability for high-speed transmission over great distances, and safety because they do not transmit elec5.5
Date Communication and Equipment
335
Table 5. 15 Specification summary of IEEE-802. 4 and PROWAY

Item
TOPOLOGY
Transmission medium
Transmission rate
Node number
Modulation method
Control method
Error control
Worst case access
time
PROWAY
IEEE-802.4
Bus
Bus
75 0 co-axial cable
75 0 co-axial cable
1, 5, 10 Mbit/sec
Phase continuous FSK
Phase coherent FSK
Broadband
TOKEN PASSING
Max. 2km
1 Mbit/sec
Max. 100
Phase continuous
FSK
TOKEN PASSING
(Baton passing)
Not data error monitoring Retransmission and
recovery by immediate
acknowledge monitoring
In case of no transmission 50 ms or less
error, it is decided by
station numbers and
working parameter value.
tric power. For these reasons they hold great promise for use as a
transmission medium in factories, and their use is spreading. However,
optical connections are difficult, the cost is high, and standardization is
slow. Thus, there is poor compatibility among the products of different manufacturers. In the past optical media has been used for special
environments and high-performance communications. Nonetheless,
these points of difficulty will be overcome steadily, and in the near future we can expect that use of optical communication media will
spread.
The optical fibers used in communications are difficult to branch,
so they are mostly applied to one-to-one communications or to
ring-type communication configurations. However, recently optical
fibers have been introduced logically to bus and star LANs by using a
multiple-branch element called the star coupler. Furthermore, RS-232
C connectors with built-in optical modems for a duplex communication
are being developed, and low-cost, convenient optical communication
is becoming possible. As a representative example of a high-performance
ring network, the structure and main specifications of YEWLINK 32
are introduced in Fig. 5.75 and Table 5.16.
5.6 Basic Components of Digital Contro/
14
)-1
5.6.1 Microprocessors
Microprocessors, which have developed from the first 4-bit device
336
YEW COM
Comput.er system
CE TUM
Operator station
Intelligent recorder
LCTL : Link controller

LCPL : Optical coupler
PSU : Power supply
Fig. 5 75 Configuration of YEWLINK 32
Table 5. 16 Specification of YEWLINK 32

Specification
Item
Ring
Optical fiber cable (GI 50/ 125)
Max. 120km
32. 064 Mbit/ sec
Max. 32
Loop, duplexed nodes
Loop-back
Node bypass
Loading number Max. 8
HF bus
Type
IEEE-488
RS-232 C
Telephone
TOPOLOGY
Transmission path
Transmission rate
Node number
Fault measures
Interface
5.6 Basic Components of Digital Control
337
..8
E
"c
c
"E
u
jjj
1- Kbi t RAM
1972 74
76
78
Fig. 5 76
80 82
Year
84
86
88
90
Trends of YLSI
of late 1960 to today's 32- bit devices, are widely used in the fields of
measurement and control. The course of microprocessor element development is shown in Fig. 5. 76. The principle parts of the 32- bit minicomputers of the 1970's are now integrated onto a piece of silicon that
is about 8 mm square.
(a) Microcomputer configuration
The architecture of microcomputer varies somewhat with the
type, but it generally consists of a microprocessor chip,* which integrates the arithmetic unit, registers, and control unit, memory chips for
storing programs and data, and a peripheral interface chip for control
of input/output devices. The chips are interconnected by buses which
obey unified electrical and logical rules. The buses include an address
bus, a data bus, and a control bus. For example, an 8-bit microprocessor has an eight- wire data bus. Bus rules differ among manufacturers,
but connection among chips from a single manufacturer is relatively
easy. A device which has the microprocessor, memory, and peripheral
interface integrated on one chip is given the special name, "single-chip
microprocessor." As opposed to this, a processor comprising several
chips such as shown in Fig. 5. 77 is called a "general purpose microprocessor" in this book.
Figure 5. 78 shows an example of an actual one-board microcomputer. This single printed circuit board also carries peripheral circuitry
for flexible disk drives, hard disk drives, printers, a clock and communication control.
Here, the term "chip" is used to mean integrated circuit chip (IC).
338
Chap.5
~~r>
~
ll
'--<D__.f ==f!;:=::::::Bu
I Q I II I
<D Microprocessor
Memory
DigitS indicale lhe number
@ Peripheral inlerface
<!) Conlrol bus
of signal lines
Address bus
Dala bus
Fig. 5 77 Configuration of microcomputer (An example of 16-bit)
Fig. 5 78 Application of microcomputer

(CPU board of YEWMAC, Yokogawa Electric Corp.)
(b) Types of microprocessors

Technology for the integration of processor elements has improved and various kinds of microprocessors have been developed for
diverse applications. Table 5.17 lists the broad categories of microprocessors and their main uses. The devices listed as examples are those
that are well known in the market, but there are many others as well.
In measurement and control, the general purpose microprocessor is
most often used because it can be combined with peripheral chips in
many ways. However, as integration advances and memory capacity increases, single-chip microprocessors are also coming into frequent use.
There are many single-chip products, which differ according to the
memory and peripheral circuitry combined in the chip. Thus selecting
among the available single-chip microprocessors requires careful analysis.
339
"';:!.....
;;:
.;;:
,kl
;:
~
;:!
.....
"';;:
~
"'.....
'""
{;
.....
Digital signal
processor (DSP)
Operational processor
Bit-slice microcessor
Single-chip
microprocessor
General purpose
microprocessor
Classification
2901, 2909 (AMD)

29116(AMD)
AS 888/889 (TI)
8087 (Intel)
68881 (MOTOROLA)
PD 7720 (NEC)
TMS 320(TI)
MB 8764 (Fujitsu)
Image processing, High speed sequencer
Co-processor of general purpose

microprocessor (floating point
arithmetic)
Voice recognition, Image processing,
Fourier analysis
JL
8096 (Intel)
8048 (Intel)
6801 (MOTOROLA)
6301 (HITACHI)
Computer I/0 device, automobile

engine control, Industrial instrument
8-bit
LAN controller, NC, Robot
546 (NEC)
HD 44700(HITACHI)
16-bit
JL PD
Electrical appliance (Microwave

oven, VTR, etc. )
4-bit
68020 (MOTOROLA)
80386 (Intel)
V 60/V 70 (NEC)
8086 (Intel)
68000 (MOTOROLA)
Distributed control system,

Measuring system, EWS,
Personal computer
16-bit
Production management computer
8085 (Intel)
Z 80 (Zilog)
6809 (MOTOROLA)
Measuring instruments, Sequencer,

Single loop controller
8-bit
32-bit
Examples
Main use
Table 5.17 Classification and application of microprocessor
For bit-slice microprocessors, 4-bit devices are predominant. However, their architecture allows multiple units to be connected to form
16-bit and 32-bit processors. Unlike other microprocessors, they can
be custom-designed for specific purposes, such as image processors
which require high processing speed, high-speed sequencers, matrix
calculations, and other such applications. The operational processor
performs floating-point arithmetic, trigonometric function calculations, and so on, and is often used as a co-processor for general purpose microprocessors.
In process control, digital signal processors perform real-time calculations on signals from measurement instruments. The central operation in digital processing is integrating. Digital signal processors are
specifically designed for this purpose, and have several tens of times
the power of 16-bit general-purpose microprocessors for special calculations. They are expected to have applications in a wide range of areas, such as voice synthesis, and voice recognition.
(c) General-purpose microprocessors
There are 4-bit, 8-bit, 16-bit, and 32-bit general-purpose microprocessors. However, the 4-bit general-purpose microprocessors are now
seldom being used, but the trend is toward single-chip devices for use
in home appliances. The 8-bit microprocessors are most suitable for
handling text characters, and are applied to peripheral I/0 devices such
as printers. In the field of control they are used in single-loop controllers and so on. The 16-bit microprocessors have powerful calculation
processing capabilities, and have broad application in distributed control systems and processors for measurement equipment. In around
1984, the 32-bit microprocessors appeared, and their power rivals the
performance of the 32-bit minicomputers developed in the 1970's. Now,
they will have wide application in the fields of production management, robot control and other fields requiring high performance.
Figure 5. 79 shows some examples of actual 8-, 16- and 32-bit general purpose microprocessors. As the number of pins increases, the package design changes to increase the pin density. Also, although the number of chip elements increases greatly as the bit capacity increases,
with improved IC technology and the application of CMOS technology,
the dimensions of the silicon chip itself have not increased significantly. General-purpose microprocessors can be divided into two main
series: the Intel series (8085 and 8086) and the MOTOROLA series
(6800 and 68000). There are small differences between the processors
of these two series, but the basic processor functions are equivalent.
Figure 5.80 shows the I/0 signals of the MOTOROLA 68000 series
16-bit general purpose microprocessor. The 23 address buses allow address specification for up to 16 megabytes of memory space. A 16 data
341
8-bit
8085 (Intel)
16-bit
68000 (MOTOROLA)
Fig. 5 79
32- bit
68020 (MOTOROLA)
Examples of general purpose microprocessors

Address bus
At to A 23
Interrupt
IPLo to IPL2
Processor status
to FC2 =
F
68000
BERR
System control { RES
HALT
}
BR
Bus a rbitration control
BG
BGACK
Fig. 5. 80 Signal wires of 68000
buses permit 16 bits data to be read or written. Bus control employs an

asynchronous method, which can match the differences in response
times of memory devices and peripheral chips. In the field of measurement and control, interrupts are highly useful. The 68000 has three
interrupt lines, which, when decoded, provide interrupts with seven
levels of priority.
(d) Peripheral integrated circuits
Peripheral integrated circuits (ICs) are most useful in the development of various kinds of equipment. When a function is developed,
whether or not there is an LSI chip suitable for the function makes a
great difference in the amount and cost of the hardware involved. ICs
are being developed for high-performance display, communications,
and supplementary memory devices. Table 5.18 classifies peripheral
ICs for 16- bit microprocessors, and lists examples of each type. The examples are broadly divided into the Intel series and the MOTOROLA
342
Table 5.18 Classification of peripheral LSI

68000
series
8086
series
DMA controller
DRAM controller
68450
8237
8208
Display control
CRT character display

CRT graphic display
LCD controller
6845
63484
63645
8275
82720
Communication control
Asynchronous method
Synchronous method
LAN
6850
6852
68824
8251
8251
82586
Timer
Interval/fixed cycle
Watch/ calendar
6840
8253
146818
Classification
Function
System control
Digital/analog input-output Digital input-output

Analog input
Analog output
6821
46508
6890
8255
Auxiliary storage control
FDD controller
HDD controller
63265
63463
J.lPD 765
J.lPD7261
General purpose bus control
GP-IB controller
8291/8292
series, but by adding some external circuitry, other series of ICs can
also be connected.
5.6.2 Memory elements and storage equipment
(a) Semiconductor memory
After the appearance of the 1 kilobit MOSRAM in 1970, the bit
density of memory elements increased by ten fold about every three
years (Fig. 5. 76), and by 1986 the 1 megabit DRAM was available for
use. In parallel with this, the cost per bit dropped very rapidly, greatly
improving the equipment cost-performance ratio.
Semiconductor memory is broadly classified into random access
memory (RAM) and read-only memory (ROM). This classification
scheme and examples of the devices are shown in Table 5.19. RAM
memory is volatile, which means that the stored contents will be lost
when the power supply is cut off. ROM, on the other hand, does not
have this problem, thus in the field of measurement and control it is often used for storage of programs and constant values. RAM is further
divided into static RAM (SRAM) and dynamic RAM (DRAM). DRAM
circuit design is complicated because a refresh circuit is required to
sequentially access the memory cells at regular time intervals. However, it is suitable for large-capacity memory devices. SRAM, as opposed
343
Table 5.19 Classification of semiconductor memory

Example of element
configuration
Classification
RAN
(random
access
memory)
ROM
(read
only
memory)
*1
*2
DRAM
(dynamic RAM)
NMOS
CMOS
256K X1-bit
1MX1-bit
SRAM
(static RAM)
CMOS
CMOS
ECL
256K X1-bit
32KX1-bit
64KX1-bit
Masked ROM
CMOS
128K X 1-bit
PROM
(programable
ROM)
Un-erasable PROM
Bipolar
Erasable PROM
EPROM*'
EEEPROM"
CMOS
CMOS
8KX1-bit
128KX1-bit
8KX1-bit
Electrically programable ROM

Electrically erasable PROM
to this, can be connected to the microprocessor by means of simple circuitry and has the additional merits of fast read and write times. The
drawback is that the cost per bit is high compared to DRAM.
Conversion to CMOS technology for the purpose of reducing the
power consumption in integrated circuits is a strong trend, and this is
particularly true for semiconductor memory devices. As the number of
transistors in the memory device increases, this trend becomes necessary in order to control the heat generated in the chip. Reduction of
power consumption in the semiconductor memory devices used in control equipment is also very important. Specifically, it would allow smaller batteries to be used in the battery backup required to preserve
stored data in event of power outages. Cooling systems can also be
made smaller.
Erasable-programmable ROM (EPROM) is a type of memory
device whose memory contents can be erased by means of irradiation
with ultraviolet light then new data can be written in its place. In contrast with this, the electrically-erasable-programmable ROM (EEPROM) can be erased electrically. Eight-kilobit EEPROM devices have
reached the practical stage. As the reliability of these devices improves, they are being used more and more frequently.
(b) Magnetic storage devices
Magnetic storage devices employ the stable nature of magnetism.
There are various kinds of magnetic storage devices, including flexible
disk drives (FDD), magnetic disk drives, and magnetic tape devices. As
shown in Fig. 5.81, magnetic storage devices are slower than semiconductor memory devices, but they are non-volatile and have large capac344
II'
In
Access time [ s ]
Fig. 5 81 Storage capacity and access time of storage devices
ities. Thus, they are used for auxiliary storage for computer systems.
Among these devices, the flexible and hard disk drives have undergone
remarkable reduction in size and cost per bit, and play an important
role in the increase of system reliability and lowering of cost.
The flexible disk drive uses a circular sheet
(1) Flexible disk drive
of plastic material as a removable medium. This allow easy handling,
and since its development in 1972, it has penetrated a wide range of
markets to become a generally-used removable storage medium. Beginning with the 8-inch disk that could hold 400 kilobytes of data, the
disk size shrank to 5.25 inches and then again to 3.5 inches. At the
same time the capacity increased to the high density of 1.6 megabytes.
Table 5.20 shows the various combinations of size and capacity of existing disks. Because of this variety, it is necessary to check the compatibility of the disk and disk drive. The trend is toward increased use of
the 1.6 megabyte disk. The 8-inch disk is mostly used for removable
media in large computer systems.
Table 5. 20
Disk size
8-inch
5. 25-inch
Kinds of flexible disk devices
Capacity (unformattep, Mbyte)

0.4
(single side
single density)
0.8
(double side
single density)
0.5
3. 5-inch
(high double
density)
1.6
(double side
double density)
1.0
1.6
1.0
1.6
345
(2) Hard disk drives

Formerly, magnetic disk drives had disk packs
or disk cartridges that were used as removable media. Recently, small,
large-capacity hard disk drives (HDD) have been developed. The hard
disks can be grouped together into a single unit inside the disk drive,
and they have gradually become the mainstream device-at first for
personal computers and now also for measurement and control equipment.
In the HDD, the disk unit is directly connected to the spindle motor. The principle of the HDD is illustrated in Fig. 5.82. The disk drive
consists of magnetic disks, spindle and motor constructed as a single
unit to rotate the recording medium, a head positioning mechanism,
and peripheral control circuits. Head positioning is accomplished by
means of either a stepping motor or a voice coil. The stepping motor
method uses a simple mechanism, and is thus used for small and inexpensive drives. The voice coil method uses a servo mechanism for positioning the head, thus allowing high-speed access and high-density recording.
Fig. 5 82
Configuration of magnetic disk device
HDDs use the same three sizes of disk as the FDD. In addition,
large disks are also used for very large capacity applications. Examples
of actual disk drives are shown in Fig. 5.83. There are also example of
5.25- inch drives that have capacities exceeding 300 megabytes. HDDs
feature a sealed construction, and thus have excellent resistance to environmental contamination compared to other types of supplementary
memory devices. Recently, small environment-proof HDDs have been
under development.
5.6.3 Display elements and devices
The man-machine interface allows interaction between people and
346
Chap.S
50 Mbytes (
Fig. 5 83
Example of fixed magnetic disk (5.25 inch)
equipment. It uses various kinds of display elements and devices which

are appropriate to the purpose at hand. Table 5.21 lists the types of display elements and devices, important aspects of their display characteristics, and their applications. The cathode-ray tube (CRT) has a wide
range of applications, beginning with serving as a computer terminal.
The plasma display is more expensive than the CRT, but it is light in
weight and has good display characteristics. Thus it is beginning to be
used to replace the CRT in some application areas. Liquid crystal displays are thin and feature low power consumption. They are often
used for portable, small display capacity devices. The electroluminescent display (ELD) has a highly luminous light-emitting surface, and
has been a subject of recent research and development. It is also used
for back-lighting in liquid crystal displays. Light-emitting diodes
(LED) and fluorescent display tubes are mostly used for the single purpose of alphanumeric character display devices. There are also such
special devices as magnetic inversion display plates and magnetic migration displays, which are suitable for wide-area advertising.
(a) Cathode-ray tubes
Cathode-ray tubes (CRT) have been used as display devices for a
long time. However, they have improved in step with progress in fine
process technology, and resolution has improved to the point of being
a mature technology. Presently, CRTs used as display devices are classified according to their scanning method into raster-scan types and
random-scan types (Fig. 5.84). Random-scan devices are further divided into refresh-type displays, in which data must be rewritten repeatedly within the luminous persistence time of the phosphors, and storage-type displays, in which the phosphors have memory characteris5.6 Basic Components of Digital Control
347
::!
"'.....
~:
~
~
;::
....
~
::!
.....
~
"';;;-
~
c:.r,
Oo
(Note)
0
0
6
6
Display
capacity
Resolut ion
6
6
6
Display
area
Compared characteristics
0
0
Thickness
Control monitor, CAD/ CAM , EWS, Simulator
Measuring instrument, Portable device
Control monitor, Production management
Instrument display, Panel display
Instrument display, Panel display
Measuring instrument, Portable device
Panel display, Tra ffic control display,
Traffic information display, Trouble
status display, Public relation,
Advertisement
Panel display, Public relation,
Advertisement
Application
Classification o f display elements and devices
: excellent, 0 ; good, 6 : not good, x ; inferior .
Magnetic migration type display
CRT
Crystalline liquid display
Plasma display
LED
Fluorescent display tube
EL
Magnetic inversion display plate
Classification
Table 5. 21
<
--
Random -scan type C RT
Raster-scan type CRT

Fig. 5 84
Scanning method of CRT
tics. Random-scan type displays achieve high-quality graphics because

oblique lines are free from the step-like distortion characteristic of raster-scan types of display. However, they have limited display speed
and they are expensive. Now, the random-scan displays are seldom
used.
On the other hand, raster-scan displays are in wide use because,
as an extension of television technology, they now make use of integrated circuits, which has resulted in increased reliability and lower
cost. The principle of the raster-scan color CRT display is illustrated
in Fig. 5.85.
Color CRT
Driver section
Display control section
Deflection
Display data
Shadow mask
Fig. 5 85
Principle of raster scan type C RT di play
(1) Color CRTs

On a color CRT screen, red, green and blue phosphor dots are arranged in a regular, sequential pattern. There are
three electron guns corresponding to these three colors (RGB). A shadow mask guides the electron beams from these electron guns such that
they strike the phosphor dot array.
(2) Color CRT drivers
To form a display on the CRT screen, external synchronization signals and a video signal are necessary. The synchronization signals (horizontal and vertical) produce deflection scanning of the electron beams. The video signal (luminance signal) causes
349
the phosphor dots to produce light.

(3) Display control
The synchronization signals specify the beginning of horizontal and vertical scanning. The scanning position is accurately determined in the display control section by means of a counter.
The display data corresponding to the scan position is read from the
display memory, and the appropriate video signal is generated. In this
way, information can be displayed on CRT screens.
(b) Liquid crystal displays
Formerly, liquid crystals were mainly used as small-capacity display elements in small devices typified by electronic calculators and
digital wristwatches. However, recent progress in thin-film semiconductor technology has improved the capacity and responsiveness of liquid crystal display elements, and it has become possible for these devices to handle the large-volume display data for computer terminals
and measurement instruments. Achievement of color capability is progressing faster for liquid crystal displays than for other thin display devices. Liquid crystal television screens 3 inches to 5 inches in size have
already become practical, and 10-inch to 12-inch, 640-dot by 400-dot
panels will soon follow.
Figure 5.86 shows the operating principle of the twisted nematic
(TN) liquid crystal. The liquid crystal lies between two polarizing filters whose polarization planes are at right angles to each other. When
there is no electric field applied perpendicular to the filters, the crystal
molecules are twisted as shown in the figure. In this state, light passes
through the filters, because the angle of light deflection is 90 . On the
oenecr ion filler
Fig. 5 86 T
J50
Crystalline liquid model
other hand, if an electrical field is applied, the crystal molecules orient

themselves parallel to the field, and the light is blocked by the polarizing filters. Employing this principle, it is possible to display information by controlling the electric field applied to individual crystals that
have been formed into a pattern.
5.6.4 Analog/digital conversion
Analog/digital (A/D) conversion, which means changing an analog
signal into a digital signal, is an important technology that is essential
to the introduction of computers to the field of measurement and control. The output signal from sensing devices such as temperature detectors and flow rate detectors are mostly analog signals which must be
converted to the digital signal form suitable for input to computers.
There are many conversion methods, and the choice of method is based
on conversion speed and conversion accuracy. Table 5.22 lists various
conversion methods and their main applications. Figure 5.87 shows the
relationship between conversion accuracy and conversion speed.
The V-F conversion type generates a pulse signal whose frequency (F) is proportional to the voltage of the input analog signal (V), and
the voltage value of the input signal is determined by counting the
number of pulses in a given period of time. The V- T method, on the
other hand, is based on counting the number pulses generated by a
standard clock within a gate time (1) that is proportional to the voltage of the analog signal (V). Both of these methods are integration
methods, and have the disadvantage of being relatively slow. However,
Table 5. 22 A/D conversion method
Classification
Use
Integration
method
V-F conversion type

High accuracy-low speed use
ex. : Charge balance type
Digital voltage meter
Panel meter
V-T conversion type
ex.: Dual slope type
Comparison
method
Feedback comparison type

Medium-high speed use
Data acquisition
ex.: Successive approximation type,
measurement
Servo-balancing type
Process control
High speed use
Non feedback comparison
Image processing,
type
ex. : Parallel comparison
Waveform memorizing
type,
device
PCM communication
Serial comparison
type
JSJ
I6
I4
I2
Servo-balancing
type
8"
g
Io
8
i:i5
Dual slope
integration type
2
Ips
Ims
Conversion time
Is
Ins
Fig. 5 87 Characteristics of AID converters
they have the merits of being inexpensive and resistant to noise.

The comparison methods have the advantage of being easily adapted to fast IC circuitry. Of the comparison methods, the successive approximation type is often used in process control and data acquisition.
There are also many ICs which can be applied in this method. The
basic circuit and operation waveform of the successive comparison
method are shown in Fig. 5.88. The conversion start signal clears the
Start signal
Stop signal
Analog voltage
approximation
type register
Comparator
'
Time
Clock
Fig. 5 88 Principle and circuit configuration of successive approximation type
352
Chap.5
register, and then the output of the D/A converter is compared with
the analog input voltage bit-by-bit, starting with the highest-order
bit. According to the result of each comparison, a 0 or 1 is set in the
corresponding bit of the resister.
5.6.5 Optical communication elements
Optical communication, which developed out of broadband digital
communication systems, has recently begun to occupy an important position as an elemental technology for the construction of total control
systems in factories. The main components of an optical communications system are a light-emitting element, a light-receiving element,
and an optical fiber, which serves as the transmission path. In addition, fiber splitting and coupling devices are also important. These devices are used to branch one fiber into several fibers, and conversely, to
combine several individual fibers into a single fiber. Fibers are broadly
classified as step-index fibers (SI), graded-index fibers (GI), and single-mode fibers (SM) , according to distributions of their refractive indices. However, in the field of control, the GI fiber is often used for
reasons of economy, speed, and so on.
Light-emitting diodes (LED), and semiconductor laser diodes (LD)
are used as light-emitting elements. The characteristics of these two
types of device are listed in Table 5.23. There are two types of
light-receiving elements. One is the PIN photodiode, which converts injected photons directly into electrons. The other is the avalanche photodiode (APD), which using the avalanche amplification effect, produces
a larger electrical output for a given input light.
Table 5. 23 Comparison between light emitting diode and laser diode
Light emitting diode
Laser diode
(LED)
(LD)
Spectral bandwidth
Output power
Coupling efficiency
to optical fiber
Lifetime
350 A
1 mW
Several %
30A
Use
Mainly used for

shortwave band.
Low, medium speed.
One million hours
Short distance
transmission.
lOmW
50%
One hundred thousand
hours
Mainly used for
longwave band.
High speed.
Long distance large
volume transmission
353
REFERENCES
1) E. Ohshima:"Transition of automation and safety in chemical plant," for safety of

High Pressure Gas Eng., 22, 4 (1985) 158-165 (in Japanese).
2) Y. Wakasa, et al.:"Enhancing reliability of CENTUM distributed process control
system," Yokogawa Tech. Rep., 27, 4 (1983) 3-12 (in Japanese).
3) S. Takigishi, et al.: "User oriented flexible operator station in a distributed process
control system," Yokogawa Tech. Rep., 28, 4 (1984) 24-31 (in Japanese).
4) T. Ogawa:"Process data highway-configuration, functions and protocols," Yokogawa Tech. Rep., 26, 4 (1982) 46-52 (in Japanese).
5) K. Matsunaga:"Digital algorithms for process control-PID control, deadtime compensation and advanced control," Yokogawa Tech. Rep., 27, 1 (1983) 38-43 (in Japanese).
6) K. Matsunaga:"Documenting process control sequence by decision tables," Yokogawa Tech. Rep., 26, 3 (1982) 44-50 (in Japanese).
7) H. Tamura:"YEWMAC, YEW's manufacturing line control system development
philosophy and overview," Yokogawa Tech. Rep., 27, 3 (1983) 24-27 (in Japanese).
8) H. Kawai, et al.:"Development policy and technology of YEWMAC 300/50," Yokogawa Tech. Rep., 30, 1 (1986) 13-22 (in Japanese).
9) Kojo Jidoka Jiten Henshuu Iinkai: Encyclopedia of Automization for Factories,
Sangyo Chosakai (1983) (in Japanese)
10) Japan Electronic Industry Development Association: A Report of Investigation for
Industrial Computer Systems, 60-A-232 (1985) (in Japanese)
11) H. Tamura, et al. :"How the YEWCOM computer fits into the YEW total system architecture," Yokogawa Tech. Rep., 27, 2(1983) 2-64 (in Japanese).
12) S. Yamamoto, et al.:"YEWCOM performance enhancement," Yokogawa Tech. Rep.,
28, 4 (1984) 17-23 (in Japanese).
13) K. Ohmori, et al.:"YEWCOM redundant system," Yokogawa Tech. Rep., 30, 1
(1986) 7-11 (in Japanese).
14) K. Sakamura: Computer Architecture, Kyoritsu Shuppan (1984) (in Japanese)
15) I. Morishita: Hardware of Microcomputer, Iwanami Shoten (1984) (in Japanese).
16) lnfor. Process., 27, 6(1986) (in Japanese).
17) Y. Nagahashi: Design of A-D/D-A Conversion Circuits, CQ Shuppan (1980) (in Japanese).
18) M. Maeda: Optical Fiber Communication, Denki Tsushin Gijutsu News Sha (1981)
(in Japanese).
354
Chapter 6
FINAL CONTROL ELEMENTS
This chapter describes the final control elements of instrumentation systems. Featured are the control valves that operate either by
signals from a controller or by manual action to control the transmission and cut-off of fluids such as liquids, gases, and steam. The controller can be considered to be the head, with the control valves acing
as the hands and feet. The control valve, which receives the control
signal, consists of an actuator that converts the control signal to an
operating variable and a main valve body that applies the work directly
to the object of control.
6.1
Types of Control Valves
There are many different types of control valves used in industry.

The type used depends on the requirements of the process involved.
The follwing descriptions are classified according to the types of operation signals.
6 .1.1 Pneumatic control valves

This control valve functions according to an air pressure signal
(0.2 to 1.0 kg/cm 2 G). The air signal can be applied directly to the valve
actuator, but frequently a positioner is used to improve response and
minimize hysteresis. Pneumatic control valves have been used for
many years, and so have become relatively standardized and very reliable. They have many advantages, such as being relatively inexpensive,
easy to maintain, and equipped with essential features to prevent bursting. This type of valve is the most popular type currently in use.
6.1.2 Electrical control valves
This control valve operates by an electric current signal (mainly 4
rnA to 20 rnA DC). In the manipulation of motor-drive valves and elec6.1 Types of Control Valves
355
tromagnetic valves, an electric pulse signal may also be used. In the

case where an electric current signal is used, the actuator is manipulated by a current-to-pneumatic, current-to-current, or current-to-hydraulic positioner.
6.1.3 Hydraulic control valves
The strong points of the hydraulic control valve are its good response to the actuator and its strong operating power. It is manipulated by means of a current-to-hydraulic positioner.
6.1.4 Self-powered control valves
These are self-regulating control valves installed in the field as
simple regulators. The process variable, such as pressure and differential pressure, is the valve's direct power source or the valve is manipulated by using the change in vapor pressure of the liquid enclosed in
the temperature sensing element. Now, because the control valve
works directly with the controlled system, it has the most important
role. Thus choosing the type of device that fits the control system and
the control purpose is the key toward achieving the success of the
instrumentation system.
This point is important and attention is concentrated on the control valves that fit diverse control requirements and have the highest
level of use as final control elements for instrumentation systems. The
focus is on:
(1) A method for choosing control valves that fit the instrumentation system
(2) Types of valve bodies and their characteristics
(3) Types of actuators and their characteristics
Figure 6.1 illustrates the construction of a pneumatic control valve
with a globe valve, which is widely used as a basic valve body subassembly. Particulars about globe valves will be taken up in Sec. 6.3, but
this shows a representaive example of a control valve.
6.2 Choice of Control Valves
6.2.1 Various conditions affecting choice
The control valve is impotrant in the process operation. Not only
are the specifications of the valve itself important, but it is also necessary to sufficiently consider other matters related to the control valve
in order for it to function as required. The following are the main
points to keep in mind when specifying a control valve.
(a) Target process
It is necessary to completely understand the control system, in356
Chap.6 Final Control Elements
....
0
~
"
u
Yoke
<
Lubricator
Packing seat
Bonnet
Gasket packing
Guide bu hing
Valve plug stem, Gland ]

Trim [ Plug, Lantern ring
Seat ring, Metal packing
Fig. 6 1 Con !ruction of control valve
eluding the control valve. Additionally, one should sufficiently understand the start-up and shut-down of the process itself, including proper conduct in an emergency situation.
(b) Purpose of use
Excluding manipulation of the process variable of the fluid itself,
there are control valves that control the cut-off and release of fluids,
mix two fluids, separate the flow into two directions, or exchange
fluids. There are also valves that control a pressure drop from a
high-pressure system to a low-pressure system. Also, there are valves
that incorporate two or more of the above functions. Therefore, the
most appropriate control valve is chosen after determining the purposes of a particular valve.
357
(c) Response time

In responding to a change in the manipulation signal, the control
valve experiences a period of deadtime before the plug stem can overcome friction from the packing and begin to move. There is also a period of operating time needed to move the required distance. It is necessary to consider the effect of these factors on the controllability and
safety of the entire system.
(d) Specific characteristics of the process
Determine in advance the presence or absence of self-equilibrium,
the range of variation in the required flow rate, the speed of response,
etc.
(e) Fluid conditions
The various conditions of the fluid can be obtained from the process data sheet, and these become the basic conditions for the selection
of the co~trol valve. The following are the main conditions that will be
used.
(1) name of fluid
(2) components, composition
(3) flow rate
(4) pressure (at both the inlet and outlet ports of the valve)
(5) temperature
(6) viscosity
(7) density (specific gravity, molecular weight)
(8) vapor pressure
(9) degree of superheating (water vapor)
In addition to the above data relating to normal operating conditions, one should also determine their respective values at the beginning and end of operation as well as their maximum and minimum values.
(f) Fluidity, special characteristics
One should determine the presence of possible dangers regarding
the nature of the fluid, corrosiveness, or slurry.
(g) Rangeability
In the case where one control valve cannot provide the necessary
rangeability (Subsec. 6.2.4), consideration of the use of two or more
valves becomes necessary.
(h) Valve differential pressure
As the rate of the valve's differential pressure decreases relative
to the overall pressure loss of the entire system, the installed flow
characteristics shift away from the inherent flow characteristics. (Refer to Subsec. 6.2.3.) The rate of control valve pressure loss in a piping
system is a complicated problem. Although it is impossible to generalize, a value for PR between 0.3 and 0.5 is usually chosen.
358
(t) Shut-off pressure
The highest value of the differential pressure at the control valve

shut-off time is important data to be used in the choice of the actuator
and in ensuring a sufficiently strong design for each part of the control valve. Designs in which the intake pressure is set equal to the maximum shut-off pressure are numerous, but this method might result in
over-specification of the valves. Thus it is necessary to consider actual
use conditions when determining the shut-off pressure.
(;) Valve-seat leakage
Determine what quantity of valve-seat leakage can be tolerated at
the time of valve shut-off. It is also necessary to know the frequency
with which the valve shut-off condition occurs. Generally, the degree
of leakage is indicated by a percentage of the value of the rating Cv of
the control valve. ANSI B 16 104-1976 is widely used for classification
of leakage and for testing conditions.
(k) Valve operation
There are two objective in the operation of the valve; the first of
these is fail-safe operation and the second is operation according to the
valve input signal. Regarding the first objective, the movement of
valve operation is in a safe direction of the process in case the input
signal and the power supply is lost. The operation is classified as "air
failure close," "open," or "lock." Regarding the second objective, the
opening and closing direction of the valve is adjusted in according to
whether the signal input to the valve increases of decreases, but the
operation is not necessarily the same as the fail-safe operation. When
the valve closes as a result of input increase, this is called direct action.
When the valve opens as a result of input signal increase, this is called
reverse action. Figure 6.2 shows combined valve action in cases where
the actuator and plug show direct and reverse action in a globe valve
~~ ~!
(1 )
Figs.
(2)
Actuator
(3)
P lug
(4 )
Valve operation
( I )
D irect
Di rect
Direct action
(2)
( 3)
D irect
Reverse
Reverse action
Reverse
Direct
Reverse acti on
( 4)
Reverse
Reverse
Direct action
Fig. 6 2 Value action
6 .2 Choice of Control Valves
359
with a diaphragm actuator.

([) Explosion-proofing
Electrical equipment used in conjunction with the control valve
must have the explosion-proof rating required by the location in which
the valve is installed.
(m) Power supply
For pneumatic power supplies, it is important to provide clean air
with water, oil and dust removed in order for parts such as the actuator and positioner to function without failure. At the same time one
must determine the actuative pressure (air motor diaphragm area x
air pressure) and capacity so as to secure enough actuative power.
(n) Piping specifications
Determine the specifications of the piping in which the control
valve is installed. The important specifications include the diameter of
the pipe, the piping standards, the quality of the material, the type of
connecton to the piping, and so on.
6.2.2 Sizing
Sizing refers to choosing the appropriate valve by calculating the
control valve's flow coefficient rating value Cv in accordance with the
conditions of the fluid passing through the valve. There are several
methods and modified methods of calculating Cv, but the one most for
practical applications will be dealt with here. We will use the easily-managed FCI method (Recommended Voluntary Standard Formulas
for Sizing Control Valves). However, there are some process conditions
that cannot be covered by the FCI method. In those cases, one should
use the calculation method that ANSI/ISA advances as a complement
to the FGI method.*
The value of Cv: This is a number representing the capacity of the control valve. It is calculated by measuring the volume of flow in US gallons in one minute through a valve whose degree of opening is fixed,
with the differential pressure held at 1 psi at a temperature of 60F.
For example, if 10 US gallons per minute can flow at 1 psi under the
differential pressure, then the value of Cv 10.
* The FCI formulas were obtained empirically based on experiments using water.
Errors will result when these formulas are used under conditions different from the
experimental conditions. Other calculation formulas can be applied (ISA Handbook)
for conditions such as fluids exhibiting cavitation or flushing, high viscosity or
very small flow rate fluids, compressible fluids when there is a large relative pressure drop, and when valves such as ball valves or butterfly valves which have a
high pressure recovery ratio are used with compressible fluids.
360
(a) Calculating Cv by the FCI method

(1) For liquids:
(6.1)
Cv=l.17Qz/Ii-
(2) For gases:

(6.2)
LlP<O.SH
(6.3)
(3) For steam:
L1P<O.SP1
W(1 + 0.001 3 TsH)

13.5v'LlP(H + Pz)
W(1 + 0.001 3 TsH)
11.7H
(6.4)
(6.5)
In these equations, Cv is the flow coefficient of the valve, G9 is the

specific gravity of a gas in the standard condition relative to air in the
standard condition, Gz is the specific gravity of liquid at the temperature of use relative to water in the standard condition, P 1 is the inlet
pressure of the valve (kg/cm 2 abs), P 2 is the outlet pressure of the
valve (kg/cm 2abs), LJP is the valve differential pressure PcP 2 (kg/cm 2),
Q 9 is the volumetric flow rate of the gas (Nm 3 /h), Qz is the volumetric
flow rate of the liquid (m3/h), T1 is the intake temperature (K), TsH is
the superheated temperature of the steam CC), and W is the weight
flow rate (kg/h).
6.2.3 Flow characteristics
(a) There are five basic flow characteristics as given below:
(1) Quick-opening (saucer shape)
(2) Square root (quadratic characteristic, V-notch)
(3) Linear
(4) Equal-percentage (geomertic rate configuration)
(5) Hyperbolic
Three of these are commonly used: quick-opening, linear, and
equal-percentage.
(b) Inherent flow characteristics
Refer to Fig. 6.3 for the characteristics exhibited where the pressure differential of the valve is maintained at a specific value.
With the quick-open characteristic, a small change in the opening
results in a large change in flow. Therefore, it is limited to on/off use.
361
Stem position
Fig. 6 3 Inherent valve characteristies

(ISA
Hand book of control valve)
For control use, the linear, equal-percentage, or other characteristic

intermediate between these is used (many of the rotating valves apply).
(c) Choice of flow characteristics
When the control valve is installed in p1pmg, the valve pressure
differential changes in accordance with change in flow, causing the installed flow characteristic to differ from the inherent flow characteristic. The installed flow characteristic changes according to the ratio of
pressure drop across the valve to the pressure drop across the entire
system. Examples illustrating the linear and equalpercentage characteristics are shown in Fig. 6.4.
If P R is the ratio of the valve pressure drop to the pressure drop of
the entire system (including the valve), the distortion of the installed
characteristic increases as PR decreases. For a control valve, PR=0.05
is the approximate limit of tolerance. When the ratio of valve pressure
drop is small, the linear installed characteristic approaches the
Theoretical equal-percentage
inherent flow chracteristic
-Travel
(a) Linear inherent flow characteristic
-Travel
( b ) Equal-percentage actual inherent flow characteristic
Fig. 6 4 Installed flow characteristies (ISA
362
quick-open characteristic, and the equal-percent installed characteristic approaches the linear characteristic.
The choice of the valve characteristic is important in compensating for the installed flow characteristic. However, determining the
characteristics and distribution of the entire system is generally difficult, so as a yardstick for practical use in the following situations, the
equal-percentage characteristic is chosen.
(1) When the pressure distribution is unclear.
(2) When LIP is large for low flow rates and small for high flow rates.
6.2.4 Rangeability
With respect to the control valve, rangeability is the ratio of the
maximum and minimum controllable flow of the valve in practical use.
(a) Inherent rangeability
With inherent rangeability, the pressure drop across the valve is
maintained at a fixed level, thus
R=
Qmax = Cvmax
Qmtn
Cvmtn
Generally, for globe valves R = 30 to 50, and for butterfly valves,

R=20.
(b) Installed rangeability
The rangeability of a valve actually installed in a piping system is
called installed rangeability. Installed rangeability changes according
to the valve's position in the system. Figure 6.5 gives an example of a
system including resistance in series with the valve (series resistance),
and Figure 6.6 shows an example of a system including resistance in
100
~il
80
c"O
-~a
=~-
- ., 60 1-
.~ 8
JPL
PR=i1Pv/i1PL
u ~
].g
u"
40
:;~
~~ 20
oo
~~
~~t~ "'
b~
>
1).2.
.;
Cl..
iO
40
60
80
Valve lift (%]

( a ) Linear characteristic
~s
=~
~ ;:;;::" "u
p.
Lll-'\...-~ ,...,
~ t::::
100
10 0
16
~I
::@:r
80
60
~/)_
II
g~ 40
. "..
=~
ll > 20
>'0
I--'
0~10
20
40
r.IJ_
i'tJ ~
JPL
PR=.dPv/JPL
~~
~~
60
~~
.r
"'
80
,;~
100
Valve lift [%]

( b) Equal-percentage characteristic
Fig. 6 5 System consists of valve and series resistance (ISA
363
10
100
80
80
-Q
LRJ
60
"'
40
E-<
80
100
60
100
40
Valve lift (%}
Valve lift (%}
_ (Flow through by-pass resistance
,\ Flow through valve at full lift
(a} Linear characteristic
( b} Equal-percentage characteristic
20
)2
Fig. 6 6 System consists of valve and by-pass resistance
parallel with the valve (by-pass resistance). In both cases, the rangeability of the control valve drops and influences controlability, so attention should be paid to the pressure distribution of the piping system.
6.2.5 Materials
(a) Valve body materials
The most commonly-used valve body material are listed in Table
6.1.
(b) Trim materials
Trim comes in direct contact with the fluid and is a changeable
part. Materials on level higher than those used in the valve body are
used to maintain trim function over a long period of time. Materials in
standard use are SUS 304 and SUS 316, but in cases such as when high
differential pressure or high temperatures, abrasion due to slurry, or
cavitation are expected, one must choose materials that have a high degree of hardness. Two methods are mainly used, namely using a hard
material and facing with a hard material. These are described below.
(1) SUS 440C
This is made of 17 Cr martensitic steel, and has a relatively high degree of hardness due to heat treatment. It is a superior
material for resisting cavitation for use with hot water.
(2) SUS 630
This is made of precipitation-hardened stainless steel
(also known as 17-4 PH). It resists corrosion to the same extent as
SUS 304, and is often used because of its resistance against to wear.
(3) Stellite
As an alloy composed mainly of cobalt, this material is
excellent because of its strength at high temperatures and its resistance to corrosion. It is widely used as a hard facing material.
364
e:
......
"'""
i1
<::>
0\
SCPH 21
SCPH 61
Cr-Mo steel
SCS 13 A
SCS 14A
SCPH 2
SCPL 1
Cast steel Carbon steel
Stainless steel casting
FC 20, FC 25
JIS
Cast iron
Material
-5 to+540
-45 to+350
-5 to+593
-5 to+650
-196to+800
A 217-WC 6
A 217-C 5
A 351-CF 8
A351-CF 8 M
0 to+250
(oC)
Operating temperature
A 216-WCB
A352-LCB
Compatible material of ASTM
Code
Table 6. 1 Value body materials (]IS B 8243-1981)
Corrosion resistance, cryogenic

and high temperature
High temperature and

high pressure
General
Low temperature
General
Main use
(c) Seal material
Control valve seal materials are used in the gasket, where sealing
is employed in the stationary section, and in the gland packing, where
sealing is used in the sliding shaft section of the valve. In both cases,
the type of fluid, temperature, and pressure must be considered in
choosing the appropriate material.
(1) Gasket
Asbestos and Teflon materials are widely used. Metal
gaskets are used for high-temperature and high-pressure applications.
Asbestos is heat-resistant and is an excellent material because of its
chemical stability. It is the most widely used of gasket materials, but
there is a world-wide trend to regulate its use because of the harmful
effect of asbestos fibers on the human body. As a substituted, packing
made of pure black lead (Grafoil) is employed because of its usefulness at high temperatures.
(2) Gland packing
Most control valves, except the Saunders valve,
use gland packing. Gland packing must keep the fluid from leaking out
from the sliding section of the valve shaft. In must also minimize slide
friction and minimize hysteresis. Because there is a conflict between
these two functions, packing material must be carefully chosen. Figure
6. 7 shows an example of gland packing installation.
(i) Teflon V packing: Of the various types of packing, this sealing
exhibits the least sliding resistance and excellent sealing properties,
and also shows excellent resistance to chemicals. It is widely used as
gland packing for control valves, but may not be used at high temperatures.
(ii) Asbestos coil packing: Asbestos coil packing is classified roughly into
Teflon V - rings
Asbestos yarn with packing lubricator assembly
Fig. 6 7 Gland packing
366
types impregnated with Teflon, which exhibit relatively little sliding

resistance and have excellent sealing properties, and types with a special alloy additive, which are used in high-temperature/high-pressure
applications. Both types require use of a lubricant grease, and application of a lubricant is usual for high-temperature use.
(d) Ceramics
Ceramics have been used recently as materials in control valves
for their resistance to corrosion and wear as well as their resistance to
heat. Ceramics are superior to metals in these areas. On the negative
side, they are still rather expensive and are mechanically weak and susceptible to heat shock. Manufacture is problematic when complicated
configurations are required, and it is difficult to use them jointly with
metals. Thus there are many points to be considered, but nevertheless
one can expect to see new uses for ceramics in the future.
6.3 Control Valve Bodies
Because the valve body comes into direct contact with the fluid, it
is necessary to construct the valve body to fit the process conditions
and purpose of use. In this section, we will show the various types of
control valve configuration and their characteristics, the valve body ratings, and methods of connecting the valve to the piping.
6.3.1 Characteristics of various types of valves
Control valves are classified as linear types or rotary types according to the movement of the main valve shaft. The various relevant
valve body types are listed in Table 6.2.
Table 6. 2 Motion types and valve bodies
Motion types of valve stem
Types valve bodies
Linear motion type
Globe valve, Angle valve, Three way valve,

Diaphragm valve, Gate valve
Rotary motion type
Butterfly valve, Ball valve,

Eccentric rotating plug valve
(a) Globe valves

The globe valve has a spherical valve body, and the center line between the intake and outlet ports is a straight line. The fluid flow
forms an S-shape within the valve. This is called a spherical valve.
According to the construction, globe valves are classified as single-seated, double-seated, or cage valves. These are the most commonly used among the control valves. The construction of this group
367
( a ) Simgle eated valve
( b ) Double seated valve
( c) Cage valve
Fig. 6 8 Construction of globe va lves
of valves is shown in Fig. 6.8.

Because of the shape of the globe valve, it is possible to increase
the pressure resistance of the valve body. The valve can be used at either high or low temperatures, depending on the quality of material
chosen and the shape of the bonnet. By choice of plug shape, the flow
can be given any desired characteristics. The rangeability differs according to the shape of the plug, but is typically limited to a value between 30 : 1 and 50 : 1
(1) Single- seated valves
These are the most basic type of control
valve; they have excellent valve shut-off ability. The quantity of leakage at the valve seat is rated with a Cv value of less than 0.01%, and
with seat-face treatment, this can be reduce to less than 0.0001%. if a
soft seat is used, it is possible to obtain tight shut-off. Because the full
differential pressure is received by one plug, the required actuative
power is greater than for a double- seated valve under the same conditions of use. It is generally suitable to use a control valve with a small
intake diameter. In addition, a cut-off valve or an on-off valve may be
used to obtain excellent shut-off.
(2) Double-seated valves
These control valves have two ports, and
quantity of leakage at the seat is rated at a Cv value of less than about
0.5%. In this respect, they are inferior to the single-seated valves.
However, the required actuative power is much smaller than for the
single-seated valve under the same conditions of use, because the power imbalance is eliminated by the upper and lower plugs. Generally, a
control valve with at least a medium-size intake diameter is desirable.
368
(3) Cage valves

Cage valves are equipped with a plug that slides
within a cage. The cage guides the plug directly, and because the guiding area is large, vibration of the plug is suppressed and movement is
stabilized when there is a high differential pressure. On the other
hand, this valve is not suitable for use with fluids that contain admixtures. Regarding the inherent flow characteristics, this valve can
be specified with a variety of shapes of the openings or windows in the
cylindrical cage. There are both balanced and unbalanced shapes; the
unbalanced shape is most frequently a single-seated valve, and the balanced shape is usually a doble-seated valve.
(b) Angle valves
The center line running from the intake of the valve body to and
the outlet forms a right angle, and the direction of the flow also
changes by 90 as it passes through the valve. Usually, the fluid moves
from the side, and then downwards. It is used to ensure that slurries
and viscous fluids flow easily, and is also employed when a drain may
not be used on the outlet side of the valve. The inner construction is
roughly the same as the globe valve, and there are both single-seated
and cage configurations (Fig. 6.9 (a)).
(c) Three-way valves
This valve has three directional openings. There is both a mixing
type that is used to mix two different fluids, and a diverting type that
channels one fluid into two directions. As the purpose of these valves
is to mix or divide fluid flows, they normally have linear characteristics and are constructed such that the total amount of flow is constant
regardless of the degree of opening (Figs. 6.9 (b) and (c)).
(d) Diaphragm valves (Saunders valves)
The shape of these valves is such that there is a weir in the center
of the valve body and the course of the fluid is pinched off by this
weir, which is opened or closed by means of a diaphragm. The course
of the fluid is simple, so the lining of the interior surface of the body
is simple and easily made resistant to corrosion. These valves are also
appropriate for use with slurries and viscous fluids. The inherent flow
characteristic is nearly linear. Control is not good when the opening is
small and the rangeability is about 10: 1 or 15: 1(Fig. 6.9 (d)).
(e) Gate valves
The opening and closing of these valves is performed by a vertically aligned disk that partitions the flow of fluid. Because diameter of
the valve inlet is the same as the inner diameter of the pipe when the
valve is fully opened, pressure loss is slight. Also, it has a high
shut-off function, so it is often used as a cut-off valve.
When used as a control valve, a seat ring with a V-shaped orifice
is chosen, and desings to improve the inherent characteristic and range6.3 Control Valve Bodies
369
( a ) Angle valve
(d) Diaphragm valve
( g) Ball va lve
(c) Mixing type

three-way valve
( b ) Divening type
three-way valve
( f) Buuerfly valve
(e) Gate valve
( h ) Characterized ball valve
( i ) Eccentric rotating plug valve
Fig. 6 9 Construction of vasious valve
370
ability have been developed (Fig. 6.9 (e)).

(f) Butterfly valves
In these valves, a disk rotates inside the valve body to open and
close the valve. This construction is relatively inexpensive, and these
valves are appropriate in cases where the diameter of the opening is
large and when the differential pressure is low. According to the pressure of the fluid, maximum torque occurs when the valve is opend to
about 70 and when this level is exceeded, the torque decreases abruptly. Thus when this valve is employed it is most commonly used at
oo to 60. The inherent charcteristic of the valve is close to the
equal-percentage characteristic. When a oo to 60 opening is used, rangeability is shout 20: 1 or 30: 1 (Fig. 6.9 (f)). Often, the butterfly valve
dose not have a seat, and at full shut-off the amount of leakage is
about 1% to 3%. To decrease the amount of leakage, various methods
of sealing using soft materials such as rubber and Teflon 411 have been
devised.
(g) Ball valves
In these valves, the valve body contains a ball which has an opening in it. The ball is rotated to open and close the valve. When a soft
seat is used, the shut-off function is extremely good and it is possible
to achieve tight shut-off. A metal seat has superior heat resistance,
but the shut-off function is inferior to a soft seat. This type of valve is
often used as a cut-off valve or as an on-off valve because of its good
shut-off and low resistance when completely open (Fig. 6.9 (g)). Inherent characteristic improvements have been devised for when the ball
valve is used for control. Such improvements include using a partial
sphere with a V-shaped notch. With this configuration, the inherent
characterisitic of the valve is close to equal-percentage and the rangeability is large (100: 1 to 300: 1).
When used as a control valve, the ball valve is similar to the globe
valve, but is used under a gentle range of temperature and pressure
conditions. It is particularly appropriate for use with slurries, including fiber slurries such as pulp liquors, because there is no obstruction
in the fluid course and there is shear between the disk and the seat
when the valve is shut off (Fig. 6.9 (h)).
(h) Eccentric rotating plug valves
The plug of this valve opens and closes the valve as it rotates
about an eccentric axis. Structurally, it is similar to the ball valve
when used for control purposes as listed above, but its use of an eccentric axis is its distinctive feature. Because of the eccentric axis, the
seat and plug do not come into contact except when the valve is shut
off.
Improvements in the shut-off function and reduction of torque are
371
being developed. The inherent characteristic of the valve is close to linear, so a cam positioner is sometimes used to produce an equal-percentage characteristic. The rangeability is relatively large at about 100: 1
(Fig. 6.9 (i)).
(z) High-temperature/high-pressure valves
Generally, in the design of valves, high temperature is defined as
350oC or higher, and high pressure is defined a pressures above JIS 63
kgjcm 2 class (ANSI class 900). Under these conditions, cavitation and
noise become problematic, and proper choice of valve structure and materials is important. Countermeasures for cavitation and noise will be
touched on here.
(1) Cavitation countermeasures
One countermeasure is to choose materials to prevent damage from cavitation. Hard materials such as Stellite facing or SUS 440C are used. Another countermeasure is to employ a valve structure that make it difficult for cavitaion to occur.
There are many cage valves suitable for this purpose from a structural
perspective.
(2) Noise countermeasures
Noise produced by a cotrol valve can be
classifided into the following three types.
(1) Noise arising from mechanical vibration of the structural parts.
(2) Noise from the flow of liquid.
(3) Noise from the flow of gas.
To eliminate the causes of the first type of noise, manufacturers
study the natural vibration of the parts (mainly the plug), and then
design the parts so as to shift the resonance point. The use of a cage
valve with a sturdy guide lessens the likelihood that such a phenomenon will occur.
In the second type of noise, caused by fluid flow, almost all the
noise results from cavitation. It is necessary to consider the fluid conditions and the structure of the valve to avoid cavitation.
The third type of noise, from gas flow, is the most difficult to predict, and dependig on the situation, the noise generated can be very
great. Noise is produced as gas passes through a contraction.
High-speed flows create turbulence in the gas. Also, a shearing force
can develop when a high-speed flow and a neighboring low-speed flow
mix, producing shock waves. However, the mechanism of this effect is
complex and difficult to analyze.
One way to lessen then noise produced is to first break down the
high-speed flows and average the speed distribution, at the same time
shifting the noise into a high frequency region which is more easily decreased. The pressure differential that occurs in the control valve is
dispersed. See Fig. 6.10 for an example of this effect.
372
An example of low noise valve consisting

of a mult-slotted cage valve and a d iffuser
installed down tream
Fig. 6 10 Low noise valve for gas flow (Fisher control Inc.)
(j) Cryogenic valves
These are for use with very low temperature fluids such as LNG,
liquid nitrogen, and liquid oxygen. Characteristics that differ from other valves are note below.
Austenite stainless steel is used. Copper alloy or
(1) Materials
aluminum may also be used. Teflon and Kel-F are used as seal materials, but care must be taken with Teflon, because of its high heat
contraction rate.
The bonnet must be sufficiently long to maintain
(2) Construction
the gland packing at ooc or above. The gap between the valve shaft
and the bore is made extremely small so that the upper part of the bonnet is filled with gas. The valve body must be designed to have as
small a heat capacity as possible, and for this purpose a welded connection between the valve and the piping is desirable.
(3) The interior of the valve should be carefull washed and all traces of
water completely remove.
6.3.2 Rating
The valve body is treated as a pressure container. Consequently,
given safety considerations, the allowable amount of stress at actual
use temperatures is determined by the quality of the materials. However, making strength calculations and valve body designs from the pres6.3 Control Valve Bodies
373
sure of the fluid each time becomes complicated and uneconomical.

Moreover, compatibility with available valves is lost. For this reason,
valves are given ANSI standard pressure ratings. The ratings are divided into classes ranging from 150 to 4 500. These are standardized (ANSI B 16.34) and are called basic rating pressures. Similarly, there are
]IS ratings, such as temperature/pressure rating of pipe flanges (]IS B
2201). These are classified into nominal pressures from 2 K to 63 K.
6.3.3 Connection to piping

Among the various ways of installing control valves on piping and
machinery are the flange method, wafer method (flange-less), welding
method, and screw method. The most commonly used connection methods are listed in Table 6.3.
6.4 Control Valve Actuators
6.4.1 Conditions under which an actuator should be installed
The actuator opens and closes the valve. It receives the control signal from the controller and generates the driving power to open or
close the valve with certainty. An actuator must provide the following.
(1) A driving power that is sufficient to overcome the friction in
the valve shaft and thrust generated in the valve by the liquid
pressure, and maintain the normal position of the valve shaft.
(2) A rigidity that overcomes the change in thrust on the valve
shaft between full shut-off and full opening and maintains the stability of the valve shaft.
(3) A responsiveness that allows it to follow changes in the control
signal swiftly and precisely.
6.4.2 Power sources
Actuators operate on hydraulic, pneumatic and electric power.
Among these, pneumatic actuators are the most widely used for the following reasons.
(1) Compared with other types, the construction is simple and breakdowns are rare. Also, a large actuating force can be obtained.
(2) It possesses intrinsic anti-explosive properties and is easily serviced.
(3) It can be installed to operate in conjunction with long-distance
electrical transmission signals by joint use with a current-to-pneumatic positioner, current-to-pneumatic converter, or electromagnetic valve.
(4) It is less expensive than other types.
However, air is compressible, so its use in high-response and
374
Table 6. 3 Types of piping connections and use

Type
Descripti on
Flanged
end
Raised face
type
Flat fac e
type
Tongue and
groove
type
~
~
Most widely served.

For below the cia 600 , this
type i genera ll y used .
ed for ca t iron va lve .
For low pressure such as ]I
10 K and less than the class 125 .
. .,6:.~,
Ring joint
type
Wa fer type
Welding
end
Ue
~
~
T his type used cases of hazardous

fl uids and high pressure.
Usuall y, va lve mounts on groove
side.
In many cases, this type is used
for high pressure uch as the
class 900 or higher.
T his type is often u ed for rotary
type valves such as butterfl y
va lves.
Socket
welding
end type
For high temperature and ihgh

pressure.
Used for 1!/,B or up to 2 B.
Butt welding
end type
For high temperature and high

pressure.
Used for more than 2 B.
Threaded
end type
T hig type used fo r non- hazardous

and low pressure fluid .
Adopted for l Y.B or up to 2B.
375
high-accuracy applications is limited. Thus the use of servo-motor and

current-to-hydraulic models continues to increase, despite cost considerations.
6.4.3 Types of actuators and their characteristics
There are many types of actuators. These are classified according
to type of valve and input signal in Table 6.4. In recent years, the rotary valve has come into wide use. There are various actuators suitable
for use with this type of valve.
Table 6. 4 Classification of actuator
.
.
. =rBDiaphragm actuator
ylinder actuator
Pneumatically-operated--LLmear moti_on
actuator
Rotary motion
Air motor actuator
Electro-mechanical
Electro-mechanical
. ___r-Linear motion----,____! actuator
actuator
~Rotary motion----lLServo actuator
(electronic type)
Electro-hydraulic
___r-Linear motion----,____! actuator
Electro-hydraulic
~Rotary motion----lLServo actuator
actuator
(hydraulic type)
Hydraulic actuator
____r-Linear moti?n~ylinder actuator
~Rotary motwn---J
Pneumatic-hydraulic
__r--Linear motion~Servo actuator
~Rotary motion---J
(hydraulic type)
actuator
(Note)
Sometimes, spring type is called single-action type as well as

springless type is called double-action type.
Representative examples of the construction of actuators and their

characteristics are listed in Table 6.5 and explained below.
(a) Pneumatically-driven type
(1) Diaphragm actuators
These are the most widely used actuators.
They have simple construction and excellent movement accuracy and
response. They are also highly dependable. Diaphragm actuators are
divided into linear types and rotary types. Both types can be installed
to provide direct action and reverse action. The constructions of these
types of diaphragm actuators are shown in Figs. 6.11 and 6.12. In the
linear diaphragm actuator, direct action occurs when the actuator
shaft is pushed out as a result of increased air pressure; reverse action
results when the actuator shaft is drawn in.
The most important parts of this actuator are diaphragm and the
spring. Neoprene rubber and ethylene-propylene rubber reinforced
with fabric, such as nylon, are used in the diaphragm to heighten pressure resistant properties. A spring with stability and a high degree of
376
::j
<..,.
;;l
<:>
1::1
;::
--"
"'
~
<:
'I
::t
o!:o..
0'1
actuator
Electro -mechanical
Pneumaticall y
operated
actuator
Discription of
operation
Electro-mechanical
actuator
Springless type
Electro-mechanica l
actuator
Springless type
Linear motion
Diaphragm
actuator
Spring type
Rotary motion
Rotary motion
(Combination of
linear motion
valve)
Air motor actuator

Springless type
Cyl inder actuator

Springless type
Diaphragm
actuator
Spring type
Type
Rotary motion
(combination of
linear motion
valve)
Linear motion
Output stem
motion
gr
1f
Schematic diagram
Small
Large
Small
Large
Large
Small
Output
Small
Large
go
60"
or
Large
Large
Small
Stroke
AC
100
200
v
v
AC
200 v
220 v
440 v
Pneumatic
). 2 to4 . 0
kg/ em'
Pneumatic
or gas
4 to 70
kg/ em'
Pneumatic
3to 7
kg/em'
Pneumatic
). 2 to 4. 0
kg/ em'
Power
source
Size
Small
Middle
Small
Large
Large
Large
Table 6 . 5 Schematic diagrams of actuator and its features
Required
(built in )
Required
Required
Required
Required
Required
or not
required
Positioner
Locking up the
stem position
Locking up the
stem position
Spring Controls
Locking up the
stem position
No control
Spring controls
Action on
power fail
"'
~
~
:i
...::s~
l:l
'">:!
;;
0.,
actuator
Hydraulic
Electro -hydradic
actuator
Electro-mechanical
actuator
Discription of
operation
Rotary motion
Cylinder actua tor

Spring type
Cylinder actuator
Springless type
Electro- h)draulic
ac tuator
Spring type
Linea r motion
Linear motion
Elect ro- hyd raulic

actuator
Spr ingless type
Servo actuator
(electronic type)
Springless type
Servo actuator
(electronic type)
Springless type
Type
Linea r motion
Rotar)' motion
Linear motion
Output stem
motion
~ ()
~-
~I
---
c;9
---
~I
Schematic diagram
Large
Large
Small
Ali<klle
Small
Small
Out put
60'
or
90'
Large
Small
Small
60'
or
90'
Small
Stroke
Hydra ulic
10 to210
kg/ em'
Hydraulic
10 to 2!0
kg/ em'
AC
100 v
200 v
200
AC
lOOV
AC
100
AC
100
Power
source
Middle
small
Middle
Middle
Small
Small
Size
Required
Required
Required
Required
Built in
Built in
Posit ioner
No control
No control
Spri ng controls
No control
(fixed )
Lock ing up the

stem posi ti on
Locking up the
stem position
Action on
power fail
IkE -- (> D---'>1~~

Diaphragm
fJ.~~~~;;~
Direct act ion type ( DA )
Reverse action type (RA )
Fig. 6 11
Linear motion lype diaphragm actuator
Fig. 6 . 12
Rotary motion type d ia phragm actuator
accuracy is needed.
The output power of the actuator is determined by the effective
area of the diaphragm and the operative pressure as well as by the
spring. The range of the spring is usually 0.2 kg/cm 2 G to 1.0 kg/cm 2 G
or 0.4 kg/cm 2G to 2.0 kg/cm 2G, but lately actuators with higher output
power in special ranges such as 0.8 kgjcm 2G to 2.4 kg/cm 2G, and
379
1.6kg/cm 2G to 3.5 kg/cm 2G are being produced. Actuators with outer

diameters (D) from about 150 mm to 650 mm are available.
When the signal pressure (operative pneumatic pressure) enters
the diaphragm case, the air pressure causes the diaphragm to move
and compresses the spring. As a result the output power shaft is held
in place by a balance between the pneumatic pressure and the opposing
pressure of the spring.
(2) Cylinder actuators
These actuators have poorer response than
the diaphragm actuators. However, they have many advantages such
as being usable with high operating pressures and achieving high output power with a small actuator. They are used particularly often with
rotary valves, on-off valves or large-port valves. The output power
shaft is either a linear motion type which moves up and down, or a rotating type. Both types are available as spring or springless models.
An example of a spring type is shown in Fig. 6.13. The operating pressure of cylinder actuators is generally above 4 kg/cm 2G, and a positioner is required to achieve proportional movement.
The output power of the springless type is far greater than the
spring type. It is also more responsive to large strokes. However, if
loss of air occurs it looses orientation. Thus, when used with cut-off
valves and so on, a volume tank and lock valve should be provided, and
Linear motion type
Rotary motion type
Fig. 6 13 Cylinder actuator (Direct action type)
380
the valve should operate in the safe direction or should be locked in

the present position.
(b) Electric actuators
The merits of the electric actuator include procurement of a power supply is easy, a strong operating force is obtained, and there is no
lag in signal transmission. On the other hand, their construction is
more complex, and explosion-proofing is need in environments where
explosion can occur. They are also more expensive than the pneumatic
types. Making the most of their merits, electric actuators are widely
used for valves with large ports such as oil tank source valves and water cut-off valves for waterworks and sewer systems, or when no pneumatic power source is available. A current-to-current power source
is built in, and devices that use a servo-motor to improve the control
accuracy are being developed. The range of uses for electric actuators
is increasing.
(1) Electric motor type
These actuators are further divided according
to their construction into those for use with valves having small-bore
intakes and those for use with large-bore intake valves. The smaller actuators are driven by commercial 100V AC power supplies. The rotation of the motor is transmitted to the output shaft through gears. As
with globe valves, linear valves use an added linking mechanism to convert the rotational motion of the output shaft into a linear motion. An
exterior view of a small motor-drive valve is shown in Fig. 6.14. There
is also a built-in current-to-current positioner type that operates by a
4 rnA to 20 rnA DC input signal. The larger models are driven by a
200V AC or 400V AC three-phase power supply. An example of a large
motor-drive valve is shown in Fig. 6.15.
Fig. 6 14 Small type motor-drive valve
381
Fig. 6 15
Large type motor-drive valve
Zero
Input signal
4 to 20 mA DC
~Power
supply
~100/IIOV AC
Q) Feedback
Output
ontrol
G) Amplifier
Fig. 6 16
Input/ Comparator
Construction of servomotor type
This type is used with relatively small bore

(2) Servo- actuators
valves. Although it does not have a very large output, it is highly accurate and has good response as it most useful merit. Figure 6.16 shows
an example of its construction.
A DC input signal of 4 rnA to 20 rnA drives the motor in a direction such that the difference between the input signal and the degree
of valve opening is eliminated. The rotation of the motor moves the
trapezoidal thread of the gear train up and down. This action provides
feedback to a differential transformer that detects the degree of valve
382
opening. The result is that the output shaft position is proportional to

the input signal. In cases where rotating motion is required, rotation is
accomplished by a worm-to-worm gear assembly.
(c) Electro-hydraulic actuators
The electro-hydraulic type combines the hydraulic pump and hydraulic equipment into a single unit. It has the following merits.
(1) It has high rigidity and operating accuracy due to use of a
non-compressible fluid.
(2) The hydraulic part is completely packaged, and the electric power supply is the only input, thus it is easily handled.
On the other hand, construction is complex, and as with the electric motor types, an explosion-proof construction is required in environments where explosion is a possibility. A linear motion spring
model is shown in Fig. 6.17, which illustrates both the hydraulic pump
and each hydraulic device.
Fig. 6 17
Direct action spring type (Single-motion, reverse-action type)
Each hydraulic circuit is connected inside the container, which is

constructed to prevent oil from leaking out. In this example, a nozzle
flapper mechanism is used with the positioner, and excellent response
is obtained.
(d) Hydraulic actuators
This type is used at high hydraulic fluid pressures of 10 kgfcm 2 G
to 210 kg/cm 2G. Consequently, a large output can be obtained from a
relatively small unit. However, unlike the electro-hydraulic actuators,
a hydraulic power source and high-pressure piping are necessary. The
piping circuit is constructed to prevent oil from leaking out. After
movement, the hydraulic fluid is returned to the hydraulic source.
383
6.5 Positioners and Accessories

6.5.1 Positioner functions
Positioners are used to allow the final control element to fully perform its function in the control loop. They maintain a precise equality
between the signal from the cotroller and the position of the final control element drive shaft, resisting external forces. As they are installed in a direct control valve, they can be influenced by vibration in the
piping, temperature, environmental gases, etc. Thus they must have a
solid construction. They are used in the following cases.
(1) The differential pressure of the valve is great and the effects of
change in fluid pressure are strong.
(2) Friction with the gland packing is great and hysteresis and linearity become bad.
(3) The controller signal and actuator signal differ.
(4) The distance between the controller and control valve is large
and response is slow (for pneumatic signals).
(5) The intake diameter of the control valve exceeds 100 mm, and
the response is slow due to heavy load.
(6) Two or more control valve operate in a split range in reponse to
one signal.
(7) The operation signal is amplified because a large operating power is required (for pneumatic signals).
(8) The pressure of the fluid has a strong effect, as with butterfly
valves.
(9) As with Saunders valves, it is necessary to improve the inherent
characteristics of the valve to obtain the required characteristics.
6.5.2 Pneumatic pressure positioners
These positioners move upon receiving a pneumatic signal such as
from a pneumatic controller. They are used particularly in high-temperature applications and as double-acting positioners in combination
with double-acting cylinders that required a large operating pressure.
The operating principle of pneumatic positioners is illustrated in Fig.
6.18
6.5.3 Current-to-pneumatic positioners
Most currently-used control loops consist of combinations of electronic controllers and pneumatic control valves. Current-to-pneumatic
positioners are used to receive the electric current signal and move the
pneumatic final control element. A torque motor or force motor that
employs a permanent magnet and coil is used to receive the signal in
384
::'
Feedback bellows
Input
relay
Controller output current
+ -
Output air pressure

(0.2 kg/em ' to 1.0 kgj cm2)
Fig. 6 19 Operational diagram of current-to-pneumatic positioner
6 .5 Positioners and Accessories
385
place of the metal bellows used to receive a pneumatic signal. Fig. 6.19
illustrates the operating principle of a force-balance positioner with a
torque motor. This positioner is equipped with a manual- automatic
mode switch, which allows manual operation in case of failure in the
signal channel.
6.5.4 Current-to-current positioners
These are used to improve the control of motor-drive valves. The
operating principle is illustrated in Fig. 6.20. The positioner receives
feedback signal Xt that is proportional to the degree of valve opening
from the potentiometer that works with the motor-drive valve. It
compares Xt and the valve control signal Xs with the two independent
comparators A1 and A 2 The valve drive motor then rotates one way or
the other according to the # 1 and # 2 relay con tact signals that are
driven by A1 and A2 The positioner controls the system so that x, and
x 5 coincide within the limits of the setpoint neutral zone.
------
Current-to-current
positioner
Setpoint
"'
Stem
position
signal
r - - -<>--+--<~-='=---t
I
I
Stem
! l___.
'
__
_L ] i on
L - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - __ J
Fig. 6 20 Operational diagram of cu rrent-to-current positioner
6 .5 .5 Accessories
Aside from positioners, there are accessories such as volume boosters and speed controllers that improve the controllability of a control
valve. The main types of accessories are described here.
(a) Volume boosters
These act to increase the quantity of air supplied and improve response. When the capacity of the control valve actuator is large, and
when the signal piping is long, the volume booster is used to increase
386
Chap.6
Final Control Elements
the operating speed of the control valve. However, because this alone
cannot improve the static qualities of the control valve, volume boosters are often used together with a positioner.
(b) speed controllers
These constrict the flow of air and so control the speed of movement. The structure comprises a check valve and a variable restriction,
and works by constricting the flow in one direction only.
(c) Lock valves
Lock valves are used to stop or divert the pneumatic circuit immediately in front of the actuator, thus forcing the control valve to move
in the safe direction in the event of air supply failure. When air pressure is restored, the lock valve automatically returns the system to its
original state.
(d) Electro-magnetic valves
These are used when it is necessary to switch the pneumatic circuit by means of an electrical signal. They are often used for the
switching of on-off valves.
In addition to the accessories listed above, there are others such as
air sets, pneumatic switching valves, volume tanks, and limit switches
that are used as needed. Examples of the various types of accessories
combined with actuators are shown in Fig. 6.21
..____
Air
supply
Input
signal
Air
r----v-- supply
.___ _ _ Input
signal
Fig. 6 21
qrt
roouollo
Air
.-----.,~- sa pply
L...---- - lnput
sigual
Positioner
Examples for using of control va lve accesseries
6.5 Positioners and Accessories
387
6.6 Self-powered Valves

Mainstream process instrumentation systems are centralized, with
the information center located in a central control room. Final control
elements perform their functions according to commands from this control room. However, to reduce instrumentation costs, construction
time, and maintenance costs, self-powered valves are still used in installations where supplementary power is unnecessary. The self-powered valve receives its energy directly from the controlled system, so
the amount of control is limited. However, the structure is simple and
economical, and when used appropriately the results are very good.
6 .6.1 Pressure-regulating valves
These are generally called pressure-reducing valves, and are normally used to maintain stable secondary pressure without affecting the
primary pressure. There are both direct-action types and pilot types
(Fig. 6.22). In the direct-action type, a diaphragm lies between a
spring and the opposing secondary pressure. Because the valve opens
and closes according to the differential pressure, these valves are used
with liquids or smallbore applications.
(a) Direct-action type
Fig. 6 22
(b) Pilot type
Pressure regulating valve
In the pilot type valve, the pilot valve, which is actuated by a

diaphragm, adjusts the primary pressure applied to a piston. Thus although the offset is small compared to a direct-action valve, this type
of valve has the demerit of not operating when the differential pressure falls below a certain level.
In addition to these two types, there is also a fixed differential
pressure control valve that maintains a fixed differential pressure be388
tween two flows. These valves are used for the atomizing of fuels.
6.6.2 Temperature control valves

This is a self-powered valve that controls the flow of steam or
cold water to maintain a fixed process temperature. Although there
are many ways of obtaining operating power from the heat, such as
the steam pressure method, liquid expansion method, and gas expansion method, the steam pressure method is the most widely used.
In the vapor pressure method, the valve is controlled by the generated vapor pressure of the liquid that is contained in the temperature
sensing section. The temperature is set by adjusting a spring in the
same manner as the pressure- reducing valve.
As with the pressure-reducing valve, there are both direct-action
and pilot types. The liquid in the temperature sensing section is ethylene chloride, acetone, alcohol, or some other liquid, chosen according
to the temperature range. It is necessary to confirm that problems will
not arise if the liquid leaks into the process fluid.
6.6.3 Flow control valves
The flow control valve combines a constriction such as an orifice
with a fixed differential pressure valve such that the differential pressure created as the fluid that passes through the constriciton is maintained at a fixed level.
The flow rate is set by turning a handle which varies the aperture
of the constriction. These valves are used mainly for heavy oil and
light oil burners.
6.6.4 Float valves
These widely used valves are operated by the buoyancy of a float
resting on the surface of the liquid. The buoyancy is amplified and
transmitted to the valve by levers (Fig. 6.23). The size of the float and
the leverage of the mechanism must be adjusted according to the valve
size and the differential pressure.
Fig. 6 23
6.6 Self- powered Valves
Float valve
389
PRACTICE QUESTIONS
6.1 Choose the most appropriate valve size for the following application from the three
possibilities listed below.
This valve regulates the flow of water. The maximum flow rate is 55 m3/h, at which
time the pressure conditions are P 1 =7 kgjcm'G and P, = 5 kg/cm'G. The inherent characteristic of the valve is equal-percentage.
1. 50 A rating c.= 48
2. 65A rating Cv = 72
3. 80A rating Cv = 100
6.2 Determine the required value for the following conditions.
Fluid: air
Flow rate: 2 000 Nm 3/h
Pressure: P1 =5 kgjcm'G, P, =3 kgjcm'G
Temperature: 20oC
ANSWERS TO QUESTIOSNS
6.1 The required value of Cv is 45.5 (from Ep. (6.1)). Normally, the control valve size is
selected such that the degree of valve opening at maximum flow is 80% or less, allowing for various types of error and ease of use as a control valve. here, from Fig. 6.5,
taking the vertical axis as flow rate =C., the approximate valve opening is obtained.
Cv = 95(%), valve opening = 95 to 100(%)
50A
Cv = 63(%), valve opening = 80 to 90(%)
65A
Cv = 45.5(%), valve opening = 75(%) (approx.)
BOA
Accordingly, in this example, item 3, 80A, the valve for which Cv = 100(%) is selected.
6.2 As L1P<0.5P1 the calculation is performed using Eq. (6.2).
2 000
c.= ---z73
1X(273+20)
j 2(6.033+4.033)
27 9
REFERENCES
1) JIP Standard: JIP-7B-64-82 (in Japanese)
2) J. W. Hutchison: ISA handbook of contnal valve, 2nd Editon, ISA (1976)
3) T. Ishii: Process control Instrumentation Systems, Denki Shoin (1973) (in Japanese)
4) C. S. Beard: Final control elements, Vo1.2, ISA (1973)
390
Chapter 7
SYSTEM ENGINEERING
In a wide range of fields today, the words "system" and "engineering" have become so popular that they are now felt by many to be overly used and ambiguous terms. In particular, on hearing the expression
"system engineering," it would be difficult even to guess the actual
work or process in question. This lack of clarity, however, can be
somewhat alleviated if we take a look at the basic meanings of these
two words. The word "system" originates from the Greek word meaning "to place together," and came to mean the coexistence of a variety
of diverse things (e.g., men, objects, money, information) utilized in a
holistic way so as to achieve a new function independent of the composite parts. In addition, the word "engineering," apart from the modern
connotations associated with Mechanical Engineering, Civil Engineering, and the like, also has the fundamental meaning of careful planning
and efficient management.
In the light of the above, we define instrumentation engineering
as design technology applied to the systemization of instrumentation facilities in order to achieve the objectives of a certain industrial plant or
process. In this regard, if we view the instrumentation manufacturer
as a specialist providing measurement and control technology and the
user (customer) and plant builder as the source of process and plant
design know-how, all objectives and functions must be clearly presented so that all parties concerned can together complete the project in
question. This chapter describes, from the viewpoint of the instrumentation manufacturer, the elements of instrumentation engineering
from the original planning stage up to installation and adjustment.
To begin with, in relation to basic system engineering, we describe an outline for plant construction as well as how project tasks
should be divided amongst the user, plant builder and instrumentation
manufacturer. We then explain the work involved at the "instrumentation system design" stage in which specification clarification meetings
Chap. 7
System Engineering
391
are held with the customer in preparation of actual instrumentation

system construction. Also described is the work involved from this
stage up to final test runs and turnover from the viewpoint of basic
design and engineering. Items related to common technology such as
instrumentation power sources and safety measures are covered as
well.
Following the above is a description of "human engineering" factors as related to the design of a control room (operations room),
which is continuing to increase in importance as the focal point of
plant operations. Next, the planning and design of instrumentation
work such as related to instrumentation installation, piping and wiring, and system startup (pre-commissioning) are described. Finally,
the methodology behind a "design review" is introduced as an effective
means of insuring product quality.
7.1
System Engineering Basics
7.1.1 Plant construction overview

In the process of constructing a plant, not only is the cooperation
of the end user necessary but that of specialists as well, such as those
in the chemical, mechanical, metal, civil, electrical and instrumentation
engineering fields. Moreover, in addition to the primary engineering
problems involved with building a plant, extensive consideration must
also be given to other influential factors such as local environmental
conditions and required manpower and living provisions during plant
construction. A flow chart for plant construction is shown in Fig. 7.1.
(a) Feasibility study (specifications evaluation)
Cost studies: All factors contributing to the overall cost of the project
are investigated, such as product marketability, raw material acquisition, site conditions, transport facilities, general construction costs,
general production and sales costs, manpower costs and money flow
planning.
Technological studies: Investigations are performed as to whether or not
current technology may be used, and if not, whether development of
new technology is feasible. In addition, other technologically-related
factors such as licensing and the advantages and disadvantages of particular processes are reviewed.
Environmental studies: From the viewpoint of public nuisance, the possible affects of liquid waste, noise pollution, offensive odors and radiation are investigated, and in general, the possibility of any negative influences on society during both normal and abnormal plant operations
is evaluated.
392
Chap. 7 System Engineering
Feasibility study
~
l
I
Construction go-ahead
t
Process basic design
Column and vessel design

Equipment design
Process
detail
design
Piping design
Pump and compressor design
Electrical design
Instrumentation design
Civil and building design
Miscellaneous equipment design
f
Procurement and manufacturing
Leveling and foundation work
Installation work
Coustruction
Piping work
work
Instrumentation work
Inspection and testing
f
Test run (Commissioning)!
t
Actual operation
Fig. 7 1 Plant construction flow chart
(b) Construction go-ahead
On the basis of the above feasibility study, a decision is made as to

whether construction planning shall proceed or not.
(c) Process basis design
At this stage, basic plant configuration and process data needed
for plant construction are determined, as listed below.
(1) Verification of design conditions
(2) Material input and output, heat input and output
(3) Plant operation method, safety control measures
(4) Floor plan and basic structural specifications
(5) Process flow diagram
(6) P & I (Piping and Instruments) diagram
(7) Process layout chart
(8) Utility plans (electric power, water, heat, etc.)
(9) Basic specifications for major equipment (pumps,compressors,
7.1
393
blowers, heaters, heat exchangers, columns, etc.)

(10) Basic piping specifications (materials, conduit paths, etc.)
(11) Basic specifications for instrumentation and electrical systems
(d) Detailed design
In detailed design, which is also called manufacturing design, specification documents, drawings, etc. are prepared in a detailed manner
on the basis of the above basic design. This will involve equipment
design (e.g., columns, tanks, pumps, blowers, etc.), piping design, electrical design, instrumentation design, and civil/floor plan design, for example.
Table 7.1 shows the allocation of engineering work from the feasibility study up to detailed design, as seen from the viewpoint of the
instrumentation manufacturer.
(e) Device procurement/manufacturing
At this stage, each device specified in the above detailed design is
either procured or manufactured. In addition, each device unit must
eventually undergo inspection to verify that it meets specifications and
achieves required performance.
(f) Plant construction
Actual plant construction is divided into the following steps: leveling and foundation work, major equipment installation, structural construction, piping, electrical wiring, instrumentation, heat insulation
work, and painting.
Although the need for overall management of the plant construction process is obvious, note that due to the complicated relationships
between the various construction steps in a modern plant as well as
the detailed planning involved in each step, a powerful management
technique should be employed such as a network analysis method or
PERT (Program Evaluation and Review Technique).
(g) Test run/system delivery
After all construction work is completed, system evaluation begiris
on a partial basis by first checking the performance of individual devices. Next, as devices are adjusted, the range of evaluation broadens,
and the performance between multiple devices is evaluated. Eventually, the entire process comes under scrutinization. In the final stage of
testing, a so called "water-run" is performed in which overall performance is evaluated by substituting water or air for actual materials. Finally, an actual test run (commissioning) is performed before handing
over the system to the end user. The above tests and evaluations
should be performed with representatives from the end-user, plant
builder, and the various electrical and instrumentation equipment manufacturers present.
394
Chap.7
System Engineering
7.1.2 System design considerations

(a) Clarifying required product quality .
In order to facilitate the design and engineering of the plant in
question, probably the most important consideration is product quality.
In particular, the product quality required by the customer (plant user)
must be accurately and fully understood so that design work can proceed in the right direction. Although the required quality is initially
conveyed in the form of specification sheets and technical meetings,
the descriptions or explanations may turn out to be insufficient or unclear and thus will need to be redefined. Accordingly, actual requirements must be thoroughly researched and specified as accurately and
realistically as possible. In this regard, it is convenient to make use of
quality development techniques such as that depicted by the diagram
in Fig. 7.2. As shown, product quality passes through several stages of
development where the quality in each stage is subject to problem analysis in relation to design and technical capabilities. In other words,
quality development starts from required quality development, breaking down each stage step by step as quality function development and
design function development. It eventually will be developed to a level
at which target product quality can be realized in products.
(b) Incorporating new technology
Once the go-ahead has been given on a particular project, it is rare
that all aspects of the technology to be employed (e.g., theory, design,
manufacturing methods) are fully understood. There are usually some
elements of a project which involve new mechanisms or procedures (unknown technology), or at the least, known technology applied to new,
untried conditions. Known and unknown elements in the technology to
be applied must be thoroughly investigated and evaluated in a stratified fashion.
In particular, development of stratified unknown technology is carried out once the new technology required has been clarified by the
quality development procedure explained above. However, in the case
of projects such as plant building, i.e., one-shot projects, which must
be met within a limited cost range and time period, some trade-offs
must be made between functionality, expense and delivery date. Thus,
in the same way as described above, depending on the progress in
design work, items requiring new development must be closely scrutinized in an organized manner and efforts made to minimize unknown
technology by applying or combining with .known techniques. Despite
such efforts, however, verification and modification under actual plant
conditions are still required. This development concept is illustrated in
Fig. 7.3.
7.1
395
""
~-
"'"'
;;;
"';;!
""l
Plant builder
0
0
0
0
0
2. Operation technique
3. Initial investment and running cost
4. Schedule
5. Site conditions
6. Miscellaneous
Basic design
1. Material and energy balance
investigation
2. Operation technique and safety management investigation
3. P&I and schedule preparation
4. Equipment and machine
layout preparation
5. Instrumentation system
design
6. Establishment of management data processing systern and process control
computer systems
0
0
Allocation
E
Proposal for instrument selection (computers,

micro-computers, devices, etc.)
Examination and proposal for optimal system
based on items 1 and 2.
Proposal for CRT displays and mathematical

model
A; Working scope desided by end user independently or by discussion with plant builder
B; Working scope decided by plant builder independently or by discussion with end user
C; Working scope decided by plant builder based
on consultation with instrument manufacturer
D; Working scope decided by plant builder depending on the proposal from instrument
manufacturer
E; Working scope decided by instrument manufacturer based on discussion with plant
builder or end user
F; Working scope decided by instrument manufacturers independently
Allocation of engineering work
Proposal of basic specificatior lS

and operation techniques
1. Production facility data
Work item
Table 7.1
'I
"'~-
"'
~
;;;
"'~
......
:"!
Instrument
manufacturer
Instruments manufacturing
design
1. Flow sheets (loop drawings
of instrumentation flow)
2. Determination of instrument
functional specifications
3. Control panel design
4. Sequence logic design
5. Control panel internal power
system design
6. Control panel internal connection diagram
7. Design of instrumentation
work drawing
8. Miscellaneous
7. Instrument specifications and

data sheets
8. Control panels and sequence
logic specifications
9. Process control computer system specifications
10. Instrumentation works specifications
11. Manuals for over-all adjustment and delivery
12. Manuals for operation and
maintenance
13. Manuals for routine inspections and emergency
maintenance
14. Miscellaneous
0
0
0
0
0
0
0
0
0
0
Preparation of loop diagrams, work sheets and

data sheets
CRT display and graphic display drawings
Preparation of sequence tables
Preparation of specifications based on the basic

data of items 2, 3, and 4 etc.
Preparation of specifications based on the documents of items 3, 4, 5, 7, and 8 etc.
Drafting of overview manual related to the instrumentation system
d iscussion
with the
customer
Engineering
technology
Simulation
EnJineering
technology
Ach ievement
of target
prod uct quality
Know- how
Fig. 7 2 Outline of qua lity development
nknow n characteristics
Investigation
Simulation
Step
Study of releva nt techn iq ues

Develo pment of new techn iques
Demon trat ion by model
quality
Step 2
Step n
Unkno wn characteristics
i
398
Fig. 7 3
Sy tern techn iques fo r ad va ncing the project
Chap.7
System Engineering
7.2 Instrumentation System Design

7.2.1 job planning
In this stage the outline of the job is first clarified, after which a
job organization is set up, a basic schedule established, and specifications narrowed down. During this time, all unclear points should be
clarified to avoid any confusion in the job, keeping in mind at all times
the slogan QCD (Quality, Cost, Delivery). The work flow is shown in
Fig. 7.4. The relationship between basic design and the work overseen
by the project leader is shown in Fig. 7.5.
(a) Work preparations
(1) job clarification
All relevant specification sheets and documents
are thoroughly studied to ascertain the scale of the job, the time required, and the level of difficulty, and to help management setup methods for dealing with the job, the target quality, probable cost
(man-hours), etc.
(2) Documentation exchange
Specification sheets and documents to
be supplied from the customer and those to be supplied from the instrumentation manufacturer to the customer are identified and listed. Examples of such documentation are described below.
(i) Documentation from the customer
(1) Purchase order (copy)
(2) Specification sheets (instrument list; instrument specifications;
process data sheets; functional specifications; specifications of
procured items, etc.)
(3) System specification sheets (system outline; control panel specifications; sequence specifications; P & I diagram; process flow
diagram; utility specifications, etc.)
(4) Computer specifications
(5) Instrumentation work specifications (factory layout diagram; facility layout; control room plans; piping diagram; duct/pit diagram; drive system documentation, etc.)
(6) Miscellaneous (customer established standards; plant building
schedule; safety standards; security standards; job organization
chart, etc.)
(ii) Documentation for the customer
(1) Project quotation
(2) Specification sheets (instrument list; instrument specifications;
instrument external view diagrams, etc.)
(3) System specifications (system configuration diagram; comprehensive flow chart; panel diagrams; functional description manuals; technical description manuals, etc.)
399
. - - - - - - - - Descripti o n - - -- - - - -- ,
I.
Understanding job outline, Establishment of a work

organization. Preparation for narrowing down specifications
2.
h ems to be discussed with the customer

Project organization (Customer and manufacturer), Scope
of work and sched ule
2-2) .
Instrumentation system, F~ctory layout , U til ities and process

data
2-3).
Determi nation instrument and function details

(Type . Ranges a nd now sheets)
Instrumentation work (Materials, Wiring or piping paths,
Scope of work, Organization. Schedule, etc.)
3.
Manufacturing instructions and commu nication of

pecification information
4.
Approval drawing (Final confirmation of specifications

arranged with customer)
5.
Modification and corrections
Inspection manual,
6.
om pan y inspection and inspection with
witness
7.
De 1ivery inspections
8.
Installation and testing
9.
Test-run manual, Training for operators and handling

complaints
10.
As-bui lt drawings and job summary
Fig. 7- 4
Flow chart of basic design work and job management
(4) Computer specifications

(5) Miscellaneous (safety organization chart; job organization chart;
minutes of technical meetings, etc.)
(3) Overseas jobs
The following items must be kept in mind while
narrowing down the specifications in the case of an overseas job.
(1) Applicable laws and regulations
(2) Environmental conditions (climate; utility concerns such as air
conditioning and power source, etc.)
400
~
Manpower
required
(Man-hours)
Preparation
Planning
Basic design
Manufacturing
Inspections
Installation
Adjustment
/-~
Work preparation
Narrowing down of specifications (Over-all)
arrowing down of specifications (Basic)
arrowing down of specifications (In trumentation in general)
arrowing down of specifications (Instruments and functions)
Main work
arrowing down of specifications (Instrumentation work)
Approval and coordination
Inspection preparation Inspections
Adjustment and preparation for runing
Stan-up
Test-runs (Commissioning)
Completion activities
Project
plan
Schedule
Main
documents
Fundamental
sepcifications
Comprehensive
now sheet
System
configuration
Fig. 7 5
Final drawings
Approval drawings
Data sheets
Comprehensive work
sheets
Inspection
manual
Operation
manual
Quality
inspection
cenificate
Final
documentation
specifications
Control panel specifications
Functional specifications
Instrument 1ist
Relationship between instrumentation de ign and job management
(3) Shipping, transport and storage conditions

(4) Documentation lists, language used and format requirements
(5) Training curriculum, language used, training place requirements
(6) Acceptance conditions, test methods
(4) Scope of services provided (contract clarification)
In order to facilitate initial design work in common areas between the instrument designer/manufacturer, the end- user, and any other third party, mutual
401
consultation must be diligently held to clearly determine the scope of

work to be performed by each party. Experience has shown that without close contact between the different groups, problems related to inadequate materials and insufficient facility tie-in at the time of instrumentation installation can easily occur at the site. Examples of areas
for which the scope of responsibility must be clarified are piping interfaces, power interfaces, major equipment location, cable purchasing/layout, document preparation, and equipment transport and
off-loading, etc.
(b) fob organization
The job organization required within the company (the instrument
manufacturer) should be considered in relation to the scale of the job
in question. In this regard, it may be necessary to contact relevant
personnel or department heads to brief them on the job and obtain
their feedback on organizing the job. At the same time, the organization agreed upon should be compared with the job organization of the
customer and related engineering companies or manufacturers.
(c) fob schedule
Schedule charts must be prepared reflecting both the schedule of
the entire project for the customer as well as that related to instrumentation design and installation. These schedules are provided to ensure that engineering work and instrument manufacturing proceed at
an appropriate pace. They may be prepared with a master plan, intermediate plan, or plans of various degrees of detail as necessary using,
for example, the PERT project management techniques mentioned
above. For design work, in particular, it is important to include documentation scheduling in addition to that for instrumentation and installation.
It should be noted that although these schedule charts usually represent the schedule of the various departments in a time-series fashion, their possible use as proposal documents to the customer should
be kept in mind when preparing them.
(d) Proceedings of technical meetings
The following describes important items which must be considered
for facilitating engineering work.
Since general specifications usually exist only for the job at the
time of initial negotiations, frequent meetings are required to arrive at
more specific technical descriptions. As can be expected, many modifications and additions will be made to these specifications at this time.
However, as technical matters usually come to overwhelm the discussions, there is a tendency to forget cost and schedule factors which are
also essential to the project. All three of these factors should be kept
in mind by all parties at these meetings. In this regard, efforts should
402
be made to plan meetings carefully, e.g., decide on the topics of discussion and exchange relevant documentation beforehand, so that all parties can arrive at the meetings well prepared. Minutes of each meeting
should be carefully taken to provide an accurate record of what has
been or has not been decided or considered to date. In this way, final
specifications should be more accurate, i.e., more inclusive of the three
important factors: technology, cost and schedule, and should also bearrived at sooner.
7.2.2 System specifications
In this stage, specifications for the overall instrumentation system
are finalized.
(a) Comprehensive flow sheet preparation
A flow sheet shows the relationship or interconnection between
the various elements making up the measurement and control loops in
the system. These elements are sensors, converters, receivers, controllers, final control elements, etc. Preparation of the instrumentation
flow sheet concentrates on representing the system as seen from an
instrumentation point of view and is based on various sources of information such as process, plant layout, and P & I diagrams. This flow
sheet becomes a basis for further design work. An example of a comprehensive flow sheet is shown in Fig. 7.6. Major symbols used in flow
sheets of this type are shown in Table 7.2.
(b) System configuration
A system configuration diagram (Fig. 7. 7) shows the basic concept
of the instrumentation system in terms of concrete system equipment
such as transmitters, control panels, digital instrumentation devices
and computer equipment. Accompanying this diagram will usually be a
complementary manual called, for example, "System Overview," which
helps to explain in more detail system concepts and makeup. This overview usually consists of the following items.
(1) System outline description
(2) Operation
The man-machine interface as it relates to the operator. Information transfer and operations are described for control room
operations as well as on-site instrument operations, both under normal
and abnormal conditions.
(3) Safety measures
Countermeasures such as redundant and analog
backup systems are described for such problems as power and air supply failures and computer and DDC system failures.
Methods used to output and transmit
(4) Information management
operation data and to manage process data are described.
(c) Overall layout verification
Based on plant layout diagrams, the layout related to control
403
LPG -+ I...N + HN:
CRUDE
L-- - - - - - - - - - RES
FUEL
Fig. 7 6
404
Example of a n instrumentation flow sheet for topping
Chap.7
System Engineering
""'
"'~
<:;
'-l
Turbine type now

meter
Volume type now
meter
Electro magnetic
now meter
Area type now

meter
Nozzle
Venturi tube
Orifice
Fine tube
Pneumatic- pressure
signal
Oil-hydraulic press
signal
Electric signa 1
C lassfication
-~
r-ff}---1
1--[[}--!
r-{D-- f
~ or
'"
---- - - --
Graphic symbol
--
Valve actuator piston type

-
Valve actuator electromagnetic type
Valve motor- electric motor type
Valve motor-diaphra gm type

(Pressure balance type)
Valve motor- diaphragm type
Ball valve
Buuerfly valve, damper louver
Three-way valve
Angle valve
Valve (general)
Classfication
~
L_ ~
C8:]
,,,
C1fJ
l><J
symbo
GraP.~
Table 7 2 Main instrumentation symbols (Selected from J IS
Local panel
Rear panel
installation
Control room
Site
0
0
8
8
8
GraP.hic
symbol
of monitor and control location is

not required
(!): General case where the distinction
Change-over switch
Where the
distinction of
monitor and
control
location is
requ ired
(!)
Transmiuer
Classfication
8204)
I CPRT
~-
iiii
~~
;~
5S
~~
~~
: Serial Printer
: Color Hard Copy
Unit
: Duplexed Fie ld
Control Stat ion
:Terminal Board
Cubicle
:Gateway Uni t
:
l .... . .. _ ,. ,. .. .. .. . . . . . . . . . . J
jCGWU
iCTBC2
'l
CFC02
'I
: Operator Station
1CO PCV : Operator Console
~
~-
.;
CGWU
~!!!
~=
~ii
II
!I
~~
EE
ii
~=
52
~~
i~
~~
!!!!!!!
~ij
Computor
r----
t COPSV
I CCHU
"';;:::!!
Computer room
Q,
CFCD21CTBC2
1"'
COPSV COPSV
Main control room
room-~
1------l~t~
Ilf:~tric
~ ~panel
----t+--J
COPSV
Sub-comrol
---
CFCD21CTBC21CFCD2 1CTBC2
Control
panel
Fig. 7 - 7 System configuration example
CFCD21CTBC21CFCD21CTBC2
~-
rooms and field-mounted instruments and accompanying facilities,

operation rooms, computer rooms, power supply rooms, etc. must be
clarified. At the same time, signal transmission routes and transmission techniques should also be checked out.
(d) Instrumentation layout in control rooms
Based on the general plans for control room, computer room, power supply room and other specialized rooms, layout of control panels
and other system devices is determined. This must be performed keeping in mind the functional layout for operation monitoring, as well as
the need for maintenance space. Refer to Sec. 7.3 for details on the
man-machine interface in control rooms.
(e) Utility interconnection
Verification of the amount, characteristics and location of the various equipment providing system utilities must be made with reference
to the power system diagram, air-conditioning plans and other utility
documentation. Such equipment usually includes power devices (e.g.,
commercial power devices, uninterruptible power devices, batteries
and power distribution boards, etc.), air supply devices for both factory and instrument use, air-conditioning devices, and miscellaneous
devices such as for pressurized oil and N2 gas. Refer to Subsec. 7.2.6
for details on power and air supply equipment used to drive instrumentation systems.
(f) Process data
The characteristics, data and desired control values related to process objects from the raw material stage up to the final product stage
must be understood and described on a process data sheet. In addition,
as these data will be used as a basis for the design of individual control
loops and their driving means, it must be as accurate as possible, incorporating any changes made to date. Examples of specific process data
items are pipe size, piping standards, type of fluid, temperature, pressure, flow rate, density, viscosity specifications, construction dimensions, etc.
7.2.3 Device and function specifications
In this stage, individual loop flow sheets are drawn up, control systems designed, and concrete specifications for devices and system functions determined.
(a) Preparation of loop flow sheets and control system design
Once the characteristics of a particular process are understood, an
appropriate control system must be decided upon. In recent years, the
use of microprocessors in distributed DDC (Direct Digital Control) and
in small-size controllers (single loop controllers) has enabled the selection of a wide variety of control functions. These range from simple PI
407
and PID control up to high-level "advanced" control. Refer to the appropriate chapter in this book for more details on this topic.
Given that a control system has been determined for a particular
control loop, a loop flow sheet can then be prepared detailing the control and interlock system as well as the interface with other equipment. In addition, since the loop flow sheet will be used as a basic
reference for instrumentation design, other information relevant to the
control loop should be prepared as well at this time. Examples of such
information are equipment list, equipment manufacturers, models,
measurement range, custom specifications, process data (physical properties, characteristics, pipe size, other mechanical data), control calculation format, inter-loop interface, interlock conditions, and man-machine interface (e.g., CRT, digital display). An example of a loop flow
sheet is shown in Fig. 7.8.
Fig. 7 8
Example of a flow sheet for anti-black smoke combustion control
(b) Signal interface list
This document lists signal characteristics, e.g., destination,

source, and signal type, for all signals transmitted between instruments, devices or other equipment making up the system. Signal interface is divided as follows:
(1) Signals between centralized instrumentation equipment such as
control panels, digital instrumentation devices and computers, and
408
Chap. 7
System Engineering
field devices such as transmitters and converters;

(2) Signals between various centralized instrumentation equipment;
and
(3) Signals between centralized instrumentation equipment and electric, pneumatic, or other facilities.
Note that "interface list" as used in this book refers to the same
type of documentation which may be called "input list," "input/output
list," etc. depending on the work involved. An example of an interface
list is shown in Fig. 7.9.
(c) Instrument list
Detailed information pertaining to all instruments employed is listed, such as instrument name, tag number, model, manufacturer and
quantity. This document is used as process data for the design and
manufacture of transmitters or as simple sizing data. An example of an
instrument list is shown in Fig. 7.10.
(d) Instrument sizing
Sizing based on process data for such elements as orifices and
valves must be documented. Since such elements are directly installed
on process equipment and piping and may come in contact with
high-temperature or corrosive process liquids, the materials used in
their construction must be carefully selected. Examples of orifice and
valve data sheets are shown in Figs. 7.11 and 7.12, respectively.
(e) Specifications of procured instrumentation devices
In the process of configuring an instrumentation system, it is frequently necessary to procure instrumentation devices from outside
one's own company. Accordingly, purchase specifications for items to
be resold must be prepared, as must prices, specification sheets and
function descriptions. The necessity of additional services such as quality guarantees, user manuals, training, and maintenance must also be
considered.
(/) Sequence control
In the case of implementing PA (Process Automation) or FA (Factory Automation), sequence control functions have become indispensable. Moreover, the demand for such functions has particularly been
felt along with the increased automation of production facilities in
recent years. For the definition and basic concepts of sequence control,
refer to Sec. 2.5. This section concentrates on how to determine sequence specifiCations from an engineering point of view.
(1) Overview of sequence control equipment
In order to perform sequence control, a variety of hardware is possible, such as relays, semiconductors, PCs (Programmable Controllers), distributed DDC, and
computers. PCs in particular have become very popular in recent years
due to the advances in microprocessors. An example of a PC configura7.2 Instrumentation System Design
409
""
""
tl"J
"'~
"'
Signal
No.
___r-
Auto mode dumper

operation switch
A ir valve
Fully closed
Air valve
Fully open
A ir contro l valve
close
Fuel oil shut down

valve close
(West)
___s---
__r---
__r--
__r-
__r-
Fuel oil shut down

valve close
(East)
__r--
____r--
COMP/ LOCAL
Change over
(3 notchs)
(2)
Signal form
Type
Signal
specification
Signal name
Combustion stop
Interface Iist
room
Control
Control
room
room
Control
Control
room
Control
room
Control
room
Control
room
Control
room
Location
Wiring
Interface list
Terminal
No.
Fig. 7 9
CUNB(I)
CUNB(I)
CUNB(I)
CUNB(I)
CUNB(I)
CUNB(I)
CUNB( I)
CUNB(I)
Panel
name
Scope
(plant bui lder side)
Spec.
Terminal
No.
CUNB(I)
CUNB(I)
CUN B(I)
main
Operation operation CUNB(I)

desk
room
main
Operation
operation C_UNB(I)
desk
room
main
LOCAL
LOCAL
LOCAL
room
CUNB(I)
6 points
Remark
6 points
6 points
6 points
6 points
6 points
6 points
ma in
CUNB(I) 6 pointsX3
operation
LOC A L
Location
Inter
face
point
Operation
operation CUNB(I)
desk
room
---
---
---
UD
Operation
panel
name
Pa nel
scope
(End user side)
SHEET NO. 7
"............
O'q
:1!
!::::!
~.
"'
"""'
1:)
:1!
"""l
SERVICE
!-1
I Ma.\ur~nl
. .........
I ............ t
rT.. ....... r-
--,
n1 [
~v 1rncror
INSTRU M ENT
,\1FOR
IMOOEl :o.UMBit
l voKOGA WA I POC-tb-t
mVT.,.p./>011<0<
I.
Fig . 7 10
SK YD-lOOtA
I Al;~rm
01
~~~:
STED-110..\t\' B
I mV T~mp./,oltal<l!
411
Pl)('-16-1
BALl BAll!
SVN' B lCSST
j'-""l.(l91SV:~:"
-------
...
.......
SPECIFICATIONS
lnp.~1
.......
2 Pt:n
1 Sd
!Qnlft 1111
.......,.... ,..,,.,.,
Otn 'pd
l r lnpul 2""
I I noubl ,,,_,,
lnsturment Jist
.....
0 -IOmV
O-I Om Y
---
I SNVD-.!OOM I
I mV rc-~II.IOf
i . :. .
fl
I o-10m vl
I RANGE~
VOKOGAW A STED-liiH...S 0 -I OmV
II)
! lSI , .......,
I '" I Ala<m ,.,,.,
Ill
::::...... :::::: .... ::l:: . : 1-m1:;;::: t~~:r7~~'cs.ll -~c~,~--
....... ld'l~ute
TAG
I No.I
---l~i LolO~~-~.C'n)ltf
l iTEM
lnstrumrol ltst
'"
... I
......
I
I
I , ll
RP
RP
RP
Rr
RP
RP
RP
! OIY I {:~'lA- I Dumet~~or'l I

IU::MARKS
.Sii (;.fT NO
~-
""""
;::!
-.
t'l'J
""~
~
"'.....
'I
tv
.....
~~
I IJ
11.1
llo v.
~ UC'I
N'll.l Loop
On ftt.~
No
Ta~
[ Q ' l)'
fOIIJ0-1 1
d.ll.t . . hl
COG
~-1100
950
'\l>.lr
Flo"' r:.ltt:
Hutd 1Seal<
Nm'hl
I. I
~(XI
420
\ hn j mmH,O)
~or
Preuurf
10
!'C)
Nor
Temp.
Flmd spectftea ttons
001
KPI
\I)CQ
''t)
100
j",_)
Humtdtt)
150A
JISI OK
f\omtn~l
M.ttena l
I SGI'
'I
dtitmcta
Conncctton
I
155. ~
1mm1
Inner
d t.1~1cr
sptli<.-.uton~
Fig. 7 11 Orifice data sheet
05
1t1 \ mJI
JTIIIt)
Spe<'tfiC
Ptpt
100
~mmlt~OI
lo-.
pre.~urc=
_j Allo,., .. blc
th<
CmmH,Oi
onfil"'e
.tc-ro'~
pr~)ure
OtO'e,.-enttJ I
j (Orrlt'f
of
tap'
T) pc:
i S L S.'OJI
SUS.104
1\1;UC'fl.ll
Ul t\
-U i)A
ISOJIO
~1DI C-
T)p<
Ortfice 'pectfi'-'<.&tton'
I Rtm.trl )
...........
"";;!
~
:::!
~-
~
:::!
-"'
'l
10
,,
I)
ll
II
biO""
Au
No.I l oop
Control
JOI
'
O't)'
i.ht
/ ht
A"
I lrool oroo I
Min.
nov.
H ~Ol
{mm
Mtn
Mu _
M1n.
na . . no ...
.. ..
~ n
No<
r-o~ t ur~
TempeSp:r;:lf~
Ifill II)'
Pipt
dtarnrt
SOAI
JISSKFF
COftii''IOfl
Nonu11..11l
I
-II
STPV
flail
Mill~!-
spifta~l1011)
\ ' lliH!'
11)'
I""""'I
Typ<
""'
X
.JSOA
J,lle'
Pon
SUS.JO.t
ISCPHll
tl~IC)
Pon
,.....,..
(')III'Mkr
-~
OA
I
2-8:
, ,,~~~
...,.,,,
'"
""~
1 I
ll(ti(Hl
Typ<
AaWIIOr
V~;l~~ Spec'lfiCiiUOM
!Bod)'/
trim I
rut I
M111t--
VviH'
Bod)
Fig. 7 12 Co nt rol va lve data sheet
I,JqJ
ll.l N1111l
I sso I so 1- 1 " I
nov.
at
PrH.Sui~
~pifi~a t ioni
P"m"'y
pt"e)1o-
Flutd
M:u..
ill
I I II
Flo""' nne
.,,.. M"'
(Nm ,
I I
FCV- ~
TJ
No.
\'ah~ doll;~~
' PP,AS
Nl<
AI X l
,t4~ XH0
''"""'
""llh
t;lNP.
n..
..... ,h
n,
Rtm.~rL~
comp.nto
1-... I
Duple system
19
Host computer
.,
EP - ROM
Printer
Remote 1/ 0
11 11111 I
Ploner
_j
Fig. 7 13
A P
system configuration
tion is shown in Fig. 7.13. PCs feature easy creation and modification
of sequences as well as highly-reliable, compact and lightweight equipment. They also feature communication functions which, due to their
ease of implementation, facilitate system construction and expansion.
Moreover, PCs have recently come to incorporate more sophisticated
arithmetic/calculation functions for achieving even more powerful control. Examples of PC functions and commands are listed in Tables 7.3
and 7.4, respectively, and an example program is shown in Fig. 7.14.
Distributed DDC, apart from the features described above for the
PC, is also characterized by being able to execute both feedback control functions and sequence control functions within the same piece of
equipment. Specifically, it can easily receive and transmit feedback control and sequence control information. For example, if a tank level
rises above its specified value, a drain pump can then be activated, or
if the concentration of a mixing bath reaches its specified value, the
stirrer can be stopped and an output pump activated. Thus, since measurement control and sequence control are very closely related, the ability to realize both feedback control functions and sequence control functions in the same controller is extremely convenient and efficient when
designing an instrumentation system.
(2) Sequence specifications
When defining sequence specifications, it
is important to first determine the operation specifications for the machines and equipment in question. Major specifications in this regard
414
Chap.7
System Engineering
Table 7.3 Typical PC functions

Items
Contents
Input/output points
From approximately 64 to 2048 points depending

on unit size
Arithmetic functions
Relay, Timer, Counter

Four basic arithmetic operations
Square-root extraction and exponential computation
PID control computation
Numerical comparison ( >, <, =)
FORNEXT
etc.
Communication functions
Remote input and output

CPU link
Computer interface
Data high way (Data communication cable)
Intelligent input and output
Numerical input and output

Analog input and output
High speed pulse counter
Pre-set counter
Registration
PID parameters
ASCII
relate to devices and drive methods used, operation and display equipment, and input and output signals. Moreover, personnel safety and
equipment and plant protection during both normal and abnormal
operations must also be incorporated into design plans.
(3) Specification description
The main methods used today for expressing sequence actions are relay ladders (development diagram), logic circuits, flow charts, time charts and decision tables, as described
earlier in Sec. 2.5. Each of these-methods has its advantages and disadvantages in regards to a particular application. Moreover, the format
of any one method will vary somewhat depending on the engineering
field, the type of process, or the equipment involved. Nevertheless,
several of the above should normally be used for any particular sequence description, and all sequence documentation should be clearly
annotated and as error free as possible in the fashion of engineering
blueprints. Examples of a time chart and a sequence signal chart as
commonly used in instrumentation design are shown in Figs. 7.15 and
7.16, respectively. Symbols generally used in flow charts are listed in
Table 7.5.
(4) Safety measures
In addition to drive methods and operation pro7.2 Instrumentation System Design
415
""'
"'
~
~;:;
~
"';;::!
'l
0'1
Or load
And load
Or not
( ~~ )
Or
[ ~~) [ ~)
r-u00(]0
[~~ )0
Logica l OR operation with the

previous condition.
Logica l AND operation with the

previous condition.
Logica l OR NOT operation.
Logica l OR operation.
Logical AND NOT operation.
l0
[ ~N~
1-
And not
Logical NOT start operation.
Logical AND opera tion.
lG
Logical stan operation.
Function
rn
r ~~
...
QD
Symbol
And
Load not
Load
Instruction
Interlock
Temporary
memory relay
Latching relay
Shift register
Counter
T imer
Out not
Out
Instruction
Function
~0
Causes a ll the rela y coils

between this instruction and the
LC instruction to be reset or
not reset according to the result
immed iately before this
instruction .
Temporary memory relay

operation .
Latching relay operation.
Shi ft register opera tion .
GJ
Down counter operation.
On-dela y timer operation.
Inverts the results of a logical

operation and then outputs them
to the specified output relay.
internal auxiliary relay, latching
relay or shift register.
Outputs the result of a logical

operation to the speci fied output
relay, internal auxiliary relay.
latching relay or shift register .
[ ~)0
[~)
Symbol
Table 7. 4 Typical PC instructions
':)
.....
""......
;!
"'"'
oq
"'~
;!
<:;
;!
...."'
"'....
;:
;!
......
t-v
8~
8! uiv l
~ ( uh ]
IIIove (MOV )
Move not
(MOV. NOT )
88
] ~
Symbol
Jump end
Jump
Interlock clear
Inst ruction
Diagnostic (FAL )
End
Invert s and transfers data

bet ween 1/0s.
Subtract (SUB)
Add (ADD)
Compare (CM P)
I Instruct ion
T ransfers data bet ween I/ Os.
Clears t he JMP instruction .
Causes all the contents of a

program between this ins truct ion
and the JME instruction to be
ignored or executed accord ing
to the res ult immediately before
this instruction.
Clears the I L instruction.
Function
80
8 !REk,]
8 B
8 (i@
8 (C ~ P
Symbol
The end of a program .
Indicates the t ype of failure or

abnom1al mode.
Subtracts data between l/Os.
Adds data between I/ Os.
Compa res data between I/ Os .
Function
Up
[sHIFT](3-Homeposition
Down
OJ[write]
[!][write]
QJ[write]
QJ[write]
[ ~t J [DCIJ [write]
(B[D[D
[ tn-
[ r.o~-1~--
[
[
[
~001t- n.. -
rooo
0002"
fo~~~~--~~-0001 0003
t-oo~~ -,:-
0000 0002
~1--1
1--1 1--1
1
,__.9001
0003
01
rooo
0002
,____. ............
t---I0003
l
0001 - -:r-..
09
fooo ooo2 ooo3

t---tt---1~ 1
0001
QJI._O_~ii--
09
~ 0002 0003
0~
'i019l!o3
Fig. 7 14
]
]
]
OOJI
An example of PC programming
cedures, equipment reliability must also be taken into account when

designing safety measures for sequence control. Starting with life safeguards as the prime consideration, the integrity of the entire system
must be insured against the effects of any accident or abnormality
which may possibly occur. Needless to say, this requires a thorough investigation into the relationship between control objects, drive conditions, and hardware reliability. Specifically, the following items should
be investigated when designing in safety measures:
(1) Power supply abnormalities (dividing the power supply system
into subsystems should be considered as a countermeasure)
(2) Power loss/release measures (affects of power loss on safety;
restarting measures after release)
(3) Reliability of sequence control equipment (backups, redundancy
418
Chap.7
System Engineering
Operator station
Mes age output
Contact
input and
contact
output
Sequence control
function
.g
Counter setting
~
-;;
.~
.,
c:
0
~
-- I
Sequence table
- ..'
Transmitter
Feedback control function
Fig. 7 15
Components of sequence control functions and their signa ling diagram
as countermeasures)
(4) Emergency halt conditions (halt conditions should be considered
within the range of central operations, on-location operations, and
device operations, respectively)
(5) Erroneous operation countermeasures (wrong switch operation,
interlock conditions)
(6) Information transfer (lamps, buzzers, sirens, CRT displays)
(7) Accessibility to maintenance check-points
Sequence specification
(5) Sequence specification documentation
sheets, which are used as a basis for specifications approval and detailed design work, must be prepared based on customer specifications
419
Lsr
c:::) Time
Remarks
rl
na l
Opera tiona~
elements
(!)
No.I
Nn.2
No.3
No.4
@>@)@
@>
@>
@>
@>
1
dz
1 ~
z
zz
---,------------ ------- --- ----- -- --- ----- --------- ------------------------- ---
Tlmer
~Timer
Time
frlmer
d
z
~
z
-;z
~
z
~
z
-:
z
1--
I
I
f-----1-.-
No.I
No.3
,_
t-h
No.I
@ Mixi ng -cold-blast cont rol valve
(!) H-Parallel blast
S-Parallel blast
old -blast control valve
Mixing cold blast valve close

HS Switching signal
Fig. 7 16 Time chart example
Table 7.5
Flowchart symbols
Symbol
Process
Represen ts the sort of any processing

fun ction. arne of the fu nctio n is written
inside th e symbol. This symbol is also
used for representing logic element AND.
Refer to No. 7.
Automatic
Manual
On li ne
Represents manual processing fu nctions related to seq uence con trol.
Off line
Rep resents manual process ing functions

not related to sequence control.
Decision
<>
420
Description
Represents a decision to separate the process flow into two or more branches dependin g on a condition. Branches are indi cated by yes, no, high, low, etc. adjacent
to this symbol.
Chap.7
System Engineering
Table 7.5 (continued)

3
Predefined process
Lamp
Usage
:0:
5
Display other than

lamps
Buzzer, bell, siren,
Used when a predefined flowchart indicated in the flow.
ITurn lamp on I
ITurn lamp off I
@
@
Represents data displayed an any common

information medium.
>
Usage
ISound buzzer I
ICancel buzzer I
@ ~
~
@
~
Example of symbol No. 1.
AND
OR
NOT
10
Timer
Example of symbol No. 1. Indicates both

"timer start" and "T minutes passed" function.
11
Switch
For instance, for changing over from automatic operation to manual operation.
12
Connector
Used to show which process on the same

page should be connected with a conditional branch, or to show the relation between
different sections so as not to complicate
the flow chart. Connection No. is indicated in this symbol.
::::r
0
13
Connector
0
14
Terminal interrupt
c::J
Same as No.l2 above, but indicates connection between different pages. Connection
No. is indicated inside the symbol and
page references may also be given beside
the symbol.
Represents flowchart boundaries.
For example, startpoint, endpoint, abort,
stop, etc.
421

15
Comment Annotation
16
Flow line
17
Cross
18
Branch
---[
Used "for descriptive comments. The dashed line is drawn from left or right to
connect comment annotation.
Represents connection of symbols and direction of flow. Arrows are not needed for
directions from top to bottom or left to
right. But, arrows must be used for directions opposite to those.
; sequence flow
; process flow (Note 1)
------ ; flow needs human intervention (Note 2)
In two flow lines cross, they should do so

at right angles.
-c
Used to branch flow line into two or more

lines in which branched lines have the
same flow meaning as before branch.
Branch point is indicated by a black dot.
-c
r-
--- ...L_ __
I
Note:
1. For example, this flow line can be used for the following process.
Level reaches upper limit--.. After valve is open and

tank level reaches upper
limit, the sequence advances to the next step.
~ Limit switch turns on
After valve is opened by
openning signal, the
valve limit switch is
closed. The sequence advances to the next step.
2. This line includes the case where the sequence cannot advance to the next
step without human operation.
IPreparation is ready I
1
D-
L--------~--4
I"
ISwitch turns on I
,.:..'-'------.
IValve open I
3. Location and tag No. may be indicated inside of symbol so that the flowchart can be used as an instruction manual for operators.
and on the results of technical meetings. Such documentation usually

includes the following:
(1) Cover sheet (table of contents, statement of objectives)
422
Chap.7 System Engineering
(2) Process overview (describing sequences covered)

(3) Process flow sheet (illustrating entire process)
(4) Power supply diagrams (total power supply system, power supply
diagrams, dividing diagrams for sequence systems, etc.)
(5) Interface list (type, conditions, specifications, etc. of input/output
signals, etc.)
(6) Safety measures (redundancy, fail-safe measures, etc.)
(7) Operation description (sequence operations, operator procedures,
flow charts, time charts, etc.)
{g) Control panel design
The main role of a control panel is to provide centralized management of the instrumentation system installed throughout the plant. A
control panel normally consists of controllers, recorders, display units,
rack mounted instruments, and operation switches.
Control panels are usually located in an environmentally-controlled room from which various system parameters such as flow rate,
pressure, temperature and tank levels can be measured and controlled
as well as from which control valves can be manipulated, process status monitored, and warning and halt signals output.
The configuration of a control panel depends on a variety of factors such as its main function, installation location, size and instrumentation system. Moreover, in the case of instrumentation systems
employing one major type of instrumentation, conditions such as the
amount of operator information, operator's actions, display readability,
emergency actions, and fail-safe operations also influence the configuration of a panel. On the other hand, in instrumentation systems
combining various instrumentation equipment such as the increasingly
popular computerized and distributed DDC devices, information display, monitoring and operation functions are becoming centralized on
CRTs. As a consequence, the role of control panels is changing. It
should also be mentioned that as additional factors are considered such
as providing an efficient and pleasant control room environment, e.g.,
instrument/furniture layout, spacing, colors and lighting, control panels complementing the overall work environment have come to be
demanded.
Thus, in the design of control panels, a comprehensive engineering
approach must be taken which includes human engineering considerations. Refer to Sec. 7.3 for more details on control rooms and man-machine interfaces.
(1) Common control panel configurations
Some common panel configurations used today are the upright type, bench type, console type,
cabinet type, outdoor type and pipestand type, as illustrated in Fig.
7.17.
423
Vertical (Open) type
abi net type
Vertical (C lo e) type
Vertical
( lo e and wa lk through) type
Console type
Bench type
Outdoor type
Pipe stand type
Fig. 7 17 Types of control pa nel s
(2) Control panel instrument layout

Although operability and maintainability is first taken into account in deciding on the general configuration of a control panel, also of prime importance here is the
layout of instruments, i.e., positioning of switches, displays, meters,
knobs, etc., in particular as they effect the operator. Determining a final layout requires thorough consultation with the customer. An example of instrument layout for an upright type of control panel is shown
in Fig. 7.18.
(3) Design materials (engineering data)
For detailed design of control panels, the following list of materials are required. During the engineering phase, as control panel plans are being drawn up, these materials should be studied and used for establishing the various levels of
design, as well as to help narrow down specifications at technical meetings.
(1) General specification sheets: General design objectives, requirements standards, regulations, etc.
(2) Instrument list: instrument name, model, manufacturer, tag number, etc.
(3) Instrument configuration diagram: diagrams showing instrument
loop configuration
(4) Instrumentation system input/output list: number of input/output
points, tag number, terminal number, etc.
(5) Control room floor plan: entrance dimensions, space allocation,
424
Chap.7
System Engineering
t---
- --
1200- - - - - - - t
2400
2 200-
f-r===
Recommended
mo unting level
j _u
iu
2050
2 150
0
.....
OOAnnunciator display
1900
1800
1700
1540
1450
1400
1020
----
----
::::::::::
9cQ
.\l
~-
(JUJ()
= =Switches
0
980
------ ------ -
800
-----------------
At least 100
At lea t 100
* U nit : mm
0
100
Fig. 7 18
Floor level
Example of an in trument layo ut on a vertica l panel
Fig. 7 . 19 Control room and control panel layout
pits, etc. (Fig.7.19)

(6) Control panel external views, instrument layout
(7) Draft of graphic displays: P&I diagrams, plots, etc.

(8) Power supply, grounding diagrams: control panel and power distribution board tie-in
(9) Sequence action description : logic diagrams, flow charts, sequence
tables, time charts, etc.
(10) Annunciator display lamp layout and inscriptions
(11) Name plate inscriptions
(4) Preparation of control panel specification document
This document is made up of the various detailed specification sheets for the con7.2
Instrumentation System Design
425
trol panel. It should be brought along to technical meetings during the

engineering phase as a reference for modification and approval. These
specifications usually include the following items:
(1) General specifications: installation location, air-conditioning requirements, weight limits, size limits, structural format, number
of panels, dimensions, paint, etc.
(2) Structure: materials, thickness, etc. of front/back panel, side panels, top panel, doors; channel base, anchor bolts, hanging bolts,
etc.
(3) Wiring: Wiring format, power supply, incorporated power supply
equipment and power distribution board, power distribution
format, wiring materials and colors, wiring standards/regulations,
etc.
(4) Annunciators: Type of display lamps and warning bells and buzzers, layout, functions (non-lock-in, lock-in, double lock-in, etc.)
(5) Sequence circuit diagram: format, power, sequencer, input contact
points, wiring format, etc.
(6) Piping: air supply, air supply devices, air header, bulkhead, piping materials, etc. (Refer to Fig. 7.20 for a conceptual diagram of
an air supply device.)
(7) Auxiliary devices, parts: internal lighting, fans, maintenance power plug, replacement parts, etc.
Supply air
Fig. 7 20 Outline of an air suppling device
(5) Control panel design and CAD/CAM tools

Up to the point that
the final configuration and instrument layout of a control panel is finally decided upon, a good number of meetings will have taken place
during which engineering plans and drawings undergo repeated correction and modification. As a result, CAD (Computer Aided Design) sys426
Chap. 7
System Engineering
terns have been increasingly used as an aid in preparing control panel

proposals and approval drawings since corrections and modifications as
well as databasing can be easily and quickly performed. Usage of CAD
tools, which not only saves on labor costs and provides neat and professional output, has increased in all areas of control panel design and documentation to the point that hand-written documents are all but obsolete. In addition, on the manufacturing side, CAM (Computer Aided
Manufacturing) systems have long been in use for sheet metal and panel cutout processes. CAD/CAM systems have already proven themselves as effective in a wide range of functions from control panel
price estimation to its manufacture. From now on, attention will be
paid to increasing labor savings and raising quality levels.
(6) Submittal of approval drawings
Once detailed design work has
been completed based on engineering data received from the customer,
the following approval drawings and related documentation must be
submitted. Control panel manufacturing normally starts as soon as
drawing approval is received.
Control panel specifications, Channel base drawing, External view
drawings, Instrument layout drawing, Instrument list, Internal device
layout drawing, Instrument loop drawing, Sequence circuit diagram,
Power supply diagram, Power switch allocation, External terminal
designation, Annunciator list, N arne plate list, Graphic display
drawings, Miscellaneous drawings, etc.
(h) Digital instrumentation
Although distributed DDC engineering is not fundamentally different from general instrumentation engineering, it must incorporate a variety of functions which conventionally have been designed and constructed separately (e.g., control panels, graphic displays, sequencers,
records, loggers, operator instructions, communications). This is in
addition to DDC control functions. As a result, engineering has become
more concentrated than in other forms of instrumentation engineering.
Since the requirements, development methodology and functions of distributed DDC are explained in detail in Chap. 5, an example configuration and main functions only are summarized here in Fig. 7.21.
(1) Digital instrumentation engineering
The main engineering steps
here are specification confirmation, basic design, detailed design, work
sheet description, hardware specification determination and customer
software design. The engineering work flow is shown in Fig. 7.22. As
can be seen, in addition to strictly engineering processes, other processes such as hardware manufacturing, system generation (software
preparation), debugging and acceptance inspection are included as well.
(i) Confirming customer's specifications: The scope within distributed
DDC can be realized must be defined based on customer specification
427
Operator station
olor hard
Printer
Mai n functions
Operati on a nd monitoring
Messages (D isplay and voice
output)
Graphics
Reponing (Events)
Logging
Engineering
(System bui lder, y tern test,
System maintenance, Self
documentation)
Main functions
Feedback control fu nction
Sequence control functions
Annunciator funct ions
Calcul ating functions
Fig. 7 21
Exa mp le of a d istri buted DDC configuration a nd mai n funct ions
1~1
omrol functio n design process
Fig. 7 22
Production process
Flow chan of di tributed DDC engineering work
sheets and the results of technical meetings. Along with these initial
specifications, which may include P&I diagrams, input/output signal
lists, desired functions, and operation scope and methods, other factors
such as operability, controllability, reliability, maintainability and safety must also be considered when determining this scope.
(ii) Basic design: Basic design is performed based on the specification
sheets for the various functions desired by the customer. These include feedback control, sequence control, CRT display functions (e.g.,
operations, monitoring, trends, graphics, operator instructions, etc.),
logging, communications (e.g., items, frequency of use, processing) and
system configuration (e.g., grouping, redundancy, additions, modifica428
Chap.7
System Engineering
84. ,,II
18:11
IIIII III
REACTOR
Isme~ I11 mkl I
Im rml
( a ) Graphic display screen
11111111
!P!Gt !DE MESSAGE PANEL

II
84. 8. 11
16:41
8.11 W"
I REACTOR-2 WAITING EMPTY COMMAND

11 1.11 11'"
I REACTOR-2 COOLING STARTED
13 8.11 11:"
I REACTOR-2 HEAT ING STARTED
II 1.11 12168
I REACTOR-2 FILLING STARTED
II 8.11 11168
I REACTOR-2 WA ITING START COMMAND
MORE
( b ) Operator guide message screen
Fig. 7 23
Example of a distributed DDC and CRT display
429
tions). Examples of CRT displays for distributed DDC are shown in

Fig. 7.23.
(iii) Detailed design: Based on the basic design performed above, detailed design is carried out on feedback control loop diagrams, sequence control flow charts, logic diagrams, time charts, and display-design drawings.
(iv) Hardware manufacturing specification sheets: Manufacturing specification sheets related to hardware factors are prepared from such documents as system configuration drawings, man-machine function descriptions, detailed input/output signal lists, and equipment layout diagrams.
(v) Work sheet description: This document is prepared based on the
results of detailed design work.
(vi) System generation: Based on the contents of work sheets, a control
system for the user is constructed through use of a system generator,
and typically, a system floppy disk is prepared for customer use.
(vii) System debugging: In the system debugging phase, individual system components (e.g., operator console, field control station) are first
checked out individually, followed by a comprehensive system-wide
test. These tests should be performed according to previously prepared
test procedure manuals and the test results recorded on prepared
forms.
(viii) Acceptance inspection: Final approval of all system functions
should be performed by carrying out a customer-attended final acceptance test (FAT) based on a previously prepared final-acceptance inspection manual. The results of the final acceptance test should be recorded on a prepared form. Figure 7.24 shows the engineering documentation system for distributed DDC. An example of the guidelines
for document transfer between the customer and manufacturer is
shown in Table 7.6.
7.2.4 Instrumentation work specifications
In this stage, the information listed below is collected, organized
and verified as needed for carrying out instrumentation work.
Plant layout diagram, Facility layout diagram, Equipment placement diagram, Equipment external views, Plant piping diagram,
Drive system documentation, Safety standards, Facility interface
list, Plant wiring diagram, Duct/pit diagram, Materials list, Work
schedule, etc.
In particular, during the course of technical meetings with the customer, the above information can be used as a basis for arriving at detailed specifications regarding the scope of instrumentation work, process and facility interfaces, instrumentation layout, wiring, piping, etc.
430
Chap.7
System Engineering
"''
~
....
lit
aq
"'[
lit
~
lit
"'
!IEDUNDANCY
\
L=RJNCTIONS
OVERVIEW
MESSAGES
TREND POINTS
LOOOING
G RAPHICS
MAN MACHINE
SEQUENCE
RJNCTIONS
CONTROllED
ITEMS
METHOD OF
OPERATION
CONTROL INSTR UMENT

FLOWCHART
DOCUMENTS PREPARED
BY CUSTOMER
INTERFACE
SPECIFICATIONS
Fig. 7 24
FilLING IN THE
WORKSHEETS
'
INSPECTION
APPROVAL
OPS : Opera1or station

T l : Techn ical information
IM : Instruction manual
GS General specifications
SE Sales e ngeener
US User work sheets
MONITORING
FORTRAN UTIL111E
STATION BUILDER
STATION BUILDER
'\=BUILDE R
Engineering documentation system for distributed DOC
USER WORK SHEETS
OPS USER
SPECIFICATIONS
IPBCIJIICA110NS
'onOI!HDAL
B:::~010N_:~AN;~~~~~r
Q!I.APHIC BUILDER
_j
\1iEM
- - - - --- _ _
PS SYSTEM
~
_ ECIFICATIONS
'
FUN~IONS ~~RK~EET ENTRY-~UIDE
\
~~ A. USER'S
DISPLAY PANEL AWORKSHEET ENTRY GUIDE
( P & I)
I GEN~L
SPECIFICATIONS(~~
OPS
TECHNICAL INFORMATION Tl.__ __ _
DOCUMENTS PREPARED
FOR EACH JOB
"'-
"'"'
;:,!
"'~
"'
.....
"""
~
1 . Explanation of functiona! specs. (by manufacturer)

2 . Explanation of any
deviations from customer specs. (by manufacturer)
3 . Request for approval
1 . Explanation of customer specifications (by

customer)
2 . Explanation of method
of operation (by customer)
3 . Man-machine interface
Display panels (proposed)
Confirmation of custamer specifications
Presentation of basic
design specifications
1 . Confirmation of whole
process from meetings
up to shipment and
start-up
2 . system configuration
3 . Interfaces with other
equipment
4 . Reasons for wanting
new system
5 . Confirmation of other
rules
System configuration
and job overview
Item
Purpose
Steps
0
0
0
0
Graphic panel(proposed)
Overview panel configuration (proposed)
Completed user worksheets
Sequence control data

Flow charts
Logic diagrams
Time chart
Verbal description
Feedback control
Verbal description
Man-machine interface
specifications
Control function flow

diagram (PID)
List of 1/0 signals
Requested specification
sheets
Proposed method of
operation
Documents that the

customer supplies to
the manufacturer
0
0
Functional specifications
Basic design specifications
Feedback control
composite loops
Sequence framework
design
J\pproved hardware
manufacturing specifications and drawings
Original user work
sheets
Factory progress schedule

Organization chart
Hardware manufacturing
specs. (for discussion)
List of documents requested
Documents that the

manufacturer supplies to
the customer
Table 7. 6 Guidelines for document transfer between the customer and manufacturer
:z
;:,:
aq
"'
;;i
""
"'....
;:,:
~
....
c:;
;:,:
""
;;i
<:!
;:;
"'....,....
t-v
:-'1
Item
1 . Confirmation of detailed functions based
on design specifications
2 . Final request for approval
3 . Documents sent to
customer before
factory acceptanc e test
Confirmation of schedule
Purpose
Confirma tion of design specifications
Steps
Documents that the

customer supplies to
the manufact urer
o
Design specifications
Feedback control loop
diagrams
Sequence control flow
chart
Logic diagram
Graphic panel specifications
Logging specifications
Documents that the

manufact urer supplies to
the customer
In regards to wiring, particular attention should be paid so that the

wiring path or wiring method does not cause noise or other adverse affects in other plant facilities, and the maintainability of such wiring
should be seriously considered. In addition, plans for actual work operations must be drawn up specifying work management, labor requirements, schedule, safety management, emergency communication system, etc.
7.2.5 Related work
In addition to the engineering work described above, the following
processes or inter-department support will also be required as a part
of system engineering or project engineering work.
(a) Coordination
In this stage, which precedes detail design, technical information
gathered as a result of basic design work (engineering) is passed on to
the manufacturing department (order entry). Specifically, the material
listed below is prepared reflecting overall job planning. Moreover, an
initial kick-off meeting should be held for presenting a job overview
and describing system specifications to detailed-design groups and related departments.
System configuration diagram, Comprehensive flow sheet, Loop
flow sheet, Power supply system diagram, Equipment/device list,
Function specifications, Procurement specifications, Instrumentation work specifications, Control panel specifications, Sequence
specifications, Work sheets, Job overview, Detailed schedule,
Work organization chart, etc.
(b) Approval documentation
After a sufficient number of technical meetings have been held
with the customer, the various specifications decided upon should be
collected together and organized in the form of an approval document.
On the basis of this document, the customer and instrumentation manufacturer should then mutually arrive at final approval. In regards to
the form of this document, although one overall project approval document is ideal, there are also cases in which, due to manufacturing
schedules and deadline considerations, multiple volumes are prepared,
such as a system volume, device/function volume, control panel volume, instrumentation work volume, etc.
After checking approval documentation, the customer returns it to
the instrumentation manufacturer. At this time, a certain amount of
corrections and modifications is normal, and in some occasions major
modifications may be required. In any case, such modifications should
be made as quickly as possible and specification revision procedures instigated within manufacturing. Moreover, since specification changes
434
may significantly effect current functions and cost estimates, the manufacturer should meet with the customer again to go over the contents
of these changes. Discussions will include negotiations as to how the
cost of such changes will be shared and how they will affect the manufacturing schedule.
(c) Inspections
In general, inspections are divided up into in-house inspections
and customer inspections.
(1) In-house inspections
The individual in charge of the basic design
for a particular project task must verify that the target quality has
been achieved.
(2) Customer inspections
A similar inspection to the one described
above is also carried out by an individual on the customer's side involved with design work or quality control. A quality inspection certificate must be prepared at this time. In addition, in the event that specification changes occur during inspection, these must be closely verified
by reviewing specification change sheets, correction directive sheets,
and the minutes of technical meetings.
(d) Shipping
Shipping and acceptance conditions are often indicated in separate
documents or by contractual provisions with the customer.
(e) Instrumentation work
Support must be provided for instrumentation installation specifications and items to be negotiated with the customer from the standpoint of a job designer.
(f) Start upftest run
At start-up time, a job designer will perform device adjustment,
device running, etc. through mutual consultation with the customer,
and if necessary, witness an actual test run.
(g) Final processing
Final design documents must be prepared which reflect all changes
made after the submittal of approval documents ap.d which conform to
the system as actually delivered to the customer. These documents
should be compiled in the same way as that of the approval documents
for consistency sake. In addition, since the customer will mainly be using final design documents for maintenance purposes, instruction manuals and maintenance manuals should be included as well. The type of
binding should also be selected so as to prevent excessive wear due to
frequent use or long-term storage. Needless to say, customer feedback
and agreement as to the form and contents of final design documents
should be obtained in advance of preparation.
435
(h) job summary

A job summary document is useful for recording the specialized
know-how obtained during the course of a job, and as such can be an
invaluable asset. This document should concentrate on describing new
technology appearing throughout the job from basic design to system
operation. Moreover, it should serve as a useful reference for future
instrumentation jobs.
7.2.6 Instrumentation drive system design
(a) Instrumentation power supply
The instrumentation power supply provides power to instrumentation equipment enabling it to collect information from throughout the
plant and to take appropriate system actions based on this information.
Thus, power to such equipment, which can be looked upon as the
plant's brain and nerve system, is of supreme importance. Moreover,
in the event of power reductions or total power loss throughout the
plant, plans must be made to ensure safety in all plant processes. Common items requiring attention in instrumentation power supply design
are listed below:
(1) Countermeasures to power outages
LMomentary outage
Extended outage
(2) Countermeasures to power fluctuationstVoltage fluctuation
Wave form distortion
Frequency fluctuation
(3) Load characteristics and levelsPriority devices
General devices
Sequencers
Computers
Miscellaneous
(4) Power capacity selection------- Plans for increase, etc.
An example of a power supply system diagram for an instrumentation system is shown in Fig. 7.25. Note that the private power-supply
generator shown here is intended for use during extended power-down
periods and moreover takes several minutes to start-up. Consequently,
plant operations must be stopped for a short time to change over to
the power supply system.
Due to the widespread popularity of microprocessors in instrumentation equipment, it has become difficult to configure a system
with equipment not having them. This situation has demanded that
power supplies for modern instrumentation be of even higher quality
than traditional systems. The following paragraphs describe power sup436
t _______________ .
Fig. 7 25
Power supply ystem diagram for instrumentation
plies for modern instrumentation use and countermeasures for momentary power loss.
(1) Uninterruptible power supply
Since instrumentation equipment
plays the role of the plant's brain and nerve system, power supply problems that can halt operations can cause major damage to plant business. In addition to power outages, since microprocessor memory protection processing is immediately initiated if power voltage fluctuation
417
UPS (Or imply invertor)
Direct connection
I
to commercia l o - + - - - - - -- + - - - - - '
power ource etc.
__j
ote : CVCF ( Cons tant Voltage
Fig. 7 26
onstant Frequency )
Uni nte rruptible power system for in trumentation
or wave form distortion is detected, even within one cycle, small

amounts of fluctuation or distortion cannot be tolerated. Consequently,
it has become common to supply power to instrumentation in the form
of an uninterruptible power supply (UPS). A typical UPS configuration
is shown in Fig. 7.26, and standard requirements for a UPS are described below :
(1) Minimal maintenance, continuous operation
(2) Extended operation during power outages
(3) Power quality (e.g. , voltage, frequency, wave form) as required
for instrumentation equipment functions
(4) High-reliability
(2) UPS operation
In the UPS circuit diagram of Fig. 7.26, power is
usually supplied to the load from the AC current terminal via the rectifier, inverter and SCR switch. However, in the event of problems in
the AC input or the inverter circuit, power will automatically be supplied from the battery (instantaneous switchover). Although the battery capacity in such a structure varies depending on the estimated
length of power-down time, a typical backup capacity is 30 to 40 minutes. Note also in this configuration that in the event of a faulty inverter, the SCR switch will automatically changeover to direct commercial power input. In selecting an appropriate UPS, points which must
be considered are rush current, peak current, and wave form distortion, as described below.
(3) Rush current
The inverter possesses drooping, or cut-off, characteristics for times of overload as a self-protection mechanism. As a
result, rush current in the load can have negative effects. Although all
electrical devices possess some form of rush current characteristics,
devices requiring particular care are those having rectifier input forms
(capacitor input forms). In particular, computers, instrumentation
equipment and communication devices usually have internal single-phase all- wave rectifier circuits in which, as shown in Fig. 7.27,
438
Chap. 7
System Engineering
Rush current wavefom
Fig. 7 '1:1
Rush current waveform
rush currents are generated due to capacitor charging when current is

applied. The amount of this current can reach ten times or more that
of the stationary load current.
(4) Peak current
A load in which a sinusoidal current flows on application of a sinusoidal voltage is called linear load, and a non-linear load
otherwise. A typical example of a non-linear load is that having a capacitor input form. In this case, the peak value of the current wave
form can become 2.5 times the effective value, and the corresponding
voltage wave form will exhibit significant distortion, as shown in Fig.
7.28.
Input voltage waveform

Waveform is distorted
-1-- -\ --
-f--+-Hr---
+--'- by now of peak current
Fig. 7 28 Peak current and voltage waveform
(5) Wave form distortion

In order to detect abnormal voltage within
instrumentation devices, peak value detection and instantaneous value
sampling are commonly used. Since, as described above, distortion will
be exhibited in the voltage wave form due to a non-linear load, systems in which instrumentation equipment have a large nonlinear load
will require the placement of a rectifier on the UPS side as a load countermeasure.
In addition, even if the effective value of an input voltage wave
form having distortion is the same as that of one not having distortion,
the terminal voltage of its smoothing capacitor will be lower, and the
instrumentation equipment may detect a power-down situation. For ex7.2 Instrumentation System Design
439
A : Ideal input voltage waveform

B : Distorted input voltage waveform
Example:
Effective Voltage across
value
capacitor
A IOOV AC
138V DC
B IOOV AC
120V DC
Fig. 7 29
Distorted input voltage waveform
ample, as shown in Fig. 7.29, even though input voltage wave forms A
and B both have an effective value of 100 V AC, the terminal voltage
of the smoothing capacitor for the B wave form will be lower, or in other words, the voltage supplied will be equivalent to an 85 V AC wave
form without distortion.
(6) Power-supply equipment output capacity
Assuming a linear load,
the output capacity of power-supply equipment (UPS) will be characterized by a sinusoidal current in which voltage and phase agree. For a
non-linear load, however, the required output capacity must be determined with due consideration given to the current wave form and
phase in addition to the power needed by the load.
Also to be considered when determining the output capacity of
power-supply equipment are such factors as equipment type, format,
and the manufacturer. Based on years of experience and actual equipment results, criteria for selection of power-supply equipment are
shown in Table 7. 7.
Note that the output capacity obtained from Table 7. 7 is greater
than the power required by the system. This factor does not represent
simply a margin for the power-supply equipment but is the power capacity necessary for normal operation in the case of non-linear loads.
In general, the power-supply equipment manufacturer should also
be consulted to help determine output capacity requirements, or to assist when power-supply equipment already procured do not satisfy output capacity selection criteria.
In addition, power-supply equipment on the host side should be
given the same consideration as described above. This also holds for
UPS when determining its output capacity as host-side power-supply
equipment, since it possesses rectifier load characteristics.
(b) Instrumentation air supply
Similar to the instrumentation electric power-supply described
above, the instrumentation air supply also plays a major role as a drive
440
:t
......
;:;
(J>
"'
aq
<;:"
;:;!
(J>
;:;
c;
;:;
.,........"'
;:;!
(J>
.....
'l
0
0
0
AVR (automat ic voltage

regulator )
MG(mot or generato r
set)
UPS (Uninterruptible
power system)
Excessive
variation
in commercial
frequency
High susceptibility
of commerc ial
line to highfrequency noise
1.5
2
3.5
Output capacity
High liability of
factor
commercial power
to momenta ry failure (service factor)
(Note).
(or where the system
equipment requires a
protectio n against a
momenta ry power
failure within 0. 5 sec.)
High
Low
Price
(X : not good; 0: good; : best)

ng the total power demand of all the system
Note: The installed capacity of the power facility is calculate d by multiplyi
to determin e the _output rating of a specific
measure
a
as
serves
factor
equipmen t by the output capacity factor. This
it is usual1y necessary to provide
specified,
value
power facility. If the output capacity factor is set lower than the
currents.
uency
high-freq
some protectio n against
Excessive
variation
in commerc ial
voltage
Isolation
transform er
Type of power facility
Table 7. 7 Criteria for the selection of power facilities .
source for the instrumentation system. Although the use of pneumatic

instruments has been decreasing in recent years, they still find use in
some areas such as processes susceptible to explosions or local control
devices. In addition, the majority of control valves are still driven
pneumatically. As for quality requirements, an air supply must be
clean and stable, and measures must be taken for pulsation absorption,
momentary power loss, and compressor breakdown. In the first two
cases, a typical solution is an air supply tank, while for compressor
breakdown, multiple compressors or switchover to a factory air supply
are often used. Another example would be the temporary use of N2
gas, or for that matter any technique which is both safer and less expensive. An example of air-supply equipment for instrumentation use
Power source
Fig. 7 30 Instrumentation air source system diagram
(1) Air supply quality

Air purity for instrumentation use must in
general meet the values listed below. Compressed air can be purified
by such devices as coolers, multi-stage filters, drain separators and
dehumidifiers.
Foreign particle diameter:
10 ,urn or less
Foreign particle density:
0.1 mgjm or less
Humidity (dew point under pressure): 10oC or less
Oil droplets:
1 ppm or less
Harmful gas:
20 ppm or less
Even if the air used for instrumentation eventually reaches atmospheric pressure, dry conditions without condensation must be insured.
Figure 7.31 shows the relationship between atmospheric pressure dew
point and dew point under pressure. If we set the dew point under
pressure to 10C, theatmospheric pressure dew point for a pressure of
7 kgfjcm 2 becomes -17C, and dry air with no condensation can be obtained anywhere in the air-supply system (the dew point may be lower
depending on the environment). Water content in the air can be elimi442
Chap. 7
System Engineering
[kgf/cm 2]
~ :/""./~'/'/'
60
""' ~~~7
C(~
50
g~ ~~":.-V//
/.
40
E 30
.,a
a.
"
~~
~ ~~~ V/
""'
;:
a.
~ -1
Cl
~ ~ ~~V/ /
0~ ~ ~ P'/": ~-"~/ /
0~ ~ l:%:'"t('V/ /
-2 0~
~
-3
~ v.,..-_....
~k':.::'
-4 oiL
-40
v ~~\.\C:.......-
o?' vv
1/1
I
-30
"0
:/" v
~ /"'/V
~'~ ~
~ L'l". . . . ' / V
20
7
~f"'h ~~/i'(/
.,~
~ ~V/
~ r.
~ ~:/2 v/v /
.'!l 0-::: ~ v~
v
.7
-..............
v v
..............
~{r<{r<Y.,'/,
/ / l:4~~~~
..............
/'
-20
-10
10
0
Atmospheric pressure dew point ["C]
Using this chart :

This chart is used for converting the dew
point under pressure to the atmospheric
pressure dew point and conversely.
(ex.) To determine the atmospheric pressure
dew point under a pressure of 7kgf
/cm 2, draw a horizontal line from IO"C
point on the vertical axis over to the
7kgf/cm 2 line, and then draw a vertical
line from this point of intersection to
the horizontal axis. The atmospheric
pressure dew point is found to be
- l?"C as shown in the figure to the
right.
'
~b
\."~-
20
E
"~
i: 1O"C 1-------,,F
aa.
0"
QL---~--~~-------Atmosphric pressure dew point ["C]
Fig. 7 31 Dew point conversion chart between atmospheric pressure

and indicated pressure
nated by dehumidifier devices, which may use a variety of methods

such as mechanical refrigeration and silica-gel. Each method has its advantages and disadvantages depending on environmental conditions,
and should be selected accordingly.
When determining
(2) Determination of air-compressor capacity
air-compressor capacity, it is common to treat it as a function of total
instrumentation system air consumption. The capacity of air-supply
equipment is expressed by Eq. (7.1):
(7.1)
443
Here, Q0 is the air-supply equipment capacity rating (Nm3/h), q is the

total air consumption for all devices (Nm3/h), and K 1 is a safety factor
(2.0 to 2.5).
In order to clarify device air consumption, we quote its definition
from JIS standards (JIS C 1801): "Given that a device is supplied with
its rated air pressure, if the amount of air flow is measured at various
points throughout its operation range, the maximum air flow value obtained is taken to be the amount of air consumption for that device."
The amount of air consumption for devices operated intermittently such as shut off valves and cylinder valves is given by Eq. (7.2):
(7.2)
Here, Qn is the amount of air consumption (Nm 3/h), Vis the cylinder
or diaphragm volume (!), P 1 is the supplied air pressure (kgf/cm 2), and
n is the number of operation cycles per hour (cycles/h).
The air reservoir (backup tank) is installed to smooth the pressure
pulsations from the air-compressor and also serves to temporarily support control valves and other equipment if air-compressor operation is
halted by a breakdown or power loss. Accordingly, air system parameters such as air consumption, backup time, and the lower limit of the
supplied air pressure must be investigated taking into account operability, maintenance and safety of the entire instrumentation system. A
nomogram for selection of air reservoir capacity is shown in Fig. 7.32.
The air reservoir capacity can be calculated by Eq. (7.3) below. A
backup time of 30 to 40 minutes is common.
(7.3)
Here, V is the tank capacity (m3), q is the total device air consumption
(Nm 3/h), T is the backup time (h), P 1 is the normal pressure = supplied pressure (kgf/cm 2), and P 2 is the minimum operating pressure
(kgf/cm 2).
7.2. 7 Other system functions (safety, failsafe and redundancy measures)
Any discussion of safety measures for an instrumentation system
must necessarily involve the reliability of instrumentation devices making up the system. If this reliability cannot be grasped quantitatively,
then it is impossible for system reliability to be predicted qualitatively
by such predictive analysis methods as FMEA (Failure Mode Effect
Analysis) and FTA (Fault Tree Analysis). A positive approach to improving the safety of each individual instrumentation device must be
adopted.
444
Chap. 7
System Engineering
Pressure difference AP =Normal pressre P1 -Minimum operating pressure P2

For example, to obtain an appropriate reservoir capacity for:
Air consumption: 60 Nm 3/h
Backup time: 30 minutes
Normal pressure (P 1): 7 kg/cm 2 and
Minimum operating pressure (P2): 2 kg/cm 2,i.e.
Pressure difference (AP): 5kg/cm 2
( I) A line is drawn vertically upwards from the air consumption axis (at 60 Nm3 /h),and intersects the backup
time parameter line corresponding to the desired backup time value at point (a) (here 30 min).
( 2) A horizontal line is drawn to the left from point (a),
and intersects the pressure difference parameter line
corresponding to the given pressure difference value
at point (b) (here 5 kg/cm 2 ).
% (3) From point (b) a vertical line is drawn downwards
f' ,o
to point (c) where it intersects the air reservoir capa"" ~ city axis. Point (c) gives the required reservoir capa~~city (6m 3 ).
~--
??::~
20
18
16
14
12
10
8
6
Air reservoir capacity m 3
80
100
120 140150
Air consumption Nm 3 /h
Fig. 7 32 Air reservoir capacity selection nomogram
Some examples of methods for improving device and system reliability such as explosion-proof and earthquake-proof techniques,
redundancy and failsafe systems are described below.
(a) Instrumentation device safety measures
(1) Explosion-proof construction
Instrumentation devices used in areas containing explosive gas or dust must not be the source of sparks
or heat which may set off an explosion. Methods used as a countermeasure to explosive gases include explosion-protected, pressurized, increased-safety, and intrinsic-safety techniques. Countermeasures to
flammable dust consist of both special and standard dust-proof configurations, while special dust-proof configurations are used for explosive dust. These are explained in more detail below.
(i) Explosion-proof technique: In order to achieve an explosion-proof construction, an enclosure must be strong enough to withstand the force
445
of an explosion caused by the chance ignition of an explosive atmosphere which has somehow penetrated the enclosure from the outside.
In addition, gaps in enclosure components such as the spindle and composition plane must be appropriately designed so that flames generated
by an explosion are cooled as they escape to the outside, resulting in a
temperature below the point at which explosive gases in the immediate
area can ignite.
(ii) Pressurized technique: In this technique, the internal pressure of an
enclosure storing air or an inactive gas is maintained at 5 mmH 2 0
above the pressure outside the enclosure. As a result, an explosive atmosphere is prevented from penetrating the enclosure.
(iii) Increased-safety technique: This is a commonly used technique for
ensuring the absence of any factors which may act as a source of ignition. Specifically, a safety factor is added to the creepage distances
and clearances and to the temperature-rise value associated with winding so as to lower the possibility of any form of ignition.
(iv) Intrinsic-safety technique: Here, a system is configured so that it intrinsically does not have the capability of igniting an explosive atmosphere not only during normal operation, but even in the event of arcs,
sparks, or overheated components due to accidental shorts, groundings, or open wires.
(v) Special dust-proof technique: In this technique, a sealed enclosure is
made dust-tight so that no outside dust can penetrate it.
(vi) Standard dust-proof technique: In this case, a sealed enclosure is configured so that outside dust cannot easily penetrate it.
(2) Earthquake-proof construction
For plants in which abnormal
reactions to earthquakes may have a significant impact on society (e.g.,
nuclear power plants), earthquake-proof construction must be applied
to instrumentation devices, in particular, to related safety assurance
mechanisms. Earthquake-proof construction should also be considered
for control panels in control rooms and outdoor cubicles.
(b) Instrumentation system safety measures
(1) Redundant systems
Due to the fact that the reliability of instrumentation devices can only be raised so far, and since devices which
are totally fault-free do not actually exist, system reliability can only
be raised further by incorporating duplicate, triplicate, or higher redundant systems.
Basic types of redundant systems and their corresponding MTSF
(Mean Time to System Failure) are shown in Table 7.8. As can be
seen, incorporation of a maintenance system (repair work is performed
at the time of a breakdown) has a significant effect on the MTSF of a
redundant system. These formulas can be easily applied to instrumentation devices. For example, assuming a device with a MTBF (Mean
446
Chap. 7
System Engineering
Table 7 8 Redundant systems and MTSF

Redundant system
MTSF
Non-maintenance
system
Parallel~
A 1
redundancy
A 2
MTSFo=
Maintenance system
i..1
MTSFo=t
MTSF=t+-t
2-out-of-n redundancy
MTSF0
- 1
- 1111
MTSF=MxMTSFo
(n-1)..1
_1_+
fl
M - n-1 n(n-1)..1
1
1
--+n-1 n
Time Between Failure) of 105 hours (breakdown rate A = 10-5/h), the

improved reliability in the case of a parallel redundant system would
be given as:
Non-maintenance system:
An improvement of about 1.5 times can be obtained from the
result of MTSF 0 ::::::: 1.5 x 105 (h)
With a maintenance system:
An improvement of about 500 times can be obtained from the
result of MTSF ::::::: 5 x 107 (h)
(assuming a repair time of 100 hours: repair rate fJ. = 10-2/h)
The parallel redundant system of Table 7.8 is often used in the
case of duplicate detection probes or relays, in which a normal condition from at least one of the two is acceptable. The stand-by redundant system is commonly employed in battery backup systems for DC
current. The 2-out-of-n redundant system consists of n devices and is
based on the principle that if 2 devices suffer breakdown at the same
time, a system breakdown has occurred. Such a system is commonly
used for parallel operation of a DC power supply box in an instrumentation system. Moreover, the configuration of n=3 is especially appropriate for 2-out-of-3 majority selection formats commonly used for shutdown (emergency halt) sequences. An example of a 2-out-of-3 selection format as used in a nuclear reactor neutron measurement system
447
Safety operation
uclear reactor
Indication. Recording.
ontrol operation
. -- - - - So ON. OFF
r+- - -- -
r+-1-- -- - -
D : Detecting element
A : Amplifier system
J : Failure judgment sy tern
M : Mean circuit
L : 2-out-of-3 logic circuit
S, 0
S, 0
, OFF
, OFF
M 12 to M 3 1 : Deviation detector
OR : OR circuit
4 :
3-channel to 2-channel changeover switch
T : Trip circuit
Fig. 7 33 2-out-of- 3 measuring sy tem with fa ilure j udgment system

(2) Failsafe measures
In a failsafe device, the occurrence of a fault
will result in measures which force equipment conditions to move in a
safe direction. Two examples of failsafe systems are described below.
(i) Alarm system: The configuration for a failsafe alarm system is
shown in Fig. 7.34. The main point of consideration here is whether
the output relay should be normally energized or energized only during
abnormal occurrences. If energized only during abnormal occurrences,
then if by some chance the power to the alarm system is shut off, the
relay will remain in an deenergized state, thus indicating a normal system condition regardless of the input signal. This configuration has
the danger that power failure cannot be noticed.
Signal
Normally energized
output reiay
Power _ _ _ _ __j
Fig. 7 - 34 F ailsafe alarm system
448
Chap. 7
System Engineering
If, on the other hand, the relay is normally energized, then if the
power is cut-off, the alarm will sound regardless of any input signal,
advising of an abnormal condition. Thus, an alarm system in which its
output relay is normally energized is considered to be a failsafe system.
(ii) Reverse-signal transmitter: In the temperature control loop of Fig.
7.35, if the output signal from a transmitter employing positive characteristics goes off, the controller will interpret this as a low-temperature signal, and will begin to adjust the system to higher temperatures. Since this configuration does not provide for any safety measures in the event of an erroneous off signal, it is not a failsafe system.
Temperature
tramsminer
Fig. 7 35 Temperature control loop and reverse characteristic tra n miller
If this control system is based on reverse characteristics, however,

an off transmitter signal will signify a high-temperature condition to
the controller, resulting in adjustment to a lower temperature state,
i.e., a safer or failsafe condition.
Thus, for systems in which high values of measured variables such
as temperature and pressure can be dangerous, failsafe measures can
be implemented by applying reverse characteristics to the early stage
of a loop, i.e., detection probes or transmitters. Conversely, if low values of measured variables are dangerous such as in the case of a tank
level control loop, application of sensors or transmitters having positive characteristics will produce a failsafe system.
(3) Distributed DDC redundant configurations
A one-unit distributed
DDC configuration usually has about 30 to 40 loops in its control domain, and about 100 total points if monitoring functions are taken into
account. Such a system obviously requires high reliability, and a common method for raising the reliability of distributed DDC employing
microprocessors is duplicate systems. Moreover, in addition to this pri7.2 Instrumentation System Design
449
~1111
~.0
OPS
OPS
II
Bus
Dual-redundant bu
:---------------------------& ---- ------&----SCN
IPS
IY,
~t
DX
PS
~~:~ : :::J
I'
''
Signal conditioner
I / 0 Nest
uu
<<
0-T
/"\
Dk:J
-------
vu
,._:;..-;.
~~
Dua l- red undant power supply
r;1r
V>
0-
iii
0-
Dual-redundant mu lt iple
in put / output card
-=
Dual-redundant internal bu
l
I
'
' t:
''
I
'
L -- ------- ----- ----------- ----- ----------- ---- --~
Fig_ 7 36
Example of a redunda nt configuration fo r di tributed DDC
mary objective of improving the reliability of a single unit system, a

duplicate system can achieve system objectives in a far more cost-efficient manner. An example of a redundant configuration for a distributed DDC system is shown in Fig. 7.36. This configuration features
the following redundancy techniques.
(i) Duplicate OPS (Operator station): By duplicating the OPS, a system
characterized by mutual backup can be configured.
(ii) Duplicate processors: With two processors, one processor can automatically take over process control in the event that the other processor breaks down.
(iii) Duplicate internal buses: By duplicating the internal bus which connects the processor and 1/0 nest, data transfer can be continued in a
normal fashion by one bus if the other one becomes inoperable.
450
Chap .7
System Engineering
(iv) Duplicate power supplies: By duplicating the I/0 nest power supply
card, one power supply card can continue to supply power to the 1/0
nest if the other one breaks down.
(v) Duplicate multipoint input/output cards: By duplicating the analog input/output cards used for multipoint control, process control can be
transferred automatically to one card without influencing the process
if the other card suddenly breaks down.
In addition to the need of deter(4) Instrumentation system interlock
mining how to effectively implement a plant "trip" as a safety measure
for process instrumentation, of more importance is how to plan for
safe intermittent plant operation in response to instrument faults. The
following describes two methods for safe operation based on an analog
instrumentation system having internal decision-making functions.
(i) Temporary locking of a final control element: Although use of this
technique may be naturally expected, it is still an emergency measure.
However, in the case where the problem is determined to be a general
one and not partial or localized, this technique is obviously not suitable. For major system faults, step-wise action for bringing about a
plant trip is necessary.
Situations appropriate for locking of final control elements can be
classified into instrument faults (abnormal signals) and loss of electric
or air supplies.
In the case of instrument faults, the type of fault can be determined by detecting abnormalities in signal absolute value, signal rate
of change, relevant process variable ratios and process variable deviation (exceeding tolerance values). It should be noted here that except
for particular transmitters and clear-cut faults such as thermocouple
open wires, it is difficult to definitely distinguish between instrument
faults and actual process abnormalities with the above type of abnormal signals. Such a situation is the basis for adopting a redundant control system. For example, in the case of duplicate transmitters where
excessive deviation between their two output signals has been detected, both transmitters can be checked during locking of the final
control element and the faulty transmitter disconnected. The control
loop can then be switched over to the operable transmitter and the
interlock cancelled.
In the case of partial electric power loss or air supply loss in pneumatically operated final control elements, locking of final control elements will naturally take place as an emergency measure; direct move
to a failsafe condition is not allowed here.
Depending on the importance of the process or control loop, some
power-loss countermeasures are separate power supplies for each loop
(e.g., battery backup), dedicated power supply for final control ele7.2 Instrumentation System Design
451
ments only (remote manual operation), and local manual operation.

As an example of interlocking during power loss, consider a pneumatic drive mechanism in which interlocking is performed by cutting
off the air pressure circuit of the diaphragm or cylinder section. In order to effect a "bumpless" return to automatic mode when the normal
air supply is restored, the position of the final control element must be
detected during interlock. In addition, controller output which is performing reset windup due to feedback must be pulled back. In the particular example of Fig. 7.37 which incorporates a pulse/pneumatic pressure converter along with a pneumatic pressure/current converter, the
current is substituted by the position signal of the final control element and is feed back. The PMD (Pulse Motor Drive) in the figure performs tracking for the controller up until the output signal agrees with
the feedback signal from the operating air pressure. Following this, automatic mode is restored, and a bumpless transfer performed.
---,
Tracking
I
I
I
--! ---r----I
I
I
i Pulse
I
Power interlock
output
I
I
:
Pressure switch
CD
Pneumatic pressure/ Current

Pul e/ Pneumatic pressure
Fig. 7 37
Interlock fo r air-supply los
(ii) Running at reduced operating levels : In the event that one side of a
group of auxiliary plant equipment running in parallel is tripped, the
flow rate or combustion rate of the associated system must be dropped
to a load corresponding to the capacity of the remaining equipment.
For example, in the case of boilers in which the rate of combustion is
determined by a boiler master signal, if a trip occurs in the water supply pump or the forced draft fan of boiler number 1, the boiler master
signal is reduced (runback) in coordination with the turbine governor
to a value corresponding to the capacity of the remaining boilers.
452
Chap.7
System Engineering
7.3 Control Room and Man-Machine Interface

The massive buildup in industrial plants and the improvements in
efficiency in recent years has also been accompanied by major changes
in instrumentation systems, as reflected by the move from analog
equipment to distributed DDC, and the rapid adoption of CRT -based
operations as a man-machine interface. Moreover, as a result of the
high-performance and high-capacity systems made possible by recent
advances in electronics and industrial materials, system monitoring
and equipment operations has also undergone major transitions, leading to an increased burden on the operator. This situation has required
designers to reevaluate the objectives and operability of control panels
as centered about the operator, and at the same time has forced them
to look at the problem from a total-engineering perspective combining
both human engineering and instrumentation engineering elements.
7.3.1 Human engineering and control panel design
(a) The role of the control panel
As the brain-center of the entire system, the control panel plays
the very important role of operating the plant in a safe and efficient
manner. To do so, however, it must possess control and monitoring
functions to act as in intermediary between man (operator) and machine. Thus, control panel design must consider operability, maintainability and safety in a way that conforms to the type of plant and its
objectives, while at the same time easing the burden on the operator
and providing for overall safe operation of the entire plant. The role of
the control panel is outlined in Fig. 7.38.
Alarm display.
Indication,
Recording
Higher of productivity
High quality , Lo w cost,
~Di versity
C/ine
Computation,
Control
Control
panels
Safetv operation,
Good environment
Field
instruments
Equipments
<D
Input
Output
Fig. 7 38 Role of a co ntrol panel
7.3 Control Room and Man- Machine Interface
453
(b) Body measurements and control panel design

Although in Japan various studies have been made on the physical
dimensions of the human body, data expressly for control panel design
has not yet been published. Here, basic body dimensions as taken from
human engineering studies are established as a basis for control panel
design, and standard configurations for control panel monitoring and
operations are proposed based on the results of extensive experimenting with these values.
(1) Operations and monitoring in a seated position (Fig. 7.39)
880 to 9llmm
Seated height
Chair to eye level
790mm
Chair to shoulder position
590mm
Chair height
410mm
Eye level
1 200mm (1 183mm to 1 213mm)
Visible high limit
so
Color-sensitive high limit
Jo
'
Color-sensitive low limit
70.
Visible low limit
Fig. 7 39
Reference dimensions for a sitting-type console
(2) Operations and monitoring in a standing position (basic body dimension
+ shoe height of 25mm) (Fig. 7.40)
Body height
Eye level
Shoulder height
Waist height
(3) Operation reach radius
( 4) Optimal field of view
454
1 675mm (1 650mm + 25mm)

1 565mm (1 540mm + 25mm)
1 355mm (1 330mm + 25mm)
845mm (820mm + 25mm)
650mm
30 degrees from eye level downward, 30 deChap.7
System Engineering
---1-T_hi.::...
gh_ h_e...;;;
ig_
ht 734 mm
600mm
1 - - - - - - - - JOOmm
---"=~.........;:"---
' - - - - - - - - - 0 mm
Fig. 7 40 Relationship between monitoring and operation at a console
grees to the left and right of body center.

(c) Monitoring and operations considerations
We now apply the above body dimensions to some actual examples
involving control panel design.
(1) Vertical type
In the case of a small-scale plant operation, occasions for performing control panel operations are generally few, and
such operations are simply done in front of the panel in a standing position. For a vertical control panel incorporating semi-graphics for monitoring, the operator will usually place himself at a position where he
can observe the entire panel, and for actual operations, will move right
up to the panel. This monitoring position is usually about 2 to 3 meters, depending on the size of the graphics. The relationship between
monitoring and operating for a vertical control panel is shown in Fig.
7.41.
(2) Bench type
In this type of control panel, although controllers, indicators and displays are installed in the vertical section, the operation
of various devices such as switches and displays in the bench section is
of primary concern. Under usual conditions, the operator will be located in a position from which he can observe the entire panel, and for actual operations, will walk right up to the bench. The control panel
should not be overly wide-at the most about 3 meters; some designs include a semi-graphic section on the upper panel.
7.3
Control Room and Man-Machine Interface
455
2200
209 4
1 918 -- - 190~ :~
Eye height
( I 565mm )
1389
;; ;
1212
C"'
!:
!~~6---970_..!_
Fig. 7 41
8.
1741
:...._-I......~~----+~!-E:==-----1r+.~>-+ I 565 ~;;~~~ ':, .?f
Relat io nship between moni toring a nd operation at a vertical pa nel
(3) Console type

Since the console type usually acts as a main control panel, plant equipment or other instruments (e.g., large-size displays) are often visible behind the console type control panel. Moreover, due to the importance of this type of a control panel, it is desirable to have all operations related to controllers, indicators, recorders
and displays within easy reach. In addition, although the operator will
stand directly in front of the console during operations, he will normally be situated at a slight distance from the panel. The relationship
between operating and monitoring ranges is shown in Fig. 7.42.
(d) Operator and instrument layout
Design of a control panel must take into consideration both the
type of plant operations involved and the operator's responsibilities.
For example, typical plant operations will usually include both close
monitoring of instrumentation system automatic functions as well as
active participation in system control based on operator decisions. Control panel configuration will be different in these two cases. In the former situation, a control panel will consist mainly of graphic displays
for monitoring of status and warning information. The latter situation,
however, is characterized mainly by bench type and console type control panels incorporating operation instruments such as controllers and
switches. The following points must be considered for instrument
layout on a control panel.
(1) If possible, the layout of instruments on the control panel should
agree with that of corresponding equipment in the plant work
area.
456
Chap.7
System Engineering
60 " olor- sensi tive limit
60"
ormal head rotation

Max . head rota tion
Visi ble range and operation range
(2) Instruments used for abnormal occurrences should be physically

separated from general operation ones, or alternatively protected
by guard structures or distinguished by color. In addition, such instruments should be located with the optimal field of view as described earlier (within 30 on either side of the normal line of
sight).
(3) In the case of two or more instruments making up a sequential
operation, they should be arranged so as facilitate such sequential
actions (e.g., left to right, arc formation).
(4) Instruments should be laid out with due consideration given to
the effectiveness of right-handed operation; controls requiring additional physical strength should also be installed on the right.
7.3.2 Control room engineering
The control room must be a safe and pleasant place to work in ;
the operator must be able to perform control system operations in an
efficient and accurate manner without becoming over fatigued. The following describes the factors that help contribute to optimum design of
a control room.
(a) Space and workability
During control room internal-layout design in which the minimum
space required must be determined, it is necessary to take into account
not only physical factors such as room size and shape, door and window location, number of control panels, pit and duct placement, and
7.3 Control Room and Man-Machine Interface
457
maintenance space, but other relevant factors as well such as system

objectives, number of operators, and emergency maintenance. Specifically, the following basic items should be considered.
(1) There should be enough space to enable operators and maintenance personnel to perform their work in a smooth and efficient
manner.
(2) In the case of several operators performing monitoring and
operations, the visibility and field of vision of each operator must
be considered, and the layout should be arranged to enable easy
communication. There should also be sufficient table or bench
area for document handling.
(3) Adequate maintenance space must be provided.
(4) Control panel shape, placement, color, etc., should be determined
with due consideration given to their psychological and aesthetic
influence on the operator.
(b) Livable environment
A livable environment is one in which people can comfortably perform their daily activities whether they be work or rest related. For a
control room, this would require not only adequate air conditioning,
low noise and low dust, but ideally a room with a relaxing color
scheme as well. In other words, proper selection of ceiling, wall and
floor color, control panel paint, graphic display background color, etc.,
can contribute significantly to achieving an optimal environment. The
following points should be considered when deciding on a control room
color scheme (see Fig. 7.43).
(1) Ceiling color
The color of the ceiling is important since it effects the overall lighting of the control room. Colors such as white,
ivory or creme which offer good reflectivity (80 to 90%) and do not tire
the eyes are recommended.
(2) Wall color
Wall colors, having some dependency on ceiling color
and windows, are not in general fixed. A common selection, however,
are non-glossy colors which prevent glare such as light green or ocher,
having a reflectivity of 50 to 60%.
(3) Floor color
The floor color should provide a reflectivity in the
range from 15 to 30%. Extreme bright or dark colors will not only tire
the eyes but will bring on general fatigue as well.
(4) Graphic displays
Since graphic displays have the function of
transmitting crucial information to the operator, they must be of a color different from background walls to enable quick and easy recognition. In addition, in order to clearly indicate symbols such as pumps
and tanks and their interconnecting pipelines, a low chromatic color is
preferable.
(5) CRT console
As the center of operations, the CRT console
458
Chap.7
System Engineering
~---- Illumin ation
Ceiling
( Reflectivity )
80 % 10 90 %
(3001x to 7501x)
Control panel
Graphic panel
(Low chromatic color)
Window
(Daylight)
C RT Console
Floor surface _ _ _ _ __
( Reflectivity 15% to 30%)
Fig. 7 . 43
--.J
Lighting and color arrangement for a control room
should be of a color which is both harmonious and conspicuous with respect to the rest of the control room. Moreover, the CRT console
should be placed in a position reflecting its role as the nucleus of the
man-machine interface.
(c) Room lighting
By maintaining appropriate illumination within the control room,
operator fatigue can be lessened and efficiency and safety levels improved. Conversely, immoderate lighting and light from outside the
control room can place an unnecessary burden on the operator in the
form of eye fatigue or chronic visual hindrance. Measures must be taken to eliminate such negative effects.
Control room lighting in general should be in the range from 300
to 750 lx, and should definitely be as uniform as possible within one
room. Moreover, lighting fixtures should be selected and installed so
that CRT displays and other indicators are not difficult to read due to
reflected light. The following points should be carefully considered during room lighting planning.
(1) In the case that CRT and graphic monitoring displays are combined on one control panel, the graphic panel should be relatively
bright, and the CRT selection dim.
(2) Lighting in the peripheral area about a CRT screen of about 300
lx is recommended.
(3) As a countermeasure to light reflected on CRT screens from
lamp sources, louver-style lighting fixtures should be employed,
7.3 Control Room and Man- Machine Interface
459
and their installation location and angle selected accordingly.

(4) Windows should not be positioned directly in front or behind
CRT screens. If for some reason this is not possible, some means
of controlling undesired brightness must be employed such as window blinds or smoked glass.
7.4 Instrumentation Work and Startup

7.4.1 Overview
In the design of an instrumentation system, the traditional course
taken is as follows: from among numerous process variables, those
which best reflect overall plant objectives are first selected, and the
loop configurations which measure, monitor and control these variables are then constructed. This is generally true today, but the instruments (e.g., sensors, transmitters, manipulators), monitoring devices
and controllers which make up these loops have undergone significant
advances in performance and digitization due to the recent progress in
electronics and industrial materials. Moreover, their handling and
operation have also undergone major changes compared to conventional devices.
For a particular instrumentation system, verification that the
above devices are performing satisfactorily is undertaken during instrumentation work and startup. Specifically, if we think of instruments
and control devices as the brains, senses and appendages of the instrumentation system, then the role of instrumentation work is to construct a central and peripheral "nervous system," and that of startup is
to verify that all functions of the system are operating normally. Instrumentation work and startup has the special responsibility of guaranteeing safe operations, optimizing performance, conserving energy,
and preventing harmful environmental effects. Needless to say, this
can only be achieved by careful, detailed planning and thorough implementation.
Just as construction of a plant involves a variety of specialized areas dealing with oil, chemicals, steel, power, gas, plumbing, paper
pulp, food, first-aid, etc., instrumentation work and startup operations
likewise consist of various components each characterized by specialized know-how. In this section, however, we will discuss only those elements common to the overall plant; particular know-how required at
each stage of instrumentation work and startup will be taken up in a
different publication.
7.4.2 Instrumentation work planning
"Instrumentation" is defined as "the outfitting of measurement,
460
Chap.7
System Engineering
control and other related equipment." Instrumentation work performs

the installation, wiring and piping of this equipment so as to enable mutual interaction and holistic functions.
Instrumentation work is distinguished from usual electric and
plumbing work for a variety of reasons. Measurement and control
equipment require more emphasis on accuracy, and they must function
using relatively weak signals. In addition, the actual form of instrumentation implementation may vary depending on actual usage objectives. The following describes the main elements making up instrumentation work planning.
(a) Instrumentation work estimation
One of the most important factors in carrying out instrumentation
work is projecting expenses, i.e., providing an estimate. In this regard,
it is common to add up the expenses of all sub-components when determining the total cost of instrumentation work.
Instrumentation work estimation is roughly divided into budget estimation and detailed estimation. In budget estimation, which is performed during plant construction planning, an overall budget structure
must be drawn up at a time when many points are naturally still unclear. As a result, total cost calculations must be performed along the
way as various conditions are established. Detailed estimation, however, is performed after the instrumentation work order has been set (at
least unofficially) and at a time when the engineering phase has already began and the specifications and drawings for instrumentation
work are essentially complete. Factors which must be considered when
performing estimation are as follows.
(1) Conditions affecting estimation
(i) General conditions: Work location, payment conditions, time period,
source of materials, special work conditions (e.g., night work, hazardous work), temporary utilities (e.g., site electric power, water).
(ii) Extent of instrumentation work: Extent of instrumentation installation; wiring, piping, painting and insulation.
(2) Estimation breakdown
(i) Materials cost: The materials required for each area of instrumentation work are summed up and multiplied by unit costs.
(ii) Labor cost: The man-hours required for each area of instrumentation work are summed up and multiplied by unit costs.
(iii) Other expenses: These include design expenses, on-site expenses,
transport expenses, supervisory expenses, safety measures expenses,
on-site out-of-pocket expenses, general management expenses.
(b) Design planning
Instrumentation work is only one of the various types of construction work (e.g., civil structural construction work, equipment installa7.4
Instrumentation Work and Startup
461
tion, p1pmg, wiring) going on more or less at the same time during
plant construction. Accordingly, it is quite important that basic design
plans for instrumentation work are submitted to the customer as soon
as possible, and that they be included from an early stage in design
plans of related construction work (structural construction, electrical
work, etc.) as well.
Specifically, instrumentation work specifications which should be
determined by this stage are categorized as follows.
(1) Basic specifications
These include work items, facility interfacing, materials required, extent and type of instrumentation work
design, and work schedule.
(2) Instrumentation room plans
Instrumentation rooms generally
consist of operation rooms, control rooms and relevant instrument and
device rooms in which plant processes are controlled, monitored and
operated. The size and layout of each of these rooms must be determined based on the number, shape, operability and maintenance of the
instruments and devices to be installed within each room.
(3) Cable layout plans
Main cable paths must be determined employing either of two methods: Overhead method which must take into account the process piping framework as well as the plant ceiling, pillar,
and wall configuration; and underground method which involves pit
layout and construction. The former is considered more appropriate
from a cost point of view. In order to negotiate the scope of work
here, the planning drawing, including structural size and cable routing,
should be submitted to each department, e.g., for the overhead method, to the structural department, and for the underground method, to
the earth works department.
(c) On-site instrumentation work control
Control at the work site is divided into schedule control, safety
control, and materials control, as described below. Each of these are essential to the smooth execution of instrumentation work.
(1) Schedule control
Since instrumentation work is only one of several types of construction work going on at the same time as described
above, it of course cannot advance at its own pace ignoring the status
of other closely related construction tasks. Schedule supervision will
thus see that instrumentation work progresses in step with other areas
by taking part in the following activities.
(i) Plant-schedule meetings: A basic plant schedule, called the "Plant
Construction Schedule," is prepared based on a master schedule provided by the customer. This schedule summarizes the beginning and completion dates of each type of construction work required for plant completion.
In connection with the Plant Construction Schedule, representa462
Chap. 7
System Engineering
tives from each construction area must hold a plant-schedule meeting

regularly to ascertain work progress and to discuss means of resolving
any problems, i.e., work behind schedule. The instrumentation work
manager should always be in attendance at these meetings to discuss
points of particular concern to instrumentation work, and he should obtain the minutes of the meetings for future reference. An example of a
Plant Construction Schedule is shown in Table 7.9.
(ii)Schedule preparation: The particular schedule for instrumentation
work must follow that of the Plant Construction Schedule and must be
based on the master schedule mentioned above. Moreover, any changes
in the work schedule must be clearly reflected on the appropriate
schedule charts after mutual discussion at plant-schedule meetings.
(2) Safety control
Out of all types of industrial accidents, it is generally felt that the highest rate of occurrence is at construction sites, as
is the severity of injuries. Nevertheless, there is also agreement that
accidents can be prevented by effective application of engineering, education, and enforcement in a combined, thoroughgoing manner. All parties involved with plant construction must take an active part in implementing safety education and making safety inspection tours, and
should strive to maintain an organized, clean and ordei-Iy work environment with the attitude of preventing industrial accidents at all cost.
The most common forms of industrial accidents in order of frequency is as follows:
(1) Falls and falling objects
(2) Collapsing structures
(3) Electrical shocks
(4) Heavy machinery (cranes, construction equipment) accidents
(5) Traffic accidents
(6) Fires
(3) Materials control
Construction materials used in instrumentation work should be kept in on-site storage lockers or in the operations office, and inventory accurately recorded. A specialist in materials supervision should be chosen to oversee such operations and to insure that no losses due to fire or theft occur. In addition, even instrumentation equipment and devices delivered to the customer must be
stored and carefully recorded for later instrumentation work. It should
always be kept in mind that since much equipment and materials are
stored for extended periods of time, unexpected shortages or losses
that are suddenly discovered can seriously impede the progress of instrumentation work.
7.4.3 Instrumentation work design
In instrumentation work design, detailed specifications are deter7.4
463
"
~-
~;;i
;;l
;:"'
"
-"
..
..
=
b
Thermal insulation
Testing
Wiring and piping
Installation
Purchasing of materials
Designing
Manufacturing
Coordination of specifications
Operation
Commissioning
..
.2
""s "
5 "
"s
"E"'
Electrical work
Piping work
Installation work
Building and structual work
Civil work
"!
Item
Process equipment design
Process design
""'~
l
5
6
...... ......
...... ...... .....
lO
ll
12
13
14
Number of month after work commencement
...... ...... .....
Table 7. 9 Plant construction schedule

15
16
17
.....
.....
18
--
19
0o
20
mined and drawings prepared so as to enable cost-efficient installation

and hookup, easy maintenance, and high-functionality of the instrumentation system. In order to achieve these goals, it is necessary to
prepare for construction work after investigating plant peculiarities
and the properties of materials to be measured while following related
standards and regulations. In addition, the interlap between the various types of work such as civil structural construction work, installation, wiring, piping, etc., must also be clearly specified.
(a) Design procedure and method
Instrumentation work design usually begins once drawings and
documentation related to plant configuration and processes have been
completed, as summarized below.
(1) Plant documentation
Includes process flow sheets, general plant
layout, plant facility layout, process piping diagram and piping standards, process devices and transmitter installation diagrams, structural
construction drawings.
(2) Instrumentation equipment documentation
Includes instrumentation equipment specifications, control-device system configuration diagram, external views of instrumentation and control devices and instruction manuals, control panel drawings, power supply system drawings.
(b) Standards, regulations and guidelines
In addition to various standards, instrumentation work must also
consider a variety of regulations and guidelines. The following are examples of standards, regulations or guidelines which may have to be reflected in design drawings.
(1) Standards
Industrial, construction, electric, measurement, lighting, painting, pressurized gas, non-destructive inspection, welding
(2) Regulations
Building, electric, electric components, pressurized
gas, anti-pollution, fire prevention
(3) Guidelines
Electric facilities technology, electric component
technology, gas and dust explosion prevention, static electricity safety
measures
(c) Design drawings and documentation
The following types of drawings and documentation may be required in order to carry out instrumentation work (actual drawings required depend on the situation).
(1) Equipment layout drawing
Shows the layout of control and monitoring equipment used in plant processes in a manner which facilitates
operation and maintenance.
(2) Field-mounted instruments location drawing
Used for verifying
the location or installation of instrumentation equipment; this drawing
becomes the basis of the instrumentation wiring and piping drawings.
7.4
465
NO.
Description
Material
Dimension
Q'ty Technical
standard
12m
CD
Pipe
SUS316
12s6 (0D)
Connecter
SUS316
12s6 X 1/ 2NPT
Nipple
SUS316,SCH40
l/2NPT
Drain pot
SUS3!6,SCH40
Union joint
SUS316
Union tee
Stanchion
SUS316
(j)
1
!2s6 X 12s6
12s6 X 12s6 X 12s6
1
1
so;
Half nipple
SGP
SUS316,SCH40
l/2NPT X50
Valve
SUS316
l/2NPT
JIS10K
Fig. 7 44 Hookup drawing of a pressure transmitter installation
A basic drawing indicating the layout of con(3) Hookup drawing

nection pipes used for conveying pressures and differential pressures
from process to field instrumentation devices; materials used are also
indicated on the same drawing. An example of a hookup drawing for a
pressure conveyor is shown in Fig. 7.44.
A system-wide drawing showing the
(4) Loop configuration drawing
various loops between field instrumentation devices and control and
monitoring equipment.
Drawings covering the en(5) Instrumentation wiring system drawing
466
Chap. 7
System Engineering
tire instrumentation wiring system, including signal system drawings,

power supply system drawings, and ground connection system drawings.
(6) Instrumentation piping system drawing
Drawings covering the entire instrumentation piping system, including air supply system drawings, pneumatic signal system drawings, utility piping system drawings, and hydraulic piping system drawings.
(7) Instrumentation main trunk line drawing
Shows the main trunk
lines along which instrumentation wiring, piping and cables are run;
also indicates the size and installation method of ducts, racks, pits, etc.
(8) Detailed instrumentation wiring and piping drawings
These drawings provide detailed information for carrying out actual instrumentation wiring and piping; they are prepared on the basis of the
above drawings.
(9) Auxiliary hardware drawings
Indicates hardware in addition to
instrumentation equipment which must be constructed during the instrumentation work phase, such as field control panels, terminal boxes,
connection boxes, etc.
(10) Wiring and piping connection drawings and tables
Provides information on equipment and device input/output terminals for performing
instrumentation wiring and piping.
(11) Work planning documents
These documents provide overall guidelines to help instrumentation work progress smoothly, e.g., work
schedules, work organization, and materials management.
7.4.4 Startup execution
Startup operations involve checking, adjustment and commissioning with witnesses of all instrumentation devices and equipment installed during plant construction, i.e., verification that the instrumentation system exhibits the functions as originally planned .
. In general, all delivered instrumentation devices and equipment
will have been manufactured and delivered under strict quality control
procedures. However, even if all devices and equipment are individually in top condition, and installation, wiring and piping work has been
performed satisfactorily, it still remains to be seen as to whether the
above have been interfaced with other facilities and utilities (e.g., electric and air supplies) to achieve a unified, smooth running system. This
type of overall functional verification is referred to as startup operations, or "pre-commissioning."
In recent years, instrumentation systems have been becoming
more complex and highly-functional as the scale of industrial plants increases. As a result, startup operations for a new plant have become indispensable, and a high-level of technology in these operations have
7.4
467
~-
"""""!
:i'
;;!
~
v,
CD
Plant
Heat
Fig. 7 45
Control room building completion

Power source equipment completion
Transport of control panels
___CJJ__J/
Preparation for commissioning
System functional test
Commissioning with witnesses

Confirmation of operation
Operation instruction
Adjustment device
- Q Relevant
work for commissioning
Instrumentation unit verification

Confirmation of installation
Confirmation of instrumentation work
Visual examination of each instrument
Confirmation of electric and pnumatic supply
Set up functions after installation
Operation test
(Insulation test etc)
and lead pipes etc.
safety devices, etc.
power source
Mechanical completion
) ~~- .. -,;o
commissioning work
([) Test on completion of instrumentation work,
@ Installation of control panels
Confirmation of instrument installation status
Unpacking and inspection with witnesses
instrument
each
for
Acceptance inspection
Plant construction process and placement of commissioning work
Preparation work for commissioning

Scheduling
Man-power planning
Documentation preparation
Confirmation of adjusting
devices
tlc::"
c::
.g
Electric work
Mechanical work
Major
Civil
foundation
work
Plant
Plant
completion
commissioning
Preliminary Commissioning Commission?'ng
start completion
arrengement for
\l
commissiomng
~
n
Completion
Preparing completion report
Return all documentation
related to customer
also come to be required.

The schedule and placement of startup operations within plant construction work is shown in the diagram of Fig. 7.45.
7.4.5 Startup operations
(a) Startup preparation
Some preparation work is required to enable on-site startup operations to run smoothly, such as schedule preparation, personnel planning, and general interface meetings, as described below.
(1) Schedule preparation and verification
The schedule for startup
operations must be prepared in conjunction with the overall schedule
for plant construction.
(2) Personnel planning
The number of people required for executing startup operations must be determined in conjunction with the
above schedule.
(3) Documentation/tool preparation
All documentation and references required for carrying out startup operations and those devices and
tools required for adjusting and verifying the operation of instrumentation equipment must be organized in advance. A checklist should be
used to insure that all necessary items (e.g., documents, procured devices, carried-in devices, calibration equipment) are included.
(b) Individual operation verification
After installation of a particular instrumentation device, its individual operation must be verified, as follows.
(1) Installation verification
Installation must be performed in a manner that does not hinder device or equipment functions and maintenance.
(2) Hookup verification
Wiring and piping must likewise be performed in a manner that does not hinder device or equipment functions
and maintenance.
(3). Visual inspection
Equipment must be visually checked for any
damages that may have occurred during shipment or for any conditions
not within specifications.
(4) Power/air supply verification
The electric and air supply systems
must be checked for proper voltage and pressure levels.
(5) Equipment setup
After device or equipment installation, some
setup procedures may be necessary before actual operation, such as removing internal "stoppers" used to prevent damage during shipment,
adding ink to recording devices, and installing radiation sources for radiation analysis devices.
(6) Individual operation inspection
The accuracy of individual equipment as well as the operation of graphic displays and sequences must
be verified by inputting test data.
7.4
469
(c) Loop verification procedures
The operation of individual loops between field devices and panel

instruments must be verified as preparation for commissioning with
witnesses.
Various types of loop configurations
(1) Loop operation verification
and operation must be verified with simulation input including computer interfaces, system functions, sequence operations, and analyzer
loops.
Based on previously prepared materials
(2) Test run preparations
and documentation, various system constants and transmitters must be
readied for operation (so that measurements can be taken).
(d) Test runs
Apart from verifying with the customer that the instrumentation
system is up and running, operation descriptions and adjustment advice are also provided during test runs (commissioning).
Whether or not the instrumentation sys(1) Operation verification
tem is demonstrating the functions desired is verified.
Operation of instrumentation devices and
(2) Operation description
equipment are explained especially as it concerns operators.
Also intended for the operators is advice relat(3) Adjustment advice
ing, for example, to PID parameters tuning and setting of various constants.
Various conditions for an(4) Condition setting of sampling equipment
alyzer sampling equipment are set at this time.
7.5
Quality Assurance
7.5.1 Engineering quality

System engineering performs the role of collecting, integrating,
and unifying numerous and diverse forms of technology. Consequently,
the task of determining quality standards in such a complex situation
can be quite difficult. In many cases, quality will be judged satisfactory or unsatisfactory merely by verbal agreement on a qualitative basis.
As a result, quality standards often come to depend on the ability and
temperament of the lead engineer and project team members. Also
coming into play here is the fact that both current technology and its
applicability as well as undeveloped technology are involved, the latter
in particular posing new and unsolved problems. Thus, in the light of
the above, the question can be asked, "What in engineering has the
most sway over engineering quality?" The following describes one approach to this problem.
Engineering capability can be defined as the amount of current
technology that can be applied, the extent to which undeveloped tech470
nology can be explored, and the level of problem solving skills. These
can be broken down as follows.
(1) Current technology
The quality and number of engineers

The quality and amount of accumulated know-how
(2) Undeveloped technology
Application know-how development capability

(3) Standardized operations
Extent to which operation procedures are documented and applied (manual preparation)
Extent to which know-how reuse is systematized and employed
Effectiveness of productivity improvement measures
As can be seen, the quality of engineering and the capability of developing new application know-how still heavily depends on the ability
and resourcefulness of individual system engineers and the project
team as a whole. Thus, overall quality can hardly be improved without
education and some form of self-improvement, and the importance of
organized instructional programs along these lines should be evident.
In regards to the reuse of previously obtained know-how, of importance here is the extent to which know-how is systematically collected
and made available as a common resource to engineers. A poorly organized system will inevitably result in individual know-how scattered
amongst engineers, eventually becoming lost with time. In addition,
productivity improvement measures such as know-how standardization
and automation should be adopted to promote efficient engineering processes.
The contribution made by each of the elements above to engineering capability is illustrated in Fig. 7.46.
Standardization, Automization, Reuse
OJT: 0
the job training
OFF.JT : OFF the job training

Accumulation
of knowhow
~ Education (OFF. JT, OJT, a nd study)

'v-I Self-improvement
Fig. 7 46 Contributi on to a company's engineering capabi lity
7.5.2
Design review (DR)
Since system engineering has the role of collecting, integrating,

and unifying various forms of technology as described above, the
7.5 Quality Assurance
471
amount of knowledge and ability employed has become enormous in

proportion to the increased complexity and scale of industrial systems.
It is thus not uncommon for an engineer or even the whole project
team to lose some control over the situation and make mistakes. This
situation has lead to the introduction of "design review" as an effective method of reducing such problems.
Design review involves an organized, detailed review by specialists
of each process stage from the beginning of sales operations to the delivery, test run and turnover of the system. Any predicted problems
which are uncovered during the review are studied in advance, and appropriate countermeasures taken. Figure 7.47 shows various types of
design reviews as applied to the different stages of a job from inquiry
to turnover, in particular from a system engineering point of view.
Note how for each design review, various methods of quality management such as quality development, FTA (Fault Tree Analysis) and
simulation are employed for resolving unclear points.
<D
Customer required quality
Inqu iry
@ Proposal
@ Award
@Turn over
Company inspect ion
@ Operati on
Witnessed inspection
Basic design
Proposa I review
<lll Detai l design review
Proposa l linal review
Basic design review
@ Commission ing
Detail design
Manufacturing
Fig. 7 47
472
<V
Instrumentatio n work a nd precommissioni ng
Startup preparation review

Overa ll evaluat ion
Sy tern engineering qual ity assurance
Chap.7
System Engineering
(a) Proposal review

After acquiring plant planning information in this design review,
strategies and tactics employed for accepting a particular order are
checked, and countermeasures to problems occurring both before and
after order acceptance are reviewed. Moreover, in regards to estimation, the contents of the estimation document are examined to see
whether customer needs can be realistically achieved, i.e., whether the
quality, cost and schedule requirements of the job are balanced in such
a way as to be mutually beneficial to both the instrumentation manufacturer and the customer.
(b) Proposal final review
Once the order has been finally accepted (at least unofficially) and
ready for execution, contract conditions and estimations must be reverified, and target quality, cost and schedule (basic design conditions)
clarified.
(c) Basic design review
In order to obtain concrete estimations of the cost and time required to meet the quality requirements of the customer, basic design
work must be checked in so far as it correctly takes into account all
relevant factors such as process data confirmation, equipment selection, general specifications, equipment interface, and total system reliability and safety.
(d) Detailed design review
At this point, the contents of detailed design (also called manufacturing design) are checked to see whether the principles and ideas established by basic design have been faithfully carried over. In particular, since it is common for various departments to be placed in charge
of different types of instruments and devices, much emphasis is placed
on whether the interface between this equipment has been designed in
a coherent, logical manner. Also evaluated is the amount of consideration given to maintenance and flexibility (system additions, modifications, etc.).
(e) Startup preparation review
Construction of the instrumentation system usually involves final
on-site procedures (installation, adjustment and test run). Indeed, it is
no exaggeration to say that the installation and adjustment procedures
give the finishing touch to the system. Consequently, those engineers
or the project team in charge of these areas must become thoroughly
familiar with all aspects of the system, and the preparations undertaken for system startup must be perfected as far as possible. The design
review in this case thus checks the extent of system comprehension,
documentation organization, task descriptions, work scheduling,
on-site organization. work conditions, and the like.
7.5 Quality Assurance
473
(f) Overall evaluation
This evaluation covers the entire length of the job from order acceptance through specifications meetings, manufacturing, inspection,
delivery, installation, adjustment, and test runs. Examples of items
checked are the level of quality achieved, i.e., does it meet customer
specifications, the effectiveness of joint meetings, and the cause of
schedule holdups. Note that in order to prevent a self-pleasing review
from being prepared, the customer should be requested to participate
either directly or indirectly (e.g., questionnaire).
(g) Objects of review and reviewers
Since applying the same depth of investigation to all job tasks is
not necessarily the most efficient, the level of review should be decided depending on the difficulty and scale of the task in question. An
example of allocating design reviews is shown in Table 7.10.
(h) Design review breakdown
Table 7.11lists the types of documentation normally required for
design reviews, Table 7.12 shows a section of a typical checklist, and
Figure 7.48 shows a typical format for recording a design review.
Review comment list
r~-tr~p~"'!":?"!"lir~~'i'o~l"!i.""'"~~~
Summary of task problem
{Filled in by person in charge)
Problem
Countermeasure
{5WIH)
p
Enforcement
Check
Recurrence
prevention
A
~~
Fig. 7 . 48 Example of a design review record
474
Chap.7
System Engineering
..,._
;::!
"';::"'
~
.....
.c
General manager'
Manager
General manager
General manager
Note;
Section member
Section member
Section member
Section member
Section member
Section chief
Manager
Chief reviewer
[Manager-
Manager
Manager
--]
Manager
Manager
Manager
General manager
Proposal final review
Manager
Manager
* : Branch manager. in case of branch office.
Manager
Manager
General manager*
Manager
Proposal final review Proposal review
General manager
General n.anager
General manager
[ Generaf manager'!
Manager
Manager
General manager*
General manager
[lliVlSlOnmanage!J
Proposal final review Proposal review
General manager'
General manager
IDivision manager~!
General manager
Proposal review
Manager
Assistant
(including system
engineer in charge)
Review respondent
(including salesman
in charge)
Reviewer
Sales headquarters
Sales coordination
department
Sales planning
department
Sales department or
branch office
Srstem engineering
devision
System order processing devision
System design diYision
Quality control
devison
Review member
All small scale jobs
(1) Highly-difficult job

(2) Medium scale job
(3) General manager or branch manager
guaranteed job
Highly-difficult job
(2) Large scale job
(3) Division manager guaranteed job
Review subject
(I)
Section
Department or branch office
Sales headquarters
Revie\\ unit
Table 7.10 Object of design review and review number
Name of document
19 Conference notes file
Oq
..-
-
~
"'....
..
Customer's specifications
Quotation
Request for quotation
Basic data file for quotation
Competition comparison table
Approval for award
17 Contract document
18 Award analysis
11
12
13
14
15
16
10 Approvals for major terms
1
2
3
4
5
6
7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
From8
In case of an overseas job the overseas

check sheet should be attached.
From 1. This also serves as a cover sheet.

From2
From3
From4
From5
From6
From7
Proposal final review Remarks
0
0
0
0
0
0
0
Proposal review
Table 7. 11 Typical documents to be prepared for design review
Job information management table

Review record
Proposal document list
Review check sheet
Comment list
Problems summary table
Job situation file
(with job execution plan)
8 Bid analysis
9 Job request sheet
No.
""'
"""
~
"')
~
~
-!>..
;:!
"'"'
~-
......
.C)
c.,.,
What item was most important in choosing the control system?

What item was most important in choosing the operation system?
What item was most important in choosing the operation method?
How can non-standard processing such as start up/shut down etc. be handled?
Was there any trade-offs between cost and reliability, maintainability, safety and product performance? (Trade-off is to make compromises between
the three essential product elements of quality, cost and delivery date.)
How did you arrive at the optimum answer for the trade-off?
Is the scope of service defined?
Are the specifications for resale products defined?
Are the specifications for instrumentation work defined?
Are the specifications for service defined?
Have past problems been investigated and countermeasures taken?
If new devices have been designed, was a check made for any infringements
of patents or utility models?
Did you apply for patents or utility models?
Are extensions or revisions considered?
Is the definition for reliability of the system clear? Did the customer agree
to them?
Are the requisitions from the customer clear and does this system satisfy
them?
How is the reliability evaluated (quality function development, FMEA and
FTA etc.)?
FMEA(failure mode and effects analysis)
FT A( fault tree analysis)
Design
Reliability
(Q)
Quality
Content
Description
Evaluation item
Basic design review items
Table 7.12 Checklist
Evaluated
score
""'
~-
~;:;;
;:;!
"';;::-
"l
""'~
Evaluation item
Content
How was the reliability certified (redundancy, failsafe and foolproof etc.)?
Redundancy(providing the alternative measures in the case of failure)
Failsafe (safe is insured if the function is lost)
Foolproof (prevention of human errors or careless mistakes)
Are environment specifications clear and does this system safety them?
Are the load factors and allowances sufficient?
Is the derating sufficient?
Are you familiar with the background for the award of this job?
Are you familiar with the production plan of this plant?
Are you familiar with the object and expected use of this system? On realizing them, what are the important points of the system?
Are points of disagreement and alternatives confirmed with the customer
and the agreement reached documented?
Are detailed discussion items confirmed and agreement reached?
Are the problems resolved for the proposal review and the proposal final
review? And if there are any problems remaining, are the countermeasures for the remaining problems clear?
Are all the specifications of the customer's requisition described on the
drafts of approval drawings?
How is the satisfaction of the customer confirmed for the requisition specifications?
Are there any problems to be resolved regarding the requisition specifications?
If problems exist: Are the causes understood?
Have countermeasures been considered already?
Have prevention procedures for the recurrence
of these problems been implemented?
Are there any undetermined items in the specifications?
If so: Are the schedules for determination of them established?
How did the prevention for specifications changes perform?
Reliability
General
(continued)
Description
Table 7.12
Evaluated
score
""'
;::!
;;!
"'"'
;;!
.0
'>!
:'1
Is the countermeasure for excess heating complete?

Is the countermeasure for excess noise complete?
Are the measures for dust-proof and drop-proof complete?
Is space for maintenance secured?
Is failure investigation difficult?
Is installation or dismantle easy?
Is inspection easy?
How much time is forecasted for maintenance?
How much is forecast for availability (reliability or production ratio)?
what extent does it safety customer requests?
Are the measures for preventive maintenance defined?
Are the PL and PS requested? What kind of measures are taken?
PL (product liability; product responsibility)
PS (product safety; safety for products)
Is the requisition for maintenance personnel defined?
Does the customer agree to them?
Is the selection criteria for spare parts and tools appropriate?
Reliability
Maintainability
To
Did you check all applicable laws and regulations for compliance?
Are the countermeasures adequate for preventing explosions?
Is there consistency between system configuration and instruments?
In the case that the control system includes new products to be developed,
does the agreement include each section concerned (such as development,
manufacturing, documentation, and postservices)?
Are the alternatives to nonstandard components investigated using standard
components?
Are the specifications, functions and characteristics of nonstandard components defined?
Are mill certificates necessary?
Design
-1>..
"""""
~;;i
"'~;;!
'l
(C)
Cost
Evaluation item
How does the system react to emergencies in the process and how does its
behavior reflect on system design?
How does the system react to emergencies in utilities and how does its behavior reflect on system design?
How does the system react to emergencies in the system and how does its behavior reflect on system design?
Has the affect of momentary and long-term failures of the systems been considered?
How was hazard-analysis (safety evaluation) carried out?
Are there any problems?
Are the countermeasures sufficient?
Safety
Is the building cost at the time of the construction award confirmed?

Was the cost verified at the start of design work?
What is the cost ratio (actual cost/forecasted cost) at the end of design
work?
Are there any improvements in cost by comparison with the cost verified at
the start of design work?
Are any improvements developed for this job applicable to other projects?
Are there any problems?
Are countermeasures defined after looking for the causes of any problems?
Are preventive procedures for these problems implemented?
Are all problems on proposal review and proposal final review resolved?
Are countermeasures for unresolved problems sufficient?
Were there any changes in specifications?
Did you estimate these changes?
Is the cost to be forecasted for future work increasing? Are countermeasures against this matter ensured? Is there any possibility for improvement? If so, what measures are effective?
Content
Description

Evaluated
score
'l
e:
;:!
"';:;"'
l
....
b,
Did the customer agree with the specifications for witnessing inspection?
Are there any items for special inspection and their preparation ready?
Are the style, content, language, type, and number of approval drawings
and as-built drawings defined?
Are the persons in charge of preparing inspection manuals with witnesses defined?
Is the curriculum of education, text books, language, instructors, place,
date, and number of trainees defined?
Are the points to be considered regarding transportation defined?
Are the points to be considered regarding storage defined?
Is the packing manual defined?
Is the language to be used in documents defined?
Quality
(Q)
Was the master schedule issued?

Are the any schedules unwritten in the master schedule?
Was the master schedule agreed to by the customer?
Was the master schedule agreed to by the department and section in charge?
What is the critical path?
Are there any problems that will affect the process?
Are there any special comments for this job?
What kind of response is made to problems that arise? Is there any plan?
Are all the problems regarding proposal review and proposal final review resolved? Are the countermeasures for unresolved problems. sufficient?
Delivery date
(D)
REFERENCES
1) Automation, 28, 1 (1983) (in Japanese).
2) Automation, 30, 4 (1985) (in Japanese).
3) A View of Engineering Industry, Toyo Keizai Shinpo Sha (1977) (in Japanese).
4) Yokogawa Tech. Rep., 19, 3 (1975) (in Japanese).
5) The Association of Japan Instrumentation Industry: Instrumentation Work Manual
6) T. Fuzita: Process Instrumentation, The Association of Japan Instrumentation Industry (1984) (in Japanese).
7) A. Kobayashi, et al.: Instrumentation Work Planning and Execution, Tokyo Denki
University Press (1984) (in Japanese).
8) Editorial Committee of Anthropometry for Ergonomics: The Measure of Man, Ningen to Gijutsu Sha (1979) (in Japanese).
9) Study for Terminology of Project Management: ENAA-1984-MANAG-3 The Japan Machinery Federation and Engineering Advancement Association of Japan
482
Chap. 7
System Engineering
Chapter 8
ADVANCED CONTROL
Although feedback control in distributed-type instrumentation systems is still centered about PID control, "advanced control" techniques
which make use of the special features of digital equipment are also
coming into wide use. Many functions which have been difficult to realize in analog devices can now be implemented relatively easy in a digital format. For example, dead-time units and also complicated loops
such as non-interacting control can be easily accommodated in digital
systems. This chapter describes advanced control as a control format
on a somewhat higher level than PID control having relatively high applicability.
8.1
Control Theory Considerations
In process control which started out with an emphasis on PID control, the control system is analyzed and designed based mainly on frequency response. This was a relatively familiar form of control even
amongst control technicians. In the 1960s, however, the state-space
method was introduced in which a system is described by simultaneous
differential equations. This method was used as a basis for the development of various optimal control theories. Following this, design methods related to the frequency domain were reconsidered, resulting in
the appearance of INA and other methods applicable to multivariable
control system design. This led to the realization that modern design
theory cannot be separated into that for the frequency domain and
that for the time domain, but rather must be developed on the basis of
a uniform theory.
An overview of the applicability of process control theories is
shown in Table 8.1. This table shows typical techniques employed for
the different categories of system expression, control system design,
control system configuration, and identification method. Although the
8.1
Control theory considerations
483
~
::I
.....
I:>.
::I
I:)
::t:.
!}
c.,
Identification
method
Control system
Control
system
design
System
description
I
.
".~,,C
c ........
r '
-------r---Sin;u,atron ----------- :
'
aOJe- diagrani :
function
; Nichols chart
: Vector diagram
; GG band
: Root locus method
I T ransfer
differential equation
Simultaneous
res.;o,;s.; .neiliOJ :
: F"re<.uer.c;.
; Transient response n1ethod
'
: Correlation method
~ .?.!'!'~!~~~- ~!'~)~i_s______ _ __;
1- PD controller
With n~nlinear band PID controller :
Sample-and- hold P!D controller
Two degrees of freedom type PlD
com roller
Feedforward control
Decoupling control
Smith controller
Learning controller
~ . ?~!f:!~~!~l! -~~t;t!~!l_e: ! ~J~. _t1"!! ...
: PHi c!in-ti-oiler :
apace trans ormataon
.
.
1 Lmear equatoon
f
L II
I
oes<:t-ibi;,g. riiiiciiol,- :
Handling
of
nonlinear
type
'
k ilsi S<itiares nieitiOJ ~

~ :"}~~!~.u.'!'. !i~!'!i.h.~ .'!'~~~'?':' __ ..
; Generalized least squares method ;

: Instrumental variable method
~--'~~~<";! ___ ___ _
: ivit.il;\aria-ble itiitifaior
' Observer
Kalman filter
: Multivariable servo system
; Internal model control
: Self- tuning controller (non- PlD type)
; l\!odel reference adaptive control
C - r'
!
:Si.UU1aifo_n_---------- :
; Pole assignment
: Model matching
<:C.,iiro!" ~
; Dynamic programming
: Riccati equation
~ o.>tiniaY
.
.
uatlon(conunuous typel Quantization
Time domain description
Table 8. 1 Over view of process control theory
Frequencr domain description
table is generally divided into frequency domain and time domain applications for convenience sake, it must be kept in mind that, as mentioned above, this is not absolute, and that a very definite mutual relationship exists between the two.
A question that might now be asked is : where within the overall
system are process control techniques applied? Figure 8.1 shows how
process control techniques can extend over a rather wide range of system related components. Although for matters of control, concentration on only control characteristics is relatively straightforward,
design work must also take into account the adjustment and maintenance of the control system. As shown in the figure, various techProduction ma nagement
1Pr
l~od~uct~i~o~n~p~la~n~n~in~g:J------T Scheduling
'Demand forecast
(Kalman filter, etc.)
Recipe management
~Linear Programmin g .
L Nonlmear programm1ng
~
'
PID ont rol

Advanced control
Stationary control
Improved PID
Non linear PID
Sample-and-hold PID
FF
Dead-time compensation
Auto tuning
Learning control
Fuzzy control
Expen system
Operation condition
change
Recipe change
Auto-stan up
Auto-stop
Intern al model control
rt
Control system dcsignl-- --r-- CA D Program

Optimal feedback control
Pole assignment
Observer etc.
Simulation program
Identification program
Dynamic characteristics
Static characteristics
Multiple- regressive analysis
Instrument maintenance
Device diagnosis
Tool con trol
analysis -
IOpentionl
Safety operation---,--Alarm display

1
L__ Process diagnosis
Expen system
Operator training-- Real time simulator
Fig. 8 1 Applications of process control techniques
8.1 Control theory considerations
485
niques are being developed for improving overall maintenance and system operation in addition to direct system control. However, this chapter is not intended to cover such a broad range of topics, and we will
concentrate on the area of advanced control as delineated in the figure.
Advanced control can be divided into control which simply adds
new functions to a traditional PID controller and that having structures different than those of PID, as summarized in Table 8.2.
8.2
Feedforward Control
Feedback control, as described in Chapter 2, is characterized by

the detection of any deviation arising between the desired variable and
the controlled variable, followed by attempts to eliminate this deviation. Thus, if no deviation is detected, no control actions are taken. If,
on the other hand, disturbance in the system can be measured and its
influence predicted, countermeasures to prevent the appearance of any
deviation can be taken beforehand. This type of control is called "feedforward control."
8.2.1 Feedforward control in a heat exchanger
We consider the example of a heat exchanger to help explain the
fundamentals of feedforward control. In the traditional feedback control format shown in Fig. 8.2, the temperature T 2 of the outgoing fluid
is the object of control. If the amount of incoming fluid is suddenly increased, the temperature T 2 naturally decreases, and the temperature
controller TC 1 is activated. As a result, the amount of steam is increased, and a new equilibrium state is eventually reached. The problem here, however, is that the temperature goes through a period of
fluctuation before setting down, as shown in Fig. 8.3. If, however, the
amount of incoming fluid is measured and the amount of steam increased in a corresponding manner, a temperature variance should not
appear.
Heat equilibrium in the system of Fig. 8.2 can be expressed by the
following formula:
(8.1)
Here, Ws is the rate of steam flow, Hs is the latent heat of the steam,
wp is the rate of incoming fluid flow, cp is the specific heat of the
fluid, T2 is the output temperature, and T1 is the input temperature.
Solving Eq. (8.1) for the amount of steam required, we get
486
Chap.8 Advanced Control
'I
....~
"'-
~
....,
~
"'<::>
:
Oo
~
Main purpose of control system
Non-PID control
Expert method
Fuzzy control
Adaptive control
Pulse transfer function control

Optimal feedback control
Sliding mode control
Countermeasure for dead time

Advancement of controlability
Advancement of tracking robustness
for. set point change
Countermeasure for parameter
change
Applied to the difficult process
in making process model
PID+Advanced control With nonlinear band PID controller Countermeasure for non linearity
Countermeasure for random noise
Sample-and-hold PI controller
Smith controller
Feedforward controller
Countermeasure for disturbance
(load change etc. )
Non-interacting control
Countermeasure for interaction
Override control
Constraint control
Two degrees of freedom PID
Optimal response both disturbance
control
and setpoint change
Self-tuning controller
Countermeasure for parametiric
change
PID control
Type of control
Table 8. 2 Various measure of process control
Distillation column
Compressor
PH control
Level control
Component control
Component control
Heat exchanger
Example
team now ratel
w,
~----.
Output temperature
(Temperature of nu id
to be controlled)
T,
Heat exchanger
w,.
T,
Inlet temperature
Fluid now rate
Fig. 8 2 Control of a hean exchanger
T ime
Fig. 8 3 Temperature change due to the load step change
T,
Fig. 8 4
Ws=
Feedforward control of a heat exchanger
ffs Wp(Tz- Tt)
(8.2)
Thus, in order to supply this amount of steam to the system, we can

employ the feedforward configuration shown in Fig. 8.4.
8.2.2 Combining feedforward control and feedback control
Although feedforward control is highly effective, it is difficult to
completely match its gain with the process. Consequently, the use of
only feedforward control is seldom employed, and combined use with
feedback control is more common. An example of such a combination
is illustrated in Fig. 8.5. In the figure , TC 1 output a is used to negate
488
T,se t
Tl
We
Fig. 8 5 A combination of feedforward and feedback co ntrol
the error in the feedforward control, and transfer functions GFT(s) and
GFw(s) compensate for dynamic characteristics. These elements are explained in more detail below.
8 .2.3 Determination of feedforward elements

The above example considers equilibrium characterized by a
steady-state situation between the fluid load and equalizing steam. In
actuality, however, the time delay following changes in the amount of
load and steam up until a change in temperature appears must be taken into account. To better comprehend the nature of this delay, we
make use of the generalized feedforward control system shown in Fig.
8.6. Disturbance D having characteristics GD(s) produces a change in
process variable PV, and a change in manipulated variable MV appears
after passing through Gp(s) . Note that manipulated variable MV is origr - - - - - - - - - - - - - .... .. - - - - - - - - - ,
: Process
.. ...... .... .... .................................. .................. ,II
: Control system
D Disturbance
::
,--------+'+'----~ (Load )
element
l
I
I
I
I
~ Setpoinl
:
I
I
I
I
value~
+
: sv
PID
'I
MV : Manipula.ed
L----'
PV : Pr~ess variable
I
L____ _______ ____ __________ !~~P!~L ________ ______ __ --- - ___ _.'
Fig. 8 6
8.2 Feedforward control
Feedforward control system
489
inally the sum of the output of feedforward element GF(s) and that of
the feedback-oriented PID controller.
Now, if we assume that process characteristics GP(s) and Gv(s) are
already known, how should feedforward element GF(s) be determined?
Since the objective is to prevent disturbance D from affecting process
variable PV, one method is to have the sum of the two paths by which
D can reach PV, i.e., the sum of Gv(s) and GF(s)Gp(s), be zero. In other
words,
(8.3)
Solving for GF(s), we get
(8.4)
If, for example, the process is characterized by first-order lag and

dead time, the denominator and numerator of the above would become
(8.5)
Gv(s)
(8.6)
and the feedforward element GF(s) on substitution of the above would

then be
G ( ) - _ Kv. 1+ Tp(S) -(LD-LP)s
F s KP 1 + To(s) e
(8.7)
However, a point of caution here. If L 0 > Lp, there are no problems,

but if Lo<Lp, it is impossible to realize, and a situation unsuitable for
feedforward control appears. If feedforward control is attempted,
then, excluding the dead-time term in Eq. (8. 7), the lead in the lead/lag
element will be somewhat overly used. The characteristics expressed
by Eqs. (8.5) and (8.6) are appropriate for measurement of step responses in an actual plant environment.
Feedforward control application examples
(1)Feedforward control in a furnace
Figure 8. 7 shows an example of
temperature control in a furnace. Here, feedforward control is performed by manipulating the amount of fuel in response to the amount
of raw materials input into the furnace.
(2)Feedjorward control in a distillation column
Figure 8.8 shows an example of feedforward control in a distillation column, in which the
amount of reflux and reboiler steam is adjusted corresponding to the
8.2.4
490
Chap.B Advanced Control
Off gas
Raw mat~
Air
Fuel
Fig. 8 7 Feedforeward control of furnace
Steam
<D
Pre-stage distillation column

Distillation column
Lead/Lag
Fig. 8 8 Control of distillation column
8.2 Feedforward control
491
rate of feed flow and its composition. Also employed here is feedback
controller TIC-I which treats the amount of steam as a manipulated
variable. In addition, lead/lag elements LL-1 and LL-2 are used for
compensation of dynamic characteristics.
8.3 Control of Dead-Time Processes
8.3.1 Dead-time processes
The period during which the effects of input applied to a process
or system have not yet appeared is referred to as "dead time." Dead
time may take on a variety of forms within a process control loop,
ranging from pure dead time as in the delay in conveying materials or
in performing analysis in an analyzer, to apparent dead time in higher-order lag systems. In any case, if dead time exists within a closed
loop, control will usually be difficult. The above can be easily understood by the following example. There are probably many people who
have experienced the unexpected difficulty of trying to adjust the water temperature of a shower (Fig. 8.9). To raise the temperature of the
water slightly, the first step is to open the hot-water faucet in what is
hoped is a corresponding amount. If, however, no change is felt after a
short period, it is usually thought that the faucet was not opened far
enough, so that it is then quickly opened even further. The end result
may be a sudden onslaught of scalding hot water. The desired temperature will never be reached in this way, and the best alternative, as
most people have learned, is to manipulate the faucet slowly in small
increments.
Hot water
Cold water
/II~
Fig. 8 9
Temperature control of shower
This situation also holds true for automatic control in a controlling

device. Using Fig. 8.10 as the basis for a slightly theoretical explanation, the following conditions related to the gain and phase delay must
be satisfied for oscillation of the closed loop to occur:
(8.8)
IGc(s) Gp(s )I= I

L
Gc(s)Gp(s)=I80
(8.9)
where Gc(s) and Gp(s) are the controller and process transfer functions,
492
Chap.S
Advanced Control
Controller
Process,
~~''T
Fig. 8 10 Block diagram of control system
respectively.
Case 1: Gp(s)=1/(1+Ts)
For a process having a first-order lag, if
we set s = jw, the phase delay is given by
L Gp(s)= -tan- 1wT
(8.10)
As can be seen, even if it is assumed that w = oo ; the most the

phase delay can be is 90. Accordingly, for control by a P controller or
a PI controller, stability, at least theoretically, can be achieved whatever the gain.
Case 2: Gp(s)=e-Ls
In this case, in which the process includes dead
time, we obtain the following using Euler's theorem:
L Gp(s)= L e-jwL= L (cos
wL- j sin wL)

(8.11)
=tan-'(_ sin wL )= _ wL
cos wL
Thus, as w increases, the phase delay becomes infinitely large, and if
the controller's gain becomes sufficiently high, instability will occur.
Since an actual process will be characterized as superimposed by both
first-order lag and dead time, its transfer function can be approximated as
(8.12)
Thus, the ratio L/T can be used as indicators of the difficulty of control. for example, the larger the value for LfT, the more difficult is
the control. Figure 8.12 shows examples of step responses to step disturbances applied to the control system of Fig. 8.11. As can be seen, if
T is held constant and L varied, longer L times result in longer settling-down periods.
Disturbance D
Fig. 8 11
PID Control system of process with dead time
493
12s
~I:~;!i:::ll llllllllllllllllll
L = S (s )
Po= 30 [%)
To= O.I7 [ min )
To= 0.03 [ min )
L = lO [s )
P s= 53 [%)
To = 0.3 [ min)
To= O.OS [ min )
L = IO [s)
Ps= 53 (%]
To = 0.3 [min )
T 0 = 0.05 ( min )
L = 20 [s ]
Ps= 83 (%)
To = 0.52 (min )
T o= 0.06 (min )
111111111111
Fig. 8 12
L = 40 (s)
Ps= 110 (%)
T o= 1.0 (min )
T o= 0.07 (min )
11111111111111111111
1111111111 111 1111111
111111111111111111
L= 20 [s )
Po= 83 (%)
r, = 0.52 [ min)
To= 0.06 [ min)
L = 40 [ s )
P o= IIO (%)
T, = 1.0 (min )
T0 = 0.07 [min )
Step re po n e to etpo int change in co ntrol y tern using the PID controller
(T = 20 [s])
8.3.2 Smith controllers

(a) Principles of Smith controllers
Two common techniques for improving the controllability of processes with large dead times are the Smith controller and sampling controller. The following will first discuss the Smith controller.
The Smith controller technique 1l was first proposed by Otto Smith.
As shown in Fig. 8.13, a process model within the control system is included, and if we exclude the dead-time, the process is expressed
simply as G(s). In order to better understand the actions of the control
system here, we employ the equivalent transformations shown in Figs.
8.14 (a),(b) and (c), disregarding disturbance D. Comparing the equivalent diagrams in (a) and (c), we can see that the dead-time factor has
been moved outside the loop. Thus, since no dead-time is present withDisturbance D
(1 -
e- '') G (s)
Compensator
Fig. 8 13
494
o ntrol system u ing Smith contro ller
Chap.8
Advanced Control
C (s)
(a)
{} Equivalent transformation
C ( s)
( b)
,0. E<Juivalent transformation

C ( s)
(c)
Fig. 8 14
Equivalent transrormation in Smith controller system
in the closed loop, it is possible to construct a controller with high gain

and fast response characteristics.
Step responses for a variety of L/T values are shown in Fig. 8.15.
The improvement over the responses in Fig. 8.12 are readily apparent.
An application example of the Smith controller is shown in Fig. 8.16.
(b) Process model mismatch
Although good control can be realized if the process model within
the Smith controller faithfully matches the actual process, in actual
systems this is nearly impossible to achieve. Expressing the process by
dead time and first-order lag in itself includes some error, as does the
values of L and T. Accordingly, the effects of mismatch must always
be considered. The effects of dead-time mismatch are illustrated by
simulation in Fig. 8.17, and those of time constants mismatch in Fig.
8.18. It should be noted here, however, that the extent of mismatch influence depends on the strength of main-controller tuning. In other
words, if the main controller experiences strong tuning, the mismatching margin becomes small, while if it experiences weak tuning, the mismatching margin becomes large. Thus, as a compromise between the
two must be reached, the extent of mismatch must be predicted and
495
12s
12s
11111111111111111111
~:::t\~L.)illlllllllllllll
~::!;l~l;.) 11111111111111111
11111111111111111111 ~::!,~tl 111111111111111
11111111111111111 1~::7;WL.] Ifllllllllllllll
1111111111111111111111
L = 5 (s )
Ps= 12 (%)
To= 0.33 ( min)
L = 10 ( s )
Ps = 12 (%)
To = 0.33 (min)
L = 20 (s)
Po= 12 (%)
To = 0.33 (min)
L = 40 (s )
P = 12 (%)
To = 0.33 (min)
Fig. 8 15 Step response of setpoint in contro l y tern u ing Smith controller

(T = 20 [s ) )
Concentration control
Fig. 8 16
B --t.__---C'Iot----'
An a pplication of Smith method in concentration control system
main controller tuning performed within this range so that instability

does not occur.
(c) Caution on applying to integrative processes
Although the Smith controller can provide good response even

with respect to disturbance, special consideration is necessary in the
case of integrative processes. In other words, even if integral action is
incorporated in a controller, offset with respect to disturbance will
arise. Referring again to Fig. 8. 13, if we solve for the output due to a
496
2.00
Lo= 28 ( s )
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.000 0
0 0
0
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.000
0
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
00
0.....
0
....
00
0
:2
00
000
00
00
....
L 0 = 26 (s )
0
0
00
oo
....
""
0
0
""
0
0
ao
0
0
00
8
0
""
0
0
0....
""
8
0
00
0
0
g 00
00
""
00
0
.....
0
0
:2
0ao
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
00
0
0
""
Lc= Tc= To= 20 (s )

PB= 60 [% ], To= 20 [s )
00
00
0 .....
0
....
00
""
00
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
00
00
00
00
00
0 0.... 0
....
:2
00
00
00
0
....
Lo= 14(s )
00
00
0
0
0
0
""
0
0
0.....
00
0
""
0
0
00
Lo= 16 ( s)
0
0
00
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.000
0
Fig. 8 . 17
Lo= 12 (s )
L 0 = 24 (s )
00
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00 8
0
0
00
00
0..... 0
0
0
0
0
00
""
00
:2
0
~
0
0
0
0
00
::e
0
....
L. = 20 (s )
(
0
0
....
0
00
0.....
0
0
""
0
0
00
00
0
0
0
N
0
0
0....
00
Re pon e of Smith controller system due to the process dead time change
8 .3 Control of Dead-Time Processes
497
2.00
Tp=28"Cs)
1.80 I
1.60 II
\
1.40 I I
\.,/' .......
1.20
1.00
0.80
0.60
0.40
0.20
00
O.OOo
0
2.00
------
ao
2.00
1.80 1 Tp=26 (s)
1.60 \\
1.40 I \
........,,
\
1.20
,_
1.00
0.80
0.60
0.40
0.20
0.00 0
00
0
"'
0
0
00
- 0
"'
0
.,.
00
0
<.0
00
ao
0
0
00
"'
-------
ao
2.00
Tp=24 (s)
1.80 I
1.60 \
\
1.40 I
\
\ ............
1.20
1.00
0.80
0.60
0.40
0.20
0.000 0 0 0
0 0 0 0 00
00
0
0
00
00
00
0
"'
00
0
;:!;
00
00
00
00
00
<.0
co
00
00
00
0
::: "' "'
---
"'
.,.
<.0
co
00
0
0
00
00
- --0
"'
0.,.
Lc=Lp= Tc=20 (s]

PB=60 [%], T,=20 [s]
Solid line-- Process output
Broken line--- Controller output
"'
Tp= 12 (s)
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00 0
0
2.00
1.80
00
00
:::::
0
0
00
00
00
"'
00
00
00
0
;:!;
00
00
00
00
00
Tp= 14 (s)
1.20
1.00
0.80
0.60
0.40
0.20
0.00 0
0
2.00
1.80 I
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00 0
0
Tp= 16 (s)
0
0
co
2.00
Tp=20 (s)
1.80
1.60 ~
1.40 \
\
1.20 \ ,
'
1.00
0.80
0.60
0.40
0.20
0.000 0 0 0
0 0 0 0
-"' - "'
0
0
.,.
<.0
00
00
-- - -
00
00
00
00
0
"'
00
0.,.
00
00
<.0
co
"'
00
00
"'
____
"'
.,.0
<.0
00
0
co
00
00
00
<.0
co
- -- - 0
"'
0.,.
00
00
"'
Fig. 8 18 Response of Smith controller system due to the process time constant changes
498
step disturbance, we obtain the following expression:

C(s)
(8.13)
where D is the size of the step disturbance.
Specifying the size of the output when t---"'oo as C00 , then from
the final-value theorem for the Laplace transform, we obtain
X{G(s)-G(s)e-Ls}D
(8.14)
Here, if Gc(s) indicates a PI controller and G(s) first-order lag, then

1 )
Gc(s)=Kp ( 1 + Tis
1+ Tis
Tis
1
G(s)= 1 + Ts
(8.15)
(8.16)
In these equations, KP and T1 are the controller proportional gain and

reset time, respectively, and Tis the first-order-lag time constant for
the process. We now get
limG(s)=lim
+1T.S =1
s~o
s~o 1
(8.17)
1 ~1ls +1T. =co

limGc(s)G(s)=lim
1 S
s~o
s~o
1S
(8.18)
lime-Ls=1
(8.19)
s~o
and consequently, from Eq. (8.7):

Ceo =0
or in other words, no offset remains. However, since for an integrative

process:
G(s)=
(8.21)
then
499
lim{G(s)- G(s)e-L}=lim( 1 -e-L)

s~o
s~o
=lim(L- Lzs
s~o
2!
+ Lasa -)=L
3!
(8.22)
Thus, even if a controller possesses integral action, some offset will remain. Several techniques have been proposed to eliminate this negative
point. In the Sawano method 2 >, even for a KpfsTP process, dynamic characteristics of the Smith compensator can be treated in first-order lag
form as K/(1 + Ts). In addition, in the Watanabe and Ito method3 >, a disturbance compensator such H(s) as shown in Fig. 8. 19 is introduced.
(Refer to the references at the end of the chapter for more details.)
Fig. 8 19 Disturbance compensation in Smith method
8.3.3 Sampling PI controller

(a) Principles of operation
The basic principle behind the sampling PI controller is that after
a certain operation has been performed, the controller waits a sufficient amount of time for the result of that operation to appear before
performing the next operation (Fig. 8. 20). Switch Sw is closed periodically for only a fixed time r, and open otherwise. If a deviation e continues to appear, controller output will be incremented in a step-wise
.
Set potnt
Measured value
8 t m___________
~e
100
_
PB
~
__1l____fL_1l_
-tTJ
Fig. 8 20 Principle of the sample-and-hold PI controller
500
R +
Sample-and-hold M
controller
I PI
~
Sample pulse
e 18
~
B.J
w---:
I
c
Fig. 8 21
!
I
Control system using the sample-and-hold PI controller
fashion, as shown. Thus, as this is an integral controller for the most

part, one might ask: why is such a controller effective for processes
having a large dead time? To help answer this question, Figure 8.21
shows as example in which, although a step change R in the setpoint
value is input, the system nevertheless settles down in a minimum
amount of time. Given that the controller's gain is correctly adjusted,
the controlled variable will match the setpoint value after only one
sampling cycle, and the system will settle down wit!:). no corrective action necessary in the next sampling cycle.
(b) Effects of parameter changes
Although the basic principle of the sampling controller is relatively simple, if controller gain or process characteristics are changed,
stabilization cannot be actived after one sampling. The effects of
changes in the controller gain on response curves are shown in Fig.
8.22. Figure 8.22 (a) shows a proper amount of gain, (b) a gain which is
somewhat too large, and (c) a gain which is too small. For these last
two cases in which the gain is not properly set, it will be necessary to
adjust it while observing the response during actual plant operations.
However, since this controller operates on a simple principle, such adjustment is relatively easy to perform, reflecting another feature of
the sampling PI controller.
(c) Application considerations
As described above, although the sampling PI controller for the
most part acts as an integral controller, it cannot be applied to an integrative-type process. The reason for this is the 90ophase delay of such
a process. Even for exceedingly low frequencies, care must be taken
since the system can be oscillatory as a 90 phase delay will easily
arise in both the controller and process.
501
1/
I I
I I
P~oc~s Jut~ut
1-- f--
PB=24 [%]
Ts=20[ s]
r= 1.8 [ s ]
T, = 8 [ s ]
5s/div
( a) Optimal gain
I
I
II
Process Output- 1--1--
Ts=20 [ s] PB=20 [%]

r= 1.8 [ s J T, = 8 [ s ]
5s/div
( b ) Case of slightly high gain
I I I
I I I
_I
I I I
Process Output f-
Ts=20 [ s] PB=32 [%]

r= 1.8 [ s]
5s/div
( c ) Case of excessively low gain
Fig. 8 22 Effect of the gain of the sample-and hold PI control system
8.4
Non-interacting Control
8.4.1 Interaction between process variables

For processes in which two or more variables must be controlled,
mutual interaction between the variables will occur. For example, Figure 8.23 shows a process in which both pressure P and flow rate F are
being controlled concurrently. If control valve Va is manipulated to con502
~ i@ vc~
p
Fig. 8 23
Flow and pressure control loop
trol the pressure, the flow also will be affected. Likewise, if control
valve Vb is manipulated to control the flow, the pressure will be affected. Thus, if F happens to be at a desired value but P is not, manipulation of Va to control P will adversely affect the flow rate, changing
what was up to now a correct value. Moreover, if we now manipulate
Vb to readjust the flow, P will be changed from its correct value.
In addition to the process described above, there are many other
kinds of processes in which interaction exisits. In the two following examples, interaction amongst all the variables exist. Figure 8.24 shows
an example of temperature control in a distillation column. Here, the
top and bottom temperatures cannot be independently altered; if either
the amount of reboiler steam or reflux flow is manipl1lated, both temperatures will be affected. In the example of Fig. 8.25, the flow of two
types of composites are mixed, and control of either one naturally
leads to mutual interaction.
Fig. 8 24 Top and bottom temperature control of distillation column
Fig. 8 25
8.4
Flow component control loop
503
8.4.2 Influence exerted by mutual interaction

In order to determine the actual extent of mutual interaction, we
investigate the problem using simulation. In the control system shown
in Fig. 8.26, process elements Cis) and G4 (s) both exhibit mutual interaction. The first chart in Fig. 8.27 shows a case with absolutely no
interaction when a change in the setpoint value is applied to the first
loop only. The dashed line in the chart represents the second loop,
which in this case is totally unaffected. If, however, we hold controller
parameters constant and raise the gain K 3 and K 4 of the interacting elements, the response gets steadily worse as shown in the following
charts. (The parameters for the controllers and process model are as indicated.)
_rD( 1)_
D( 2)
_.+~-I
D ( 3 )-....,+~---1
Fig. 8 26 Block diagram of control system for the interactive process
8.4.3 Expressing the degree of interaction

The degree of interaction will naturally differ from process to process. Depending on the degree, the appropriate manipulated variable/controlled variable combination must be selected, and whether or
not to incorporate non-interacting elements into the system must be
decided, as discussed below. The following considers methods for estimating the degree of interaction.
(a) Relative gain
The two-variable processes shown in Figs. 8.23 and 8.24 can be expressed as shown in Fig. 8.28, where g11 , g 12 , etc. indicate transfer functions. To begin with, let's focus our attention on the relationship between manipulated variable u1 and controlled variable y 1 For the case
in which the gains of controllers 1 and 2 are both zero, i.e., the loops
504
Chap.S Advanced Control
&;
v,
~
-..
::!
,.,.
~-
"'
;:!
~
I
;:;
,.,.
""
Oo
0.00
00
00
00
00
0
0
"" "' .... "'
00
0
<-
>
0
0
00
0
"'
0">
:=:
Time [ s]
00
00
00
00
0
0
- ""
00
0
0
00
0
0
0
"' .... "'
00
=:r,~-
I
0
<-
>
00
0
0
0
"'
0">
:=:
Time [ s]
0
00
00
0
I) =I
K,=l
K,=l
K,=0.4
K.=0.4
D(2) =0
D(
D(
"" "'
0
0
0
.... "'
0
0
>
0
0
<-
0
0
"'
0">
Time [ s]
0.00!
0.50
1.00
!.50
00
00
00
0
0
0
0
"" "' .... "'
00
>
00
<-
00
"'
0">
:=:
Time [ s]
00
00
00
Response of loop with interaction (3)
(Setpoint change is given No.I loop)
00
//'~\.. ..__;//,~'\.,~--'/..-~'.~.__.//~',,~
2.oo.--------------~
o -o.5or
"
E"
:=:
Response of loop-with interaction (2)

(Setpoint change is given to No.I loop)
0
0
0.50
2)=0
K,=l
0.00:
K,=l
0 -0.50~
.r,
K,=O
,/ \ \'-,//I.r',,........__,/.....-~....... .......__..~..
/
K,=O
-1.00:
PB,=P/3,=65 [%)
Tu = T,= 18 [s)
D( 1) =1
1.00
!.50
2.oo.--------------~
Fig. 8 27 Simulation for the loop with interaction
Response of loop with interaction (l)

fr
0.00
:I
"-0.50~
0.50
1.00
!.50
2.00
Response of loop with non-interaction

-2.000
0
-1.50
-1.00
-0.50
"
E"
0.50
1.00
!.50
2.00
2)=0
K,=l
K,=l
K,=l
K.=l
D( I) =I
K,=l
K,=l
K,=l.2
K,=l.2
D( I) =I
D(2)=0
D(
Fig. 8 28
Block diagram of two-variable control loop
are open, if we denote the relationship between u1 and y 1 as y 1/u 1lopen,

then clearly
.1!1._
U! open
(8.23)
= gn
Next, if we close only the second loop and denote the transfer function
between u, and y1 as ydu 1lc10 se, we obtain the configuration in Fig.
8.29. Consequently, this transfer function becomes
(8.24)
Examining this equation, the second term on the right arises due to
interaction, and the larger this term becomes relative to the first term,
the greater the interaction. If we now take the ratio between the
above expressions for the open loops and closed loops, we can define
the relative gain ,{ 11 as follows:
1/.l_l
U1 open
1/.l_
/IiiI
(8.25)
1 _
Ui close
Fig. 8 29 In case of No.2 loop is closed in two-variable control ioop
506
In other words, as this value indicates the degree of interaction as received from the closed loop, a situation with absolutely no interaction
will correspond to An= 1, and the degree of interaction will be indicated by the amount that An differs from 1.
In the same way, the relative gain for the other combinations of
Uc+Y2. u2~Y1 and u2~Y2 can also be defined as A21 A12 and A22 respectively.
A12
A21=
U12
Uuk2U22
UI2
1 +k2U21
U21
U21
Uuk1g22
1+k1U12
g22
U12kt!J21
1+k1Uu
(8.26)
(8.27)
(8.28)
If we now consider the gain of the two controllers to be sufficiently

high so that k1 and k2 ~oo, Equations (8.25) to (8.28) can be rewritten
as follows (assuming gn to g 22 represent static gain only):

Au
A12=
A21=
A22
UnU22
UuU22- U12U21
(8.29)
UI2U21
(8.30)
UI2U21
UnU22- U12U21
(8.31)
UnU22- Ut2!J21
UnU22
UIIU22- UI2U21
(8.32)
On examining the above equations, the following relationships between

the four interaction coefficients appear:
Au= A22
(8.33)
AI2=A21
(8.34)
Au+ A12 = 1
(8.35)
A21 + A22 = 1
(8.36)
These relationships are useful when solving for interaction coefficients in an actual process. Specifically, in the case of two inputs and
two outputs, once An is determined, the other three follow from the
above relationships. In addition, the relationship between process variables and interaction coefficients can be expressed as follows:
8.4
507
(8.37)
This is referred to as the relative gain matrix. For a process with no

interaction, the above becomes
U1
U2
Y1110
Y2
On the other hand, a relative gain matrix of

Y1
Y2
U1
U2
0.8
0.2
I02Q8
indicates that y 2 should be controlled by u1 and Y1 by u2 The relative

gain matrix is thus useful for determining the proper combination of
controlled variable and manipulated variable. This will be described in
more detail later.
(b) Interaction coefficients
Relative gain as described above has been used to express the degree of interaction based on static characteristics. This is not the only
method, however, for defining the degree of interaction. The following
equation, which includes dynamic characteristics, defines the interaction coefficient A. Here, a higher value of ..l. indicates a higher degree
of interaction .
..t(s)=
(8.38)
This equation can be used for graphically determining system stability.

8.4.4. Controlled variable and manipulated variable combination
As described above, the relative gain is a useful device for determining the proper combination of controlled variable and manipulated
variable. Ignoring for the time being dynamic characteristics and considering only static gain, the relative gain for a process in which
9n = 0.8, 9!2 = 2, 9 21 = 1, and 9 22 = 0.5 is given by
0.8X0.5
0.8X0.5+2X1
508
O17
.
Thus, since ..i 12 = A21 = 1- An= 0.83, the relative gain matrix becomes
Yt
0.17 0.83
Yz
0.83 0.17
The closer the relative gain is to one, the smaller the amount of interaction from the other loop. Thus, in this example, y 2 should be controlled by Ut. and y 1 by u2 However, it must be noted here that the relative gain expresses only static gain without taking into account dynamic characteristics at all. Thus, the stability of a system cannot be
immediately judged based solely on the size of the relative gain, which
should be treated as only one of several possible criteria. As described
above, the interaction coefficient equation (Eq. (8.38)) considers dynamic characteristics and can be used to draw a Gershgorin band to enable
a graphical determination of process stability. Although this method
will not be described in detail here, the reader is referred to the references at the end of the chapter. 4l
8.4.5 Non-interacting control

As described in the sections above, processes having mutual interaction are in general difficult to control. To control such processes,
one technique is to fine-tune controller parameters, and another is to
add elements to a system which eliminate interaction. The latter technique is described here. In the example of Fig. 8.30, by adding elements c1 and c2 to the system, interaction can be effectively removed.
The following relationships are present in the system of Fig. 8.30:
Ct][mt]
[Ut]=[1
Uz
Cz1mz
Ut2][U1]
[Yt]=[gu
Yz
Uzt 922 Uz
(8.39)
(8.40)
Substituting Eq. (8.39) into Eq. (8.40), we get
(8.41)
8.4
509
Process
Decoupling element
y,.
y,
y,
Fig. 8 30 Non-interacted control loop
Here, if we let
(8.42)
UnCI+ U12=0
(8.43)
U21 + C2g22= 0
then the matrix in Eq. (8.41) is made a diagonal matrix. In other

words, by letting
Ci=- U12
Un
(8.44)
(8.45)
the above matrix becomes
.r
]~
[:
U12U21
,., -0 g,
g,- @.U.
[ ::]
(8.46)
Un
Accordingly, for mutual interaction to be eliminated, the control

system must be equivalent to two independent loops, as shown in Fig.
8.31.
When performing non-interaction control, however, there are
some situations which must be given special attention. For example,
consider the situation in which
(8.47)
In this case, the gains gu-g12g2dg22 and g22- g12g2dgu in Eq. (8.46) both
become zero with the result that control becomes impossible. Thus, in
such a situation, this method cannot be used to create a non-interact510
y,
Fig. 8 31
Non-interacted control loop and its equivalent loop
ing system.
Another situation to be considered is one in which 911 and 922 are
transfer functions as the denominator of the non-interacting terms of
Eqs. (8.44) and (8.45). Therefore, if these transfer functions are higher
than the order of the numerator 912 and 921, non-interacting control
cannot be realized.
As for methods of actually implementing c1 and c2 , dynamic characteristics can be ignored and static gain only compensated, or assuming
dynamic characteristics, a first-order lead/lag element at the most can
be applied, after which the time constant can be adjusted.
8.4.6 An example of non-interacting control

Figure 8.32 shows an example of non-interacting control in the
case of a distillation column. As can be seen, a non-interacting element
has been added between TIC at the bottom and the reflux line. Of note
here is the fact that interaction elimination elements do not necessarily
have to be applied in both directions. In the distillation example here,
addition of a non-interacting element on only one side is sufficient to
obtain satisfactory results.
8.5
Self-tuning Controller
8.5.1 Overview
A PID controller in which each parameter does not have to be set,
or a controller in which parameters automatically adjust to changes in
8.5
511
~ Temperature
controller
~ Composition
0 controller
Flow controller
Level controller
Bottoms
Fig. 8 32 Advanced control of distillation column
Table 8. 3 Types of adaptive control
Changes in process
characteristics
System
Example
Predictable
Open loop adaptive control
Gain scheduling adaptive

control
Unpredictable
Closed loop adaptive

control
Model reference adaptive

control (MRAC)
Self-tuning adaptive control
(STC)
PID type
Non-PID type
process characteristics , provides extremely convenient features. Such a

controller is called a self-tuning controller, or STC for short. STC is
one type of adaptive controller. Table 8.3 classifies adaptive control formats.
If for a certain process, the functions relating process variables
and change in process characteristics are understood, then controller
parameters can be changed accordingly. This type of format is called
512
Chap.S Advanced Control
gain-scheduling adaptive control and is characterized by open-loop

operations. In contrast, if process characteristics cannot be predicted,
they must be continuously re-estimated until parameters are adjusted
satisfactorily. This format is generally divided into model reference
adaptive control (MRAC) and self-tuning control (STC).
Figure 8.33 shows the structure of gain-scheduling adaptive control. Here, the influence exerted by the environment on plant characteristics are measured, and con troller characteristics are changed so
that the response of the closed loop will agree with desired characteristics. On the other hand, in the MRAC system shown in Fig. 8.34, a
closed-loop response model is incorporated as a reference, and controller parameters are changed so that closed-loop response characteristics
will agree with those of the reference model. The type of controller
used here is not necessarily a PID controller, but may be of a much
more complex configuration. Figure 8.35 shows the structure of STC.
The main feature here is an on-line identification device which estimates plant characteristics based on plant input and output. The type
of controller here is PID or controllers of higher order.
At this point, let's consider where STC can be applied. In addition
to its original intention of responding to changes in process characteris-
Setpoint +
Fig. 8 33 Gain scheduling adaptive cnotrol
Fig. 8 34 Model reference adaptive control system
8.5
513
Desired cha racteristics
1Adaptive
mechanism 1
.-J Ide nti fication J- - >::--{ Com roller
Plam
I
Fig. 8 35
Self- tun ing controller
tics, it is also employed to make tuning operations more simple. These

two general applications are summarized below.
(!)Responding to changes in process characteristics
Startup time / shutdown time
Load changes
Internal-state changes (e.g., batch reactor)
Non-linear characteristics in final control elements
Elapsed-time changes
(2)Tuning simplification
Elimination of operator idiosyncrasies
Reduction of plant startup time
8.5.2
Gain-scheduling control
As described above, gain- scheduling control is used when the functions relating process variables and parameter changes are know. Use
of the gain-scheduling technique for temperature control of a reactor
is shown in Fig. 8.36. Given that reactor temperature and process gain
are related as shown in (b), then if the controller gain is changed as
shown in (c), the closed-loop gain can always be held constant.
Figure 8.37 shows an example of calorie control for a mixed gas.
The system here mixes three types of gases B,C and L having different calorie ratings to obtain a mixed gas with a certain desired calorie
rating. In particular, C gas or L gas is mixed in as a ratio with respect
to B gas. Thus, in order to perform calorie control, the C/B ratio undergoes feedback control in which a Smith controller is applied. However,
as dead- time varies depending on flow velocity, the dead-time incorporated in the Smith controller will have to be likewise varied. Accordingly, gas flow must be measured after the gases are mixed to enable
dead-time calculation and appropriate compensation. In addition, the
gain in calorie variation with respect to C/B ratio operations is given
514
Chap.B
Advanced Control
Temperature controller (SLPC) /
\.
Manipu lated variable for
hot and cooling water
Discha rge
( a) Reactor gain scheduli ng control system diagram
c
;;
...
01)
..... ..!!
oc;
c!::
-.. c0
0"'
- Reaction
temperature
( b ) Process gain characteri tic
-Reaction
temperature
( c ) Characteristic of F(X)
Relati on hip between process gain characteristic and function generator F(X)
Fig. 8 36 Gain scheduling control of reactor
by the following formula:

~
K= d(C/B)
Qc-Qs+~Qc-~L?L/B
(1 + CB+ L B 2
(8.48)
Here, Qc,QB,QL and QM represent the calorie rating of the C,B,L and
mixed gas, respectively, and C,B and L the respective gas flows. By
calculating this gain by computer, the proportional gain of the controller can be changed.
The above examples of gain-scheduling adaptive control, though
not particularly interesting from a theoretical point of view, find wide
practical application.
8.5.3 Self- tuning controller (STC)
To date, various types of STC formats have been conceived, with
8.5
Self- tuning Controller
515
I
I
$
C Gas
I
L--------1
Smith's dead-time compensation

L Gas
K= ~ = (Qc-QB)+L/B(Qc-QL)
d(C/B)
(l+C/B+L/B) 2
Fig. 8 37 Gain scheduling of calorie control
four major ones listed below.

(I) Ultimate-sensitivit y method 5 l
(2) Limit cycle method 6l
(3) ARMA model (successive identification) method 7 l
(4) Expert database method
In the ultimate-sensitivit y method as shown in Fig. 8.38, an auxiliary controller is incorporated in parallel with the main controller. As
the proportional gain of the auxiliary controller is gradually raised, a
critical oscillation having small amplitude is generated. PID parameters
can then be set from this critical oscillation using the Ziegler-Nichols
method.
For the limit cycle method, on the other hand, an on-off controller
is incorporated in parallel with the main controller, as shown in Fig.
8.39. Here, a small-amplitude limit cycle is generated whose period and
amplitude are used to estimate the process dead time and time constant. Based on these estimations, PID parameters are then determined
using the Ziegler-Nichols method.
The ARMA model method is not described here. The following sec516
Chap.S
Advanced Control
Main controller
Ku=Ku'+Kp
Where, { Ku': Gain of sub-controller
KP: Gain of main controller
KP=aKu
:. KP= l~a Ku'
Auto-tuning system applying the ultimate-sensitivity method
Fig. 8 38 System utilizing for the ultimate-sensitivity method
tion takes up the expert database method.

8.5.4 STC based on the expert method sl
The expert method is characterized by observation of the
closed-loop response to disturbance or a change in the setpoint value,
followed by tuning of PID parameters based on previously established
rules. As shown in Fig. 8.40, the configuration for this method is divided into response-curve analyzer and PID paramter tuning sections. In
the PID parameter tuning section, the observed waveform is classified
into a set of basic response curves and compared with a collection of
similar curves stored in the knowledge base. Based on the results of
this search, a fixed rule is then applied to modify PID parameters.
(a) Control target
As STC based on the expert method is still not a firmly established technique, we continue our description using as an example the
design incorporated in the Yokogawa Electric Corporation YS-80 series
single-loop controller. Targets for closed-loop responses are divided
into four types as shown in Table 8.4. Determining a control target
8.5
517
+
Setpoint -
( o)
F~F"IT
(1)
(o)
r : Period of limit cycle

lJ :Amplitude (p-p) of limit cycle
F : Amplitude of ON-OFF generator output
k : Proportional gain of main controller
7; : Integral gain of main controller
Fig. 8 39 Limit cycle method>
Self-tuning function
r--------------------
r
I
Process
Fig. 8 40
518
Functional brock diagram of self-tuning controller
Table 8. 4 Control target type

Target
Performance criteria
Overshoot : Non
Overshoot : 0
Overshoot : Small ( 5 %)
minj I e I tdt (IT AE)
Overshoot: Medium (10 %)
min
Overshoot : Large (15 %)
minje 2 dt (ISE)
j I e ldt (IAE)
e ; control deviation
based on the amount of overshoot has been found to be a very convenient method in actual use.
(b) Tuning rules
Figure 8.41 shows typical closed-loop response curves. Figure
8.41(E) shows an optimal response curve, while (A) shows one having
no oscillation at all; since the latter implies a delayed response, both P
and I parameters are made small. In (B), since the oscillation of the response curve has a long period and thus slow stabilization, P is made
small and I large. In (C), however, the high-amplitude damped oscillation calls for a large P and a large I. Finally, for the response curve in
(D) in which the damped oscillation has a short period, P is made large
and I small. Although these examples show the direction of parameter
change, the response curves must be closely analyzed and their difference with target response curves determined to obtain the actual
amount of tuning needed.
{ P--+ Larger
I---+ Shorter
{ P--+ Smaller
I---+ Shorter
response
curve
{ P--+ Larger
I---+ Larger
{ P--+ Smaller
I---+ Larger
Fig. 8 41 Response curve patterns and tuning rules
8.5
519
Disturbance
Process model
PV
Process model: e ' 08/ (1 +lOs)
Fig. 8 42 Simulation test configuration diagram
Converging response
p =70.2
{ I = 13.6
0=2.7
OV
+
Converging response
p =70.2
{ I = 13.6
0=2.7
Initial { P = 320
response I = 5
0=0
Converging response
p =70.2
{ I =13.6
0=2.7
0
50
Time [s]
100
Fig. 8 43 Results of simulation
520
(c) Simulation test
A control system in which disturbance can be repeatedly applied is

shown in Fig. 8.42, and observations of convergence behavior in this
system are shown in Fig. 8.43. Starting with three different initial response curves and PID parameters, it can be seen that the PID parameters converge to the same values after the several tuning actions in
each case.
8.5.5 STC application considerations

STC, though an effective means of control, is not a cure-all for
process control; control objectives must be thoroughly considered and
the most appropriate method selected. In particular, the following points should be considered at this time.
(1) Has the reasons for introducing STC been clarified?
(2) Can gain-scheduling control be used?
(3) Is on-line identification necessary? Can off-line identification be
used?
(4) Is identification. signal marking feasible? Size, type?
(5) Can STC catch up with the speed of process change?
8.6 Optimal Control
8.6.1 The meaning of "state"
This section introduces the field of optimal control based on the
concepf of "state feedback." First of all, however, the meaning of
"state" must be clarified, and although a thorough study of the subject
would fill many pages, a short review will be covered here.
Assume a process defined by the following transfer functions:
Y(s)
lJ[S')
(8.49)
(1 + TIS )(1 + T2S)
Rearranging this equation, we get:

Y(s)
1/s
1/s
(J[S')=---r; 1 + 1/T1s . T2. 1 + 1/T2s
(8.50)
The above two equations correspond to the block diagrams of Fig.

8.44. In the figure, the 1/s elements are integrators, situated between
input signals .X 1 and x2 and output variables x1 and x 2 , respectively. The
relationship between the signals in the figure can be given as follows:
8.6 Optimal Control
521
r----x-,--~~
X2=y
~ ~-----o
Ill (Equivalent)
Fig. 8 44 Brock diagram of 1/(l + T1s)(l + T2s)
(8.51)
(8.52)
y=xz
At this point we define the following:

Y=[y],
X=[xt, xzF,
A=
r- i.
1
Tz
-~1'
U=[u]
B=[ } .or.
1
C=[l, 0]
Tz
where the symbol "T" indicates matrix or vector transposition. Based

on these definitions, Equations (8.51) and (8.52) can be expressed as
{ X=AX+BU
Y=CX
(8.53)
(8.54)
Equation (8.53) is called the state equation, Equation (8.54) the output
equation, and X simply the state. Note that in contrast to the transfer
functions of Eq. (8.49) which only express the relationship between process input and output, the above state equation expresses the state of
the process as well.
8.6.2 Integral optimal regulator
Introduction of the concept of state as described above has
brought internal process variables into the picture as well. If the values of these variables are used in configuring feedback control, even
better control characteristics can be expected, as discussed below.
Given the system in Fig. 8.45 with setpoint value Yr. the deviation
522
Yr
Fig. 8 45 Integral feedback control
and corresponding integral are then defined as follows.

e=y-yr
(8.55)
q= jedt
(8.56)
Substituting the above into Eqs. (8.53) and (8.54) we obtain

(8.57)
Here, if we let
then
(8.58)
The control objective here is to determine the control input U which
minimizes the performance function expressed by the following 2nd-order function at Yr = 0
Here, Q and Pare weighting functions. The above is referred to as a

LQ (Linear Quadratic) problem, and is generally solved using the Riccati equation with respect to K as follows:
KBeP- 1BreK- A reK- KAe- Q =0
(8.60)
Using this K, control is determined as

8.6 Optimal Control
523
(8.61)
U=-P- 1BerKX
For the system in Fig. 8.45, we thus get
u(t)=k1X1 + k2X2+ k3q
= k1X1
+ k2X2 + k3 j<Y-Yr )dt
(8.62)
which is called an integral optimal regulator. As for a means of solving

Eq. (8.60), the following equation can be solved numerically.
~~ =K(t)BeP- 1BreK(t)-AreK(t) -K(t)Ae-Q, K(t)=O

(8.63)
8. 7 Kalman Filter
8. 7.1 Kalman filter formula
The Kalman filter is used in a wide variety of fields. It is expressed as a set of iterations estimating the current state of the system, and it is suitable for on-line computerized calculations. The formula for this filter is presented below in a discrete data system.
For a system described by continuous state equations such as Eqs.
(8.53) and (8.54), such equations can be converted to a discrete data system as shown below (derivation omitted). In particular, if we let
P=e
AT
(8.64)
(8.65)
we obtain
(8.66)
{Xk+l=PXk+ QUk
Yk=CXk
(8.67)
Given that input noise Wk and observation noise V k is applied to the

system (Fig. 8.46), their variance can be denoted by Gaussian white
noise Wand V, respectively. The problem for the Kalman filter is to
Fig. 8 46
524
Block diagram of system
Chap.B
Advanced Control
measure input/output values Y 0 , Y1 ,-, Yk+h U 0 , U1 , .. , Uk and then

perform an optimal estimation of Xk+1 The formula for doing so is
given below without proof or derivation.
(8.68)
Here, Xk+l and Xk denote the estimate of X at sampling times k + 1
and k, respectively, and :Yk+l the measured value of Y at sampling time
k + 1. In addition, gain Kk +l can be determined from the following formulas.
Mk+l=P(I- KkC)MkPT + QWQ
Kk+l=Mk+lCT(CMk+lCT + V)-
(8.69)
(8. 70)
The Kalman filter can be implemented on a computer by starting

off with initialized values and then calculating the above three equations at every sampling time, i.e., by performing iterations. The term
C(PXk + QUk) in the brackets on the right side of Eq. (8.68) represents the estimate of Y at sampling time k + 1 based on the estimate at
sampling time k; this follows from Eqs. (8.66) and (8.67). As a result,
the entire expression in brackets is the difference be~ween the actually
measured value and the estimate of Y. Thus, Equation (8.68), the estimate of Xk+l at time k + 1, can be interpreted as consisting of a corrective term obtained by multiplying this difference by gain Kk+l Gain
Kk+l is given by Eq. (8. 70), in which Mk+ 1 denotes the estimation error
and V the variance of the observation error. Accordingly, if the observation error is large compared with the estimation error, Kk+l is made
small. In other words, since for a large obserbation error the reliability
of measured variable :Yk+l is low, strong corrective action is not taken
in Eq. (8.68). Estimation error Mk+l is given by Eq. (8.69). Since the
first term on the right of the equation takes on a form in which the expression(/ -KkC) is multiplied by the previous estimation errorMk, it
can be seen how the larger Kk becomes, the smaller (/ -KkC) and
hence Mk+l become. This can be interpreted as follows: the greater
the previous correction, the closer we are to the true value, which
means a smaller estimation error.
8.7.2 Application to the parameter estimation problem9l

A certain water management office has come to use the Kalman filter to predict the daily amount of water demand. Water usage changes
in a complex way depending on a variety of factors such as weather,
season, and day of the week. Specifically, a different formula is used
for each of the four seasons, as follows:
8. 7 Kalman Filter
525
Spring: Qp=aut2+a21t+ as1r+a41

Summer: Q P = a1zt2 + azzt + aszr + a4z
Fall:
Q P = a1st2+ azst + assr + a4s
Winter: Q P = a14Ss + a 24 Syf + a3 4t + a44r + as4 Qy + aa4
(8.71)
(8.72)
(8. 73)
(8. 74)
In these equations, QP denotes the 24-hour total demand increment, t

the temperature at the time of prediction, r the weather factor increment (r= 1 during rain and 0 otherwise), Syf the amount of snowfall
during the previous 24-hour period (mm), Ss the accumulated amount
of snow (mm), and Qy the actual amount of demand for the previous
24-hour period.
Setting up of the above equations probably included some elements
due to experience. Nevertheless, to the extent that we assume them to
be correct, the Kalman filter can be applied for determining their coefficients. For example, if we rewrite the equation for Spring as
Y=[Qp]
(8.75)
X=[au az1 as1 a41F

C=[t 2 t r 1]
(8.76)
(8.77)
then the system equations become

{
xk+l=xk
(8. 78)
Yk+l = cxk+l
(8. 79)
and the problem is to estimate Xk+l To apply the Kalman filter formula, we substitute
P = I (unit matrix)
uk = o
Qw= 0
to arrive at the following three equations:

{
Mk+l =(I- KkC)Mk
(8.80)
~k+~=~k+iC~CMk+lc__r + V)- 1
Xk+1=Xk+(Yk+l- CXk)
(8.81)
(8.82)
Calculation of these three equations once every day enables the coefficients to approach more correct values. Use of this method for estimating water demand showed effective results, as shown in Fig. 8.47. 9 l
526
30000
::>.
..
......__.
0 6/20
Fig. 8 47
-----
,~
6/30
,~
;:..
/'"'"-------- ',
7/10
Actual value --Prediction value -
.-"'---
7/20
Month date
J-..
7/31
rvv""
8/10
..,
~
'
8/20
Result of total daily demand Prediction for a distributing reservoir
8.8 Other Forms of Advanced Control

For cases in which disturbance is repeatedly applied, or in which
changes in the setpoint value are performed in a periodic fashion, the
results of previous control actions can be stored and used as a basis for
parameter correction. Such a method is referred to as "learning control." Moreover, parameter correction based on simple exponential filters are gradually coming into use.
Testing is also being performed on fuzzy control, a form of control
which applies fuzzy set theory. The object here is to incorporate within the control system the experience possessed by a skilled operator,
experience which naturally includes elements of fuzziness. The extent
of ambiguity is expressed through use of "membership functions."
Fuzzy control is appropriate for systems in which preparation of a process model is overly difficult. In all, it is expected that the development work in this field will lead to even more advanced forms of control based on expert systems applied artificial intelligence.
REFERENCES
1) 0. ]. M. Smith : "A controller to overcome dead time," /SA], 6, 2 (1959) 28-33.
2) Nozawa : "Control for astatic control object including dead time," Automatic Control, 7, 5 (1960) (in Japanese).
3) K. Watanabe, et al.: "Disturbance rejection of Smith predictor control system,"
Trans. Soc. Instrument & Control Eng., 19, (1983) 187-192 (in Japanese).
4) M. Araki : "Computer-aided-design of multivariable control systems: Further developments of the inverse Nyguist array method," Systems & Control, 26, (1982)
218-227, 353-362, 489-497 (in Japanese).
5) T. Kitamori : "Study on an adaptive control system based on the ultimate sensitivity method," Trans. Soc. Instrument & Control Eng., 6, 6 (1970) 537-545 (in Japanese).
6) A. Sumi : "Adaptive control system applied with limit cycle," Trans. meeting for
Study of Analog Tech., 12, 8, (1972) 39-47 (in Japanese).
8.8 Other Forms of Advanced Control
527
7) K. J, Asti'om, et al.: "Theory and applications of self-tuning regulators," Automatica, 13, (1977) 475-476.
8) Y. Sakai, et al.: "Enhanced YEWSERIES 80," Yokogawa Tech. Rep., 30, 3 (1986)
9) H. Kamei, et al.: "On-line water demand predictions with Kalman filter," Yokogawa Tech. Rep., 25, 4 (1981) 36-40 (in Japanese).
528
Chapter 9
CONTROL OF PROCESS UNITS

(Application I)
The purpose of this chapter is to describe the basic elements of

representative process units used in the process industry. This material should serve as a basic reference when planning control systems for
such process units.
9.1
Overview
Analog instrumentation is usually characterized by independent application to each control loop, with integrated configurations rate. The
introduction of distributed control .systems, however, brought about
the widespread use of coordinated control systems as typified by feedforward control in continuous processes and totally automatic systems
in batch processes. The application of selective control for protecting
equipment and ensuring safe operations has also come to be widely
adopted.
Thus, in order to effectively apply these new forms of control devices, it is felt that a thorough understanding of the various factors affecting process units is now more than ever indispensable. Examples of
such factors with relevant control techniques indicated in parenthesis
are; principles of process operations and control objectives, adjustment
of equipment load (throughput control)*, relationship between disturbance and manipulated and controlled variables (feedforward control),
load variation and change in process gain (PID optimal setting), particular process-dynamic characteristics (application of advanced control
methods), equipment constraints (selective control), operation economy
(energy conservation, optimal cost control) and startup and shut down
operations (labor-saving control).
At present, knowledge related to the above is mostly obtained by
* The
9.1
contents in the parenthesis show the relevant control techniques.
Overview
529
experience, namely by plant operation technicians. This knowledge,

however, is not enough and a quantitative understanding is also considered to be essential. It is in this regard that the further development
of chemical engineering geared to process operation and control (plant
operation engineering)!) is eagerly awaited.
Accordingly, the application examples in this chapter, although by
no means exhaustive, aim to provide the beginner with a deeper understanding of process unit control in the light of the above considerations.
9.2
Control of Fluid Transport Processes
9.2 .1 Pump control

(a) Centrifugal pump capacity control
Small motor-driven pumps often employ a control method in
which a control valve incorporated on the discharge side can be manipulated (Fig. 9.1). Capacity control of this pump can be illustrated by the
curves shown in Fig. 9.2(a) in which one type of curve indicates pump
characteristics (flow versus discharge pressure) while the other indicates total pipe-pressure loss including that of the control valve. Thus,
if we consider the flow rates at two points where these curves intersect, we can see that as we increase the pressure loss in the control
.' . "'{y"
tF? -,'
Ps
'
Ps Static pressure difference

between tanks
4PLo to 4Pu ---- - Pressure loss in each piping section
Fig. 9 1 Pressure balance in centrifugal pump line and its capacity control
530
Chap.9 Control of Process Units
Pump characteri
cic
Pressure loss
1 in total piping
0..
Pressure
loss in
piping
Pressure lo
in piping
Sla .
._
Suo tic
~e
praau~
~--~
0~------------~~~~
-
f low rate
(a ) Discharge valve operation
loss
F, F, Fo
Flow rate F
( b) Pump speed opera! ion
Fig. 9 2 Ca pacity control of centrifuga l pump
valve from points 1 to 2, the flow rate decreases.

Pump characteristics are in general given by the following expression:
(9.1)
where P is the discharge pressure, n the pump speed, F the flow rate,
and K a constant.
Based on the relationship above, an energy-conser vation control
technique can be devised for a large-type pump. In this devise, as
shown in Fig. 9.2(b), flow rate is controlled by changing pump characteristics, which is achieved by altering the pump speed without a control valve. In order to employ this technique as effectively as possible,
the ratio of dynamic pressure loss* to total pipe-pressure loss must be
large and the time during which low-load operations are performed
must be sufficiently long with respect to the total operation period.
This can best be clarified by performing calculations for a dimensionless model as shown in Fig. 9.3(a) . The curve for pump characteristics
in this figure represents a typical example of Eq. (9.1) and is given by:
P=l.2-0.2F2
(9.2)
With F= 1 and the ratio of dynamic pressure loss to total pipe-pressure loss as/, the total pipe-pressure loss P L is then
The pressure loss changing in proportion to the square of the flow rate.
9.2 Control of Fluid Transport Processes
531
ur-----..:P_=~I.2 - 0.2 F'
100 .---..---,,---,,-----,-----.---.
~ sor-~~~r-~r-~-_,
_ _,
~
~6or-~~~r-~~~-~-~
1:!
c:
.g 40 1--1-....::::lill"""'-'
g 2or--r-~r-~~~~~~_,
~
Flow rate F
0
o~
.-~o.7s-o~.6~~~-L---L~~
Flow rate F
( b ) Relationship between now rate a nd energy
( a ) Dimensionles model for comparison
conserva tion rate
(j i partial rate of dyna mic pressure loss

in total loss)
Fig. 9 3
omparison in energy co nsumption between va lve control and speed control
PL=(l- /)+ /F 2
in which (1 - /) is static pressure loss.

In the same figure, if the flow is decreased from a state of total
load (1) to a flow F, we can then compare the power W 1 required when
using control valve regulation with the power W2 required when adopting pump speed control. Expressing pump power as (discharge pressure) x (flow rate), W1 and W 2 are then :
Wt=(l.2-0.2F2)F
"'2={(1 - /)+ /F 2}F
If we now define the energy-conservation rate M as
M =100 ( Wt- "'2)/-1

then from the two expressions above:
(0.2+ /)(1- F 2) X 100

1.2-0.2F2
(9.3)
Figure 9.3(b) shows the results of calculations based on this equation.

The affect that changes in F and f have on the energy-conservation
rate M can be quantitatively understood by these curves.
(b) Capacity control of a constant-volume pump
Capacity control of rotary pumps using gears or vanes can be per532
Chap.9
Control of Process Units
~ -- --1
I
I
I
)---.-~ :f---. To process
Relief valve for protection

(a)
apacity control by speed adjustment
F,
Flow rate
( b ) Change in pump charateristic
Fig. 9 4 Capacity control of con tant volume pump ( 1)
formed by the speed adjustment technique shown in Fig. 9.4 or the bypass-valve adjustment technique shown in Fig. 9.5.
9.2.2 Compressor control
(a) Capacity control of a compressor and pressure loss in piping
Although capacity control of a compressor within a process is in
principal not very different than that of a pump, some differences do
exist as shown in Fig. 9.6. Here, pressure controller (PC)* is often
used instead of flow controller (FC) and a change in capacity can be instigated from either the feed side or demand side via selection of the
respective controlled variables; suction pressure and discharge pressure.
The dynamic pressure loss of compressed gas in the piping can be
obtained by the following equations. Here, the symbols V, F and M correspond to those in the figure and represent volumetric flow rate,
mass flow rate and molecular weight, respectively.
Equivalent to LC (level controller) in the case of fluids.
9 .2 Control of Fluid Transport Processes
533
( a ) Capacity control by bypass valve opera ti on

Bypass valve opening Pump characteri tic
Middle Small (Constant speed)
L a~ge
'
1:1.,
Pressure loss in
\
tota l process piping '
.:..
t
k------------ Fr ----------~
Flow rate
(b) Bypass valve opening and now rate in each part
Fig. 9 5 Capacity control of constant volume pump (2)
r -G.}o,R- -------Re~~~--1
: Direct :
I
p,
P,. T,
''
"------ --- - -
Constant pressure
in demand side
Compressor
F. V
P,, T,,M
Fig. 9 6
Capacity control of compressor in proces
V -_RgTtF
MPt
F=cJ(H- H)
Mj{
(Rg: gas constant)
(9.4)
(c: flow coefficient of piping)
(9.5)
From these two equations, we arrive at: 2 >

V= cRgTt
JMT2
534
J(H-Pt
Pt +l)(H-Pt
Pt _H-Pt)
Pt
Chap.9
(9.6)
Total piping characteristic
Volumetric flow rate V
Fig. 9 7 Compressor characteristic and total piping characteri tic
The above equations are used to obtain the relationship, as shown

in Fig. 9. 7, between the curves for total pipe-pressure loss and compressor characteristics (turbo type). Here, the flow rate V can be controlled by changing compressor characteristics (e.g., altering speed) as
in the case of pump control.
(b) Turbo compressor characteristics
The following rules (fan rules*) are established for turbo compressors having a relatively low compression ratio and handling a gas with
a low molecular weight (such compressors are also referred to as fans
or blowers depending on the compression ratio value):
(1) The flow rate is proportional to compressor speed
(2) The discharge pressure is proportional to the square of the compressor speed
(3) The shaft power is proportional to the cube of the compressor
speed
Note, however, that characteristics different from these will appear as the compression ratio becomes large.
Changing of turbo compressor characteristics with the objective of
capacity control can be achieved by changing either compressor speed,
suction vane opening or suction valve opening. Figure 9.8(a) shows a
case in which characteristics change by varying compressor speed.
Also shown in the figure is a surge limit line which defines the operable part of turbo compressor characteristics. Within this limit, discharge pressure falls along with a decrease in flow rate, exhibiting
so- called negative resistance characteristics. This gives rise to cyclic
fluctuation in flow rate and discharge pressure, which cross over the
limit line. This pulsation is referred to as "surging," and its negative
affects can be countermeasured by implementing appropriate opera-
These rules also hold for pumps.
535
Surge limit line
1 Surge limit line
100
~~~
~
100
-:--- -~":;:%
50
~~"
10
~~~
0::
50
50
v (%)
V (%)
(b)
90'
,'~
/ ~.
0 o~----~5~0------~IO
~O
( a ) Characteristic change by speed variation
,l=:::i1::::
I
----~
100
haracteristic change by various opening

of suction vane
Fig. 9 8 Characteristic of turbo type compressor
tions such as 1) increasing the flow rate as the pressure loss on the discharge side decreases, or 2) decreasing compressor speed, so that operation continues only up to the point where crossover would occur. In
particular, a large compression ratio can result in immediate damage to
a compressor due to the shock on entering the cycle. The anti-surge
control method described in following paragraphs is indispensable in
such a case.
Figure 9.8 (b), on the other hand, shows a case in which characteristics are changed by varying suction vane* opening under conditions
of fixed compressor speed.
Although the varying of suction valve opening also changes characteristics in a very similar way, the range of adjustment is narrow compared to the former techniques, and moreover since energy loss is
somewhat large, this technique is usually adopted only for compressors
having relatively small volume.
(c) Anti-surge control for turbo compressors
The position and form of a surge limit line will vary depending on
gas composition and suction temperature. Assuming, however, a compressor for which these conditions do not change and moreover in
which speed is fixed, then by incorporating a bypass flow operation as
shown in Fig. 9.9(a), flow rate will not drop below the minimum set value and surging can be prevented. On the other hand, for adjustable-speed compressors in which the minimum set value is not fixed, if
we lower the set value along the surge control line** corresponding to
a reduction in speed (Fig. 9.9(b)), then if the process flow rate falls be* Installed on the compressor body and manipulated by a control signal.
** A line drawn parallel to the surge limit line at a certain allowable distance.
536
Bypass
Minimum set value
-~ -;
Aftercooler
I
I
----<1--- v---1!
To process
(a ) Antisurge control at con tan t speed
/ !
t
<D
Minimum set value
Surge limit line

Surge control line
@ Piping characteristic
( b) Modification in minimum set value at variable speed
Fig. 9 9
Antisurge control on unchangeable cond ition in suction
low the fixed minimum set value, useless power can be reduced without bypass flow.
In the above control systems, it is assumed that flow controllers incorporate an anti- reset-windup mechanism and control valves (positive
plug*) possess linear characteristics and rapid opening/closing capability so that control actions can be performed as fast as possible.
Considering now the case in which gas composition and suction
temperature vary, if the compression ratio P 2/P 1 is relatively small (under 3), then the surge limit line can be approximately given as :2 >
H-Pt ~ MV
----p:~ Ct-r;2
(9.7)
where c1 is a machine-dependent constant.

In addition, the following equation holds for an orifice incorporated on
the suction side:
v7.ffi
V=cz (1i'!f
(9.8)
The valve opens (safety side) in the event of less air.
537
where c2 is an orifice constant.

From the above two equations, we get:
h=~=
C(H- P.)
CtCz
.
(9.9)
Thus, as shown in Fig. 9.10, by selecting a coefficient* Cc having a val

ue larger than C as determined by the above equation and by controlling
the bypass valve on the basis of h =Cc(P2 - P 1) surging can be prevented.
Compressor
Fig. 9 10
(d)
Antisurge control in case of change in suction condition
Capacity control of a rotating positive-displacement compressor
Compressors of this type, although employing a variety of techniques such as Roots, vane, Nash, or screw, all possess flow and discharge pressure characteristics similar to positive-displacement
pumps. Thus, the methods employed for capacity control of pumps
may also be applied here.
(e)
Capacity control of a reciprocating compressor
This type of compressor usually contains several pistons and cylinders in which the compressor capacity of each cylinder is controlled
through use of a clearance pocket actuator and unloader actuator (Fig.
9.11). The unloader, which manipulates the suction valve, can set compressibility to zero by keeping the valve in a open position. The clearance pocket, on the other hand, can increase or decrease the volume of
compressed space by completing opening or closing its operation valve,
i.e., the amount of compressed gas delivered within the full piston
stroke can be controlled in a two-stage fashion. Referring to the figure, if the clearance pocket valve is completely closed, the piston will
move up to position A and the pressure within the cylinder will rise up
to P 2 The volume of gas corresponding to residual piston stroke 1 will
then be delivered. If the clearance pocket valve is now completely open* To provide a safety margin.
538
learance pocket actuator
Unloader actuator
Fig. 9 11 Operation principle of unloader and clearance pocket
ed, the piston will move up to position B with the internal pressure
again reaching P 2 and the volume of gas corresponding to residual piston stroke 2 delivered.
The difference in delivered volume (at suction conditions) for
these two situations can be determined by the following expression,
where v specifies the volume of the clearance pocket: 2>
v{( P2)''
P,
7_
1}
'
=specific heat at constant pressure

specific heat at constant volume
Figure 9.12 4 > illustrates how the capacity of a fixed-seed reciprocating compressor can be adjusted in 5 stages when both unloader and
clearance pocket actuators are used.
j. .
.:;: c
--+-:--c1_ea~rance pocket
~
~
Clearance pocket
t:l
75%
~ bl
.,.
Clearance pocket
"!"
25%
100%
1: c
Clearance pocket
50%
T-
J
~
Clearance pocket
C]
0%
Fig. 9 12 Multi-stage capacity control of reciprocating compressor
539
9.3 Control of Heat Transfer Processes

9.3.1 Control of heat exchangers
(a) Heater control by steam valve adjustment
Figure 9.13(a) shows a temperature control process in which steam
flow G decreases proportionally with a reduction in flow F in response
to a high-load situation. As shown in (b) of the figure, the pressure P
inside the vessel is reduced at this time since the heat load becomes
small.* Thus, if the steam feed pressure P s is kept constant, the pressure drop across the valve increases. As a consequence, if the control'
valve possesses linear characteristics, the open-loop gain for temperature control increases considerably. To compensate for this, a valve
having equal-percentage characteristics should be adopted instead of
one with linear characteristics in practice.
F. T,
Drain
( a ) Temperature control process
load F (%)
( b ) Change in control va lve pressure d ifTe.renoe by
load change
Fig. 9 13 Heater control by steam valve
In such a process, although the value of pressure drop across the

valve at maximum load often presents a problem at the design stage, 7
to 10% of the feed pressure (absolute) is considered a sufficient pressure drop. It should also be noted that if the load decreases remarkably, the valve will move to nearly complete closure, and since the Cv
value becomes discontinuous in this state, an on-off line operation will
occur. In addition, in the case of a steam heater with the set temperature around 100C, internal pressure P will approach atmospheric pressure with a decrease in load and the steam trap will cease to function.
The temperature difference between the steam within the vessel and the process
fluid may be small.
540
Chap.9
(b)
Heater control by drain valve adjustment
Heating capacity can also be controlled by manipulating drain

flow,* as illustrated by the system in Fig. 9.14 (a). With this technique, there is no interference with drain discharge due to load reduction as described above, and a wide range of load variation can be accommodated. In addition, since valve size here is considerably smaller
than that in steam control, a definite savings in equipment cost can be
realized. However, control response in this technique is somewhat inferior to that of control by a steam valve. Referring to the block diagram of this process in Fig. 9.14 (b), the following expressions are obtained for determining the transfer function expressing the heat transfer rate corresponding to drain flow manipulation .
.Q(tl_
r
GJS)= Ts+1'
(9.10)
It can be seen that this process has a first-order capacity delay.

Process flow
Drain
G, ( a ) Temperature contra I process
G ( s)
L-----------1 j_
r
1-------
G, (s) : Drain flow rate
Q ( s) : Heat transfer rate for condensation

G (s): Steam flow-in rate
: Over-all heat transfer coefficient on
condensing surface
iJ T : Average temperature difference between
steam and process fluid
Wh(s): Amount of drain in vessel
A ( s) : Effective heat transfer area for

condensation
: Latent heat of steam for condensation
( b ) Heat rate response for drain flow operation
Fig. 9 14 Heater control by drain control valve

The value of the overall heat transfer coefficient of the condensation surface is
much larger than that of the drain soaking surface.
541
The time constant in the above expressions is usually a few minutes in most equipment (refer to practice question 9.3). Therefore, a
controller having a narrow proportional band (20 to 30 %) with an integral time of about one minute and having a control valve with a positioner shall be used in this technique. Although the closed-loop natural
oscillation cycle here is quite long compared to control by steam valve
adjustment, it can be used with almost no interference in heat control
for equipment such as distillation column reboilers.
(c) Control of fluid-to-fluid heat exchangers
We first consider a case of heat exchange in which the flow of a
high-temperature fluid is controlled and a low-temperature fluid is
heated to a set temperature (Fig. 9.15). It is assumed in the following
description that the only disturbance is the flow rate and temperature
of the low-temperature fluid, and that the inlet temperature of the
high-temperature fluid is invariable. In this process, if the inlet temperature of the low-temperature fluid drops or if its flow rate increases, temperature controller TC will act to increase the flow rate of
the high-temperature fluid. This results in a large average temperature differential between the two fluids which in turn increases the
amount of heat-transfer (although the increase in flow rate brings
about an increase in the heat-transfer coefficient which aids control,
this is exceedingly small compared to the effect exerted by the change
in average temperature differential).
However, since the effect of increasing the average temperature
differential by increasing the flow rate of the high-temperature fluid
is limited when the exit temperature of the high-temperature fluid apTn "
-:!c __~ C, (hSpecilic)

.r __ -.
ear
'
Fu
Tn
Tc,
TH ,
Constant
Con tant Unva ri able
Fe T"
C (Specific)
<
Tn
TH,
Temperature
CASE I. Tc, Decrease
Tm
Unvariable
Temperature
CASE I . Fe Increase
hea t
Fig. 9 15 Te mperature control for fluid-to-fl uid heat exchanger
542
Chap.9
proaches its inlet temperature, rapid saturation tends to occur. Thus,

in heat processes similar to this one*, gain characteristics for the
change in heat-transfer rate due to flow manipulation become extremely nonlinear, presenting a problem to control in not a few cases.
The following relationship holds for the variables shown in Fig.
9.15: 3 )
(9.11)
We consider a particular process in which the variables in the above
equation are as follows : Te1 =92 C, Te 2 =200 C, THl =260 C, T" 2
=230C, Fe=50 kg/min, F"=180 kg/min, and CH=Ce=l. If either TCJ
or Fe varies, this equation can now be used to calculate the desired
change in F" to maintain Tc2 at 200 C (Fig. 9.16).
In addition, the sensitivity (gain) of the change in Te 2 due to adjustment of F" can be obtained by differentiating Eq. (9.11) as follows:
aTe,
aFH
TH,- Tc,
FcCc
((FcCc/UA)+(1/2){1 +(FcCciFHCH)]2. 2FH 2CH
Using this equation, Table 9.1 shows the results of calculating the
change in gain for changes in Fe in the above example. Just how remarkable this change is can be seen by the values listed. Although this
tendency can be somewhat relaxed by adopting an equal-percentage
600
500
400
.'
60 0
''
'
'
''
:'
50
'
=-
:'
.. 300
~ 20 0
100
00
'
'
''
''
''
''
'
\.
'
I 0
Tct ['C)
100
:----
20 0
( a ) Disturbance Teo
Fig. 9 16
20 0
-v
20
40
Fe [ kgf min )
60
( b ) Disturbance Fe
Change in heating flow rate for process side di sturbances
Such as in condenser adjustment by cooling water, or stirred- tank temperature ad

justment by jacket-coil cooling water.
543
Table 9.1
Process gain change of fluid - fluid heater for heated flow disturbance
Fe (kg/ min)
FH (kg/ min)
~~~ ("C/kg/min)
55
50
555
180
97
0.006
0.055
0.166
45
fJ Tc2 / (
fJ~H) x 100CC%)
3.33 X 10- 2
9.90x 10- 2
16.1 X 10- 2
control valve, examining the last column in the table reveals that such
compensation is not sufficient.
(d) Temperature control by heat exchanger bypass
Temperature control as performed by the bypass configuration to

the heat exchanger shown in Fig. 9.17 exhibits excellent response compared to the control techniques so far described. As a result, this technique has come to be frequently used. However, if the bypass valve
opening changes due to variation in process flow rate or the temperature controller (TC) setpoint value, control gain will also change resulting in degraded control. In addition, since the parallel-piping flow resistance of the heat exchanger and bypass valve changes, disturbance
will be applied to process flow F. These undesired affects can be alleviated by using a three-way valve instead of a two-way one, as shown in
Fig. 9.18.
Steam
--{0
___ j
F,P,
'
'-+F--=.":....:T._,--rP~~Ti!!..o_To process
Drai n
F., To
r.;:;::.._
I
I
~--- "\!:_Sr -----'
Fig. 9 17 Temperature control of heater by process bypass valve
The following equation holds for the two-way valve of Fig. 9.17.
ToF= TtFt + T2F2
If we now designate the bypass valve Cv value as c1 and the Cv value equivalent to the flow resistance on the heat-exchanger side as c2,
the following equations also hold.
Ft=actJP;-Po
544
Steam
r----..,
'
''
'
' ~f-t---+..J
..:.P.:.J
F~""=7'
- ,-
-F,
F-
iI
F,
T,
L--------------9------J
.
F,
'
''
(a)
ase of di vening type three-way valve

Steam
,.1-fC\-.,
:~ I
'
'
F------~--~
F,-~-F
f,
F,
T,
(b ) Case of mixing type three-way valve
Fig. 9 18 Temperature control of heater by three -way va lve
Fz = acz./ P;- Po
F= Ft + Fz= a(ct + cz)./ P;- Po
where a is a constant. From these equations, we arrive at the following:

(9.12)
It can be seen from this equation that if T 1 is invariable, then T 0 will
not change as long as T2 remains constant. In regards to T2 , since it
will approach steam temperature Ts if heating capacity of the heat exchanger has a sufficient margin, its amount of change is small. However, it can also be seen that c1 becomes small if T1 decreases with T 0
fixed, or alternatively, if set value T 0 is raised while T 1 is fixed.
The affect that change in c1 exerts on T 0 can be obtained from Eq.
(9.12) as follows:
~~ =
- ( Ct
~2cz)z ( T2- Tt)
(9.13)
This equation shows how process gain will vary due to change in c1
545
even if T 2 - T1 remains constant.

We next investigate cases in which a three-way valve is used instead of the two-way bypass valve above. For ease of description, we
will consider cases in which flow resistance of both the heat exchanger
and bypass can be ignored.
Since the sum of the Cv values for both ports (linear characteristics) of the three-way valve is constant, we can obtain equations for
the valve characteristics of each port as follows:
cl=(1-l)cvr}
(9.14)
c2=lcvr
where Cvr is the above sum and l is the valve lift (full stroke = 1).
Since the relationship in Eq. (9.12) still holds with the above conditions, we substitute Eq. (9.14) into that equation to get:
To= T1+(T2- T1)l
. . aaro =
T2- Tl
From these two equations we can see that if T1 and T 2 are fixed, T 0 is
determined only by land process gain is constant.* In addition, the piping flow resistance does not change but takes on the constant value Cvr
according to l.
In actuality, however, bypass piping resistance and heat exchanger
flow resistance do exist to an extent and adjustment of the manual bypass valve can be performed to make the two equivalent. Although this
situation is somewhat different than the ideal one described above, the
same results are generally obtained and control is greatly improved
over the two-way valve bypass.
Although the above descriptions have dealt with bypass temperature control of heating processes, the results obtained completely apply to cooling processes as well.
Heating furnace control
(a) Control system configuration
Furnace control, as depicted by the system shown in Fig. 9.19, is
performed by integrating various inter-related control functions. The
following describes each of these control functions.
(b) Total heat control
Primary disturbances to the furnace process are the amount and
quality of the feed fuel. Such disturbances, however, can be suppressed by keeping the total amount of heat input into the process con-
9.3.2
In actuality, F does cause a change in T 2 to some extent.
546
Chap.9
Draft control valve
F.B. : Feedback signal

F.F. : Feedforward signal
FD F : Forced draft fan
IDF :Induced draft fan
Fig. 9 19 Configuration of furnace control system

Summing up calorie
r---- ----- ---~------ -----------i

1
.-------,~
1
1
l1
I
~~Total heat
00~~
L----
setvalue
r-~
I
I
DT-J
~=~:ngHtA~TA~pA~
Fuel gas. A
PC -,
r-
High
-4---'
I
J~
Qc_-~V
"6__
t_
:
_Input of heating
value
Minimum~---- ---Air flow control B
--.-Air flow control A

.----~ Minimumcl>set value
I : 1----il 1----~c
Process heat demand
To burner
Pf
I
L---
u
1
I
I
I
High
~---.,
~-,
:
I
I i
E!,J
~select ~I
[Psi fs
Fuel gas B
Fig. 9 20 Total heat control system at fuel gas mixing
stant, as shown by the system in Fig. 9.20. Here, a process residual

gas A whose amount and quality vary and a stable supplementary gas
Bare combined as fuel to the furnace. Gas heating value Ht (kcal/Nm3)
is either calculated from the value detected by an analyzer (e.g., process gas chromatograph, density meter) or obtained from a gas calorie
meter. Pressure controller A (PCA) and pressure controller B (PCB) in
the figure maintains the minimum value of the burner inlet pressure in
all situations and prevents dangerous conditions from arising by
flame-out (Reber to item (f) below).
(c) Fuel and air flow control
The flow rate of fuel supplied to the furnace is determined by the
amount of heat required by the process. In order that the fuel is
burned as completely as possible, however, an amount of air A slightly
547
more than the theoretical amount of air A 0 determined by stoichiometry must be supplied. This ratio, A/A0 , is called the excess air ratio
and is usually denoted by m. An optimal value for m is determined by
the type of fuel and combustion equipment and combustion within the
furnace is controlled with mas a desired value.
In addition, the relationship between the concentration of 0 2 in
the combustion gas and m is given by the following well-known equation:
0.2l(m-l) XlQ0::::::0.2l(m-1) XlOO (%)
0.21a+m
m
Here, a is a constant dependent on the type of fuel, e.g., 0.5 for methane, 0.33 for light oil, 0.2 for heavy oil, and 0 for carbon.
Thu.::., Lased on this equation, the value of m can be controlled
through use of an oxygen analyzer. In general, if the furnace heat load
is above 50% of the total load, the optimal value of m can be considered constant. On the other hand, for that below 50%, m tends to increase rapidly. Accordingly, the concentration of 0 2 in the combustion
gas (a desired value) may also have to be increased depending on the
load (Fig. 9.21).
A widely used fuel-air ratio control system5l is shown in Fig.
9.22. The basic function of this system is to set the fuel flow rate
based on the process heat-demand signal while at the same time setting the amount of combustion air by multiplying this signal by an appropriate ratio g1
This ratio, which must be a value promoting optimal combustion,
receives feedback correction based on the output from the controller
governing 0 2 concentration in the combustion gas. The setpoint value
for 0 2 %, moreover, is automatically set according to the heat load
(fuel flow rate) for the reason described above. In addition, during transient periods when the load varies, the occurrence of black smoke due
8
~6 i--o,m=1.4
\
------'
~24
m=l.ll
0o
20
40
60
Heat load [%)
80
100
Fig. 9 21 Relationship between heat load and optimum oxygen

concentration in combustion gas
548
Process heat demand

I
I
I
~--------------{
t !------------------;
r-----------1..----- -----,
'
!Plus~
1
1
l )
'/
FA ,>-, F,:
,------'
'-----,
L----"f
I
!
:
!
:
i---;!-----'
Q
~
I
o,c
\
cBIASV HighI
1
tn
combustion gas
High
select
n.
- 0 2 Content
~nd
low hmtt
.........,,__-=
I
:aQ-J
L:,T ___ J
lq,={0.5+K(a-0.5)}X2
~
1 .o'w Minimum
~~
set value
L----~---~
t-~
~--->
0:
I
~
e:V
- K,
~- Minus _
bias
'
+K, ~
Low
;-- bias - select
-K,
~oad
~~
To burners
<=!:.->
1
-4:~~it--Fuel (Constant calorie)
Air
Fig. 9 22
Fuel-air ratio control system
to inadequate combustion air and heat loss caused by excess air are prevented by adding limiting functions (dash lines in the center of the figure) to the flow controller (FC) setpoint values for fuel (FSV) and air
(ASV). The relationship between the various values in the figure is
given by the following expressions in which FE is the calculated fuel
flow rate corresponding to the air flow:
(excess air limit FE-K2 )<(FSV)
<(black-smoke generation limit FE+ K 1)
(black-smoke generation limit FA- K3 ) < (ASV)
< (excess air limit FA + K4 )
(d) Furnace draft control
Draft within the kind of furnace shown in Fig. 9.19, which commonly employs a forced draft fan, is controlled by a draft control
valve. In such a configuration, draft control can be improved by adding
the manipulation signal of the air flow control valve to that of the
draft control valve via a time-delay factor.*
* Feedforward
control.
549
(e) Heating tube control

Piping within a furnace is formed by many branched parallel pipes. As liquid within the pipes are heated and then evaporated, it becomes necessary to control the flow in each pipe in order to uniformly
distribute the feed flow.*
Although in most cases the furnace piping outlet temperature can
be kept constant by controlling the amount of fuel, this is not necessarily easy to achieve because of the large time delay in feedback control.
However, feedforward control from process flow to fuel flow can be effective when the process flow varies frequently. In addition, if disturbance is present on the fuel side, it should be suppressed by employing
the countermeasure such as described in item (b) above.
(f) Safety control
In order to prevent problems such as burn-out of heating tubes,
damage to the furnace structure, or explosions in the combustion chamber, an emergency operation system can be installed as described in Table 9.2. Sequence control is also appropriate for automating startup
operations and for securing safety in general.
9.4
Control of Distillation Processes
9.4.1 Binary-component distillation column control

(a) Process and related equipment
The following describes the process and functions of the continuous distillation equipment shown in Fig. 9.23.
(1) Functions
The feed liquid consists of two components, one having a low boiling point and the other a high boiling point. After passing through a preheater, it enters the distillation column at a middle
stage where it becomes mixed with liquid falling from the upper column section (enriching section). The resulting liquid then itself falls to
stages in the lower column section (stripping section) coming in contact with vapor rising through the column. At each stage, residual liquid in upper stages are heated by vapor from the lower stages and boiled. In particular, since the low-boiling point component can be evaporated from the liquid by this vapor, the concentration of this component in the liquid will decrease for stages closer to the bottom of the
column. The major liquid component at the bottom of the column thus
becomes the high-boiling point one, and part of this becomes a bottom
* In case
controller
turbance,
flow rate
amount.
550
of fluid flowing uniformly at constant rate in each pipe without the flow
(FC), then if the flow should happen to decrease in one pipe due to disvaporization will 'begin increasing the flow resistance and decreasing the
even more; flow rate in other pipes will thus increase by a corresponding
Table 9.2 Emergency operation for furnace protection

Factors of operation
Reason for operation
Flow in heating tube decrease Buring out or coking of heating tube

Outlet temperature of heating Buring out or coking of heating tube
fluid rise
Supply pressure of fuel low
Differential pressure between
fuel oil and atomizing fluid
Increase
Decrease under the limit
Flame-out (trigger on explosion in furnace)

Back fire in fuel gas
Flame-out
Incomplete combustion
Draft pressure rise (depends

on stopping of fan etc.)
Internal gas flow out of furnace
Temperature of gas after

combustion rise (depends on
bursting out of heating tubes
etc.)
Furnace gets damaged
Emergency stop button
Operater make decision depending on

circumstances
Emergency operation
Note
Fuel cut
For some kind of fuel, down flow piping of shut

off valve should be purged by N, or steam
Feed cut for heating tube
In order to prevent heating tubes from over

heat, steam etc. should be flowed
Steam purge for stack
This operation is necessary if the draft pressure

rise on IDF stop and exhausting of combustion
gas is still required
product while the remainder is reheated by a reboiler to generate rising vapor.

On the other hand, while the vapor nsmg through the enriching
section comes in contact with falling liquid (internal reflux), the concentration of the low-boiling point component gradually increases in
the same way as above. As a result, the vapor which eventually reaches the top of the column (overhead vapor) consists mainly of the
low-boiling point component and is cooled within a condenser to
become liquid again. A part of this liquid then becomes distillate produCt and is removed, while the remainder is returned to the column as
reflux, and falls down inside the column.
(2) External material balance
The following equations are obtained
9.4
551
Overhead va
Feed
Renux drum
ft
0 - Distillate product
'----------,~
II
vapor
Reboiler
J---
Heating steam
Trap
....._....,Drain
L------------'~~1 -Bottom
Fig. 9 23
product
Outline of continuon di tillation equipment
for inflow and outflow of the system in Fig. 9.23 during steady-state
conditions.
F=D+B
( m~terial balance of low-boiling)

pomt component
(9.15)
(9.16)
from which we obtain:

(9.17)
Here, z is the feed liquid concentration (low-boiling point, the same as

follows), Yn the liquid phase concentration at the nth stage (distillate
product concentration) , x 1 the liquid phase concentration at the 1st
552
stage (bottom product concentration), F the feed liquid flow rate, D

the distillate product flow rate, and B the bottom product flow rate.
(3) Control variables and disturbances in continuous distillation procesThe object of distillation process operation is to separate the
ses
supplied liquid in the most economical way and to produce a distillate
or bottom product having the desired concentration (specific value).
Product concentration thus becomes the main controlled variable and
the flow rate and concentration of feed liquid becomes the main disturbances which must be suppressed. In general, since distillate and
bottom product concentration change quite slowly in responding to
these disturbances or other manipulated inputs, and in particular,
when load changes occur frequently, the adoption of feedforward control for handling such disturbances should be considered.
The suppression of thermal disturbances to the distillation column
is also important for stable operations. To begin with, if the enthalpy
of the feed either increases or decreases, the amount of vapor rising
from the feed plate will correspondingly increase or decrease and the
amount of liquid falling from this plate will change conversely. Accordingly, control is carried out to stabilize preheating conditions of feed
liquid (if the temperature of the feed liquid is the same as that at the
feed plate, vapor condensation or generation will not occur). Likewise,
if the reflux temperature increases or decreases, the amount of overhead vapor will increase or decrease and conversely the amount of liquid falling through the column (internal reflux). Thus, in cases when
such temperature change is large and needs to be stabilized,* an internal reflux flow control system can be adopted in which reflux flow is
either increased or decreased to compensate for changes in its temperature.
(b) Vapor-liquid equilibrium and column separation
(1) Vapor-liquid equilibrium of a binary-component solution
We consider an equilibrium situation in which a mixed solution consisting of two
components having a high and low boiling point is brought to a boil under constant pressure. In such a system, a fixed relationship holds between the concentration of the two components in their vapor and liquid phases depending on the type of components. This relationship can
be measured by the equipment shown in Fig.9.24 (a) under atmospheric
pressure conditions. The results of measurement are shown in Fig.9.24
(b) for a benzene-toluene system. For most mixed solutions in such an
equilibrium state, the concentration of the low-boiling point component in vapor phase (y) is higher than that for the same component in
liquid phase (x), which is the basis of separation as used in distillation.
For example, in the case of an air-cooled condenser.
9.4 Control of Distillation Processes
553
Thermometer
Co ndenser
ooling
water
12or-------------.
Benzene- Toluene system
11
1----x -~..;..__-
70
Heater
0.4
0.6
0.8
( b) An example of vapor- liquid equilibrium
( a ) Mea uring dev ice
Fig. 9. 24
0.2
Benzene conccmration [Mo le fraction ]
Va por-l iquid equilibrium under atmospheric pressure
(2) Relative volatility

The ease or difficulty of separation in the distillation process can be understood by the size of relative volatility as
described below. To begin with, the relative volatility aAs of a mixed
solution composed of a low-boiling point component A and a high-boiling point component B is defined by the following equation:
~(1-y)
(9.18)
x (l-x)
Relative volatility aAs is usually greater than 1 and separation by distillation becomes easier as this value increases. Moreover, as shown by
Table 9.3, relative volatility tends to increase generally as the pressure
decreases for a fixed-components solution. Thus, the energy required
for constant separation decreases for distillation performed under lower pressures.
(3) Total reflux operation for binary-component distillation
A continuous distillation process operation in which the respective flow rates
Table 9.3 Change in relative volatility of benzene-toluene system depends
on pressure
Total
pressure (mmHg)
860
760
660
560
554
Benzene
boiling point CC)
Relative volatility a
83.9
80.1
75.4
70.3
2.58
2.60
2.66
2.72
Chap.9
F= O
B= O
Fig. 9 25 Total ren ux operation
for the feed liquid, distillate product, and bottom product are zero and
in which rising vapor Vr generated by heating is completely condensed
into reflux L is called a total reflux operation. The following investigates such a system using the simple example shown in Fig. 9.25.
In this example, since each plate is in a vapor-liquid equilibrium
state, the following equations can be obtained where a indicates the
average relative volatility* for a stage.
ii
Yafp -ya~
xa/1-xa
Multiplying these three equations by each other, we get :
!S .
V/p1- x.g1~ . f20a

12r-- --
x./
Ya/(1- YJ}
~/(1 x.J
(9.19)
The following equations are also obtained from the material inflow and
outflow of the low-boiling point component at the top plate and bottom:
Vry2+ Lya= VrYa+ Lxa,
VrY+ Lxa= VrY2+ Lx2
Since in total reflux operation Vr = L, it can be seen from the first

equation above that y 2 =x 3 and from the second equation that y 1 =x 2
Applying these relationships to Eg. (9.19), cancellation of some of the
numerators and denominators can be performed giving us:
* Since in general the relative volatility differs somewhat between the top and bottom of the column , an average value is taken.
9.4
555
iia
Ya/(I- Ya)
xd(I-xl)
This relationship can be extended to a nth plate configuration in the

same manner. Specifically, if we designate Yno and x10 as the concentration of overhead vapor and bottom liquid, respectively, we obtain:
(9.20)*
(4) Separation of distillation
In order to express the degree of separation for continuous distillation equipment under operation, the column
separation Sis patterned after Eq. (9.20) and defined as follows:
(9.2I)
For example, under operational conditions represented by Yn = 0.95 and
x 1 =0.05, column separability S becomes 361. Although a number of
equations are known for which S can be calculated from related variables, the following ShinskPl equation is the most concise and easy to
use:
S=[ JI +I/{( ~r/D-I)z}
(9.22)
where Vr is the flow rate of rising vapor.

Examining this equation, it can be seen that in the case of total reflux operation for which D is 0, it reduces to Eq. (9.20). Note also that
the value n in Eq. (9.22) is not the actual number of column plates, N,
but is the effective number of plates (n=NE) obtained by multiplying
N times the plate efficiency** E.
(c) Control system for binary-component continuous distillation
The control system for a particular piece of distillation equipment
will in general be determined by target and disturbance conditions.
Since it would be impossible to cover all of the various cases here, a
few representative examples are described.
We
(I) Control of distillate concentration during changes in feed flow
consider a system in which the concentration of the distillate product
is controlled by manipulating its flow under conditions of changing
feed flow rate and constant heating steam flow rate.
Referring to Eq. (9.22), if Vr is constant and z, a, and n all invariable, then S will clearly depend only on D.. Moreover, by rearranging
* Fenske
equation.
As vapor-liquid separation in each plate is not performed under ideal conditions,
correction is necessary based on efficiency.
**
556
Eq. (9.21), we get:

Yn
(9.23)
S-yn(S-1)
If we now eliminate x1 between this equation and Eq. (9.17), then if Yn
is kept constant, we can see how D/ F depends on S, or in other words,

on D. A fixed relationship thus holds between F and D. Figure 9.26
shows several curves representing this relationship based on calculations of successive approximation by computer with the parameters
shown.
0.6
Condition _ _-+----!
z=0.5
=1.52
n=l8
0.5
Fig. 9 26
1.0
F (t/h)
1.5
Relationship between each flow rate in feed and distillate

when concentration of distillate is controlled
Based on the above description, it can now be seen that the distillate concentration Yn in relation to disturbance F can be maintained at
a fixed value by setting distillate flow D according to the relationship
D= f(F) as shown in the figure. Figure 9.27 shows two types of control
systems based on this conclusion employing (a) feedback control, and
(b) feedforward control. Note that in the latter case, a dynamic compensator element L/L (Lead/Lag) and feedback corrective operation must
be included.
(2) Control of bottoms concentration during changes in feed flow
In
this system, the concentration of the bottom product is controlled by
manipulating its flow under conditions of changing feed flow rate and
constant reflux flow rate. Approaching this system in the same manner as with the previous description, we rewrite Eq. (9.22) under
steady-state conditions of Vr = L + D and D = F- B:
(9.24)
9.4
557
----~
o n tan!
~
onstanl
onslanl
~
I
I:_ __ __ ____ J
L--------..J
( a ) Feedback co nlrol system
Fig. 9 27
( b ) Feedforward control system
Concentra tion control of di tillate for feed flow change
From this equation it can be seen that S depends only on F- B.

If we now eliminateD between Eqs. (9.15) and (9.16), we get:
(9.25)
In addition, from Eq. (9.21):
Yn
(9.26)
1+x,(S-l)
If x, is held constant in these two equations, then B/F depends on

S, or in other words, on F- B, and a fixed relationship between F and
B must hold as B = g(F).

Figure 9.28 shows two control systems based on this result.
(3) Control of distillate or bottoms concentration during changes in feed concentration
We first consider a system in which the concentration of
distillate product is controlled by manipulating its flow rate under conditions of changing feed concentration and fixed feed and heating
steam flow rate. With these conditions in mind, we first eliminate x 1
from Eqs. (9.17) and (9.23) and see that D is dependent on S and z. In
addition, since from Eq. (9.22) S depends on z and D, we can now eliminate S from these last two equations to obtain a relation between z and
558
Chap.9
F - ---1
.___ __
{ll<l- D
ons tan t
F - D----1
z
:
onstan1~
Vr
--t--
( a ) Feedback control system
( b ) Feedforward control system
Fig. 9 - 28 Concentration cont rol of bonoms for feed fl ow change
D: D = <P(z). With this result, a system exactly like the feedback control one in Fig. 9.27(a) can be realized. In order to realize a feedforward control system, however, the feed flow rate in Fig. 9.27(b) must
be substituted by a concentration signal, and the function f(F) by the
relation D = <P(z) described above.
Considering next a system in which the concentration of bottom
product is controlled under conditions of changing feed concentration,
we can draw an analogy from Fig. 9.28 in the same way as above (Refer to practice question 9.4).
(4) Control of distillate and bottoms concentration during changes in feed flow
Assuming that feed concentration z is invariable, then in order
to keep distillate and bottom product concentration Yn and x 1 at fixed
value, the ratio D/ F as expressed by Eq. (9.17) must be held constant,
as must the value S from Eq. (9.21). In addition, if S is constant, then
from Eq. (9.22) the ratio Vr/D must also be constant, which in turn
means that V/F is constant as welL
Based on the above discussion, we examine the control system 6l
shown in Fig. 9.29. Although by employing a concentration controller
for both the distillate and bottom product, feedback control for both
the distillate flow rate and heating-steam flow rate is possible, controllability can be improved even further by adding two feedforward paths
from the feed flow rate. Note also that due to the mutual interference
9.4
559
r-- --
___
vT
..--..;.:_
____,
iF. F.
~-~IF
. Feed
~ Constant
-rSF.
r ,----~
-(f.9
Distillate
Vr
-f-
~!"""G--I)IQ-to-(
: Steam
I
I
I
z. :
B
Bonoms
L____ _-~--------clfJ
F. B
Fig. 9 29 Concentration control of both distillate and bottoms for feed

flow change (Case of high purity in bottoms)
between the two concentration controller loops in this system, it is recommended that the bottom product exhibit high purity (e.g. , 0.999)
while that of the distillate product be relatively low (0.8 to 0.9). If the
opposite case is needed, a system can be configured in which the distillate concentration is controlled based on reflux and the bottom concentration is controlled by manipulating the bottom flow rate.
In either of the above two systems, mutual interference is low, so
that no problems should arise in practical use. 3 )
(d) Control of distillation column pressure
Stable operations in distillation equipment require column pressure
to be controlled in a steady-state fashion. Although various types of
control systems for achieving this have been processed, 7 ) this section
presents only several examples which have come into wide use, all characterized by the condenser using cooling water.
Referring to the system in
(1) Inert gas injection and purge technique
Fig. 9.30(a), if the pressure in the column should happen to drop, inert
gas is injected to maintain pressure, while if the pressure rises, gas in
the column is purged to the outside, thus lowering the pressure. Two
560
, fpC\- -------,t
,XY
- ~-------------- -,
:~
lnen gas :
...'
I
I
L--- - - - - Distillate
(a) Injection of inert ga
_____
(b) Hot gas bypas
,....._,.JP<:L-----------,
_,:
' x:::r
'
I
I
I
I
(c) Submerged in condenser
Fig. 9 - 30 Pressure control at distillation column
control valves are employed here operated on a split-range basis. Despite some negative features such as increase in vent-gas flow rate and
solubility of inert gas in distillate, the high responsiveness of this technique has come to be widely used in pressurized columns operating
near atmospheric pressure.
(2) Hot- gas bypass technique
In the example shown in Fig. 9.30(b), if
the amount of vapor from the top of the column increases causing the
pressure to rise above its setpoint value, the pressure control valve
will be moved in the closed direction (in the opposite direction if pressure drops below setpoint value) decreasing the amount of bypass vapor. Consequently, the amount of condensation on the surface of the
liquid in the reflux drum decreases, which in turn causes the surface
temperature to drop and the pressure in the drum to fall. As a result,
liquid in the condenser tends to flow into the reflux drum increasing
the condensation heat-transfer area in the condenser and raising the
speed of vapor condensation. This finally causes the pressure in the column to drop back towards its setpoint value. In an actual situation, if
the condensed liquid at the condenser outlet is sub-cooled* to about
4 oc, the control valve pressure differential will become about 0.2 to
Cooling to a temperature even lower than saturation temperature; bypass vapor

flow rate is governed by this temperature.
9.4
561
0.3 kg/em. Because of the excellent response of this technique, it has

come to be applied in many pressurized columns.
(3) Submersion of condenser heat transfer area by condensed liquid
As
shown in Fig. 9.30(c), column pressure is controlled in this technique
by manipulating the amount of condensed liquid flowing out of the condenser thereby increasing or decreasing the condensation heat-transfer area.
(4) Condenser cooling water adjustment technique
This technique
should be easy to understand without the aid of an illustration. In particular, referring to the earlier description of fluid-to-fluid heat exchangers, it should be clear that if the heat load drastically increases,
the sensitivity of pressure control by cooling water adjustment will decrease significantly. Another disadvantage here is that as the cooling
water flow drops during times of low load, scale deposits from the heat
transfer surface are produced by low flowing rate. Thus, this technique requires careful consideration on application.
(5) Floating pressure controf2>
During times of fixed load, it is possible to set the value of the distillation column pressure even lower if
the temperature of condenser water drops. Since a lower column pressure is directly related to energy conservation as described in (2) of (b)
(about 2% reduction in energy for every 1oc drop in cooling water
temperature 3 > ), maintaining a column pressure as low as possible by
making the condenser always operate fully corresponding to the cooling water temperature in the condenser can be quite effective (floating
pressure control). On the other hand, as previously described, the column pressure must be kept as close as possible to a steady-state value.
An example of a control system satisfying these conditions is shown in
Fig. 9.31.
In the figure , the function of the pressure controller PC is funda-
Fig. 9 31 An example applied for floa ting press ure control
562
mentally the same as that in Fig. 9.30 (c). If cooling water temperature
drops, then if the column load is fixed, the cascaded setpoint value of
the condenser's level controller (LC) becomes large. However, if the
pressure controller (PC) setpoint value is lowered via setpoint controller (SVC) output, the liquid level can be kept around normal low levels
(20%). In addition, since a radical change in column pressure disrupts
column stability, slow integral action* is adopted for SVC. Such operation is considered adequate since changes in water temperature themselves occur slowly. Moreover, since the amount of heating vapor must
be decreased corresponding to the drop in column pressure, the column
pressure signal is applied to the heating steam flow setpoint value as a
feedforward signal.
(e) Loop enhancement in a distillation column control system
Up to this point, we have only presented basic configurations for
distillation column control systems. The following discussion examines
ways to enhance system control performance in an efficient manner.
(1) Enhancing the distillate concentration control system8 >
Concentration control for the systems shown in Figs. 9.27 and 9.29 can be greatly improved by adding local feedforward control as shown in Fig. 9.32.
Here if K = 0, we have simply level controller/flow controller (LC/FC)
cascade control. If K = 1, however, a change in distillate flow D is compensated by an equivalent change in reflux flow leaving the liquid level
unchanged. The effects of distillate flow manipulation thus reach the
column as changes in reflux without a time delay. Moreover, if K > 1, a
lead element is employed to compensate for the time delay required for
reflux manipulation to effect changes in concentration, thus improving
controllability even further. On the basis of the above, it has been reported that the oscillation period of the concentration control loop can
be reduced from 5 hours to 30 minutes.
(2) Distillation equipment constraints and control
The operation range
Fig. 9 32
Improvement in concentration control of distillate 8 l
An integral time of about one hour is selected.
563
..8.
3. Reboi ler
capacity
.
:~
2. ondenser
capacity
>
1. High limit of col umn

differential pressure
....0
Feed flow rate
t
-
Column pressure P
Fig. 9 . 33 Constrained condition for load of distillation column
of distillation equipment is constrained by limits in the column pressure and in the handling capacities of various parts with respect to rising vapor flow as shown in Fig. 9. 338 ) To begin with, if the flow velocity of rising vapor in the column becomes excessive, liquid will likewise rise with vapor (flooding), creating problems for distillation. However, since the column pressure differential is proportional to the
square of the vapor flow velocity, the flow velocity limit can be determined from this relationship. For example, in the event that a concentration control system like the one in Fig. 9.29 requests an excessive amount of heating steam, the control configuration shown in Fig.
9.34(a) can halt the request if the column pressure differential goes
above its limit.
Next, referring again to Fig. 9.33, we see that with a rise in column pressure, condenser handling capacity increases while reboiler
handling capacity decreases. This is due to the fact that as the pressure rises, the temperature in each vessel rises, with the temperature
differential of the condenser increasing and that of the reboiler decreasing. In addition, handling capacity in either case will drop remarkably with a rise in cooling water temperature in the condenser, and
with an increase in heat- transfer resistance due to surface fouling in
the reboiler.
Points 1, 2, and 3 in the same figure show the limits of rising vapor flow determined from the handling capacity of the elements under
a particular operating pressure. In this example, the allowable upper
564
Chap .9 Control of Process Units
sv
Maximum value allowable
!-------~La~-~~~~r----~~f Col umn differentia l

pressure
- - - - - - - - - - ...,.
~--- FC
I
I
:
I
---.
'
:External feedback
I
.z
1
fc. . ..
----------------1~~
Steam --~-i 'l--------olcr--
L____
'.Concentration
ofbonoms
( a ) High limit control for flow rate of heating steam "'
SV
10%
~~y _________
sv
90%
r-
Maximum value
2 ~ ~ r:!!;' , .~lowabl
r---
-~
- -----
~~--,
PY'i{~~~~ -- -i
\_ ..._ __ 'm.t ......../
xi~~:rnl
.
.w
r-
Feed
reed back
.----~----~
---;!~
,
r--}:-- --
l ___ - - - - - - - - - - ,
(D Low selecter
~-~-J
-----l'ot---~,00....-
(b) Capacity-limit control of distillation column
Fig. 9 34 Constrained control of distillation column
limit of rising vapor flow will be governed by reboiler performan ce 3.

Moreover, since feed flow of the column correspon ds to rising vapor
flow, the control system shown in Fig. 9.34(b) will set the feed flow to
the governing value determine d by the current handling capacity of
each section. In this way, distillatin g capacity will always be kept at
its maximum attainable limit.
In the same figure, the upper limit of column pressure is obtained
from the output of differenti al pressure controller .JPC, and the capacity limit of the condense r and reboiler from the output of respective
control valve controller s VCl and VC2. Thus, by setting the feed flow
565
. 90%
V I
ave opening , o
j, -~~
PV2--
"f' "i'
VC- 2
,_~------
rt-t.
VC-1
4PC
1External feedback
I+Ci,(-;llJ~~rY.--bL__t--i;;,::c.T_W-_-:
'
Low selecter
~ FIC
I
L----------,-
I
'
:._--~~"!:~a.!._Ie~<!b!-~~--J
'l
1
~:~
I
Feed
If m 2 is selected, the following equation is obtained

ms(s)
b'2 (s)
100
(1+
_l_ )
l.QQ_
TIS
Pc
1+T,s
Pc= Proportional band, T,= Integral time
Fig. 9 35
=-p;'
1_
1 ___
Controller with external feedback used in selecter control
controller with the minimum value from the output of each controller,
the above objective can be realized.
Each controller in Fig. 9.34(b) having external feedback is configured in the manner shown in Fig. 9.35. 9 ) As can be seen from the inductive equation in the box, only the controller chosen by the selector takes on P +I action, while the remaining controllers apply bias to selector output by P action only. Since at the time of signal switchover by
the selector the signal must cross over the setpoint, the control deviation of the P action controller becomes zero and switchover is performed in a smooth and reversible manner.
9.4.2 Multi-component distillation column control
(a) Two-product distillation column control
The most common form of distillation involves feed liquid which,
in addition, to heavy and light "key" components, includes a relatively
small amount of heaviest and lightest components (non-key components), all of which are separated in one multi-component distillation
column. Such a column is also used as a basis for forming a multi-column train system.
Concentration break(1) Material balance and product composition
down in each section of a two-product distillation column and concentration distribution throughout the column are shown in Figs.
9.36(a) and (b). As shown , the lightest (heaviest) component is ordinarily completely separated into distillate (bottom) liquid. We thus obtain
the following equation:
(9.27)
zuF=yuD
566
Chap.9
Condenser
I
I
I
I
I
I
I
y,.
Y<
Y
F
z,.
%<
'I
'
'zu
Feed plate ,~m-----;h
r---..J
Zh
z..
""
%h
%hh
( a ) Concentration of each part in column
1.0
0.4
0.6
Concentration
( b ) Concentration distribution of
each component in column
Fig. 9 36 Two products distillation column for multi-component system
In addition, since Y11 +Yz +Yh = 1, we get:

(9.28)
Moreover, since the relation zzF=yzD+xzB holds for the light key component (l) after material balance, we use it along with the simple relation B=F-D to obtain:
(9.29)
Substituting this equation into Eq. (9.28), we arrive at
Zz-xz(l-zu)-yz(zu+zz-xz)
(9.30)
Equation (9.30) expresses the relationship between x 1 and yh (key-component separation) given feed compositions Zzz and z 1 and provided that
distillate key-component concentration Yz is a stipulated value. Figure
9.37 shows the results of applying this equation to a debutanizer column. It can be seen how, if heating steam is increased to lower bottom
butane concentration Xz, distillate pentane concentration yh will increase. Accordingly, since Yh +Yzz is fixed, Yzz will decrease, and from
Eq. (9.27), it follows that a greater amount of butane product D will be
obtained from a specific composition of feed liquid. However, the ex9.4
567
Butane product
specification
Butane Y< 86.2%
Butane concentration in bottom product x 1 [%)

Fig. 9 37
Relationship between xz and Yh at debutanizer
tent to which Xt can be lowered is determine d by considerin g the

trade-off between the cost of increased heating steam and more (distillate) product.
In the same manner, a relationsh ip similar to Eq. (9.30) can be derived for bottom liquid. We omit the derivation here and present only
the equation:
Xt
(1- Yh)(zu + Zt + Zh- Xh)- (zu + Zz)(1- Xh)
(9.31)
zh-yh
In this case, the relationsh ip between Yh and Xt is obtained given zu, Zt

and zh and the stipulated bottoms key-comp onent concentra tion xh.
(2) Relationship between light key-component concentration and light
non-key component concentration
Eliminati ng the ratio F/ D from
Eqs. (9.27) and (9.29), we obtain:
Ordinarily, since Xt (bottom liquid impurity)

lowing approximation holds: 8 l
< Zt
and
Xt
< Y~>
the fol(9.32)
From this, it can be seen that since both componen ts change while
maintaini ng a proportio nal relationship, control of multi-com ponent distillation can be approxim ately treated as that for a two-comp onent system.
(b) Concentration control in distillation column train
A process in which feed liquid consisting of N componen ts is
sequentially separated by (N -1)* distillation columns has found wide
By using more than (N -1) columns, intermediat e products may also be separated.
568
Chap.9
D,
~,1/tb
F,--...-J
Aooozo
Boooz
Co ooze
Doooz
Do ooMain component
Fig. 9 38 Separation for four components system using three columns
application. An example of economic optimal control of such a process

is described below. 10 l
(1) Process description
Figure 9.38 shows a system in which three
columns are used to separate an initial feed liquid consisting of four
components: A, B, C and D (from light to heavy components in that order). Here, the concentration of each feed component i.s represented by
Za, zb, Zc and zd, the distillate and bottoms flow of each column by D
and B, and their respective concentrations by y and x, with subscripts
1, 2, and 3 corresponding to columns I, II, and III, respectively. The
following conditions are assumed for the distillate product of each column: Y1a must be held constant, Y2b > 0.955, Y2c < 0.015, Jgc must be
held constant and Y3b/Y3c (:::::::: x2b/X2c)* = 0.03.
In addition, the relationship between the economic value V of each
component unit amount is assumed to be Va < Vb < Vc.
Since the distillate compositions of columns I and III are already
set, only column II becomes the object of economic operation. As
shown in Fig. 9.39(a), y 2 b of column II's distillate concentration is fixed
to 95.5 %, the minimum value satisfying specifications. Of the remaining 4.5 %, however, if the amount occupied by least-expensive component Y 2a is extended, the economic value V2 for D2 unit amount increases, with the amount of change given by
L1Vi=(y2a Va+ Y2c Vc)-(0003 Va+0.015 Vc)
But, since an increase in Y2a causes a decrease in Y2c, the amount of

steam, which is required to separate B and C in column II, increases,
resulting in an increase in cost C2 per unit amount of D 2
On the other hand, an increase in Y 2a allows an increase in Xw (re-
Refer to Eqo (9o32)o
9.4
569
/d
Economic gains in product Ll V2
A~AC~--------+----------b~--- Increase of separation energy costs

in column II LlCz
l------+---7'"'---r--;t'-7!Total increase of separation energy
costs AC2 - LlC1
Decrease of separation energy
r--------::;;;f'~---=::.;~~~:::::=- costs in colnmn I AC1
(Economic gains) -(Increase of
1--~~....--\:;~~~--+-+-----:-:- energy costs) AVE
3.5
y2.x 100
4.0
[%)
1.0
YcX100 (%)
1.5
0.5
( b ) Change in economic gains
Fig. 9 39 Relationship between bottoms composition of column II and

economic gains
fer to Eq. (9.32)), so that the heating steam in column I can be decreased, thus lowering steam cost C1 slightly.
In general, the following approximation holds between the amount
of rising steam and the concentration in a distillation column. 11l
Vr -;:;;;a log---1F
XtYh
We can now obtain the increment C1 and C2 from the above equation.
LJC-;:;;;c{3log X!a -;:;;;c{3log Yza
X!aB
0.03
AC ~
,(.!
z~ca
1
0.015
og 0.045-Yza
Here, c is unit cost of steam, a and {3 are constants, and x 1an is the refer
ence.
From the above considerations, we can obtain the net economic value
increment .JVE per unit amount of D2 as the following equation.
LlVE=Ll~-(LlCz- LJC,)
Figure 9.39(b) illustrates the various relationships described above.

570
The value of Yza for attaining economic optimal operation (referred to

as Y 2a0 ) can be obtained from these curves.
(2) Economic optimal control system
The logical configuration of a
system for economic optimal control of the above three-column process is shown in Fig. 9.40. The basic loop controls section and product
quality controls section perform concentration control by feedforward
mechanisms such as those in Fig. 9.29. The economic optimum control
section, on the other hand, calculates the setpoint values of each concentration satisfying various given conditions using relevant values. In
the case of using floating-pressure control, since the change in column
pressure here affects the amount of required heating steam flow, column pressure P 2 must also be taken into account. (Refer to Subsec.
9.4.l(d)(5).)
Values and costs
r---
r-'::-L--L--__,-1-_ _L,x 1 a'/X1 o' ( ""yza/Yzb)
To quality controls
in column I
I
I
I
I
I
Dz/F,
Yza
'+ Yzc ,Yza ,
Setpomts
Xzb '/xzc, ("" y,bjy,c)
Column pressure P,
D, Specifications
I
I
I
"--~---From analysis
IFeed
Fz D,' Q,' Set points

:rorward
I
~~~--------~
L_ ----
Process variables---.l.-Ll....L..J...J
(F,L, T, P)
Fig. 9 40
L...LL...LL..L--
Manipulating variables
(CV)
System construction of most economical control at column II
(c) Control of a side-stream distillation column
An occasionally used distillation process is one in which side

stream is employed in order to obtain three or more products from
multi-component feed liquid in one column. This type of process is
characterized by lower equipment cost and energy savings.
(1) Side-stream concentration control
In most cases, side stream implemented at a plate above the feed plate draws liquid in order to
reduce as much as possible the concentration of light components. On
the other hand, side stream below the feed plate draws vapor with the
object of reducing the concentration of heavy components. In Fig.
9.41(a), if we increase the flow of side-stream liquid, internal reflux below the drawing plate decreases, and as a result heavy-component con9.4
571
From
__( - ---- reOu x drum L C
.-' --{V- --r--@--,
. F. F.
~
I.
ReOu x
:r--w--;i:, ti...J
~ Ratio
Side strea m liqu id
s ~ ------ ~
~I
Oj
r.A :
...
..@.v__ _E.,!3, __ __ j
,..-- ~ -J
Feed
( a ) Side strea m liquid cont rol
( b ) Side stream vapor cont rol
Fig. 9 41 Side tream contro l of d istill ati on column
cent ration within rising vapor increases. Consequently, by reducing

side- stream flow when the temperature controller (TC) measured value becomes higher than the setpoint value, heavy-component concentration in the side stream can be controlled. Moreover, in order to
accurately maintain concentration at a certain level, the temperature
controller (TC) setpoint value can be corrected by the output from the
concentration controller. Finally, column conditions can be stabilized
by incorporating feedforward control in the side steam based on variations in reflux.
Considering now the process in Fig. 9.41(b), if the flow of
side- stream vapor is increased, the amount of rising vapor above the
drawing plate decreases and the concentration of light components in
falling liquid increases. Thus, by reducing the flow of side- stream vapor when the temperature controller (TC) measured value becomes less
than the setpoint value, light- component concentration in the side
stream can be controlled. In addition, feedforward control based on the
amount of heating steam is useful for maintaining column stability as
572
in the case of reflux above.

(2) Distillation column control with little side stream
In a multi-component distillation column, the concentration distribution of certain components in the lightest or heaviest component does not change as
simply as that shown in Fig. 9.36; maximum values may be exhibited
at an intermediate plate in the column. In order to lower these concentrations in distillate or bottom liquid, a small amount of side stream
can be drawn at an intermediate plate having such a maximum concentration. In this case, side-stream flow is controlled by keeping it at
a fixed ratio with the feed flow.
9.5 Control of Reaction Processes
Reaction equipment, in which feed material (reactant) is subjected

to chemical reactions in order to obtain a product of higher value,
plays a central role in chemical plants. Although there are various
types of reactors depending on such factors as phase of feed materials
or products, heat-transfer method and operational mode, this section
will take up the essentials of controlling a stirred-tank polymerization
reactor and a gas-phase catalytic reactor. These two types of reactors
are widely used and are good examples to understand reaction process
control.
9.5.1 Control of a stirred-tank polymerization reactor
(a) Continuous polymerization process and temperature control
(1) Process overview
Although the main material fed to the reactor
is a monomer,* a solvent may also be supplied simultaneously in the
case that the reaction is to be performed within a solvent. In addition,
a small amount of catalyst may be supplied continuously to promote
the reaction, a process which can greatly effect reaction speed and
polymer quality. An additive may even be provided "in order to control
the molecular weight of the polymer product.
Monomer polymerization is a heat-generating reaction in which
reaction temperature exerts a great influence on product molecular
weight. Consequently, reaction temperature must be accurately controlled (e.g., to within 0.5oC) employing any appropriate method
which removes the generated heat (see Fig. 9.42). Moreover, since it is
necessary to polymerize as much monomers as possible under conditions of limited reaction speed, it is common to set retention time**of
reaction liquid to a large value (several minutes to several hours).
* If
*
producing a copolymer, two or more types of monomer may be fed.

(reaction system volume)/(liquid volumetric flow)
573
IO"C
Reaction temperature
Fig. 9 42 An example for innuence of reaction temperature

to polymer molecular weight
A variety of heat-removal tech(2) Control of reaction temperature

niques are available for temperature control, depending on actual circumstances. For example, a sensible heat technique may be used in
which cooling water flows through a reactor jacket or internal coil, or
in which reactor substances are circulated externally through a cooler.
Latent heat techniques may also be used, such as cooling vapor- generating monomers and solvents in a condenser.
Figure 9.43 shows an example of such a reactor in which both of
the above techniques are used. Here, reactor temperature is adjusted
by varying the temperature of the water circulating in the reactor jacket. Specifically, water temperature is controlled only by a proportional
action controller (proportional band about 20%), with the desired value
set by the output from the reactor temperature controller (PID action)
as a cascade control. In addition, since the amount of heat removed by
-1
I
I
I
I
Fig. 9 43 Temperature control sy tern in polymerization process
574
jacket cooling is not totally sufficient, the amount of circulating gas is

adjusted based on the output from the valve controller while keeping
the cooling-water valve at an appropriate opening; this enables the
amount of heat removal from condensation to be increased or decreased. Also incorporated here is a steam heater in the circulating water system which has the role of bringing reactor substances up to reaction temperature at startup.
(b) Quality control in continuous polymerization
(1) Problems in control
The polymerization reaction process is controlled with the object of producing a variety of polymers each requiring a certain amount and level of quality as stipulated by the end user.
Since the number of various polymers in demand is usually large, product specifications switchover must take place quite frequently (e.g.,
every day or every several days), at which time control conditions
must be changed accordingly. However, since on-stream quality measurement is usually impossible, lab analysis is often employed. Form a
control point of view, this can be the source of a large dead time (several hours). Moreover, since the large retention time described above provides a significant time delay in operation results, quality feedback control becomes exceedingly difficult, resulting in long product-specifications switchover times. In addition, as odd products formed during
switchover are mostly off-specifications, the reduction of switchover
time is a prime economic consideration, and adequate control measures
for alleviating this problem are eagerly desired.
(2) Steady- state quality contrott 2 )
Taking up the problem of quality- switchover response, we first consider quality control in a
steady-state situation. In the process shown in Fig. 9.44, product quality P1 and P2 can be controlled by adjusting the concentrations of addi-
Monomer
t-----(:lcl....._
00
adjustl~~~------fDc~
Manna!
.___ ._t>IO-'........,..__..___
for A C 2 set point L_ __ _ __ _ __ _ _____j Sample
Fig. 9 . 44
P>
Polymerization reaction process and its basic control system ,
575
tives I and II respectively within the reactor liquid. Thus, if the following formulas can be determined, a means for performing quality control under steady-state conditions can be obtained.
P~=f~(Kl, T, C, Cz)}
(9.33)
Pz= fz(Kz, T, C, Cz)
Here, Tis the reaction temperature, CI and C2 are the mole concentrations of additives I and II, respectively, and KI and K 2 are calibration
coefficients based on the results of lab analysis.
In more detail, values CI and C2 in the figure are controlled by adjusting additive flow by respective concentration controllers. Moreover, as C1 and C2 are desired values corresponding to specified values
PI and P2. they can be calculated by functions (9.33), and by making
them setpoint values for each controller, the desired quality can be
achieved.
(3) Predicting the quality valuei 2 )
For the sake of clarity, the following description considers only the change in product quality PI resulting from varying the feed flow of additive I (the description applies in
the same way to P2). Referring to Fig. 9.45, the reaction is in
steady-state before changing the flow of additive I, and only the polymer product corresponding to product quality Pio (black dots in the figure) determined by function (9.33) is generated at a speed Rp. If feed
flow of additive I now undergoes a step increase at time 0, reaction
conditions change in a short time, and from this time on, only the polymer product corresponding to new product quality Plf (white dots in
the figure) is generated at the same speed. Thus, at any arbitrary
time, outlet product quality Pu (the same as that of substances within
the reactor if mixing is complete) will reflect an average value due to
the mixture of generated products corresponding to product qualities
P10 and Plf If we designate the polymerization holdup volume within
the vessel as Vp, Pu can be determined by the following mixed-process
equation,* by which we see that it varies by the first-order capacity
lag of time constant VP/Rp.
Vp~
(9.34)
Rp. dt +P~t=Plf
In this equation, VP is obtained from the value resulting from multiplying the reactor substance volume by corresponding liquid density. RP
is the value obtained by dividing the rate of heat removal (obtained by
multiplying the temperature difference between the cooling water inlet
and outlet by the corresponding flow rate and specific heat) by the po-
A similar equation holds for Pu, p,..
576
Chap.9
Additive 1 (Cha nged)
Catalyst
Outlet product
Fig. 9. 45
hange in product quality by time
lymerization heat (kcal/monomer kg). Product quality plf, being that of

polymer substances generated in the reactor at the current time, can
be determined by substituting actual measured values into function
(9.33). Thus, predicted product quality Pu for the outlet product at the
current time can in turn be obtained from Eq. (9.34). An example of a
system for performing quality prediction is shown in Fig. 9.46. Here,
constants K 1 and K2 are constantly being calibrated so that the predicted product quality agrees with the actual value as close as possible.
Specifically, every time lab analysis results are obtained, process variables corresponding to the time that data for the analysis was sampled
are searched for in a historical memory file. These values are then consolidated and values which best conform to the actual situation are calculated for K1 and K2 In this way, if the current value of the outlet
product quality can be predicted, feedback control can be performed
without difficulty and applied to the reduction of product- specifications switchover time as described next.
Feed
[g:l =:-:::.::::
u ( t) = T Process data
Ro - ---- -
v. - -----
(p.(t))
Polymerization 1--:::----:------Product c::;>

reactor
P> ( t)
Actual property
(U nknown)
1=====:::::1 11t=======~> (~: ~:l )
Current property
predicated
Fig. 9 46
Prediction system of polymerization product property
577
(4) Product specifications switchover control

For a reactor in which
polymer substances having quality P10 and P20 are generated, we consider operations in which a change in process conditions results in a switchover to product qualities Pu and P2f As described above, if functions
(9.33) can be determined, by setting values C1 and C2 to Cu and C2f corresponding to new quality values, the generated substances will
achieve new product quality after an elapsed time as shown in Fig.
9.45. However, in order to complete switchover in as short a time as
possible, a control scheme as illustrated in Fig. 9.47 13) is highly effective. Here, when switching outlet product quality from P10 to Pu, initial value C~> rather than being set to value Cu corresponding to product quality Pu, is set to a value C1 m corresponding to P1 m, which is sufficiently greater than Pu As a result, the product quality Pu of the
outlet product (a mixture of polymers corresponding to product qualities P1o and P1 m) rises quickly. If this value Pu can be predicted momentarily by the method of Fig. 9.46, then if C1 m is switched over to Cu at
the time B where Pu reaches Pu, subsequent outlet product quality
will remain constant at Pu
CP1mD
r--:r,~---..-..------------
1~
/.,.. ..
I,PI: B,'
1./'
0
Plf
--------
---
Fig. 9 47 Switching control of product-specifications
As can be seen from Fig. 9.47, the larger the value P1 m is made,
the more switchover time can be reduced. However, the step width
here cannot ordinarily be made extremely large due to a separate quality limit associated with the p 10 , P1 m and Pu mixed substance produced
shortly after switchover point B.
{c) Control of a batch polymerization reactor

A batch reaction system consists of various processes such as initial charge, pre-treatment, reaction, post-treatment
and discharge, and is repeatedly started and halted in a high-frequency, cyclic fashion. Total automation of this kind of system is being rapidly adopted due to the application of distributed digital control, which
has come to be characterized by a functional synthesis of feedback control and process sequence control.
578
------!!
><: 50
...
~t - ......
-"
1:
700 ~
600 ~ ::;;
-
<.>
500 ~ c,
400:
-; o
~e
0"
,.
0 o~~~~4~~+6~~8_J~
"
0
u
Reaction time [h)

Fig. 9 48
An example for reaction of batch polymerization process
Figure 9.48 shows an example of the reaction process for suspension vinyl chloride polymerization. Although the percent X of monomer converted to polymer is increasing with time here, its rate of
change (reaction speed) dXjdt exhibits a peak at a certain point. In addition, this rate of change is proportional to the rate of generated reaction heat. Furthermore, the overall heat-transfer coefficient U between jacket cooling water and reaction liquid decreases* as polymerization advances. Accordingly, since in this kind of batch polymerization reaction a wide range of heat disturbance usually exists, temperature control planning must carefully take this into account.
(2) Temperature control
Temperature control of batch reactions is
basically the same as that of continuous reactions described before.
However, as the reduction of total cycle time in batch reactions is
closely coupled with raising productivity, those particular processes involved with initial temperature raising must be performed as rapidly
as possible. A point of concern here, though, is the danger of runaway
reaction due to rapid acceleration of the reaction if the upper limit of
internal reactor temperature is exceeded; the rate of generated heat
surpasses heat removal capability resulting in a significant rise in temperature.
If the initial temperature raising process is controlled by standard
PID action, the described danger may occur by excessive temperature
overshoot caused by reset-windup. The methods shown in Fig. 9.49
can be used to prevent this problem. In Fig. 9.49(a), the heating steam
manipulated variable is kept constant until the deviation reaches a specified value. On reaching the switchover point, fixed bias is decreased
and after returning to an appropriate manipulated variable, the process
is switched over to PID control. In Fig. 9.49(b), on the other hand, the
manipulated variable is reduced stepwise along with the decrease in
According to the rise in produced liquid viscosity.
579
Setpoint
Change over point
(Lock up)
,- ... .._.~-
Constant manipulated
variable
variable
....Manipulated
-----------
PID Control
( a ) Manipulated variable is held in constant until changing over PID control

Setpoint
Change
- PID Output
Manipulated variable x 2 [%]
Manipulated variable x 1 [%]
o~"E { ~
pomt
Manipulated variable 100%

1-------f----I--------Time
Changing over manipulated
PID Control
variable
( b ) Manipulated variable is reduced gradually until changing over PID control
Fig. 9 49 Control without overshoot in rising temperature at batch reaction
deviation.
Although control after attaining reaction temperature is essentially the same as that for continuous polymerization, heat removal over a
wide range is necessary due to extensive heat disturbance as described
above. It should also be noted that since variation in the rate of generated heat and heat-transfer coefficient changes process dynamics, control parameters are automatically updated with advance in polymerization so as to maintain optimal control settings over the entire process.
9.5.2 Control of a gas-phase solid-catalytic reactor
(a) Multi-stage fixed-bed reactor control
A reactor made up of multiple catalyst beds is
considered as shown in Fig. 9.50. Here, feed gas reacts adiabatically in
a catalytic bed and turns hot gas,* which after mixing with cooling
gas, experiences a drop in temperature. This gas then passes to the
next bed where a similar reaction is repeated. With this type of reactor, it is common to control temperature, concentration and pressure
in each stage within desired values with the object of maximizing the
In the case of an exothermic reaction.
580
r----------- --------,
r---------- ----t-1--------.
r---
SV
L_-<t
I
sy
Q ---l i: sv
--y
! l_~
--
I
I
1
I
I
Purge gas
operation etc.
\r-11----....:....---L,--------Prod uced gas
: sy
~--
L--~--l.-----J--j~!~Feed
gas
Compressor
Fig. 9 50
Control system of multi-stage reactor
-Reacting process
---Mixing and
cooling process
~
0
~
(.,)
c::
.g
~
c
"uc::
2.
Temperature T ["C)
Fig. 9 51
Change in temperature and concentration at multi -stage reactorl 4l
reaction rate. An example of a methanol synthesis reactor in which

temperature and concentration in each stage vary is shown in Fig.
9.51.14)
In this process, reactor feed gas having methanol concentration*
Since feed gas includes recycled gas (see item (e)), a small amount of methanol
exists.
581
C0 enters the reactor with a temperature of T 11 where it undergoes

reaction in the first catalyst bed. The temperature of the gas subsequently rises as does the concentration, which becomes C1 After
mixing with cooling feed gas, however, both temperature and concentration fall, which marks the beginning of another reaction in the
next catalyst bed. This process is repeated until the end of the reactor
where the concentration at the outlet is c4.
In general, the reaction speed in a reversible exothermic catalyst
reaction under constant pressure (amount of generated product per
unit time and catalyst unit volume) will at first increase along with the
rise in temperature. In time, however, the rate of increase decreases
so that reaction speed eventually reaches a peak, after which it then
begins to decrease. In addition, if the concentration of the generated
product is high, reverse reactions will come to predominate and reaction speed will decrease. This tendency is illustrated in Fig. 9.51 by
the equi-reaction-velocity curves (broken lines). If the temperature of
each catalyst bed is set so that reaction speed approximates the maximum reaction velocity curve connected with the points of maximum
reaction velocity as close as possible, the process can approach maximized reaction rate. Reaction operations in this figure nearly satisfy
this condition.
(2) Control system
In order to achieve stable optimal operations in
the reactor of Fig. 9.50, the control system will set each temperature
controller (TC) to the corresponding temperatures shown in Fig. 9.51.
In addition, in order to suppress as much as possible transitions in reaction temperature and concentration patterns due to system changes
such as the amount of feed gas, the ratio between feed gas flow and
cooling gas flow is kept constant by manipulating cooling gas temperature. Furthermore, reaction pressure is kept at a specified optimal value by pressure controller (PC). (Refer to item (e)).
(b) Multi-tubular reactor control
In multi-tubular heat-exchanger type reactors
as shown in Fig. 9.52, each tube is filled with catalyst and heat generated in a reaction tube is removed through a heat-transfer medium, by
which reaction temperature can be controlled. Such reactors are often
used for reactions characterized by large reaction heat and narrow optimum reaction temperature range, and by catalytic time which must be
adjusted as needed. In regards to the tubes employed here, temperature distribution in the radial direction of each tubes must be kept uniform, and consequently small apertures are usually selected. For temperature distribution in the axial direction, however, a peak appears
about halfway down the tube, as shown in the same figure. This distribution is affected by reactant gas flow, composition and temperature,
582

Feed water
CD Temperature
oncentration
onsecutive reaction
@ ingle rever ible reaction
<D
Product gas
Fig. 9 52
Distribution of axial temperature and concentration

in catalytic tube of multitubular reactor
and if catalytic activity deteriorates with time, the maximum temperature value will be less, flattening out the curve. As a countermeasure
here, temperature of the heat-transfer medium can be raised (by raising the pressure controller (PC) setpoint value), thus raising the temperature level of all reactor tubes and maintaining reaction speed.
Also of interest here is the concentration distribution of product
components within a pipe, which varies according to the type of reaction. For example, for a single reversible reaction, concentration will
increase smoothly as shown by the dashed curve in Fig. 9.52. For a consecutive reaction as shown in Fig. 9.53, however, the concentration of
object product B will peak out as the chemical reaction which changes
B to C begins to advance. In the former type of reaction, temperature
will be controlled so as to maximize yield per unit time of the object
product. In the latter type of reaction, on the other hand, both yield
and selectivity (the percent of reactant A converted into object product
B) must be raised and temperature controlled appropriately with these
-----------8
.~
~
;:
!lc:
u
Inlet
Cata lytic tube length
Outlet
Fig. 9 53 Cha nge in concentration of reactant and products

at consecutive reaction
583
two requirements in mind. In general, if yield is improved, the selectivity tends to become lower.
(2) Temperature control
The number of catalyst tubes in a
multi-tubular reactor is proportional to the size of production and is
thus generally quite high (several thousand in large-scale reactors),
therefore, it is very difficult to make all of the tubes in uniform state.
In addition, since temperature distribution in tube's axial direction varies due to a variety of causes as described above, temperature within a
catalyst tube usually cannot be controlled directly. Instead, indirect
methods for keeping the temperature of the heat-transfer medium constant is adopted. In particular, in addition to the example of Fig. 9.52
in which natural circulating heated water is used as the heat-transfer
medium, a forced-circulation process is also possible. Moreover, the
process shown in Fig. 9.54 is often adopted, in which heat-transfer mediums having high- boiling points such as oil, dowtherm and molten
salt are forcibly circulated to remove generated heat by the sensible
heat technique. In any of the above processes, however, it is important
to suppress reaction runaway. This can be achieved, for example, by
making circulation flow as large as possible and the temperature differential between inlet and outlet small, by employing a heat-transfer medium with a large overall heat capacity, or by making the temperature
differential between the heat-transfer medium and catalyst as small as
possible. Each of these countermeasures improves the degree of temperature self regulation in the face of disturbance, thus contributing to
reaction stability.
Finally, because of the danger of reaction runaway or explosion associated with stopping circulation flow, reactant gas must be immediately shut-off according to a lower limit of circulation flow or an upReactant gas
_..,.. ____
,
'
I
I
'
li:UIE!rn\Kr!l>t
'
1'
Cooling water
Fig. 9 54 Temperature control and safty control of multi-tubular reactor
584
per limit of maximum catalyst temperature in some representative

tubes.
(c) Fluidized-bed reactor control
(1) Process and temperature control
In the vertical-cylindrical type
of reactor shown in Fig. 9.55, a granular or powdery type of solid catalyst is employed with reactant gas injected from the bottom, and the
catalyst is made to react while a fluidization state is maintained. Since
the catalyst section of the reactor undergoes severe agitation here,
there is hardly any temperature distribution in the radial or axial directions, which makes it possible to maintain a uniform reaction temperature across the entire catalyst. Moreover, reaction heat can be removed by incorporating multiple cooling tubes within the catalyst fluidized bed. Here, the amount of heat so removed can be sensitively
changed by raising or lowering the temperature of the cooling fluid
because of a large heat-transfer coefficient.* Referring again to the
figure, pressurized water is passed through the cooling pipes and heat
is removed via steam generated from reaction heat; reaction (catalyst)
temperature can thus be controlled by manipulating the setpoint value
for saturated steam pressure (corresponding to heated-water temperature) according to temperature controller output.
(2) Catalyst fluidized-bed adjustment
One special feature of this type
of reactor is that catalyst material may be added or removed as needed
along with gas flow during actual operation. Consequently, catalyst filling, catalyst activity and height or condition of the fluidized bed can
be adjusted so as to maintain optimal conditions for reaction. As an aid
for achieving this, the following relations can be used which include
the measure values AP1 and AP2 shown in Fig. 9.55:
hz= L1Pzh
L1Pl 1
Here, We is the amount of catalyst filling, SA is the cross-sectional

area of the fluidized bed, and p is the density of the fluidized bed.
In addition, since with experience we can. guess the abnormal state
of the fluidized bed (such as gas channeling or slagging) from the behavior of the AP recording pens, such information is effective for stable operation.
(d) Preparation of reactant-gas mixture
(1) Mixed reactant gas
Reactant gas in gas-phase catalytic reactors
as described above is, in most cases, prepared by mixing several components. The mixing ratio, although set to an optimal value for reactor
operations, can present problems if varied over a wide range. In partie-
* By
direct contact between solid catalyst and heat-transfer tube.
585
----....._..
Raw gas
Fig. 9 55 Control of flu idized-bed reactor
<D
H I Incr ease
H I Decrease
2H ,
Decrease
@ 0 2 lncrea e
Oxygen
I 00 % 1L.J.'-'----'--:~-'---'--'--'-........,'---'-~
Fig. 9 56
c,H.
omposition of mixed gas a nd fl ammable range
ular, dangerous situations such as catalyst deterioration , overheating

and explosions can occur. Accordingly, gas mixing ratio should be reliably controlled during process operations.
The triangular coordinate system shown in Fig. 9.56 can be used
to represent the respective concentrations in a three-component gas.
The example shows the point A indicating the composition of feed gas
required for the following reaction in which ethylene dichloride is formed from ethylene, hydrochloric acid gas and oxygen. Also shown is
the associated flammable zone.
586
Chap .9
Concentrations are read from this system as follows. Taking hydrochloric acid as an example, we draw a line through point A parallel
to the axis opposite the triangle vertex corresponding to hydrochloric
acid 100%. The intersection of this line with the HCl scale gives us
57%, as shown. Concentrations of the other components can be determined in the same manner.
A control system
(2) Mixture preparation control for a reactant gas 15 >
in Fig. 9.57.
shown
is
9.56)
(Fig.
gas
mixed
above
for preparing the
setcontroller
changing
by
Here, load adjustment is first performed
flow
the
however,
this,
tings for hydrochloric acid gas flow. Following
rates of oxygen and ethylene are set so as to obtain an optimal gas ratio using the output from respective hydrochloric- acid ratio units. Although change of ratio settings are actually performed by an operator,
a system can be adopted in which settings which mistakenly move the
system in a dangerous direction can be refused and an alarm sounded.
Reliability of the system can be improved as follows. As show in Fig.
9.56, as increase in the flow rate of oxygen moves mixed- gas composition towards the oxygen 100% point, the dangerous flammable zone is
entered. Thus, in order to prevent a sudden increase in oxygen flow
due to an erroneous drop in output from a faulty flow detector, a
redundant detector is incorporated so that the highest value from the
two is used as controller input. In addition, if oxygen flow should at-
Eihylene
<D
Ratio unit I
Ratio unit 2
Ratio unit 3
Low selector I
----{:lf:I----+--+-
<V
- To reactor 1
I
Low selector 2
Low elector 3
High selector
+ a Bia
I
I
I
I
~ Load
Hydrochloric
acid
_ ,sv
:~~~=~~-J
~--- -- -- --- --
Fig. 9 57
..
lt l ltl
-- --- - ---- ~ -----J
Milling and afety control system of raw gases
587
tain its set point value +a due to a defect in the flow controller FC3 of
the main control loop, auxiliary flow controller FC4 will take over manipulation of the pressure controller PC1 control valve in order to maintain this value. Moreover, if the system experiences a sudden jump in
load from a low-load situation so that the FC3 control valve quickly
opens, the PC1 setpoint value will be lowered immediately in the load
change and prevent a momentary massive flow of oxygen.
We next examine the case in which ethylene flow rate is decreased. Here, if the composition of the mixed gas moves away from
the ethylene 100% point, the danger zone is entered (Fig. 9. 56). As in
the case of oxygen above, a redundant detector is employed to prevent
a drop in ethylene flow due to an erroneous rise in output from a
faulty flow detector. In addition to this countermeasure, oxygen flow
is set using the lowest value output from ratio units 1 and 3. This is
done to prevent the oxygen flow setting based on output from ratio
unit 1 from exceeding its allowable upper limit in regards to ethylene
flow rate.
Note here that increase or decrease* in hydrochloric acid flow presents little danger, as shown in Fig. 9.56.
Although the control system of Fig. 9.57 provides sufficient protection of the reactor, incorporation of emergency shut down equipment
for each feed gas will assure reactor safety. In this case, it should be
clear from the above discussion that in such an interlock circuit, oxygen must be the first to be cut-off, followed by ethylene and then hydrochloric acid provided that confirmation of oxygen cut-off is received. In this regard, although duplicate shutoff valves for oxygen
may be the first to be considered as a means for implementing redundancy, the designer must be careful of "common" faults (e.g., drive
power failure) which can just as easily degrade system reliability.
(e) Control of a recycle reaction system
(1) Materials balance
Since in many reaction processes, pas&ing materials through a reactor only once does not totally complete a reaction,
a recycling type of process has come to be often used. Here, unreacted
gas remaining after separation of produced materials is mixed at a
proper concentration with reactant gas and returned to the reactor. A
problem here, however, is the presence of inert gases in the reactant
gas which do not contribute to the reaction, or inert gases which may
be generated by a side reaction. In either case, inert gases will tend to
accumulate due to recycle operations resulting in a drop in react gas
partial pressure and a slower reaction. As a countermeasure to this,
some of the recycle gas can be purged so as to keep the concentration
However, a drastic decrease in hydrochloric acid is dangerous.
588
CD--:i
Raw gas
, _ _ .._..._,__:.."'---{
l:t--.....
I
~------------------ --------~
C0, + 3H, - CH,OH + H,O
Component
F+ R,
F
co,
0.665
\ .000
(Unit mol)
s,
p,
Rt
0.400
0.065
0.330
H,
1.993
3.000
1.200
0.194
1.010
lnen gas
Jl.Q1l..
0.082
0.082
Jl.Q1l..
0.070
CH, OH
0.600
H,O
0.600
Total
t' Iow. rate
rauo
2.670
4.082
2.882
0.271
1.410
0.101
0.528
Fig. 9 58 Material balance and control in recycle reaction system
of inert gases within a fixed limit.

Figure 9.58 shows the flow of materials in each section of a recycle system in which methanol is formed from reactant gas composed
mainly of hydrogen and carbon dioxide (mole ratio 3) and raw gas contains inert gases (0.012 mole). In this system, 60 % of the feed gas
reacts after one pass through the reactor, and since a certain amount
of inert gases brought into the system is now expelled as a component
of purged gas, concentration of inert gases in the system is held constant.
Given that the composition of reactor feed gas
(2) Control system
has been determined, the reaction ratio (conversion ratio) for one pass
through the reactor can be controlled in a stable fashion by the method
described in item (a) above. The problem of interest here is to maintain
the composition ratios shown in Fig. 9.58 even during a change in raw
gas flow by controlling the flow of circulation gas and purge gas. In
particular, for a high- pressure reaction system having a large gas hoi9.5 Control of Reaction Processes
589
dup, controlling each flow rate by feedforward control based on the ratios listed can be quite effective against fluctuations in raw gas flow.
In regards to purge flow, feedback correction can be employed using
output from a concentration controller in order to keep gas composition in the system fixed. Alternatively, since an accumulation of inert
gases brings about a rise in system pressure, a pressure controller can
be used instead of a concentration controller with the same objective in
mind.
9.6 Other Process Control
This section describes typical examples of three types of processes, refrigeration, evaporating and drying, which are frequently
used in a variety of applications.
9.6.1 Control of refrigeration equipment
Examples of equipment used
(a) Control of refrigerant compression
for cooling process fluid and two control configurations for keeping
the fluid at a constant temperature are shown in Fig. 9.59. In configuration (a), refrigerant evaporation pressure is held constant by controlling the amount of compressor suction and as a result evaporation
temperature becomes constant due to a fixed functional relationship
with pressure. Consequently, by setting the level controller (LC) via
cascaded output from the temperature controller (TC), the area for
heat transfer by evaporation can be increased or decreased as needed
and process fluid temperature controlled. This technique can accommo-
Fig. 9 59 Control of refrigerant compression cycle
590
date a wide change in heat load (process flow, temperature) and has relatively good control response. Form the standpoint of energy efficiency, however, configuration (b) is definitely superior.
In configuration (b), the evaporator level is always kept at its highest position in order to make use of the maximum evaporation
heat-transfer area. Here, process outlet temperature is controlled by
setting evaporation pressure controller (PC) accordingly using cascaded output from temperature controller (TC).
Comparing the above two configurations, if we assume equal heat
loads, configuration (b) is more suitable for holding higher evaporation
temperature (pressure) than (a). Accordingly, the discharge pressure/suction pressure ratio of the compressor is always smaller in configuration (b), which means less power required by the compressor.
However, the larger the amount of refrigerant stored in the evaporator, the worse control response becomes.
(b) Feedforward control in refrigeration equipment
In configuration (b) above, if a more accurate control system is required with the object, for example, or preventing process fluid
freeze-up, feedforward control can be used.
Symbols for designating process variables are
those shown in Fig. 9.60. Since the heat removed from the process
fluid is equivalent to that supplied to the refrigerant via the evaporation heat-transfer surface, the following approximation holds:
~------'1:~- -- -EJ-- .?...:. ___
\ ~-~
:F. F.
;F.B.T'
Fig. 9 60
L- -- ~
VA
J...
r'\!:;T----- To compressor
rR-Ie-.-fr-ig-er-an-1 -e-va_po....,r~a:~mpressor
dr iving actuator
suction site
~-T~'~+-~7
Feedforward control in changed -pressure /Constant-level type

refrigerating process
591
Fc(T2- Tl)::::: UA (T~- Tv)~(T2- Tv)

Tv:::::t<T1+ T2)-
~5t (T2- T1)
(9.35)
Here, c is the process fluid specific heat and UA is the product (overall
heat-transfer coefficient) x (area) for the evaporation surface.
In addition, the fixed relationship P v = <p(Tv) generally holds between evaporation temperature and pressure.
(2) Feedforward control system
Referring to Fig. 9.60, given values
UA and c and using input values Th T 2 and F, value Tv can be estimated
using relationship (9.35). Consequently, a value Pv corresponding to
this Tv can then be provided as a setpoint value to the pressure controller (PC). In addition, the gain along the feedforward path requires feedback correction from the output of TC, the process outlet temperature
controller. Note also that the dynamic compensation element L/L consists of a delay element for synchronizing variation in F with the time
delay associated with change in outlet temperature, and a lead element
for negating the time delay from P v manipulation until its effect on
outlet temperature. In general, since variation of inlet temperature
with time is similar to flow rate in that it does not change rapidly, dynamic compensation for T 2 can be omitted.
9.6.2 Evaporator control
(a) Process description
Figure 9.6l(a) shows the most basic configuration of an evaporator
in which feed liquid, which includes non-volatile components, is heated
to separate volatile components. In order to cut down on the amount of
heating steam, however, the multiple-effect evaporator in Fig. 9.6l(b)
reuses vaporized steam from the previous stage. The kind of configuration is commonly used in large scale operations.
Using the symbols shown in the figure, the following relationship
holds for the process variables:
(9.36)
In the case of the multiple-effect evaporator, this becomes Fv
= FVI + Fv 2 + Fv 3 In addition, if we designate the non-volatile concentration of the feed liquid and concentrated liquid as C, and Co, respectively, we obtain from the balance of materials:
(9.37)
The following approximation holds due to the fact that the amount of
heat supplied by heating steam is nearly equal to the amount of evapo592
r-
i-------------------- ---------------- --,
@-
-*--~-Q) Fv
Feed liq uid

Concentrated liquid
( a ) Single-effect evaporator
'
:
'*"---t--@ Fv, :
:
<D
r PC l
Q) Va porized steam
@ Steam
Drain
To vacuum system
( b ) Multiple-efTect evapora tor ( Forward feed type)
Fig. 9 61 Control system in case of changes in feed flow rate
rated heat accompanying separation of the solvent from the feed liquid:*
Fv::::::K.G
(9.38)
Here, Ks is a constant dependent on the process; it becomes large as

the number of stages in the multiple-effect evaporator is increased.
If we now eliminate Fo and Fv from the above three equations, we
obtain:
c~ (Co-C)F;
~
KsCo
(9.39)
Or by eliminating F; and Fv:

(9.40)
From the above discussion it should be clear that these equations represent general relationships regardless of the type of evaporator. The relationships serve as a basis for configuring an evaporator control system, as described next.
* Feed
liquid is normally preheated close to the internal temperature.
593
(b) Control system

The object of evaporator control is to maintain the concentration
of non- volatile components in the concentrated liquid constant. During
this time, the process receives disturbance in the form of feed liquid
flow variation (product amount variation) and change in the concentration of non-volatile components. A basic control system for such a situation is shown in Fig. 9.61(a), while the feedforward control system
for the multiple-effect evaporator in Fig. 9.61(b) is used to improve
controllability. In the latter case, since the time delay required for a
change in feed liquid flow to affect concentrated liquid concentration is
relatively small compared to that for heating steam to effect corresponding concentration (in particular if retention time of each vessel is
short), a time lead element incorporated in the L/L dynamic compensator is used.
The multiple-effect evaporator in Fig. 9.6l(b) is also characterized
by parallel flow, i.e., steam and liquid flow in the same direction (parallel-flow type). As a result, however, vessel pressure as well as temperature is lower than that of the previous vessel. This low temperature,
in turn, results in increased viscosity of the concentrated liquid (e.g.,
caustic soda aqueous solution). As a countermeasure to this, the counter-flow multiple-effect evaporator shown in Fig. 9.62 can be used to
keep the temperature of the concentrated liquid high. Although control system configuration in this case is essentially the same as that
for Fig. 9.61, since time delay relationships are opposite that of parallel flow, dynamic compensation for such time delays will be different.
Another type of system to consider is that in which the amount of
available heating steam varies, which necessitates a corresponding
change in feed liquid flow. Such a control system is shown in Fig. 9.63.
Although a single-effect evaporator is shown here, a multiple-effect
evaporation (parallel flow) can be applied just as well. The feedforward
r-----,
'
Fig. 9 62
594
~,
~--~ - - To
vacuum system
Counter -now type mut iple -efTect evaporator
_ Vaporized
r-fPC\-}.
1 "'C;r
f v. team
r --------- llt ----
- To vacuum system
----:--- EJ} ----- i --------~oo
r-c: ~c
Co - CI
r-
I
I
~ ---
"'mc-J~--r-- ~~~~ ~i1

'""'
Hqld~
o.,;,
c, L--- - - - ----J
:C
~Coo~"'""'
llqld
Fig. 9 63 Control system u ed with va ria ble team fl ow rate
mechanism here, which is based on relationship (9.40), becomes even

more effective in a multiple-effect configuration. Note also that the
L/ L dynamic compensator employed here acts as a time delay element.
In each of the above control systems, the means of measuring
non-volatile concentration may employ a variety of forms. For example, a density meter may be used in which the measured value corresponds to a concentration value, or by making use of the relationship
between boiling point and non- volatile concentration, liquid temperature can be measured to give the concentration. However, in the latter
case, since boiling point itself will vary depending on vapor pressure,
and automatic pressure compensator becomes necessary which converts the boiling point to that under standard pressure. In addition, an
alternative method which has come to be widely used is to measure the
temperature differential between evaporation vapor and liquid to obtain a corresponding concentration value.
9.6.3 Drying process control
(a) Control of a batch fluidized-bed dryer
Although there are many types of batch dryer configurations, the
fluidized-bed type has come to be adopted in a wide range of fields.
Since this type is also demonstrative of drying processes in general, a
typical example will be described here.
In the equipment shown in Fig. 9.64, a granular
(1) Drying process
or powdery substance for drying (referred to as material below) is
placed in a vessel having a perforated bottom tray and is dried by having heated air pass through it for a certain length of time. During this
time, the particle water content (H 20 kg/non-water material kg) decreases in the manner shown in Fig. 9.65(b). In addition, the particle
water content rate of change (drying speed) shown in Fig. 9.65(c) bee9 .6 Other Process Control
595
Dew point
,..----- -,
1-::::::::::t----J~S,team:
___ ...
I
I
-------------------------Fig. 9 64
''
-- ~
ontrol of batch fluidized-bed dryer
t
A
Time
A~
( a) Outlet air temperature
j ~
CD
Critical water content
"'Wbd'm .,., oooo<
Time 1
( b) Particle water content
AfiB \C1i "'
3 [_
w Material
CD
pre-heatmg pen
.od
Constant -ra te dryin.g perio_d
@ Decreastng-rate drymg perood
"" 1""
D E
0
Fig. 9 65
Time I
( c ) Rates of drying
Batch fluidized -bed drying process
omes constant shortly after material pre-heating (constant-rate drying

period). The reason for this is as follows. As water comes to cover the
surface of the material, material temperature reaches equilibrium with
wet-bulb temperature T w. since heat transferred from the passing air
is used completely for water evaporation. This gives rise to constant
596
evaporation, and as a consequence, air outlet temperature To during

this period is constant but slightly higher than T w. as shown in Fig.
9.65(a).
As the drying process advances, moisture transfer from the interior of the material falls behind evaporation so that some of the material
surface dries, raising its temperature. Accompanying this is a small
temperature differential arising between the hot air and material
which decreases the amount of heat transfer and lowers the drying
speed (decreasing-rate drying period). As a result, air temperature To
at the dryer outlet gradually rises at this time and approaches inlet air
temperature Tt (at completion of drying, outlet air temperature
should, in principal, be equal to inlet air temperature). The material
water content w is given by the following equation in the above situation:2l
(9.41)
Here, k is a constant for a particular type of material and amount and
constant air flow.
Given a process in which the value k in Eq.
(9.41) above is obtained from actual results, if we now designate the desired final water content of the material as w* and the corresponding
outlet air temperature as n, the following relationship is obtained
from this same equation:
(2) Control system 2l
1 (constant value)
e wfk-K
( 9 42)
Thus, since T6 can be determined from measured values T 1 and T w.

the current value of To can be checked to see whether or not the de-
~ Js t---=-t---t---1---::
<.!
301---"'9------1----1-;
25L-~L-~L-~7~0-~-~-~~
Fig. 9 66
Tt (t: )
Relationship between dry-bulb temperature and
wet-bulb temperature of heated air
9 .6 Other Process Control
597
sired material water content has been reached. However, as direct

measurement of hot air wet-bulb temperature is technically difficult,
the curves to get T w can be drawn up instead based on easy-to-measure inlet air temperature Tt and atmospheric dew point Td, using general humidity curves. From the relationships obtained, T w can then be
surmised, as shown by the example in Fig. 9.66. In addition, these relationships can be approximated by the following expression:
(9.43)
where Tw indicates the estimated value of T w
A control system based on the above ideas is shown in Fig. 9.64.
Specifically, air temperature T 1 is kept constant, and from this value
and atmospheric dew point Td, the value of T w is determined from the
above expression. In addition, using Eq. (9.42), desired value T& corresponding to w* is calculated. If the actual measured value of To should
reach this value, an alarm is sounded announcing the end of drying
operations.
(b) Control of a continuous fluidized-bed dryer
Although there are also many configurations for continuous
dryers, we select the fluidized-bed type as an example for the same
reason stated above for batch systems.
Referring to Fig. 9.67, a granular or powdery
material having water content wf is supplied to the dryer with a mass
Air
Steam
Fig. 9 67 Con tinuou Ouidizedbed dryer and its control system
598
flow (non-water based) of W, and drying is performed through contact

with hot air supplied to the fluidized bed with a mass flow (dry based)
of FA. Since the mass of material in the vessel here is kept constant
through control of the fluidized- bed pressure differential (see Subsec.
9.5.2(c)), material whose water content decreases due to passage
through the vessel is expelled to the outside at a flow rate equal to
feed flow.
In this dryer, the heat required for water evaporation is completely supplied via heat transfer from hot air (adiabatic type).
Since mixing in the fluidized bed is completely closed in, water
content w of material within the vessel is considered to be equal to water content Wp of expelled material. In addition, provided that material
quality is invariable and due to the fact that hot air flow FA is nearly
kept constant, the following equation similar to Eq. (9.41) above can be
written:
(9.44)
(2) Control system16 >
The object of control here is to maintain product water content Wp ( = w) constant in the face of variation in feed values (Wand wf) or change in atmospheric moisture content (Td). A problem here, however, is that from a control point of view, Wp is technically difficult to measure directly. However, if control is performed based
on the following equation corresponding to Eqs. (9.42) and (9.44), objectives can be achieved without direct measurement.
T,
Dryer process
Fig. 9 68 Block diagram of continuous fluidized - bed dryer control system
599
(9.45)
Here, T wz can be determined by Eq. (9.43).
A control system satisfying the above conditions is configured as
shown in Fig. 9.67. Here, desired value Wp* for product water content
is input into the system as the value K. Also of importance here is that
the path of f(t) in the figure makes up a positive feedback circuit, (see
also Fig. 9.68). However, since it exists in parallel with another negative feedback circuit in the drying process, if the gain of the former
becomes larger than that of the latter, the control system will become
unstable. To stabilize this situation, an appropriate time-delay element
j(t) can be incorporated to bring the gain of the former lower than that
of the latter at the operating frequency.
PRACTICE QUESTIONS
9. 1 A centrifugal pump used for plant cooling water is operated in a piping system
for which the ratio of dynamic pressure loss to total pressure loss at times of 100%
load is 75%. If load percentage changes during the year depending on the season as
shown below, determine the rate of energy savings when employing pump speed
control.
Load Percentage
Period
Load Percentage
Period
July-August
December-February
100%
50%
September-November
March-June
85%
85%
9. 2 Explain the reason for adopting a three-way control valve in bypass control of
steam heater.
9 . 3 Some steam heaters may adjust the heating rate by a drain valve. Determine the
response time-constant for the change in heating rate due to drain flow manipula
tion. Assume the following conditions in the heater:
2.5kg/cm', 137oC
Working steam
Condensation latent heat
522 kcal/kg
Total condensation heat-transfer surface
100m'
488 kcal/m 2 hoC
Overall heat-transfer coefficient
433 kg
Total heat-transfer area submersion drain amount
74C
Process-fluid average temperature
9. 4 In a binary-component distillation column in which the feed flow is constant, but
feed composition changes, illustrate a feedforward control system for keeping the
concentration of bottom liquid constant under conditions of constant reflux flow.
9 . 5 Discuss the performance required for the temperature control system used in a
stirred tank type batch polymerization reactor.
9 . 6 Given a gas-phase fluidized-catalyst reactor having a cylindrical shape of diameter 3 m, total pressure differential over the catalyst bed during operation of 1000
mm, and pressure differential at the pressure taps installed at a height of 2.5 m
600
Chap.9
within the catalyst bed of 450 mm, calculate the total weight of catalyst filling the
reactor and the height of the fluidized catalyst bed.
9. 7 Given the case of compressor-refrigeration equipment using ammonia as a refrigerant in which process fluid is cooled from -3C to -7C; the level of ammonia
liquid within the evaporator is about halfway, and its temperature when the effective evaporation heat-transfer area is half is -10C. If the liquid level is now raised
and the heat-transfer. surface is completely submersed, what will the refrigerant
temperature value become? Assume heat load is constant.
9. 8 A triple-effect evaporator may be used to concentrate the sugar content of an
aqueous solution from 25 wt% to 85 wt%. In this case, since the flow of heating
steam will vary, it is desirable to implement feedforward concentration control to
increase or decrease the flow of concentrated liquid accordingly. What would be a
satisfactory value for the ration between the two flows? In addition, in regards to
dynamic here, consider whether lead or lag elements should be incorporated. Assume in this process that for every 1 kg of heating steam, 2. 7 kg of water is evaporated.
9. 9 In an adiabatic continuous fluidized-bed dryer using heated air as a heating
source, name the major disturbances involved with trying to keep the water in the
product constqnt.
9. 1 The rate of energy savings for each period can be determined from Fig. 9.3(b)
(/= 0. 75) and averaged as follows:
M =(0.62)(3/12) + (0.25)(7/12) + (0)(2/12)::::::0.330(%)

9. 2 Please see Subsec. 9.3.1(d).
9. 3 Substituting the various values into Eq. (9.10), we get T:::::O. 75 h = 4 min.
9 . 4 Using Z instead of Fin Fig. 9.28(b) results in a feedforward control system. Moreover, from Eqs. (9.24), (9.25), and (9.26), the relation B = q:>(Z) is now used in place
of g(F).
9. 5 Please see Subsec. 9.5.1(c)(2).
9. 6 From the equations in Subsec. 9.5.2(c)(2), We = 7069 kg, h 2 = 5.56 m.
9. 7 Substituting the first condition (heat-transfer area 0.5 A) into Eq. (9.35), we obtain Fc/UA = 5/8. Using this value, Tv = -7SC.
9. 8 From Eq. (9.38), Ks = 2. 7, and using this value and the concentration values in
Eq. (9.40), F0 /G = 1.125; delay element. (see Subsec. 9.6.2(b) regarding dynamic
compensation.)
9. 9 Feed flow and water content, and atmospheric humidity (dew point). (See Subsec.
9.6.3(b).)
Answers to Questions
601
REFERENCES
1) T. Takamatsu : "Roles of operator in chemical plants," Kagaku Kogaku, 49, 11

(1985) 845-848 (in Japanese).
2) F.G. Shinskey : Energy Conservation through Control, Academic Press (1978).
3) F.G. Shinskey : Process-Control Systems, 2nd Edition, McGraw-Hill (1979).
4) B.G. Liptak : Instrument Engineers' Handbook, vol, II, Chilton Book (1970).
5) Ichioka, et a/. : " Application status of microcomputer on combustion control," The
Energy Conservation, 33, 5 (1981) 6l.(in Japanese).
6) F.G. Shinskey : "The material-balance concept in distillation control," The Oil and
Gas Journal Quly 14, 1969).
7) T.G. Chin : "Guide to distillation pressure control methods," Hydrocarbon Processing
(Oct. 1979).
8) F.G. Shinskey : Distillation Control-For Productivity and Energy Conservation,
McGraw-Hill (1977).
9) P.S. Buckley, W.L. Luyben, and J.P. Shunta :Design of Distillation Column Control Systems, ISA (1985).
10) G.F. Fauth and F.G. Shinskey : "Advanced control of distillation column," Chern.
Eng. Prog. Qune 1975).
11) F.G. Shinskey : Controlling Multi-Variable Process, ISA (1981).
12) G.C. Ardell and B. Gumowski : "Model prediction for reactor control," Chern. Eng.
Prog. Oune 1983).
13) A. Mori : Transition Operation for Preparation of Polyolefin Having Different
Melt Index, Lade Open Gazette 142007 (1980). (in Japanese).
14) Osaki, et al. : "Development of new type reactor as well as energy conservation and
scale-up on methanol plant," Plant and Process, 9 (1984) (in Japanese).
15) J.T. Illidge and J, Wolstenholme : "Hazard of oxyhydrochlorination," Loss Preven
tion, 12 (1979).
16) F.G. Shinskey : "How to control product dryness," Instrumentation Tech. (Sept.
1968).
602
Chap.9
Chapter 10
INSTRUMENTATION TO MANUFACTURING
INDUSTRIES
(Application II)
As was described in Chap. 9, the various kinds of processes used

in manufacturing industries are generally comprised of processing units. Different industries, however, employ different operational methods. Thus, the instrumentation and control methods will also vary reflecting those differences.
Our thinking with respect to instrumentation has undergone much
change in recent years; the concern used to be operational supervision
and automatic control for the sole purpose of mass production. Today,
however, such considerations as safety, quality, labor savings, and energy conservation are at least equally important in the development of
instrumentation. Another significant development that has become increasingly noticeable with the introduction of digital devices and systems, is the rapid evolution toward so-called total factory automation
(FA) with the capability for small production runs of a diverse range
of products.
How extensively is instrumentation employed today? While this is
not easy to determine precisely, the Japan Electric Measuring Instruments Manufacturers' Association (JEMIMA) has come up with an
instrumentation ratio, which they define as follows:
t rumen t a t"10n ra t"10 (IR)
ms
investment
in instrumentation
.
.
mvestment m plant and egmpment
= .
Investment in instrumentation includes investment in industrial instruments, process computer systems, and instrumentation work used in
productive processes (analyzers used for experimental purposes in laboratories, for example, are excluded). Investment in plant and equipment
includes all costs associated with equipment and facilities, but not including land.
Although the instrumentation ratio is only calculated triennially,
it is nevertheless apparent that the ratio is increasing every year. In
Chap.lO
Instrumentation to Manufacturing Industries
603
I I I
11
~
0
i<'
10
~
c
Actual result r-HPrljec!ed
Manufacturing
industries only
.g
~ 8
"E
.~
"'-...
~z
,
,r-.V
''
'
I ~I
,'f.
-"--..
\
Sum
manufacturing
and non-manufacturing
industries
'
1975
I
I
1980
Year
1985
Fig. 10 1 Progress in instrumentation ratio
fact, as is shown in Fig. 10.1, the ratio jumped sharply from 6 to 10%
between 1976 and 1983.
This chapter is intended to provide a broad overview of how instrumentation is applied to such diverse but representative industries as
oil, steel, and electric power given their very different and unique operational and control requirements. With the limited number of pages we
have available, this will be no more than a cursory introduction. The
authors also acknowledge that they are affiliated with corporations
that manufacture the types of instruments described. While a certain
amount of bias may be reflected in the presentation, it is nevertheless
hoped that the information presented in this chapter will serve as a useful reference.
10.1
Instrumentation Applications in the Petroleum Industry 1l
-el
10.1.1 The petroleum industry and instrumentation

Among process industries, the petroleum industry serves to illustrate one of the most advanced in terms of process automation.
Although the existence of oil has been known since before Greek
times, active exploration for underground deposits and widespread use
didn't begin until the latter half of the nineteenth century. In the present century, control over oil resources was an important concern in the
two world wars, and of course today the importance of oil is universal604
Chap.lO Instrumention to Manufacturing Industries
ly recognized. In the post-war years, the petroleum industry has continued to make enormous strides. It is no exaggeration to say that the
rapid economic recovery of Japan and the West after the war, and the
evolution of society as we know it today was made possible by the
ready availability and use of oil.
However, the knowledge that oil is a limited natural resource and
the abrupt price increases of the 1970s (the co-called oil shocks in 1973
and 1978) led to an industry-wide effort in Japan to conserve energy.
This effort was highly successful and, in fact, has contributed to the
huge oversupply of oil that prevails on the world market today. Diversification of energy sources has also made much progress. It is estimated that by the year 2000, the world's dependence on oil as an energy source will have fallen by about 30%.
From the standpoint of instrumentation, the petroleum industry is
unique in that it is a constant-load continuous process and is basically
fairly easy to control. As we leave the era of mass production behind,
however, and enter the age of diverse feedstocks and value-added products, it will be necessary to re-evaluate our approach to instrumentation.
In the earliest days when oil first started to be used, distillation of
crude oil was carried out in a single kiln. Successive separation of the
various distillates proceeded while the operator kept a close watch on
the temperature. Instrumentation of the time didn't go beyond simple
measurement.
By the turn of the century, the demand for oil had increased tremendously, which led to the development of more advanced distillation
units and continuous processing. Instrumentation systems were also improved so they could be applied to control as well as to measurement.
Control logic also saw much improvement at this time thus enabling
even more advanced process control devices. This in turn made it possible to increase the scale of processes to unprecedented proportions.
Today, equipment economy is not the only consideration. To cope with
the greater diversity of crude feedstocks and the more stringent requirements for product quality, instrumentation and equipment manufacturers today are working closely together to develop and improve
systems. It goes without saying that a major support for this development is the state of contemporary instrumentation technology. Looking ahead, the challenge will be to achieve even more precise levels of
control by optimizing process equipment through the use of microprocessors and process-application computers.
The petroleum industry includes many activities encompassing
everything from extraction to refining and blending. In fact, if we regard the industry in its widest sense, shipment of petrochemical prod10.1
Instrumentation Applications in the Petroleum Industry
605
ucts should also be included. We will organize this chapter according

to a major distinction made by the industry itself-the distinction between on site and off site. On site is used to designate equipment that is
actually involved in the distillation, separation, and polymerization of
petroleum products, whereas off site refers to processing and/or handling that is done away from the refining complex such as blending, shipping, and processing orders for petroleum products. Subsection 10.1.2
will consider the application of instrumentation to an on-site topping
unit. Subsection 10.1.3 will then discuss off-site instrumentation, especially as pertaining to a distribution control system for petroleum products.
10.1.2 Topping unit instrumentation

The processing of crude oil into various finished petroleum products is referred to as refining. The starting raw material is crude oil
which is subjected to various refining processes including separation,
refining, and reforming (also called cracking) to produce a range of value-added products : gasoline, kerosene, gas oil, and fuel oil. Typical oil
refining processes are shown as a block-flow diagram in Fig. 10.2. The
-;:
:I
Crude
oil
-[ 1 - - - - - t- --1
o.
_::::;~-;::=
(!. Straight
gas oil
L __
Low ulfur
fuel oil
Middle sulfur
fuel oil
Bunker fuel oil
Home fuel oil
Fig. 10 2 Typical petroleum refining process
606
Chap.lO
Instrumention to Manufacturing Industries
oi l
team
' - - - - --1 Main
tower
I -- - t
1----1
Residue oil
F uel oil Steam
Fig. 10 3
Flow diagram of topping unit
initial processing of crude oil is carried out in a topping unit, such as

shown in Fig. 10.3.
Crude feedstock is continuously supplied to the topping unit. It is
routed through a heat exchanger where it picks up heat from side refluxes and product streams, and is then sent to a -desalter. The desalted crude, after recovering heat from even higher temperature liquid, is
brought up to a constant temperature (300 to 350oC) in a crude oil heat10.1
607
er, and then charged into a topping unit. The topping unit functions to
separate the crude oil by fractional distillation into components of differing boiling ranges. Typically, the crude will be distillated into
naphtha, kerosene, gas oil, and topped residue. The lighter fractions in
the main tower are still mixed together, so they are sent to a stripper
for further separation.
In the past, the various topping unit parameters such as temperature, pressure, and level were separately controlled with discrete devices for each control loop. In recent years, however, this approach has
generally been replaced by distributed control systems (DCS) using
microprocessors. In the remainder of this section, with continued reference to Fig. 10.3, we will consider the process controls that are required regardless of whether a DCS or a non-DCS approach is adopted.
Finally, an example of advanced control using a DCS will be described.
A steam-driven turbine pump is generally used to move the crude
oil from the crude storage tank to the main distillation tower. A governor adjusts the output of the turbine in response to the load on the
pump. The operating principle of the governor is to adjust the flow of
steam with a throttle valve to control the speed of the pump.
The crude is pumped continuously to the desalter, where inorganic
salts are removed from the crude by chemical or electrical desalting.
The separated salts are then withdrawn from the desalting tank according to a tank level indicator control (LIC-1) called a displacer.
Since the displacement-type level gauge is described in detail elsewhere in this volume, here we will only note that this type of gauge
has many applications at petroleum plants; most commonly they are
used as fieldmount indicating controllers and transmitters.
Of particular importance in the desalter unit are temperature control to assure complete separation of water and oil and pressure control (PIC-1) to assure the light fractions in crude aren't vaporized. After desalting, the crude is heated in tubular heaters up to around
350C. The fuel for the heaters is regulated by temperature controllers
(TIC-1) that use thermocouples mounted at the outlets of the heaters
for the detecting elements. Generally a mixture of gas (methane, ethane) and fuel oil is used for the fuel; the gas facilitates combustion and
. the fuel oil controls the temperature.
Control of the combustion system is essentially the same as that
used for a boiler system. When fuel oil is used for the fuel, a displacement-type flowmeter is used. Fuel oil has to be atomized for optimal
combustion, which means that a differential pressure controller
(PdiC-3) is required to maintain constant pressure. Optimizing the air
input in the fuel is also important. Today, this is often done by measuring the oxygen content in the exhaust gas emission using a zirconia ox608
Chap.lO
ide analyzer (Al-l), and making any adjustment necessary with an airflow compensation controller. Finally, as a safety precaution, the combustion is monitored with a flame detector coupled to a mechanism for
shutting off the fuel supply if the flame goes out.
Crude oils are extremely varied in terms of the hydrocarbons they
are composed of. They can nevertheless be turned into a diverse range
of petroleum products suited to different applications as long as the
crude can be separated into blends of specific boiling ranges. Referring
again to Fig. 10.3, it can be seen how constant temperature is maintained at the top of the main tower by regulating the volume of side reflux
returned to the tower with a temperature controller (TIC-2). The lightest fractions in the crude are withdrawn from the top of the tower and
passed through a condenser. The non-condensable component of the
gas is sent to a gas recovery unit and the naphtha fraction is sent to a
stabilizer section by means of a level controller (LIC-2), where it is further separated to produce LPG (liquefied petroleum gas).
In the topping unit, the fractions are condensed in the order of
boiling ranges. In the middle section of the tower, a constant volume
of the fractional streams is withdrawn as regulated by each flow controller (FIC-4, 5, 6). In the side stripper, stripping steam controlled by
steam flow controllers (FIC-8, 9, 10) strips the low-boiling point fractions. To assure good resolution of the fractions, side reflux is returned to the column. At the same time heat is recovered to heat the
crude, flow control is maintained (FIC-1, 2, 3) based on constant load
conditions in order to achieve good fractional distillation of the various
product streams.
Low-volatility residue oil (topped residue) accumulates at the bottom of the topping unit, but the lighter components can be extracted
from this bottom residue by blowing steam through it. Then, in order
for the topped residue to be withdrawn and charged to the vacuum distillation unit and desulfurization unit, a control loop to eliminate any
external disturbance on the down streams must be implemented. The
conventional approach has been to use a nonlinear controller (dead
band) for this purpose; within a given range, equilibrium would be
maintained within the tower and the down streams would not be adversely affected by any external disturbance. Figure 10.3 shows a cascade control loop configuration linking a nonlinear controller (LIC-3)
and a flow controller (FIC-7) for eliminating external disturbance.
Now we will consider some of the special control features that are
available with modern digital control systems (DCSs).
(a) Crude oil feedpump control
Feedpump control has traditionally been maintained not by adjusting the pump rpm, which was held constant, but by controlling the
10.1
609
load flow of the crude with a pressure control valve at the outlet of
the pump. With this method, only the pump discharge pressure and
consumption of steam has to be increased to compensate for the pressure loss of the pressure control valve. An alternative scheme that
conserves energy is to leave the pressure control valve in the constant
opening alone and regulate output by controlling pump rpm.
Figure 10.4 shows a design that enables both direct pressure control and rpm control. When the pump is first started and the load characteristic of the turbine is still poor, the pressure control valve is directly controlled by a loop consisting of the discharge pressure controller (PIC) and pressure control valve. Then, when the pump gets up sufficient speed for rpm control, the system switches over to rpm control
through a steam flow control valve regulated by a cascade controller
linking discharge pressure controller and rpm controller. This figure
shows a loop for application to a continuous control system. In actual
implementations, however, to prevent any external disturbance from
being introduced by transfer lag on the control, it is combined with sequence and sampling control.
Fig. 10 4 Control diagram around crude oil feed pump
(b) Heat control of the topping unit

To bring about an enrichment of the vapor in the more volatile
components than could otherwise be accomplished with a single distillation, the typical distillation tower is designed to return reflux to the
tower at two to four points to countercurrently contact the rising vapors.
The heat load of the side reflux is the difference between the heat
quantity of the feedstock entering the process and the heat quantity of
the products leaving the process, and is used for preheating the incoming crude. In order to control the side reflux heat exchange, it is necessary to indirectly determine this exact heat quantity. Conventional analogue controllers are incapably of handling this type of task. With a
610
distributed control system (DCS), however, the heat quantity can be

easily calculated according a computational function, simulating the
reaction taking place in the tower. It is not surprising, therefore, that
DCSs are being increasingly applied to control panels for monitoring
and adjusting the heat volume of distillation tower reboilers and heat
exchangers.
Heat quantity control is discussed in detail in another contribution
to this volume. Here we will be concerned only with the safety function that is implicit in such control.
A heat quantity control loop for a distillation tower is shown in
Fig. 10.5. Here, TI represents temperature at the inlet and outlet of a
heat exchanger, LITI is temperature differential, and QIC is a calorie
controller. The difference in heat between the inlet and outlet of the
heat exchanger is transferred to the crude on the shell side of the exchanger, thus maintaining constant reflux load in the tower. In this
system, the heat quantity Q is given by
where H(T1 ) is the enthalpy at the inlet, H(T2 ) is the enthalpy at the
outlet, F is the flow rate, and p is the density.
Distillator
Fig. 10 5 Heat control diagram for side reflux of distillation tower
A calorie controller calculates the heat quantity based on signals

detected at three points, and outputs compensation signals adjusting
the setpoint value of the secondary reflux flow controller to control
the side reflux load.
Even with a system as uncomplicated as this, if any one of the detection points ceases to function normally, control capability is lost.
The DCS thus provides monitoring over the several detection points
10.1
611
and safety interlock circuits that can be switched to maintain viable

control in the event a detection point malfunctions. Figure 10.6 illustrates interlock for heat quantity control.
The interlock monitors input signal wires breaks (lOP) from two
temperature gauges. If a break occurs on one of the input wires, the
secondary side flow controller to reflux changes from cascade mode to
automatic mode.
One additional interlock is provided for use in the event of operational error. Referring again to Fig. 10.6, when the calorie controller is
set to manual mode, in this case too the secondary side flow controller
changes from cascade mode to automatic mode. To avoid any adverse
affect on the secondary side as a result of operational error, this interlock is applied at the controller for the final output terminal. These
safety interlock features are more or less mandatory to assure the safety of the equipment and the reliability of the system as a whole. The
control loop, therefore, should be given very careful consideration
regardless of how complicated it becomes.
FICAUTO
FIC AUTO
IO P :
CAS :
AUTO:
MAN :
Breaking of input signal wire

Cascade mode
Automatic mode
Manual mode
Fig. 10 6 Safety interlock logic diagram
(c) Topping unit control incorporating analyzers

In Subsec. 10.1.1, we described how the rapid progress in energy
conservation has helped create the oil surplus we have today. Nevertheless demand for intermediate distillates such as gasoline, kerosene, and
gas oil has continued to grow. This has led refineries around the world
to convert their facilities to increase the output of those intermediate
distillates for which there is such a strong demand.
One measure, for example, with the objective of boosting the recovery rate of designated topping unit fractions is to mount on-line analyzers to side stream outlets. To illustrate the principal, we will consider an analyzer mounted on a kerosene distillate line (refer to Fig.
10. 7). The figure illustrates flash point control for kerosene. The flash
612
Chap.lO
Fig. 10 7 Control diagram of kerosene on-line analyzer
point is the lowest temperature at which combustion begins and continues in a substance; the safe handling of petroleum products is obviously contingent on the constancy of this temperature value for a given
product.
The conventional method for determining the properties of petroleum products is to take a sample of the product back to the laboratory
and analyze it. Since the analysis had to be repeated several times over
a day's time to assure the reliability of the results, a great expenditure
of time was required to stabilize the process after any alteration of the
running conditions or the kind of crude oil. On-line analyzers are extremely important for maintaining a continuous operation, and, in fact,
they are even mandatory when modifying running conditions.
Recently, the durability and reliability of these analyzers have
been enhanced to the point they can be used on-line. An example will
serve to clarify how control is effected. When the kerosene flash point
raises above the setpoint, this results in insufficient lighter fractions
in the kerosene which in turn has the effect of reducing the setpoint of
the temperature controller at the top of the tower. As the temperature
controller at the top of the tower lowers the distillate temperature,
this signals the need for more reflux to be returned to the kerosene.
This raises the setpoint of the temperature controller, and the volume
of reflux (FIC) is increased. Thus, as the temperature in the upper section of the tower is reduced, the temperature gradient within the tower becomes large. Since the temperature of the extracted kerosene distillate is lower, the distillate contains more lighter fractions.
Since the standard for kerosene involves limits on the heavy-side
as well, about 95% of distillation point analyzers can also be employed
10.1
613
for heavy-side control. The increasingly widespread use of on-line instruments, including pour-point analyzers and gas chromatograph analyzers in addition to the instruments already described, have made
refinery facilities more efficient and have contributed to improved
product quality.
10.1.3 Off-site instrumentation
In contrast to on-site, direct-production processes carried out at
oil refineries (distillation, cracking, polymerization, etc.), supporting facilities such as handling and the physical distribution system of petroleum products is referred to as off-site.
Given the concerns of off-site instrumentation, it mainly involves
processing and control of data accompanying the physical distribution.
Systems configured out of panel instruments and process computers
have been available in the past, but with the appearance of distributed
control systems and high-performance mainframes, the trend is toward much more comprehensive management of data.
The production activity at refineries generally follows a monthly
production plan from headquarters. It is only natural, however, that
the actual production management structure of each individual refinery proceed in accordance with the local flow of production activity.
The various activities that fall under off-site-the presence or absence
of a marketing network, the blending of half-finished or finished products, land and maritime shipping, etc.-will vary greatly depending on
distribution facilities and/or mechanisms that are available. Instrumentation methods are also quite different depending on the type of
operation. Nevertheless, control of distribution data is the primary concern; increasingly the recent trend is toward consolidated systems that
incorporate production management.
(a) Operations control system
The functions of this system are to control; production plan data,
receiving and shipping plan data, quality control data, incoming/outgoing shipping capacity, tank storage capacity, etc. Tank storage capacity is especially critical and must be checked every two hours.
(b) Operational control system (maritime blending)
The main function of this system is to control blending operations
aboard tankers while the product is being shipped. Five subsystems of
the operational control system will be briefly described.
(1) Order processing subsystem
This subsystem receives and processes information from external sources that is necessary for the operational control system. After the data is verified, it is compiled as a
data file. The information contained in the data file is necessary for
the smooth operation of all the other subsystems. A summary of the or614
Chap.IO Instrumention to Manufacturing Industries
Table 10.1 Order processing information

Input data item
List for assignment of
ships and handling of oil
Date, Product, and Quantity
Sailing schedule
Date, Customer, Retailer, Delivery area, Ship name,

Product, and Quantity
Received oil properties
Quality control standard value
Production schedule
Load of equipment, Rundown, Target properties and

schedule of bonded oil
Distillation schedule
Feed tank No., Rundown tank No., Feed amount

and change-over time
Crude oil tank handling

schedule
Crude oil tank operational plan
Maintenance work
Date, Work items and work period
Operational data
Feed amount to each equipment, Distillate amount

and product
Others
der processing information is presented in Table 10.1.

(2) Handling schedule subsystem
The task of this subsystem is to determine the optimum tank configuration for distribution based forecasts in one-week units. The following factors are especially relevant:
data received from the order processing subsystem, production plans,
equipment operation plans, maintenance work plans, the inventory situation for different products, quality control data, etc.
The tank capacity projections for different products are derived
from production volume data available on a per-day basis (volume of
product produced from all operations, volume produced for each product). The weekly forecast is then calculated based on the one-day units.
Figure 10.8 shows the logic flow that is used to determine the optimum handling tank. Note that in the handling tank, or handling berth,
logic there are a number of factors such as individual refinery rules
and conditions, external conditions (e.g., maritime conditions), and so
on that vary tremendously and make it extremely difficult to drive an
optimized calculation. Thus, human experience and judgment have
been incorporated into an expert-system program (the so-called tank,
berth selection support program) that is now available.
(3) Operational schedule subsystem
Once the determination has been
10.1
615
Checking various
decision logic
Fig. 10 8 Flow chart of handling tank
made in the handling schedule phase that the oil of a particular refinery is to be transported, the object of this system is to formulate a
schedule either on the same day or by the next day, of oil moving
work units. The two objectives of the logic-both realized by scheduling the minimum time needed to perform the various tasks involved-are to reduce the work time. This is done not only to conserve
labor, but also to improve product quality by minimizing the time the
product is exposed to contamination.
Once the work is actually under way, external elements often
intervene making it necessary to depart from the original schedule. To
cope with these inevitable alterations in the work schedule, the schedule is kept on an interactive basis through a CRT.
Specific scheduling tasks that fall under the purview of the operational schedule subsystem include loading, unloading, blending, transfer productivity, mixing, and scheduling documentation.
(4) Tank-inventory control subsystem
This subsystem manages all
data pertaining to tank inventories.
The system monitors the following areas; tank inventory calculations (current and revised inventory calculations), confirmed, cyclic,
and effective tank inventories. This subsystem also maintains a tank
capacity chart.
(5) Quality control subsystem
The main function of this subsystem is
616
Chap.lO
to handle sample testing (from the initial test request to the issuing of
the test report) to establish the properties of each lot. Some of the areas covered include loading and unloading operations, oil moving between equipments and tank productions operations. This subsystem
also performs inferential calculations of oil properties and ratio calculations to assist blending operations.
Figure 10.9 shows how the quality control subsystem is related to
the other subsystems.
r----
Quality control subsystem
Inferential calculation for

amount of oil in tank
and properties of oil
I
L__--1----
Sample test, Blending ratio,

Inferential calculation for
blending properties
-"'
c:.a
-;;oo
0.
E
gl
'0
n"'
0
t:
2
1;;
.5
Fig. 10 9
.5-g
-;]]
.!! .s.o~ 5:C:C
0.
u . . rJ
E ~8t.-. "E tE:
0
gl
;:g8_
~ 0 - 0
'0
.!!
Handling schedule
sub-system
'!
0
0.
1:!"'
~~
- u u
~ ...
''~' t::~
c..
=-"'
o..=>o. 8.]-g
~~ ~
~
" -<>
~B ~
Operational schedule
sub-system
Operational control system
Relationship between quality control subsystems and each subsystem

at terminal
(c) Operational control system
The task of this system make sure the various operations are conducted safely and go smoothly based on operational booking information obtained from the upper-level operational management system.
Here we will describe two components of the system, automatic
lineup and job control.
(1) Automatic lineup
Lineup is the task of determining the optimum
route to move the product from a point of origin to a point of destination.
Automatic lineup represents the off-site system as a network of pipelines and devices. As can be seen from the outline of automatic lineup shown in Fig. 10.10, the pipeline network is shown with device numbers and node numbers allocated to pipeline branching points. From a
given starting point, a line is found to a node number. Assuming, for
example, that our objective is to get from Tank 1 (the starting point)
10.1
617
Berth- I
No, of paths=2
Examples of selected lines
0-0-0 -0-0- 0-@- @-@

0-0-0 --@ -0-@ -@-@
Fig. 10 10 Outline of automatic lineup
to Berth 2 (the destination point), one of two lines could be selected.

Then, the lineup conditions of the two possible routes are evaluated: to what extent are the pumping equipment and flowmeters already
being used along the two lines? Is there a difference in capacity between the two lines? And so on. Finally, one line is selected based on
the results of the evaluation.
(2) JOB control
Another important off-site concern is moving oil
around: shipment out of a tank, blending and moving the product to a
holding tank, shifting oil from one tank to another. This type of operation that involves movement of oil from one place to another is referred to as an oil handling operation. The unit of oil handling operations is the JOB, which is specified by a sequentially assigned number
(called a JOB No.). Every movement of oil is thus designated and handled according to a JOB No.
JOB control thus consists of centralized control over the string of
yard devices (flowmeters, pumps, valves, etc.) that correspond to a selected automatic lineup covered by a given JOB No. JOB control also involves such functions as start and stop sequences, alarm monitoring
and emergency shut-down procedures, and collection of data. A system diagram of JOB control is shown in Fig. 10.11.
Blending operations while en route aboard ocean-goin g tankers
(maritime blending) calls for especially tight control. Today, even sulfur-content control (a feedback control process) and viscosity control,
tasks that used to require conventional box-type meters, have become
routine with the DDC and sequential functions that are available with
618
DDC
G) Batch information
Start/Stop
Open/Close
(!) Blending ratio
Fig. 10 11 System diagram of JOB control
F :Total flow rate

S :Total composition
F; : Component flow rate
S, : Composition of component
s,F,
---------s,F,
S,F,
F=F 1 +F,+ + F,
SF= S1F1
+ S1 F, + + S, F.
Composition control system by blending
J.l :
Mean value of S,
Total flow rate ratio control system by optimum distribution
Fig. 10 12
10.1
Composition control by blending
619
distributed control systems.

Merely maintaining the sulfur content of the fuel oil at a fixed
level is not enough; some version of the blend composition control
scheme shown in Fig. 10.12 is necessary to obtain a stable, standard
product.
In maritime blending, the DDC algorithm can be set not only for
fixed-volume batch control but also, by adjusting the setpoint appropriately, for flow-rate control. Also there are two control algorithms;
which algorithm is selected depends on the blending requirements of
the particular lot under shipment.
The maritime blending control algorithm has a number of addition
functions:
(1) temperature compensation (ASTM compensation, and primary
and secondary compensation);
(2) non-reset integrator (uncompensated value, compensated value);
(3) emergency shut-down function in response to anomalies; and
(4) a simulation function.
The determination of specific gravity when making temperature
compensation adjustments has especially been improved by making this
a software function. The older semi-fixed method produced a mean specific gravity, but now it is possible to determine the specific gravity of
every load at the time of shipment so the weight can be precisely regulated.
(d) Land shipping control system
This system maintains operational control of land shipping in fuel
oil and LPG tanker trailers and rail tank cars.
Generally, there is a higher level order processing control system
that passes contract information to the overland shipping control system. The specific functions of the overland shipping control system
are then to issue dispatch slips, verify the orders, pass control information to the lower level control system, and compile data on the shipments. In land shipping, especially when long hauls are involved, rail
tank cars are preferred because volume shipments are more economical. Oil companies generally have swap agreements in effect so they
can exchange oil to save transportation expenses. The trend is away
from shipment of oil in large-scale (20 k/ capacity) truck trailers
because the transportation costs of this mode are comparatively high.
Fuel-oil and LPG tank trailers are dispatched early in the morning
and can accommodate only 4 k/ (formerly, only 1 or 2 kl) compared to
20 k/ for the large-scale tanker trailers. A fast, large-capacity computer is essential for this system to keep up with the frequent trips, the
numerous control items involved, and the rapid accumulation of control data.
620
Chap.lO
Since operating conditions change quite frequently, the operators

must be left in charge of collecting and issuing dispatch slips. A general upgrade in system performance is strongly desired that would enable immediate execution of dispatch slips and a hatch control function
for safety. In the overland shipping control system, as in the other systems described, safety measures must continue to be developed that
are effective with fewer personnel. Basing operational information on
feedback from the operators and interlock control are particularly important. Interlock control confines and minimizes the danger to shipping outlets and field depots in the event of a system failure.
In LPG tanker shipping systems, commodity weight has generally
been determined by subtracting the tare weight from the actual
weight. This method is being rapidly replaced by DDC weighing techniques, which are far more accurate.
10.2 Instrumentation Applications in the Iron- and Steel-Industry 7 l
-lsl
10.2.1 Overview of instrumentation in the iron- and steelmaking process

In periods of high growth, the steel industry has generally sought
to achieve economic growth through production expansion by increasing the scale and speed of equipment, and through the introduction of
continuous processes. The first oil crisis in 1973, however, caused a
fundamental re-evaluation of this conventional wisdom. New demands
emerged for conservation of resources, energy, and labor; for reduced
costs and improved quality. These demands have served as a powerful
impetus for the development of equipment that consumes less energy;
for processes that are simpler, faster, and continuous; and for equipment that is direct coupled, faster, and that permits continuous processing.
Instrumentation has also been responsive to the changed environment and the new demands; as a result, a diverse range of instruments, sensors, and system devices has been developed. The point has
been reached where the post-oil crisis objectives of lower costs and improved quality are all but unrealizable without the benefit of the instrumentation that is available today. In fact, it is no exaggeration to say
that the reliability of instrumentation exerts a major influence on productivity. Two developments in particular have contributed to the
elevat:on of instrumentation equipment to this important role: the
emergence of digital control systems through the development of
microcomputer technology and improved sensor technology. For example, digital instrumentation systems have been developed that satisfy
operational needs by providing centralization of a diverse array of functions together with manipulation and supervision. It is not surprising
10.2 Instrumentation Applications in the Iron- and Steel-Industry
621
that in recent years these systems have been adopted to most processes.
(a) Overview of iron- and steelmaking
Figure 10.13 shows the production flowchart for a typical ironand steelmaking process. As is shown in the figure, the main process
divisions are extraction of iron from iron ore in the ironmaking plant,
production of steel and steel slab in the steelmaking plant, and production of various semifinished and finished steel products in the rolling
plant.
(1) Jronmaking plant
While the blast furnace is the central unit of
an ironmaking plant, other equipment is also necessary because the ore
must undergo various preliminary processing to make it more desirable
feed for the blast furnace. For example, to equalize the quality of the
feed ore and achieve stable operation of the blast furnace, a sintering
plant is necessary to fuse or frit pulverized iron ore, limestone, and
coke breeze together into uniform-size agglomerate called sintered ore
or sinter. Also requisite is a coke oven . Coke produced in this unit serves not only as a thermal agent to promote fluxing of the iron ore and
limestone but also is an indispensable reducing agent. Sintering and
coke play a further important role in creating the optimum permeable
burden which permits a high rate and also a uniform gas flow within
Steelmaki ng pla nt
lronmaki ng plant
Sintering machine
c-~--~
lro n oreL b- -
L:=:J
omes one
..
Pelleuzong
plant
Ingot
maki ng
~-~Coa l
Coke oven
Scrap iron
Con vert~
- l}~g6~ ~
Bl ast fur nace
Iron o re
Sintered ore
Coke
Limestone
:l~
Electric arc
furn ace
Fig. 10 13
622
}#'
Slabbing
mi ll / Blooming
mill
J~ -
Main products
Rolling pl ant
~ ~~~~!pile
- l....oo"
Section
Ba r
Billet
-B
~oom
~b ~b
Continuous cast ing
equipment
Structural steel
roll ing mill
Deformed b a r
- ~-Wire
- IWirerod mill
........R..._Piate
'--8P!ate mill Hot rolled
Hot-strip mill
sheet
~
.---
Hot COil
--l Hoop
__j
Cold-rolled sheet
Cold-coil
~ Galvanized sheet
Cold-strip mill
Welded tube
man ufacturing equipment
Steel casting
~- Welded tube
~- Seamless pipe
Seamless pipe manufact uring equipment
' - - - -- - - - - - Steel casting
products
'---~-
An exa mple for production flowch art of a typical iron- and

teel mak ing proces
Chap.lO
the blast furnace. In the blast furnace proper, the pre-processed sintered ore and coke are charged at the top of the furnace along with such
auxiliary material as limestone. Hot-air blast introduced from the
tuyeres burns the coke to heat, reduce, and melt the charge as it descends to produce pig iron.
With a high carbon and impurities content, pig iron has a hard,
brittle characteristic. For this reason, most pig iron is sent on to the
steelmaking plant to be processed as steel. Finally it should be noted
that the blast furnace gas (BFG) and coke oven gas (COG) are valuable
heat sources that are reused in the plant.
In the steelmaking process, the carbon content
(2) Steelmaking plant
is reduced and the impurities are removed from the pig iron to produce
steel, a material that is malleable yet strong and can be readily
worked. The center unit in the steelmaking process is the converter.
Traditionally, top-blowing oxygen was generally used in the converter, but more recently bottom-air-blown and combination blowing (top
and bottom) converters have been developed. The combination-blowing
converter produces a homogenous-quality steel very quickly with a
low level of impurities, and has thus come to be widely employed in
the industry. Gas generated in the converter during blowing is recovered and is employed as a valuable heat source elsewhere in the plant.
The liquid steel produced in the previous process step then undergoes secondary refining to achieve the precise qualitative requirements
of the finished product and also to reduce the amounts of hydrogen, oxygen, and nitrogen gases which have an adverse effect on product quality. The two principal secondary refining methods are vacuum and injection; which method is used normally depends on how the finished
steel is to be used.
In the next step, the liquid steel is teemed into molds as ingots or
continuously cast to form semifinished shapes (e.g., slabs, blooms, or
billets). In recent years, more and more use is being made of continuous casting because of its advantages over conventional casting into
ingots. These advantages include simplified processing, reduced investment in equipment (the need for heavy rolling-mill equipment is eliminated), conservation of energy, and increased yield. In Japan, for instance, the ratio of steel tonnage processed by the continuous slab-cast
method has reached 90%.
The most recent advance in process technology is a method to bypass the reheating furnace altogether so that semifinished slabs can be
sent directly from the continuous casting machine to the rolling plant.
Rolling mills can be divided into two fundamental
(3) Rolling plant
types: hot-strip mills in which the hot-roll processing occurs at a high
temperature and under pressure, and cold-reduction and temper mills
623
in which the cold-roll processing is carried out at normal temperatures

(structural changes occur at temperatures below about 720C). In this
area too, rapid advances have been made in the development of rolling-mill processes that are direct-coupled, faster, and continuous. The
procedure of processing one coil at a time has been largely superseded
by continuous rolling and continuous annealing equipment, and this
has had a very beneficial effect on quality and yield.
(b) The trend in instrumentation equipment
Iron-and steelmaking processes are fundamentally batch or
semi-continuous processes. Research nevertheless continues in an effort to simplify and speed up processes in order to conserve energy
and labor. Especially in recent years, in response to the demand for
small-lot production of diverse products, not only the enhancement of
equipment and operation technologies but also the development of advanced instrumentation equipment such as process computers, digital
instrumentation systems, and sensors has come to assume an extremely important role in the iron and steel industry. Also, as communications technologies have become more sophisticated, this has
spurred the trend toward increased integration and centralization of
discrete equipment units. Integrated control systems based on optical
communications and other advanced technologies have also continued
to grow in importance in last few years.
10.2.2 Blast furnace instrumentation

A blast furnace is a tall refractory-lined furnace in which iron ore
is indirectly reduced by coke to produce pig iron. Iron ore, sinter,
coke, and other auxiliary materials are charged at the top of the furnace, and hot--air blast introduced through tuyeres burns the coke. CO
gas given off by the burned coke reduces iron oxides (e.g., FeO) to produce pig iron, which is removed through a taphole. Impurities contained in the ore are removed as slag through slag holes.
The blast furnace is often regarded as a symbol of iron- and steelworks. The trend has been toward larger furnaces, and in recent
years, a number of giant blast furnaces have been constructed with capacities in excess of 5 000 cubic meters. Figure 10.14 shows the system
layout of a blast furnace and auxiliary equipment.
As in other industries, the principal roles of instrumentation in
the iron and steel industry are equipment maintenance, stabilization
and supervision of operations, and data collection.
The overall instrumentation system of a modern steelworks can
comprise as many as 700 to 800 loops, but, of these, at least 80% are
control loops for supervisory control. In the case of conventional analogue instrumentation, the instrument panel must be capable of super624
Chap.lO
lnstrumention to Manufacturing Industries
Furnace top
bleeder valves
To gas
holder
Fine coke conveyer
si nter
bins
Ore weighing hopper

.
mter Ore discharge conveyer
conveyer
Fig . 10 14
System layout of bla t furnace equipment
vising the entire loop in order for the operator to maintain a comprehensive grasp of the work situation. As a result, the instrument panel
for such systems generally ranges from 10 to 15 meters in width which
must be accommodated in a fairly large control room and attended by
numerous personnel. With the rapid development of digital control instruments, operator- system interaction is carried on through a CRT;
essential data respecting the work can be quickly and easily brought
up on a CRT and acted upon. This type of arrangemen t needs fewer
operators and thus the control room can be reduced in size.
Once a problem or accident has occurred within a blast furnace,
major repercussio ns are felt by all other processes. Thus, the importance of maintaining a safe and stable blast furnace operation cannot
10.2 Instrumentation Applications in the Iron- and Steel- Industry
625
be overemphasized. Redundant implementation of all critical loops is

mandatory. An uninterruptible power supply is also essential so that
operations can be resumed quickly and safely in the event of a power
failure.
One outcome of the recent availability of digital instrumentation
devices is that they are being increasingly applied to continuous and sequence control to achieve greater organic integration. Sequences such
as hot-stove switchover and raw-material outputting, conventionally
controlled with electrical devices, are frequently put under digital control today.
The configuration of a typical digital instrumentation system is
shown in Fig. 10.15.
The remainder of this section will survey the main types of equipment associated with blast furnaces and describe the various control instruments that are applied to each unit.
(a) Instrumentation for raw-material mixing and charging equipment
Raw-material mixing equipment controls the mixing and feeding
of iron ore, sinter, coke, and other materials and conveys the mix to
where it is charged at the top of the furnace. Considering the direct
link between process conditions and product quality, high precision
and reliability is of the utmost importance. A process computer is
therefore used to regulate the recipes and proportions of the constituents in the mix, and the mix feed is controlled through sequential processing.
The top-charging equipment of the furnace can be either of the
bell type or the bell-less type. The bell-less type has become much
more pervasive in recent years, because with that type, the stock-line
level at which raw material is charged into the furnace can be freely
set at any point or thickness. The bell-less type is also considerably
lighter than the bell type. In terms of instrumentation, however, the
requirements of the two types are almost identical. One difference is
that in the case of the bell-less type the seal valve is susceptible to
damage from the raw material if the valve isn't closed after all the material is discharged from the hopper. With this design, therefore, a sensor is mounted on the hopper to detect load cutoff.
(b) Blast furnace instrumentation
Blast furnaces are basically cylindrical vessels with a slight
pear-like bulge about one-third up from the bottom. Thermal expansion of the material causes it to be pushed out against the walls of the
interior of the furnace with considerable force; this, coupled with the
bulge in the lower central section of the furnace, prevents the material
from sliding down inside the furnace. As iron is formed, the material
shrinks greatly in volume. For this reason, the bottom section of the
626
Chap.lO
c;:;:,
......
.:!
:i'
1:>
.......
"':;;
....
;::;
"
;:&::..
-....
i:i
....
;::;
"'
$1
::!
....
"'
:i'
i:...,
Furnace body monitering

Bla~t control
Water leak detection

(Tuyere) (Sonde) (Profile)
Furnace body cooling

Water treatment
Special instrumentation
Raw material weighing
Hot stove (Blast)
UJ
;...
2"'
""
"
<
~u
<C
IMicro computer room I
~ "'0
o_g
...~
i5.
.~
"0
High -pressure steel gas

Gas cleaning
PW
TRT
(f) Top pressure recovery
turbine
Distribution board
CCIIIJ
I
I
~
II
,.,
biiiii;:;
I I ll-. 1 11 16
,..,.., .I
COPSV
(!)
Ill
Blast furnace
(1)
HF - BUS
:.I
llj
COPCV
Furnace Casting Casting Hot

bonom noor 2 noor 4 stove
Typical configration of blast furnace instrumentation system
ILocal area I
(j)
II
COPSV
Hot
stove
desk
IControl room I
T RT
desk
Legend
COPSV : Operater station
COPCV : Operater console
CCHC : Color hard copy unit
CPRT : Serial printer
CFCD 2 : Duplex field control l :tation
CFCS 2 : Field control station
CFMS 2 : Field monitering stati, n
..'_I_ I
Hot stove CTBC 2 : Terminal board cubici e
II
II
,.,
High
top-pressure
Blast
operation
furnace
desk COPCV desk COPCV
CFCSt CfCDI CfCDI CTBCI CfCDl CfCDl CfCDl CTBC! CfNS! CfNSI
J.
1-
CPRT
Fig. 10 15
::E"
] ~
!l
"'
il"c
(!) Hot stove (Combustion)
<D
Cohesive
Fig. 10 16 Shape of blast furn ace and reaction tatu in ide
furnace is made narrower. Figure 10.16 shows a cutaway diagram of a

blast furnace in reaction.
Due to the extremely high temperatures and pressures that are
reached inside the furnace, it is difficult to control the reaction directly. Thus, indirect monitoring is used by regulating certain parameters
and variables. The more important of these variables would include :
(a) composition and weight of the raw material ; (b) temperature,
weight and composition of the hot metal and slag; (c) volume of blast;
(d) pressure of blast; (e) temperature of the hot blast; (f) pressure at
the top of the furnace; (g) temperature inside the furnace and temperature distribution; (h) upper furnace gas flow and composition. While
this describes the situation up to present, we would note that new
instrumentation techniques and special sensors are rapidly being developed that make it possible to control directly more and more aspects of
the furnace reaction. A modern blast furnace is depicted in Fig. 10.17
showing the locations and functions of the most important sensors.
Recall that hot blast enters the furnace through openings at the
base of the furnace called tuyeres. If the tuyeres malfunction allowing
cooling water to enter the furnace, it is not only the furnace that is adversely affected. In the worst case, a steam explosion could result endangering the lives of personnel. To make sure such a scenario does
not actually occur, it is necessary to detect for tuyere cooling water
leaks. A number of conventional methods of detecting for tuyere leaks
are available including:
628
Chap.lO lnstrumention to Manufacturing Industries
Sen ors
Q) Top gas analyzer
Venical sonde
@ Top thermography device
,,
/ 1
urnace top profi le meter
Detection functions
~ Top gas constituents

2
Temperatu re and gas constituents in venica l

d irection
,//,
Mechanical sounding type
1,}'
Laser or .u:wave sounding type , ...~{?

.,-- - - - - - - - ---.,.......,...,_,
,, "
/
........ ~ ...
I
I
I
I
I
_/-'-'/
'
. . .. . . ,"
at bottom part of cohesive

Hei ght and circumferent ial distribution of
ventilation
Sampling of burden constituents such
Radia l distribution of tem{M=rature and gas
consti tuents on shaft
Distribution of temperature and gas
constituents at root
Height a nd circumferentia l distribution of
thermal load
di sti rbution of branch b last
'
Top cross sonde

Top oblique moving sonde
Throat thermometer
Circumferent ia l distribution of dropping

frequency on unreduced ore
\
tuyere and deadman

\
\
'
Shaft top sonde (Diametor sonde)
@ Layer thickness gauge

(Magnet type, Electric resistance type}
@ Skin flow meter

@ Shaft micro-pressure gauge
@ Shaft static and differential
pressure ga uge
Shaft bottom sampler

Shaft bottom sonde
@ Root sonde
Raceway size
Distribut ion of tempera ture a nd gas
constituents in raceway
Ind irect temperature
measurement of raceway
Temperature distribu tion on brick of hearth

wa ll
@ Shaft stave thermometer

@ Branch blast volume meter
@ Tuyere micro -pressure ga uge
Tuyere coke sa mpler
Raceway phenomena
measuring device
@ Tuyere sonde
@ Tuyere-nose flam thermometer
@ Furnace bottom thermometer
Fig. 10 17
Functions a nd locatio ns of blast furnace sensors
629
(1) monitoring the temperature of the tuyere cooling water outlet;

(2) monitoring H2 with a gas analyzer at the top of the furnace;
(3) monitoring CO and C0 2 gas in the tuyere cooling water outlet;
and
(4) relying on the judgment of an experienced operator.
The basic problems that are shared by all of these conventional
methods is that they are indirect, they involve a certain amount of
dead time, and they yield less than satisfactory detection accuracy. An
alternative direct method is to continuously compare the volume of supply water against the volume of discharge water. An advantage of this
method is that it can be applied to larger furnaces with more tuyeres
that require a faster cooling-water flow rate to satisfy more stringent
cooling efficiency requirements. Using an electromagnetic flowmeter
for the detecting element, digital instrumentation can provide the following functions:
(1) automatic compensation and control of the flowmeter;
(2) alarm supervision of actual and differential flows;
(3) collection and recording of actual and differential flow data.
To assure a safe operation, the temperature distribution inside the
furnace must be closely monitored. This is done with numerous thermocouples mounted on the inner walls of the furnace. The large-scale
blast furnaces of recent years are equipped with anywhere from 300 to
500 sensors which are supervised by a remote scanner.
(c) Gas-cleaning equipment instrumentation
Blast furnace gas (BFG) is captured so it can be used as a heat
source elsewhere in the facility. However, BFG is loaded with dust
which has to be removed before it can be used. Here, a brief description of some of the more common kinds of gas-cleaning equipment will
be provided.
A dust catcher is a device that separates out the courser-sized
dust particles by reducing the gas flow rate. Venturi scrubbers, usually configured in pairs, are gas-cleaning devices in which liquid injected
at the throat of a venturi is used to absorb the dust from the gas flowing through the venturi. A septum valve situated between the two
scrubbers is generally used as a final control element for pressure control at the top of the furnace, but septum valves are extremely noisy
when they open and close. Ring-slit scrubbers are being substituted
for septum valves in many contexts because they are so much quieter.
Ring-slit scrubbers are generally installed between two venturi scrubbers. Regardless of the configuration, a hydraulic drive system is used
for the actuator.
For venturi scrubbers, the main instrumentation function is water-level control. This is usually accomplished in one of two ways: ei630
ther with a gas-seal function or by monitoring scrubbing discharge. In

the past, the latter method had been much more common due to the
harsh environment a level-detecting element has to operate in. Recently, however, a differential pressure transmitter with diaphragm seal
has become available that offers a method of direct detection. These elements are mounted in pairs for redundancy with the ability to switch
over to the second element if the first element fails. The actuator has
to be designed so it is immune from clogging; usually, a butterfly
valve is used. In this environment, measures to enhance abrasion resistance are mandatory given the extreme abrasion from dust and slag.
Here too, a hydraulic drive type actuator is used.
(d) Instrumentation for high top-pressure operation equipment
It is obvious that pressure must be controlled at the top of blast
furnaces to maintain a safe, stable operation. Most modern blast furnaces operate at high pressure in order to realize two objectives:
(1) The volume of hot-air blast can be increased without increasing
the gas flow rate. This reduces the volume of dust particles and
helps to prevents hanging and channeling from occurring in the furnace. (Hanging describes a condition when the raw material in the
upper part of the furnace becomes stuck and only the material in
the bottom portion of the furnace descends. Channeling refers to a
state in which the raw material collapses and the gas blows past
it.)
(2) Operating at high pressure also makes it possible to slow the gas
flow rate, thus prolonging the contact between gas and ore; this
improves the overall reaction in the furnace and enables the ratio
of coke to be reduced.
An essential instrumentation task is top pressure control. The
loop that performs this function is the most important in blast furnace
operation. Considering its importance, the top pressure detection element is usually mounted in duplicate or triplicate for redundancy with
the ability to switch between elements. Moreover, since so much dust
is present in the upper part of the furnace, back pressure is measured
based on N2 which is intermittently purged from the vicinity of the detecting element.
The control element for top pressure control consists of a septum
valve and a ring-slit scrubber. More recently, in an effort to conserve
energy, it has become quite common to install equipment to recover energy from the furnace gas. The equipment is configured in such a way
that energy recovery and pressure control are accomplished at the
same time.
Given the instrumentation available today, it is rare for abnormal
pressure to develop in blast furnaces. In the event pressure does reach
631
a critical level (i.e., due to operator error, control equipment failure,

and so on), it can be quickly dissipated via a bleeder valve to prevent
the furnace from sustaining any damage.
Since the hopper and furnace top pressures must be equalized, raw
material is charged to the hopper in batches.
(e) Hot stove equipment instrumentation
In order to achieve fast and effective combustion in blast furnaces,
great amounts of hot gas are required. This requirement is met with
hot stoves, a specially designed type of heat exchanger. Heat is stored
in bricks by application of combustion gases. After the combustion is
turned off, cold blast is brought into direct contact with the preheated
brick, and heat is transferred to produce hot blast. A typical blast furnace is supported by 3 to 4 such stoves, which are rotated (e.g., heated
for a period of 3 hr and on-wind, or exhausting for 1 hr) to provide hot
air to the furnace on a continuous basis.
Hot stove instrumentation has three main functions; switching system control to regulate the alternating heating and blast phases of the
several stove units, combustion temperature control for the heating
phase, and blast temperature control for the on-wind, or blast phase.
The hot stove instrumentation functions are shown schematically
in Fig. 10.18.
Sequential contro
for change over
shutoff valves
Fig. 10 18 Hot stove instrumentation functions
Combustion temperature control can be broken down into fuel gas

calorie control and stove dome temperature control. Combustion control consists of (a) combustion gas flow control to regulate applied
heat, and (b) air/fuel ratio control to regulate combustion temperature.
The temperature of the stove dome can be controlled by regulating the
above two factors. To optimize the combustion in order to conserve
fuel, it is essential that the exact proportion of air is delivered to the
632
Chap.lO
fuel mixture for perfect combustion. This can be controlled by sensing

the amount of oxygen (0 2) in the exhaust gas. Moreover, as stoves approach their target heat storage temperature, exhaust gas temperature
increases, which is also a waste of fuel. Thus, exhaust gas temperature
is monitored to maintain efficient combustion temperature control. For
dome temperature detection elements, the most common arrangement
is to mount two thermocouples (one is a spare) and a radiation pyrometer with switchover capacity between elements. Zirconia oxide analyzers are generally used as oxygen analyzers for the exhaust gas.
Blast temperature control is based on a signal from a temperature
sensor mounted at the end of the hot-blast main pipe. Based on the sig. nal output, hot blast is mixed with cold blast to produce the required
blast temperature. There are different schemes for providing the blast
to the furnace, but two common methods are: the single blast method
where the blast is delivered by a single stove, and the staggered parallel method in which the blast is delivered by two stoves at a time on a
cyclic or staggered basis. Controlling blast temperature is extremely
important for blast furnace operation. A spare blast temperature sen-sor is therefore mounted along with the active sensor with switchover
capability.
A change-over sequence controls a system of shutoff valves to
regulate alternating steady-state periods of heating, blast, and rest. It
also controls the periodic transitions between these steady states. Until recently, this control was maintained by electric equipment, but
now instrumentation configurations are more common.
To monitor stove temperature, a typical stove is equipped with 40
to 50 sensors mounted on the shell and throughout the stove. The sensors are then supervised with a remote temperature scanner.
(f) Furnace cooling equipment instrumentation
This system is necessary to prolong the life of the refractory brick
used to line the blast furnace. Two types of systems are encountered;
the plate cooler and the stave cooler methods. With the migration toward higher pressure operation, the stave cooler method has come to
predominate. With this method, steel tubes cast into cast iron are installed in the shell to cool the furnace. Either natural-flow or forced
circulation evaporative cooling is piped through the tubes. Forced circulation systems are far more common today.
Instrumentation regulates a number of important variables in the
connection between the deaerator storage tank and the stave drum.
For example, monitoring the characteristics of the pure water used as
coolant is particularly important; dissolved ion density and pH of the
cooling water are monitored, alkalinity of the coolant is maintained to
prevent tube and stave corrosion. Staves are vertical strips made of re10.2 Instrumentation Applications in the Iron- and Steel-Industry
633
fractory brick that are placed edge to edge to form the lining of the
furnace. A float switch is mounted near the top of each vertical stave
system to detect for either damage to the staves or disrupted circulation of coolant.
(g) Auxiliary fuel injection equipment instrumentation
The principal fuel used in blast furnaces is coke. To reduce the
proportion of coke, however, it is generally supplemented with other
fuels. In the past, heavy oil and tar were commonly used as supplementary fuels, but since the oil crisis, every effort has been made to phase
out petroleum-based fuels and substitute coal-based ones. Injection of
pulverized coal has become the most common method in recent years.
Operation of pulverized coal injection (PCI) requires facilities to
stock the coal, pulverized it, store the pulverized coal, and finally inject it. The pulverized coal enters the blast furnace by being injected
through the tuyeres.
One thing that must be carefully considered in the instrumentation for the PCI is the prevention of explosions. Since the sensors must
be in direct contact with the pulverized coal, they are mounted inside
the pulverizing equipment. Control elements such as shut-off valves
and control valves that come in contact with the medium carrying the
pulverized coal must be designed to withstand extreme levels of abrasion. Moreover, while pulverized coal must be continuously injected
into the furnace, other units in the overall operation are often run on a
batch-cycle basis. Thus, the pulverized coal injection must be organically integrated with sequence control and continuous control.
(h) Furnace blower equipment instrumentation
Shutting down the blower is equivalent to shutting down the furnace, and therefore an auxiliary blower is always standing by. For a
single-furnace operation, two blowers are configured, and for a
two-furnace facility, three blowers are usually provided. Since blast
furnaces require a stable supply of blast, turboblowers are equipped
with instrumentation to maintain a constant volume of blast for normal operation. When switching between stoves, constant blast pressure control is applied to maintain the delivery pressure and assure a
constant amount of blast is delivered to the furnace. Moreover, since
the pressure and volume of delivered blast changes depending on the
furnace load, a centrifugal compressor is used, which introduces the
danger of surging. Surging prevention control has to be implemented
to assure this will not occur.
To maintain a constant volume of blast, flow sensors are mounted
on the gas main that delivers blast to the furnace. Flow control is effected by calculating a nominal flow value and then making any adjustments in temperature and pressure that are necessary to make the ac634
tual flow correspond to the nominal value. Final control is maintained

by adjusting the angle of the stator blade on the turboblower. The controller employs an external feedback mechanism to regulate constant
blast pressure.
Constant blast pressure control is automatically triggered by a signal that a particular stove is on-wind. The constant blast pressure control uses the discharge pressure value immediately prior to switchover
as the setpoint value. Then, when a signal is received that a stove is
about to go off-wind, control is switched back to constant blast volume
control. The controller is based on external feedback (i.e., when constant blast volume control is on).
Surging is another problem associated with blowers. Surging characteristics of each blower are determined so surge prevention can be
designed into the lines. For control, a blow-off valve is opened to keep
from entering the surging zone, which can be determined by the relation between measured discharge pressure and intake blast. If the controller indicates a large deviation from normal, the control output signal to close the blow-off valve overshoots. Since the control output signal must react quickly in the event the control zone is entered, the controller is equipped with an anti-reset function.
While this section has covered the fundamentals of blast furnace
instrumentation, there are aspects of the reaction mechanism that are
not fully understood even today. Blast furnace technology in the future will evolve toward a more competitive process that is attuned to
the natural resources situation, environmental impact, and economic
conditions. It goes without saying that the role of instrumentation in
this development is extremely important. Even greater efforts must be
applied toward the development of more sophisticated instrumentation
systems and sensors.
10.2 .3 Continuous casting equipment instrumentation
(a) Overview of the equipment
Referring again to Fig. 10.13, after the pig iron (hot-metal) is
tapped from the blast furnace, refined in a converter and then conveyed by ladle to a continuous casting machine (an alternative route to
the continuous casting equipment is via an electric furnace, as shown
in the figure). The continuous casting equipment forms the molten
steel into semi-finished shapes, e.g., slabs, blooms, or billets. This is
done, as the name says, on a continuous basis.
Figure 10.19 shows an overview of a continuous casting machine.
Molten steel from the ladle is poured via a tundish into a water-cooled
copper mold. Once the molten steel has been poured from ladle to tundish, it is teemed into the mold by means of an immersion nozzle. In a
635
""....::!
::!
"'
....
o::
....
;::
:::;.
I:)
::!
.;.,
::!
<:;
;:!
;::
....
"'....
Ladle
Swing tower
~~
'\.V~
Pinch roll
Mold
Cooling water
I M I I M rr--
Fig. 10 . 19 Outl ine of continuous casting machine
Electromagnetic induction 1
/stirrer
--y--~~~ Mold
~ Tundish
~~~
.......
~
.......
continuous casting operation, the tundish serves a dual role; it serves

as a buffer during cut-over to the next full ladle and also aligns precisely over the next mold before pouring. The inset in Figure 10.19
shows a tundish with two nozzles; this arrangement is for a
two-strand continuous-casting machine. Pure copper is the material of
choice for molds because of its good heat-conducting property. Water
flows into the mold cooling the molten steel so that a solidified shell
forms on the surface. Secondary cooling water then solidifies the rest
of the steel. The solidified cast product is then continuously removed
from the molds through a pinch-roll system, cut into appropriate
lengths at a cutting station, and transported to the rolling mill.
In recent years, continuous casting has been applied to virtually
all types and grades of steel. Even more recently-applying newer
high-speed, high-temperature, defect-free continuous casting technologies-dramatic advances have been made in new continuous process
technologies such as hot charge rolling (HCR), which makes efficient
use of sensible heat and sends the high-temperature slab directly to
the reheating furnace, and hot direct rolling (HDR), which does away
with the need for reheating. Ultimately, these developments are expected to lead to a sheet casting method that greatly simplifies or eliminates intermediate rough-rolling processes.
(b) Aspects of the instrumentation system
Continuous casting control systems are basically well established,
but new techniques have been developed in the areas of high-temperature slab technology for HCR and HDR (e.g., high-speed casting and vapor cooling), and to realize better quality output (e.g., electromagnetic
induction stirrer and electromagnetic breaker). Also, while it is quite
common to use process computers for continuous casting control systems, as supervisory functions have continued to multiply, the trend is
to shift as many control functions to the instrumentation as possible.
Instrumentation systems have also become increasingly complex. If a
function such as spray cooling tracking (described in detail below)
were given over to control by instrumentation, for example, such an
implementation would require an operation data base and high-level
general purpose language functions.
With respect to system configuration, strands are recognized as
separate units with control stations set up for each strand. The most
common arrangement is for shared control functions to be processed
by the independent control stations. Moreover, although there are
more billet strands than slab strands in operation, since the spray-cooling equipment for slabs is heavier than for billets, more control loops
are required per strand. One final point must be noted. if instrumentation is going to be used to control such functions as cutting, mold
637
width alteration, or equipment diagnosis (e.g., break-out prediction,

mold oscillation monitoring, etc.), so-called digital instrumentation control systems are inadequate by themselves because they lack sufficient
calculating speed and data memory capacity. For such applications, an
integrated computer system such as YEWMAC or YEWCOM developed by Yokogawa Electric Corporation is mandatory.
(c) Instrumentation overview
An overview of the main continuous casting process functions is
shown in Fig. 10.20. Here we will be primarily concerned with mold
molten steel level control and spray cooling water control.
The quality of a continuous casting
(1) Mold molten steel level control
product is largely determined at the mold; especially to reduce the amounts of harmful nonmetallic inclusions, it is extremely important that
the molten steel be maintained at a constant level. Quite a variety of
molten steel level meters have been developed for this purpose, as
shown in Table 10.2. In recent years, however, the general trend has
been to move away from radiation type meters toward electromagnetic
(esp., detecting coil) type meters.
To control the level of the molten steel, while the volume of metal
teemed into the mold is controlled by the stopper-rod nozzle or sliding-gate nozzle of the tundish, the cast slab has to be withdrawn from
the mold at a varying rate. However, since the optimum withdrawal
speed depends on the type of steel being cast and the dimensions, it is
not desirable from the standpoint of level control that the withdrawal
speed be frequently altered. This is, of course, assuming the withdrawal speed change is used in the level control function as the feedforward
value.
Operator intervention has necessarily been involved at the beginning and end of a pour, but some modern implementations have even
included autostart (flow-in) and autostop (flow-out) functions in the
instrumentation system. Especially with respect to autostart, however,
the detection range of the molten-steel level meters that are available
today is still inadequate, and thus there is room for improvement in
this area.
Once the molten steel has been te(2) Spray cooling water control
emed into the mold, a solidified shell forms on the mold surface. By
the time the slab has passed through the spray (secondary) cooling
zone, it is completely solidified. The objective of mold cooling control
therefore is to make sure sufficient layers, or shells, of molten metal
have solidified and the slab formed by the time it reaches the
pinch-roll. The task of the spray cooling water control is even more
critical; assurance that the still partially molten slab is completely solidified. The setpoint of the volume of spray cooling water is set in pro638
Chap.lO
Mold cooling water control
Mold liquid level control
Mold powder feeder control
Molten steel weight control

in tund ish
I.
IC
.:l
:i'
~I
Ro ll load measurement
Roll gap and

alignment measurement
Mold oscillation
monitori ng
and piling
weight measurement
marking control
Optimum culling control
Casting speed control
Automatic casting control
(Sato, et al, Kawasaki Steel Giho Vol.l4, No.3, 1982)
10 20 O verview of continuous casting process functions
' - - - -- -f measurement
Slab surface temperature
'I
Fig.
1 measurement in tundish
. . r - - - - 1 Molten steel temperature
~ 1'-1
L____J:-,.
Flow-out slag detection
Spray water comrol
Break-o ut prediction
Electromagnet ic brea ker

(EMBR )
Stirring control
;:s
I:)
"'
;;:.
:.:;
-"'
::-:
i
-
"'
:i'
<:::::>
......
type
"'
~-
....
"''<
~
~
~-
~
;:
'<
~
::!
....
0
::!
~-
::!
"'....
Laser type
Thermocouple
type
"""
Th<nnocoupl<
Tempcratun:
old
::~~~
Mokl
Scm<,
"'"""'
r-~ny
~Follow-up
Scn-o--mcchanism
lkit1n1 col
Mold
.,~
Liquid level tracking
type
Attached detecting coil

type
type
Fixed detecting coil
Mokll ~n,eoB
Radiation type
Electromagnetic
type
Laser light, directed to the surface of molten steel,

produces a reflection from the surface. The level
of molten steel can be scaled by measuring the reflection time.
Eb = & ax X0.6
Level of molten steel is estimated from the temperature distribution pattern shown by the thermocouples installed in mold wall.
r - ray source is installed in mold wall and sensor

(scintillation counter) detects the radiat ion of r - ray
through molten steel to measure the level of molten steel.
Link arm or telescopic mechanism is followed by

servo device so that distance L, between the surface of molten steel and the detecting coil, is kept
constant. The level of molten steel is shown by the
amount of movement of the coil.
An eddy current is generated in the mold by primary flux from and oscillator. The level of molten
steel can be measured by detecting the secondary
flux caused by the eddy current because the eddy
current reflects the level of molton steel.
Level of molton steel can be measured by amplify.

ing a detected change in impedance due to the
change of distance between the coil and the surface of the molton steel.
Table 10.2 Comparison of different types of molten steel level meters
~
.......
portion to the casting speed, and the coefficient of proportional equation is determined by computer based on feedback from the surface
temperature of the slab. While it is correct to say that the volume of
spray is held proportional to the casting speed of the process, the proportional coefficient must be adjusted for various conditions. The
three main variables that cause the proportional coefficient to be adjusted are discussed below. (Refer to Fig. 10.21)
Event occurance poinlS
( a) Mold widlh change
( b) Different kinds of steel
(c) Mi cellaneou events
Fig. 10 21 Spray con trol tracking
(i) Mold width (shape) : Even with the same kind of molten steel, if
the cross-sectional shape of the slab changes (mold change), the volume of spray also has to change. Since mold widths can be readily adjusted today, the flow coefficient is successively adjusted for the
different width slab that passes through the spray cooling zone.
(ii) Different kinds of steel: When different kinds of molten steel are
poured or even when the molten steel is poured from a new ladle,
identical casting conditions must be maintained while the same kind
of steel is being used. What appears to be a continuous slab, is actually made up of sections of different steels. Thus, at the boundary
where different steels pass through the spray cooling zone, the flow
coefficient must be successively adjusted.
(iii) Miscellaneous events: For example, when the immersion nozzle or
tundish are being replaced, injection of molten metal from ladle to
tun dish must be stopped momentarily. When this occurs, withdrawal
from downstream molds must also be temporarily stopped. After replacement is completed, pouring and slab withdrawal are resumed.
The volume of cooling water is constricted during the interruption,
but even so a section of slab receives excess chilling. Thus, the
spray cooling control must make an adjustment recognizing the
641
points where such events have occurred when the operation starts
up again (see Fig. 10.21). The essential point is that the slab must be
subjected to the same processing even if the casting operation has to
be temporarily halted to accommodate miscellaneous events.
This need to recognize different situations and adjust the quantity
of spray as the affected slab passes through the spray cooling zone is
called spray cooling tracking. The conventional method of handling
this tracking function has been to use a process computer and to adjust
the instrumentation set points as necessary. In recent years, however,
as digital instrumentation systems have become generally available,
plants where this function is handled exclusively by instrumentation
have increased. In such cases, sufficient operational data must be incorporated in the system to manage and adjust the amount of spray for
any event that might occur. At the same time, every spray control
loop must be equipped with a setpoint control (SPC) function.
(d) Instrumentation considerations
If cooling water for either molds or
(1) Cooling water maintenance
that the whole continuous casting
mean
can
this
spray is interrupted,
of the water control sysreliability
The
down.
shut
be
operation has to
attention in the design
close
especially
given
be
tem therefore must
is made for unlikely
provision
that
important
stage. It is equally
example, the design
for
events that could occur in the control system;
to insure an adsupply
water
should incorporate a system of values and
equate flow to keep the operation up and running.
Electromagnetic flowmeters are generally selected for this purpose because they perform better in cooling water environments and
operate with little pressure loss. To obtain precision cooling, gas-atomized cooling (a mixture of air and water) is applied to the slab as a
secondary cooling. For the air source, a vortex flowmeter is used
because it offers an especially wide range for this application.
As digital control systems have found more and
(2) Control cycle
more applications, the discrete-time control has commonly been based
on a standard 0.1-to-1-second control period. For the mold molten-steel level control, however, which is a feedback control function,
the control period has to be under 0.2 seconds to achieve satisfactory
results. Also, to control mold width adjustment and cutting equipment,
a control period of 0.05 to 0.1 second is required.
10.2.4 Instrumentation for an electrolytic galvanizing line
(a) Equipment overview
Slabs produced in a continuous casting machine are then processed
by rolling into various shapes depending on how the finished product is
going to be used. In this section, we will be primarily concerned with a
642
line set up to apply zinc to the surface of cold-rolled strip (i.e., surface
treated steel sheet) to produce anticorrosion steel sheet. Such facilities
are generally referred to as an electrolytic galvanizing line (EGL).
Figure 10.22 summarized the more common types of surface- treated steel sheet. Especially in recent years, numerous types of
electrolytic alloy galvanized steel sheet have been developed that satisfy the anticorrosion standard commonly known as the Canada code.'
The basic equipment that goes to make up a typical EGL is shown
in Fig. 10.23. The steel sheet enters the line in the form of coil, and
then undergoes continuous galvanizing. Here we will provide an overview of the main EGL equipment functions (excluding the entry and deSurface treated steel sheet
Galvanized steel sheet
Galvanized sheet (JIS G 3302)

Hot dip galvanized steel sheet
(Non JIS specification)
Alloyed hot-dip galvanized steel sheet
Hot-dip zinc . aluminium alloy galvanized steel sheet
Electrogalvanized steel sheet (JIS G 3313)
Electrolytic alloy galvanized steel sheet
Aluminized sheet (JIS G 3314)

Terneplate
Tinplate (JIS G 3303)
Tin-free steel (JIS G 3315)
Coppered steel sheet
Pre-painted steel sheet
-c
Pre-coated galvanized sheet

Welding pre-painted sheel sheet
Fig. 10 22 Classfication of surface-treated steel sheet
Entry section
<D
Pretreatment
Payoff reels
Entry shear
Welder
@ Entry looper
Plating section
Electrolytic cleaning
@ Pickling
IJ) Chemical treatment
@ Tension leveller
Post-treatment section
Delivery section
Delivery looper
Inspection table
@ Delivery shear
@ Tension reel
Fig. 10 23 Typical EGL equipment
* Canada distributes a huge quantity of salt on its road during winter to keep
them ice - free, therefore Canada's anticorrosion standard is especially rigorous.
10.2 Instrumentation Applications in the Iron- and Steel- Industry
643
livery sections).
(1) Pre-treatment section
Electrolytic cleaning: Oils adhering to the surface of the steel are removed in this section. A solution of sodium hydroxide is used for the
electrolyte.
Pickling: In this section, rust that has accumulated on the surface of
the steel is removed. A weak acid solution is the cleaning agent.
(2) Plating section
This is the section where the actual electroplating takes place. Note that most lines are not limited to a single plating
material; various types of metallic coatings can be applied to steel
sheet by changing the plating solution.
(3) Post-treatment section
After plating, the sheet metal undergoes
further surface finishing in this section. For example, a phosphating
treatment might be applied to the sheet as an undercoat prior to painting, or a special organic resin might be applied to surfaces that will
not be painted.
The equipment for the plating and post-treatment sections varies
considerably because the specific configuration largely depends on how
the surface-treated steel sheet is going to be used.
A schematic of a typical plating section is shown in Fig. 10.24.
Plating solution (an acid solution containing ions of the plating metal),
circulates between the circulation tank and the plating bath. As the
sheet is introduced to the plating bath by the contact roller, the sheet
acts as a cathode and positive ions of the plating metal are deposited
on the surface in a solid state. Essential equipment associated with the
circulation tank includes: (a) ion-supply equipment to replenish ions of
the plating metal in the circulation tank, (b) impurity removal equipment to eliminate contaminants that would have an adverse effect on
the plating, and (c) plating solution exchange equipment that is needed
Fig. 10 24
644
Conceptual configuration of plating sect ion
to clean the plating bath and circulation tank when changing over to a
different plating solution.
(b) Unique demands on EGL instrumentation
Two characteristics of EGLs create special demands on the instrumentation. First, the plating equipment must accommodate various
plating solutions (from strong base to strong acid), and second, a great
number of valves and pumps are involved in switching between plating
solutions. As far as the instrumentation is concerned, this means that
(1) special care should be taken in selecting types and materials for
fieldmount devices that will hold up well in a harsh environment, and
(2) with so many pumps and valves, it is necessary to clearly define
and compartmentalize which monitoring, operation, and interlock functions are to be handled by instrumentation and which are to be handled
by electrical equipment.
This second point especially should be given very careful consideration because how the question is answered will have a major impact on
the number of digital input and output points and also on the sequence
volume. Note too that if the plating solution cutover sequence is automated, this will also require considerable sequence volume. This is assuming the present trend continues of concentrating the monitoring,
operation, and interlock for all devices in the human-machine interface
of instrumentation.
(c) Instrumentation overview
In contrast to ironworks equipment, it is difficult to standardize
the instrumentation for application to EGL. This is particularly true
for the plating and post-treatment sections, where each steel manufacturer has developed individual product lines that require unique facilities and instrumentation. In our treatment here, however, we will only
be concerned with instrumentation that is common to different applications.
(1) Electrolytic cleaning unit
With respect to the circulation tank,
control must be maintained over temperature, concentration and solution level.
For detecting the concentration of sodium hydroxide solution, usually an electromagnetic concentration meter (specific conductivity meter) is employed. However, since it is not a pure solution, the objective
is not so much to obtain an absolute concentration, as it is to derive a
concentration range that can be used as a standard for the control
limit. Since fast response is not required, the concentration can be adjusted by on/off control of the solution and water.
Quite a number of level-control methods have been devised. One
typical technique is to charge a fixed proportion of water and solution
once the low level limit is reached.
645
Concentration control and level control tend to counteract one another, but precision is not overly important in this context. The only
essential thing is that the concentration control must be cut off as a
safety measure when the upper level limit for the equipment is
reached.
(2) Pickling unit
Basically identical to electrolytic cleaning unit, the
main control parameters for the circulation tank are temperature, concentration and solution level.
The conventional method of obtaining an approximate value of concentration cis by employing the first-order equation
c = K 1 d+K2 p+K3
where d is the reading from an electromagnetic concentration meter, p
is the reading from a density meter, and Kit K2 , and K 3 are constants.
Here too, however, as was previously the case when we measured the
concentration of sodium hydroxide solution in the electrolytic cleaning
unit, one cannot expect to obtain absolute concentration. More recently, automatic titrators and a gamma-ray concentration meters have
begun to be applied as on-line analyzers for concentration control.
Level control too is essentially the same as described above for the
electrolytic cleaning unit. One minor difference is that when automatic
mixing is applied to the solution, the exothermic reaction of the sulfuric acid must be factored in when setting up the sequence.
(3) Plating section
In addition to the continuous-control functions
that are more or less the same as in the other sections, there are two
control functions that are unique to the plating section: plating current control and contaminant removal control. Also, as was mentioned
previously, relatively large-scale sequence control programs are required to switch back and forth between single- and double-side plating and to control plating solution changeover.
Control of zinc plating solution has conventionally involved regulation of the pH value and the density of solution, as shown in Fig.
10.25. As can be seen in the figure, the characteristics of the plating
solution shift from the upper right in the figure toward the lower left
as the plating proceeds (i.e., from high density and pH to low density
and pH). Control is thus necessary to restore the solution to its original condition by supply additional ions of the plating metal.
On-line analyzers such as automatic titrators and fluorescence
x-ray analyzers have been widely applied to most electrolytic alloy
galvanizing processes; they generally work by adding either acid or
metal ions as required to restore the solution.
(d) Instrumentation considerations
(1) Fieldmount devices
An obvious but important consideration is
646
Chap.lO
I
I
r+
f.--1
"'-
Ho
-~
Lo
---t-~kAddltlonal
waq---r---
Dissolution
---i-c:
advan~~ating
progress
Additional
acid~D--~---
I
I
-pH
( Within solid-line border=control range)

Within chain-line border=control limit
Fig. 10 25 Concept of concentration control for plating solution
that fieldmount devices are constructed out of materials that are not
going to be corroded by the strong solutions with which they come in
contact. This corrosion factor is one reason purge-type detection elements have been so pervasively used for detecting level and density.
When the purge tap (also known as a bubble tube) is made of resin, it
is important to avoid long lengths (say, longer than 2 .m) of unreinforced tube. A splash proof tube should also be installed to guard against
rough agitation in the circulation tank. Finally, the bubble tube should
not be subjected to heavy loads.
In using a purge-type density gage, if the change in density is so
small that an elevation function alone doesn't produce sufficient accuracy above zero on the differential-pressure transmitter, one solution is
to use a pressure adjusting tank (reference column), as illustrated in
Fig. 10.26. With this kind of arrangement, the differential pressure detected by the gage is given by
~--------~----+-~--AS
Splashproof tube
+---As
.--""'-tt-th.L----,
PA, PB :
0
PA
Density
Pressure adjusting tank
Fig. 10 26 Purge-type density measurement using pressure adjusting tank
647
JP
= PAh.u-(PAhAH+ PBhB)
= JhpA- PBhB
and it is possible to eliminate the bias value PBhB.

One final consideration is that if bubbles or turbulence are present
on the surface of the solution, ultrasonic level gages cannot be used.
(2) Plating solution changeover
A sequence must be implemented to
make sure the solutions are kept separate when changing over to a
different plating solution. This involves verifying that valve A is fully
closed before opening valve B.
(3) dumping the plating bath solution
If sheet steel is left submerged
in a strongly acidic solution for any length of time (e.g., in the pickling
and plating sections), there is a danger of the sheet being melted. To
prevent this from occurring, an alarm should be implemented that is
triggered a certain interval (say, 20 to 30 minutes) after the process is
stopped with the steel sheet left in solution. Finally, as a precaution in
case the instrumentation system malfunctions or fails, provision should
be made so the plating bath can be easily dumped manually.
10.3
Instrumentation Applications in the Power Industry' 6)
10.3 .1 Overview
The power industry consistutes a large-scale system encompassing
hydroelectric, thermal, and nuclear power generation plants; transmission and substation facilities to distribute the power; and of course the
factories and households that ultimately determine the power consumption load. A particular characteristic of the industry is that, while
demand for power fluctuates, no means of storing electricity in commercial quantities has been devised, and as a consequence, generated
power has to exactly respond to consumption demand. The implication
for power plant instrumentation is that control results have to be extremely precise and responsive.
Today, hydroelectric power represents only about 6 percent of the
power supplied in Japan and will therefore not be considered; our examples of power-plant instrumentation will be taken exclusively from
thermal and nuclear power plants.
10.3.2 Thermal power plants
Conventional thermal power plants generate electricity through
the combustion of coal, petroleum, and natural gas. Since these fuels
are all derived from fossilized organic material laid down over the past
600 million years, this type of power plant is also commonly referred
to as a fossil-fuel plant, a designation that clearly differentiates it
648
from nuclear power plants'.

The working medium to effect the energy conversion from fuel to
power in a thermal power plant is steam. A boiler produces steam, the
steam turns a turbine, and the turbine turns an electric generator. After the steam has performed its work in the steam turbine, it is converted back to water in a converter and returned to the boiler. An
overview of a typical thermal power plant is shown in Fig. 10.27.
CD
CV
Coal yard
Chimny
Coal conveyer
Coal bunker
Boiler
Dust collector
(J) Ash handling system
Turbin-generator room
Transformer
Sub-station
@ Circulating water intake

@ Circulating water discharge
Fig. 10 27
Bird's-eye view of a thermal power plant

(Courtesy Mitsubishi Heavy Industries, Ltd.)
10.3.3 Boiler control

Fuels burnt in the boiler impart thermal energy to the boiler water and the flue gas produced is discharged into the atmosphere
through a chimney. Feedwater in the boiler is evaporated by the heat
generated by the burnt fuel and leaves the boiler in the form of steam
(Fig. 10.28).
The fundamental control system of the boiler is designed to maintain a mass balance of water and to maintain an energy balance by
burning fuel in the boiler.
These two basic objectives are realized through three major control systems in addition to the steam temperature control system.
These are:
The quantity of working fluid is kept
(1) Feedwater supply control
constant (mass balance). Drum water level is the control target for
drum type boilers.
Steam evaporation is adjusted to meet the
(2) Combustion control
* Some thermal power plants burn diesel oil for fuel, but since their output is
small and their control systems relatively simple, they are also excluded from cons
ideration here.
10.3 Instrumentation Applications in the Power Industry
649
Flue gas
flow
Steam
now
Boi ler furnace
I\
:I '1
Air Fuel Feed

now now water
now
Fig. 10 28
Energy and mass flow in boiler
load (energy balance). Steam pressure is controlled at the boiler outlet

(superheater outlet).
(3) Steam temperature control
The steam temperature at the superheater (and reheater) outlet of the boiler is kept constant (emergency
balance). Steam temperature at the superheater outlet is kept constant
(this is also true of the reheater outlet in the case of reheat boilers).
However, since boilers exhibit complex nonlinear behavior, they
are in fact quite difficult to control. To achieve good control results,
boiler control systems have become extremely large and complicated.
Application of adaptable estimate control methods based on advanced
control theory has contributed to this development. The analog control
systems of the past are no longer adequate for coping with present- day control problems.
Recent progress in microprocessor techniques, has led to the development of digital control systems (DCS) that greatly extend automation for thermal power plant control. Moreover, since the quantity and
quality of information required by CDSs has increased markedly, computer terminals and associated software have also come to play and increasingly important role for the man-machine interface in thermal
power plant control. Figure 10.29 shows a t ypical CRT operator console.
Boilers used in thermal power plants are classified into two types:
the circulation type (also known as the drum type, because they employ drums) and the once-through type. The following explanation of
boiler control systems is based on the circulation type.
650
Fig. 10 29
RT Operatcr conso le
(a) Automatic combustion control (ACC)
The purpose of boiler combustion control is to maintain constant

steam pressure at the boiler outlet so that the fuel flow satisfies steam
evaporation requirements. Fuel combustion in the boiler furnace requires that an adequate amount of air be supplied to the furnace. Combustion control can thus be broken down into fuel flow control and air
flow control.
(1) Fuel flow control (steam pressure control)
In fuel flow control, as
shown in Fig. 10.30(a), steam pressure is detected and compared
against a setpoint value, and the fuel flow is controlled so as to minimize the difference between the detected pressure and the set pressure. The steam pressure signal is called the master signal, and the
control signal that is compared against the setpoint value is called the
master control signal. Since boilers contain water of a large specific
Main steam
pressure
Steam
flow
Steam
pressure
Fuel
flow
ignal
Set point
Master control signa l

( a ) Master control circuit
Fig. 10 30
10.3
( b ) Fuel now control circuit
ucl now control system
Instrumentation Applications in the Power Industry
651
heat and a large amount of steel, there is a time lag (called the time
constant of boiler heat capacity, or boiler time constant) before
changes in steam flow (or fuel flow) are reflected in steam pressure
changes (see Fig. 10.31). This means that if fuel flow is only controlled
by a master signal, control deviation tends to increase. As shown in
Fig. 10.30(b), steam pressure control can be improved by adding a
steam flow signal to the master control signal as a feedforward signal
and by adjusting the fuel flow beforehand to meet the change in steam
flow. The master control signal is transmitted to the fuel flow control
(fuel flow control valve, coal feeder and so on) to control the fuel flow.
The amount of fuel is measured as a result of control and compared
with the master control signal as a feedback signal.
10
-~~I
~~
""
<:1
r--
<}
!:"~~
~~"-' I'-...
3".?~0);6At 9ooiih26ouw f=:-
0o 100
Fig. 10 31
zoo
r tr
300 400 soo Goo 100 soo 9oo
Boiler evaporation. [t/h]
Boiler time constant
After Ito and Fukada et a!.

( "Dynamic characteristics of
boiler'' in Mitsubishi Industries
Technical Report, Vol.l, No.I.
(2) Air flow control (jueljair ratio control)

Oxygen (air) is necessary
to burn fuels, and combustion conditions are greatly affected by the
air/fuel ratio. Combustion can be thought of as oxidation of fuels, and
the air flow needed for theoretical oxidation (combustion) is called theoretical air. In actual combustion in boilers, however, complete combustion is not achieved unless slightly more than the theoretical air flow
is supplied. The ratio of theoretical air flow to actually supplied air
flow is called the excess air ratio and is expressed by p. In actual boiler operation, the excess air ratio is found by the following equation,
where the percentage of 0 2 in the exhaust gas is measured.
p
= actual air flow/theoretical air flow

= 21/(21 - 02) (%)
where 21 is volumetric ratio of oxygen in air.

In boiler operation, when p decreases, heat loss due to unburned
fuel increases. When p increases, on the other hand, heat is lost up the
652
Chap.lO
1_..------ I
i'
'\
/
/IO_ .
....J:_hermal efficiency
/ 1 ptlmum 1
-1Ezone
,.:
'---. __
'\
III
-----.._
'
:Total
1heat loss _... .... .,..
~~
' K'------r-I . . . - . . . . . . .
Heat loss due t ',

the insufficient
firing
: ,
CD
Fuel/ Air ratio

Increase CO and smoke density
Increase NOx. SOx
I \Oss
y..e?-'I
-t------
!J) Over excess air

Extreme low excess air
Low excess air
Fig. 10 32 Relationship between excess air ratio and combustion status
chimney, with a resulting decline in boiler thermal efficiency. This relation is shown in Fig. 10.32. From this figure it is seen that maximum
boiler efficiency ranges from 1.02 to 1.10 of p. for oil-fired boilers (1.20
to 1.30 for coal-fired boilers).
The basic air flow control system is shown in Fig. 10.33(a). In actual boiler operation, the change of air flow supplied by forced draught
fans (FDF) in the course of a load change is retarded than the change
of fuel flow (due to the response speed lag of the control vane and air
flow lag). Hence, when the boiler load increases, p. decreases in the
form of a transient with the possibility of black smoke shooting up.
Most cross-limit combustion control systems such as shown in Fig.
10.33(b) use a selective relay of low and high signals, to ensure a constant and sufficient supply of air even in the event of transient conditions during load changes.
The master combustion control signal is compared with the air
flow signal by a low signal selective relay and whichever value is
found to be smaller is transmitted to the fuel control valve. Thus, even
though the boiler load increases and a signal indicating a fuel increase
is received, the output of the low signal selective relay does not increase because the air flow still has not increased. In the case of a high
signal selective relay, the rise of the master control signal together
653
Steam
now
team
pressure
Fuel
now
Air
now
ra1 io setpoim
(a )
Steam now
comrol valve
Air now con1rol vane
undamemal system of air now con1rol
Steam press ure
Air flow
Fuel now
Fuel/ Air ratio

set point
<D
Low signal selective relay

High signal selective relay
control valve
Air now comrol vane
( b ) Cross limit air now control system

Exhaust gas CO
<D
Exhaust ga 0 2
Steam now
Dead
To air flow controller

(c) Low 0
air now control system performed with set valve correction using CO value
Fig. 10 33 Air flow control system
654
Chap.lO /nstrumention to Manufacturing Industries
with load increase is conveyed to the air flow control vane and the air
flow increases first. Then, the output of the low signal selective relay
increases to correspond to the increase in air flow, thus resulting in an
increase in fuel flow. A bias signal is added to the air flow, and fuel
flow signals of each selective relay define the fJ. fluctuations during the
change. Contrary to the other operation described, air flow cannot be
decreased with load decreases, until after the fuel has been decreased.
Recently, reliable combustion gas analyzers (0 2 , CO) have been developed, so that air flow control employing 0 2 values to perform the
setting and CO values to correct the setpoint values is also used (Fig.
10.33(c)).
(b) Feedwater control (FWC)
Feedwater control is used to keep the quantity of water in the boilers constant. In drum type boilers, the water level in the drum is used
as an index of the amount of water in the boiler. Feed water flow is
generally controlled by feedwater control valves or by the speed of
feedwater pumps. As shown in Fig. 10.34, however, there is a transient characteristic (termed the reverse response) of the water level in
the drum* such that the water level rises transiently as steam flow increases. Feedwater control that only relies on detecting the drum water level, therefore, cannot produce good results. To eliminate the reverse response effect, steam flow is used as a disturbance compensation signal, and feedwater flow is used as a feedback signal. This type
of feedwater control system is shown in Fig. 10.35. Because this conLarge
Steam flow
Time
Fig. 10 34 Dynamic characteristics of drum water level
* As the turbine load increases, the steam flow from the boiler increases and the
boiler pressure decreases. This causes steam bubbles in the generating tubes to increase due to the self boiling of saturated water. Consequently, the drum water
level rises transiently with the boiler water being pushed up into the drum by increased steam bubbles. Conversely , when the turbine load decreases, the opposite OC
curs; feedwater flow increases, boiler water in the generating tubes cools, and the
steam bubbles decrease. As a result, a transient drop in the drum water level occurs even if the feedwater flow increases.
655
Main
Feed
Drum
water level water now steam now
etpoint
Feed water control valve
Fig. 10 35
Feed water control sy tem
trol system uses drum water level, steam flow, and feedwater flow as
control elements, it is termed the three-element control system. During
boiler startup, the flow measurement of both the feedwater and the
steam flow are inaccurate, and therefore only the drum level signal is
used for this drum level control system.
(c) Steam temperature control (STC)
Changes in steam temperature can have a great impact on turbine
performance as well as on boiler and turbine service life.
Main steam temperature control
(1) Main steam temperature control
is, for the most part, performed by spray water injection into a desuperheater installed at an intermediate point in the superheater. In this
case, to compensate for the response lag due to thermal inertia in the
superheater, and the transfer lag due to steam flow, the temperature
of the main steam before the spray water injection is detected for cascade control (Fig. 10.36(a)). In the case of large- capacity boilers that
operate at high pressures and temperatures, the above lags become
larger. This necessitates high- temperature resistant superheater tube
materials and the installation of more than two spray point stages.
Since steam temperature exhibits a functional relation with boiler
load (Fig. 10.37), steam flow is used as a lead signal to obtain better
control results (Fig. 10.36(b)). *
Temperature control by spray
(2) Reheat steam temperature control
* The control of steam temperature is extremely difficult for a number of reasons;

(i) the time constant of a heating surface is large and incorporates a transportation
lag, and (ii) the characteristics of heat transfer depends on the cleanliness of a heating surface. Therefore, a predictive control system has been developed that works
quite well. The system performs advance control by predicting changes in steam
temperature from boiler conditions (boiler load, cleanliness of heating surface and
so on) using a mathematical model of the boiler in a computer.
656
Desuperheater
etpoint
<D
Primary superheater
Secondary superheater
Spray control valve
( a ) Fundamental sy tern of steam temperature control
Primary
desuperheater
Secondary
desuperheater
Q) Primary superheater
Secondary superheater
Final superheater
Primary spray
control valve Secondary spray
control valve
( b ) Steam temperature control system (Cascade control)
Fig. 10
36 Steam temperature control sy tern
water injection is not used* for reheat steam because this would degrade the efficiency of the Rankine cycle.
The control of reheat steam temperature, as shown in Fig.
10.38(a), (b), and (c), includes gas recirculation, burner tilting (the aim
of both is to determine the heat absorption increase in the reheater)
and gas distribution damper methods (by changing the heat distribution between the superheater and reheater). Here we will explain the
gas recirculation control system (Figs. 10.38(a) and 10.39), since this is
the most widely used method.
When a portion of the low temperature gas at the economizer out-
* A spray control system is installed for the reheat system, but it is only used in
emergencies.
657
c,.
10
~ 90 -""'
:1'
"C
i1
8.
60
50
40
il
.s::
50
r-- ........-A ......... ...-I.e:._
B~
60
70
:-.......
............
90
80
!00
Boiler load (%)
A: Radiation superheater, B: Convection superheater, C: Radiation/Convection superheater
Fig. 10 37 Temperature characteristics of superheater
(Reprinted from JSME mechanical engineers handbook :

power plant engineering (1987))
(a) Gas recirculation method
Fig. 10 3
( b) Burner tilting method
(c) Gas d istribution damper method
Reheater steam temperature control methods
Main
steam flow
Reheater out
steam temperature
GRF
motor current
Set point
GRF
Control d umper
Spray
control valve
Fig. 10 39 Reheater steam temperature control system
let is blown into the furnace bottom by a gas recirculation fan (GRF),
the gas temperature at the furnace outlet rises due to the reduction of
radiation heat absorption in the furnace. By coupling this with the heat
658
Chap.lO
absorption increase due to gas flow speed increase while passing

through the reheater, the steam temperature at the reheater outlet can
be increased.
Because the steam temperature at the reheater inlet depends on
turbine load, a steam flow signal is used as feedforward signal as an index of the lead. As gas recirculation (GR) flow increases at low load
and the ratio of GR to combustion air flow becomes excessive, an unstable combustion condition occurs. Therefore, programming control
system to decrease GR at low load is used in conjunction with a low selection relay to prevent overload of the GRF motor due to drops in GR
gas temperature.
(d) Automatic burner management system (ABMS) and boiler protection
system
Because the operation of burners have such a crucial role in overall boiler operations, and because misoperation of burners accounts for
such a large proportion of serious boiler mishaps, most boilers are
equipped with an ABMS to improve safety and reduce labor requirements. This system, including the furnace purge system before
light-off, is so important for boiler operation control that it is designated the furnace safeguard supervisory system (FSSS).
(1) Automatic burner management system (ABMS)
The quantity of
fuel atomized at the burner is proportional to the pressure of the fuel
entering the burner. If the fuel pressure at the burner inlet reaches
the limit defined by the delivery press of the pump, the fuel pressure
required during the fuel flow increase can be reduced by increasing the
number of burners. If, on the other hand, in the case of a reduction in
fuel flow, the fuel pressure at the burner inlet drops below the limit
due to throttling by the fuel control valve, this adversely affects fuel
atomization and combustion efficiency deteriorates. The fuel pressure
therefore should be increased to the desired level by reducing the number of burners (Fig. 10.40).
Number
of burners
Upper limit pressure
r-----~-r~~~--~~~
Number of burners decrease

Lower limit pressure
---Fuel now
Fig. 10 40 Relationship between number of burners and fuel pressure
659
In the ABMS, burner inlet oil pressure is detected from the relationship of boiler load (fuel flow) and the number of burners to determine the necessary number of burners. Burners are sequentially' fired
according to a command signal from the ABMS (Fig. 10.41).
Fig. 10 41
Burner operation block diagram
Boiler protection systems safeguard

(2) Boiler protection system
against a number of dangerous operating conditions, and include a master fuel trip (MFT) system to rapidly cut off an entire fuel source (Fig.
10.42), a furnace purge system to prevent a detonation of unburnt
gases or fuels remaining in the furnace or flue gas passage when igniting the boiler, (Fig. 10.43) and a fuel leakage testing system to detect
for fuel leaks in the shut-off valves (Fig. 10.44).
(e) Other controls
Boiler plants employ numerous other control systems, but here we
will mention only two; furnace pressure control and steam air preheater temperature control.
With low-pressure fuels such as gase(1) Furnace pressure control
ous fuel or pulverized coal, the pressure in boiler furnaces is controlled
* ABMS and FSSS are open loop controls that are controlled by sequential logic,
whereas the automatic control mentioned before is a closed loop control.
660
Chap.l 0
Unit trip push button

Both FDFs stop
Both IDFs stop
All flame out
MFT Operate
Condenser vacuum low
MFT Reset
Air flow smaller tham 25% of MCR

Drum water level low
Furnace pressure abnormal
Unstable combustion
Reheater protection fault
Fig. 10 42 Block diagram of master fuel trip system

Torch fuel shutoff valve close
Warm up fuel shutoff valve close
Heavy oil shutoff valve elose
All pilot torch inlet valves close
IAll
pilot torches shut down

flame detectors no-detect
Torch fuel header pressure normal

Warm up fuel header pressure normal
IHeavy oil header pressure normal

All FDFs running
IAll
IDFs running
All GRFs running

Air flow larger than 30 % of MCR
Fig. 10 43 Block diagram of furnace purge system

Air flow larger than 30 % of MCR
All main burner inlet valves close
Timer
2 min
r-o""'i-1-sh-u-to-ff-va-lv-e..,
open for 15 sec. 1---t--i-o-t-...,-,
Main oil header pressure normal

Main burner control valve open
Fig. 10 44 Block diagram of fuel leakage testing system
at a level somewhat lower (i.e., 5 to 10 mmH 2 0 lower) than atmospheric pressure by induced draught fans (IDF). A master control signal is
added to the furnace pressure control signal as a feedforward signal
(Fig. 10.45).
661
Furnace pressure
IOF l nlel da mper
Fig. 10 45
Furnace p re ure control y tem
(2) Steam air preheater (SAH) temperature control

Air preheaters
(AH) are used to improve combustion and to increase boiler efficiency
by heating combustion air using boiler flue gas. As the boiler load decreases, however, flue gas temperature falls and so3 in the exhaust
gas corrodes the AH. To prevent AH corrosion, a SAH is installed in
the air duct at the AH inlet to prevent the AH temperature (represented by an average air temperature at the AH inlet and gas temperature
at the AH outlet) from falling below the dew-point temperature of S0 3
(Fig. 10.46).
SAH Temperature
con1rol va lve
Fig. 10 46 Steam ai r prehea ter temperature cont rol system
(f) Supercritical pressure boiler control
Supercritical pressure boiler* control can be explained by envisioning a tube ; water is fed into one end of the tube, where the water is
heated, and steam is produced from the other end. In this simple system, the evaporation quantity (corresponding to steam pressure) depends on feedwater flow. When feed water flow is kept constant and
the firing rate is increased, the end point of evaporation moves toward
Critical condition of the steam is 225.65 kg/cm 2 and 374.l5C.
662
Chap .lO
Jnstrumention to Manufacturing Industries
to the feedwater inlet, thus expanding the superheating area. The

steam temperature is controlled by the firing rate, but since the response is slow, it is used in combination with temperature control by
spray.
Figure 10.47 compares the above control systems with drum type
boilers.
Once through type boiler
Drum type boiler

Main steam pressure
Fuel control
Superheater spray
control
Drum water level - -- Feed water control
Main steam temperature
Feed water control

Main steam pressure Main steam temperature - Fuel control + superheater
pray control
:-----
r-- - ---.,
I
-~- -----
------------,
I
I
I
I
'------,I
I
I
~----=---:--i
Fuel
CD Secondary superheater
Prima ry superheater
Fig. 10 47
BFP
-~
@ Furnace
@ Furnace
Economizer
Comparison of control sy tern drum type boiler VS once through type boiler
10.3.4. Turbine control

The steam turbine plant consists of a turbine as the main machine,
a condenser, feedwater heaters, a deaerator, boiler feed pumps, a circulating water intake system and a demineralized water treatment system. The controls for some of these systems will be considered here.
(a) Turbine governor control
The generator is driven by the turbine. If the supplied energy
(the energy inherent in the steam) does not correspond to generator
output, turbine shaft torque will be insufficient (or excessive) and rpm
will fall off (or accelerate). This is illustrated in Fig. 10.48. If turbine
load increases from P 1 to PR and rpm decreases from NR toNs, and the
aperture of the steam governing valve adjusts from A to B. This
663
1
tl)
''
Ns
--------- }------ ---------1
'
p,
Fig. 10 . 48
p
Output
haracteristic of turbine peed governor
Governor
Governor motor
..
Weight
Fig. 10. 49 Operation principle of peed governor
brings the output in line with load PR, and equilibrium is restored.
Then to bring the rpm back up to NR, the governor motor controls the
steam governing valve so it adjusts to point C (refer to Fig. 10.49).
This relationship between turbine output and rpm is called speed
regulation, and generally ranges from 4 to 5% for utility thermal power plants. Speed regulation, Rs, is given by
Rs= NoNRNR x 100 (%)

Here, N 0 is turbine rpm under a no- load condition, and N R is turbine
rpm under a rated load condition.
The above explanation applies to how the turbine main machine
control device (turbine operating governor) regulates the functions of
speed and load. Other essential control functions of the operating governor are speed control and emergency shutdown in the event it is nee664
essary to dump the load.

There are two principal types of operating governors. Mechanical
hydraulic (MH) operating governors, which have been in common use
for a long time, and electron hydraulic (EH) operating governors,
which have come into fairly wide-spread use recently and are better
adapted to the increased scale and more diversified operation of modern turbines.
(b) Condenser hot-well level control
The condenser hot-well level is determined when exhaust steam

from the turbine and feed water to the boiler are in balance. However,
when the condenser hot- well level falls due to blow-off from the boiler
or when extraction from the system disturbs the balance, the make-up
water control valve (CV - 1) operates to return the hot- well to its normal level. When the level becomes abnormally high, the spill-over control valve (CV-2) operates to return the condensate to the make-up water tank (Fig. 10.50).
CV - 3
IDeaerator level
control valve
Turbine exhaust
ondensate
recirculation
control valv
j___lj
low
Boi ler feed

water pu~e
pressure
feed water
heaters
le
CV - 4
-
Make-up
water tank
BFP
CV - 2
Condensate spill-over
control valve
Fig. 10 50
Level control system around the condenser and the deaerator
(c) Deaerator level control
The deaerator holds approx. ten minutes of maximum continuous

boiler evaporation feedwater and removes oxygen or carbonic acid gas.
The deaerator also serves to cushion fluctuations in boiler feedwater
during load changes.
Deaerator level control is performed by the deaerator level control
valve (CV- 3), but when the minimum water flow necessary for the air
ejector or gland condenser (for cooling leak-off steam escaping from
665
the shaft seal of turbine) is in short supply, the condensate recirculation control valve (CV-4) operates to maintain sufficient water flow.
CV-4 is sometimes directly controlled by a condensate flow meter, but
in the arrangement shown in Fig. 10.51, the deaerator level control signal is set at a split range to operate CV-3 and CV-4 in reverse.
Deaerator level
Fig. 10 51
Control characteristics of the deaerator level control and

the condensate recirculation control
(d) Turbine and generator protection system
In the event the turbine or generator begin to operate abnormally,

all steam valves supplying steam to the turbine are abruptly shut to
stop the turbine and thereby protect it and the generators. A block diagram of the system is shown in Fig. 10.52.
Manual stop
Turbine over speed
Turbine bearing abnormal
Condenser vacuum low
IBearing oil pressure low

IExcessive shaft vibration
EH governor power lost
ISpeed signal lost
EH governor oil pressure lost
lGenerator internal
defect
Generator overcurrent
Main transformer internal defect
House service transformer internal defect
Bus defect
Turbine trip
Fig. 10 52 Turbine-generator protection block diagram
666
Chap.lO
10.3.5 Power plant system control

Electric power systems generally attempt to mm1m1ze frequency
deviations in order to maintain the speed of motors and the accuracy
of electrical equipment such as computers. Electric power generating
systems in power plants, on the other hand, are directly connected to
the power demand load. Such systems must accurately control the generation of electricity in accordance with load changes, as well as work
to enhance the efficiency of the entire power plant. Various measures
are implemented to facilitate such functions.
(a) Coordinated boiler/turbine operation
To deal with load changes, the power system controls the output
of the generator in accordance with fluctuations in the load. In other
words, the control adjusts steam flow into the turbine, and a change
command is given to each power plant from a load dispatch control center which controls the overall electrical generation of the total network. In order for each individual power plant to respond efficiently to
this change command, the following operational control methods are
implemented.
(1) Sliding pressure operation
In power plants, boilers are normally
operated at a constant pressure, and the load is controlled by manipulating the turbine governing valves. Therefore, except for the case
when the steam governing valves are wide open, a drop in turbine efficiency occurs due to the throttling action of the valves (Fig. 10.53). In
the sliding pressure operation method, main steam pressure is set approximately in proportion to load (Fig. 10.54). This way, the governing
valves can be almost completely opened regardless of load, and a drop
in turbine efficiency due to the throttling action of the governing
valves can be prevented. In a sliding pressure operation, the boiler/turbine combination that yields the optimum pressure condition is determined. That value is then used as the master signal setpoint in the program.
(2) Coordination of boiler/turbine control
The control of boilers and
,-, 100
~ ~
- ; \-1
>
GV-1
.
01)
.5
-tl "
bQ
c c
c c.
0
>
0
Turbine load [%)
Turbine load [%)
Fig. 10 53 Turbine governing valve opening and turbine efTeciency

(In case of four GV)
667
100
0
O utput (%]
Fig. 10 54 Set pressure of main steam at the sliding pressure operation

Table 10 . 3
ontrol
method
o ntrol method of boiler/ turbine block
Bo iler - follow mode
Turbine- follow mode
Load comroi Turbine

(Govern or)
Main steam " Bo iler
(Fuel)
pressure
co ntrol
Load conlroi .. Boiler

(Fuel)
Main steam .. .Turb ine
(Governor)
pressure
control
Control
y tern
Fu nda mental
comrol
function
Load comrol.. Turbine

(Governor)
Boiler
(Feed water.
Fuel)
Main steam ... Boiler
(Feed water,
P<'essure
Fuel)
control
turbines in thermal power plants includes boiler-follow, turbine-follow, and boiler/turbine coordinate modes (Table 10.3).
In the boiler-follow mode, a megawatt demand signal (MWD) is
transmitted to the turbine governor to control steam flow into the turbine, and steam pressure in the boiler is controlled by ACC. This mode
is used in large-heat-capacity drum boilers and exhibits good power- output response during load changes. Operating conditions in this
mode, however, are very hard on the boiler.
In the turbine-follow mode, the MWD is transmitted to the boiler
master control signal, and the governor controls steam flow into the
turbine to maintain constant steam pressure. In this mode, the turbine
load is changed through the time constant of the boiler, and therefore
the response is slow. An advantage of this mode is that the boiler can
be operated in a stable condition. This mode is widely applied in Eu668
rope where once-through boilers have been widely adopted.

The boiler/turbine coordinate mode is intermediate between the
other two modes already described and is employed with thermal power plants where large once- through boilers are used.
(b) Load variation and its control
The period and amplitude of load variation in an electric power system is shown in Fig. 10.55. Short period load variations of less than
about two minutes are absorbed by the turbine governor control and
load self-balance. (A good example of this principle is provided by an
ordinary motor. As the cycle increases, the speed of the motor also increases until eventually the increased load and speed settle at a proper
level). For slightly longer load variation periods of from two to twenty
minutes, automatic frequency control (AFC) is applied to control the
generation of electricity to maintain the frequency. For even longer
load variations, which are predictable to some extent, economical load
dispatching (ELD) is applied to dispatch demand to each power plant so
as to minimize fuel costs and transmission losses.
~
"'
c...l
.9
;;;
;:::
>
"0
"'0
..J
Sel
-Large
Variation cycle
Fig. 10 55 Assignment of power generation control
10.3.6 Nuclear power plant overview

(a) Nuclear reactor overview
While the use of water power and combustion of fossil fuels to
generate electricity has quite a long history, it was only in 1960 that
the first nuclear power plant began operation in Japan. At present,
there are more than 35 nuclear power plants in operation, and approximately 25% of the electricity consumed in Japan is generated by nuclear plants. A comparison of these three modes of power generation is
shown in Table 10.4.
The operating efficiency of pressurized water (PWR) and boiling
water reactors (BWR) in Japan is already extremely high at over 70%.
In the coming years, the objectives are to boost this performance even
higher and to move toward a more standardized type of light-water
reactor. Also to be noted is the dramatic progress that has been made
toward the development of a commercially viable advanced thermal
669
V)
;;fr
....
't
~-
't
.;,
"'
....
<:;
;;!
V)
;;-
<;::,
~
......
0'1
Fast breeder
reactor
Advanced thermal
reactor
Light water
reactor
Nuclear binding force

3.2 X
Binding force between atoms

6.4 X IQ l 9 (J)
3%
97%
1.5%
0.7%
97.8%
Uranium 235
Uranium 238
Plutonium 239
Uranium 235
Uranium 238
Approx.20%
Plutonium
Depleted uranium
(Blanket fuel Approx.80%
Uranium 238)
Thermal neutron
Fast neutron
Fuel
Thermal neutron
Neutron contributing
to fission
Heavy water
Light water
(Ordinary water)
Moderator
Sodium
1.2
0.7
0.6
Light water
Light water
Con ve rsion
ratio
Coolan t
}0" 11 (])
Energy y ield on fission of uranium etc.
Nuclear power generation
Oxidation reaction of carbon

molecule.
C + 0 2 -+C0 2 + heat
Table 10.5 Comparison of various types of atom ic reactors
2. 9 x 10"23 (J)
per lOOm drop
Universal gravitation
Kind of force
Energy yield per

molecu le
Change in potential energy of water

molecule (from high potential to low
potential)
Energy source
Thermal power generation
Energy from various kinds of power generation
Hydraulic power generation
Table 10.4
_Cooling water
(Sea water)
. 1 .
C1rcu
attng
pump
CD
Fig. 10 56
Steam generator
Reactor coolant pump
Pressurized water reactor (PWR)
sluice way
CD
Reactor vessel
Recirculation pump
@ Control rod
Suppression pool
(Suppression chamber)
Fig. 10 57
@ Cleanup system
Turbine
Boiling water reactor (BWR)
reactor (ATR) and fast breeder reactor (FBR). These four types of
reactors are shown schematically in Figs. from 10.56 to 10.59, and a
comparison of some of tqeir key features is summarized in Table 10.5.
Structurally, nuclear and fossil-fuel power plants have a great
deal in common; both generate electricity using steam- driven turbogenerators, and condensers and water-supply heaters can likewise be
found in both types of plants. Where they radically differ, of course, is
in their heat source: in a nuclear plant, energy is generated from the
671
Inlet tube
Reverse current valve
@ Downward tube
@ Refueling equipment
@ Steam isolation valve

@ eed water tube
Feed water pump
@ Outer concrete wall
<IJ
ontainment vessel
Control rod driving mechanism
Control rod guide tube

@ Seal plug
Fig. 10 58
Refuel ing equipment

Steam drum
(j) Upward tube
@ Lower header
@ Inner concrete wall

@ Main team tube
@ Steam turbine
Generator
Co ndenser
@ Sea water for cooling
Adva nced thenna l reactor (A TR) ( Prototype "Fugen " )
fission of heavy nuclides, and a conventional thermal plant uses the

chemical combustion of fossil fuels such as bunker oil, gas, or coal.
The rest of this section will provide a broad overview of considerations pertinent to the control of nuclear reactors.
(b) Nuclear power
(1) The atomic nucleus
The nucleus of an atom contains two kinds of
primary particles, protons and neutrons. Both kinds of particles have
roughly the same mass, but protons carry a positive electric charge
and neutrons have no charge. The portion of the atom outside the nucleus contains only negatively charged particles called electrons. In a
normal neutral atom, the number of electrons orbiting the nucleus is
exactly the same as the number of protons in the nucleus. Thus, the opposite charges of protons and electrons cancel each other and the atom
672
Chap.JO
Instrumention to Manufacturing Indust ries
cv
(j) Contai nment vessel
Outer concrete wall

@ Reactor vessel
CV
Cont rol rod

Intermediate heat exchanger
Pri mary main circulati ng pump
@ Feed water pump
@ Secondary sodium
@ Secondary main circul ati ng
pump
Primary sodium
@ Air cooler
Fuel
@ Super heater
Blanket
@ Super heat team
Steam generator
ondenser
Turbine
Generator
irculati ng water pump
@ To sl uice way
@ Cooling water (Sea water)
Fig. 10 59 Fast breeder reactor (FBR) (Prototype " Monjy u")
is electrically neutral.
Hydrogen, the lightest element that occurs in nature, has only one
proton and one electron. Chemically, the nucleus is the same as a
deuterium atom with one proton and one neutron in the nucleus, orbited by one electron outside the nucleus. The number of protons,
which is unique for each element, is called the atomic number, and the
total number of protons and neutrons in a nucleus is called the mass
number because it approximates the total mass of the nucleus.
The heaviest naturally occurring element is uranium with an atomic number (the number of protons) of 92. From this, we know that the
number of electrons is also 92, but the number of neutrons varies.
0. 7% of natural uranium has a mass number of 235, and 99.3% has a
mass number of 238. Atoms of elements that have the same atomic
number but a different mass number are called isotopes. The two uranium isotopes mentioned above are written 2 ~~U and 2 ~~U. 2 ~~U is inherently unstable and tends to break down to a more stable form, emitting radiation in the process.
(2) Nuclear fission
If a neutron is absorbed by the nucleus of an
10.3
673
inherently unstable zg~u atom, this causes the nucleus to fission, giving
raise to large amounts of energy and emitting two or three neutrons in
the process. One condition for fission to occur is that the velocity of
the neutron has to be slowed down so that its energy distribution is
close to the Maxwellian distribution in the material in which it is
found. Slowed-down neutrons are called thermal neutrons. If they are
not slowed down, or thermalized, the neutrons tend to escape, or are
dispersed, making them unavailable to fission.
There are many possible ways zg~u might fission, but a typical
reaction is
zg~U
+An
---+
zggu ---+
1 ~gBa
+ g~Kr + 3An +Energy.
Ba and Kr designate fission products, and An represents a neutron.

Compared to zg~u, zg~u almost never fissions.
It was mentioned that the percentage of fissionable zg~U constitutes only a very small fraction of natural uranium. Isotope separation is
therefore used to boost the proportion of uranium-235, to produce
so-called enriched uranium. Modern water reactors use uranium with a
zg~U concentration of 2 to 3 percent.
(3) Chain reaction
When excess neutrons produced by a fission reaction set off other fission reactions, this provides the basis for a
self-sustaining chain reaction. The energy of a nuclear reaction is
much greater than that of a chemical reaction, as is apparent from Table 10.4. For example, the fissioning of 1 kg of zg~U would yield approximately 2 x 1010 kcal of heat. To produce an equivalent amount of heat
from the combustion of oil or coal would require 2 000 kl and 3 000 metric tons of these materials, respectively.
(4) Criticality
To support a self-sustaining chain reaction, the uranium core of a reactor must be optimized so that the number of neutrons
available for absorption and the number of neutrons that escape to the
surrounding environment are precisely balanced to keep the chain reaction barely going. This optimal state is termed criticality. Criticality is
not an absolute, but depends on the multiplication factor k, defined as
the ratio of the present neutron population to that of the previous neutron generation. Thus, while the reactor operates at a steady power
level, k = 1. Criticality is characterized by a constant number of neutrons generated per second and a constant number of zg~U atoms that
undergo fission per second.
(5) Nuclear reactor
A nuclear reactor is a device where the chain
reaction we have described can proceed safely under controlled conditions. To maintain a state of criticality in the reactor, the absolute
number of thermal neutrons must be controlled, and the number of
those that fission must be kept constant. This is accomplished using
674
rods made of cadmium or boron, materials that have high adsorption

cross sections for thermal neutrons. These two materials are generally
referred to as control materials.
To moderate or slow down neutrons form the high velocities at
which they are created in the fission process, a moderator is introduced to the reactor. The most common moderators are light and
heavy water.
A very obvious requirement for fission reactions to occur in the
reactor is the nuclear fuel itself; uranium or plutonium are typical
fuels.
To prevent neutrons from leaking out of the core, reflective material is applied around the reactor core.
Ultimately, the nuclear energy produced by nuclear reaction is converted to thermal energy in the nuclear fuel. The coolant must therefore lower the temperature of the fuel. Light or heavy water around
the core commonly serves as both coolant and moderator.
(6) Reactor control
The control rods are the main instruments of
reactor control; they are gradually pulled out of the core until criticality is reached (k = 1). In a typical reactor, the control rods can be manipulated over a multiplication-factor range from 0.96 to 1.30.
By way of illustration, suppose that in bringing the number of neutrons up in the reactor, the control rods are withdrawn a bit too far so
the multiplication factor exceeds 1 and the number of neutrons begins
to escalate exponentially. To restore the desired level of neutrons, the
control rods are pushed inward until k = 1. Of course, this procedure
could also be applied in reverse.
We have stated that as the control rods are withdrawn, the number of neutrons and hence the number of fission reactions increases.
Energy released from the fissions, however, raises the temperature of
the coolant, and this causes a decline in reactivity. Hence, there is a
homeostatic tendency toward equilibrium (reactivity is a measure of
the extent of fission reactions).
10.3.7 Pressurized water reactor control system
(a) System configuration of a PWR
The reactor system of a PWR power plant consists of the reactor

coolant system (see below) and various support equipment. In the reactor coolant system, heat generated in the core is transferred from the
primary to the secondary system via steam generators (tube-and-shell
type heat exchangers), thereby generating steam in the secondary system which is delivered to the turbogenerator. The support systems can
be categorized as those involved in the routine starting, running, and
shutting down of the reactor, and those implemented to deal with po10.3 Instrumentation Applications in the Power Industry
675
:;::=====
Concainmena vessel
Main
Main
steam
relief
Conden er
ondensa tc
scorage cank Mocor dri ven

a ux iliary
feed wacer
pump
Turbine driven
auxiliary
feed wacer
pump
Fig. 10 60
676
System layout of a
[ Boron heat regenerati"e y tern )

(Broken line indicates the Oow route by which the boron is added)
========
~
===========
~-- - ,I
Coo!ant .
Coolant
cation
I
demtnerahzer
demineralizer
t
I
r ,
1Heat
I :egenerative
Prim ry pure
water tank
100
t exchanger
I
I
______ _.
I
I
water
demineraliz.er
Spent fuel pit
skimmer filter
Containment spray cooler

on~generative
cooler
W asle evaporator system
etc.
typica l PW R Atomic reactor
10.3 Instrumentation Applications zn the Power Industry
677
tential emergencies.
The system layout of a typical PWR plant is shown in Fig. 10.60.
(1) Reactor coolant system (RCS)
The principal function of the reactor coolant system is to remove heat from the reactor core and transfer it by means of steam generators to the secondary system where it
is used to drive the turbines. The major components of the system are
the reactor vessel, and 2 to 4 coolant units (called heat transfer loops),
each containing a steam generator and a reactor coolant pump. A
1100-MWe-class PWR is equipped with 4 such loops.
The coolant system is installed within the containment vessel and
consists of (a) a number of heat transfer loops (each containing a steam
generator and a reactor coolant pump), (b) the reactor vessel, (c) main
coolant pipes connecting loops and reactor vessel, (d) pressurizer, (e)
pressurizer safety valve, (f) pressurizer relief valve, (g) pressurizer
spray valve, and (h) pressurizer relief tank.
Aside from the nuclear core, the reactor vessel also contains (a) a
core-support assembly, (b) control-rod clusters, (c) thermal shielding,
and (d) in-core assemblies. The steam generator is installed above the
core so that even if the external power source supplying the reactor
coolant pump were to fail, surplus heat from the core will continue to
be removed through natural circulation of the coolant. For the coolant
pump, a vertical-operation mixed-flow type pump with sealed shaft for
leakage control is used. The typical pressurizer is a vertical vessel
equipped with a spray nozzle located at the top where vapor is condensed and electric immersion heaters installed at the bottom of the
vessel. The pressurizer is also equipped with a safety valve and pneumatic relief valve.
Pressure of the reactor coolant system is controlled by the pressurizer at 157 kgjcm 2 The immersion heaters in the pressurizer are divided into a proportional group and a back-up group. If there is a loss of
pressure, the proportional-group heaters are activated first to restore
thermal equilibrium. If equilibrium still cannot be restored with just
the proportional heaters, the backup heaters are also turned on. If pressure raises above the set value, pressurizer spray volume is adjusted as
necessary to maintain operating pressure. If the pressure variation is
more than can be corrected with the spray valve, the pressure relief
valve is opened as needed. A final defense against surplus pressure is
provided by the pressurizer safety valve. The purpose of the pressure
relief tank is to hold water to condense steam emitted by the pressurizer.
(2) Chemical and volume control system (CVCS)
The primary functions of the eves are as follows:
(1) maintenance and control of the volume of coolant in the reactor
coolant system;
678
Chap.lO
(2) regulation of boric acid concentration in the coolant;

(3) removal of fission and corrosion products in the coolant;
(4) addition of corrosion-neutralizing chemicals to the coolant and
control of water quality;
(5) extraction and recovery treatment to enable reuse of boric acid
and coolant; and
(6) sealing water supply for the reactor coolant pump seal.
Chief components of the CVCS are: water purification and makeup
system, boric acid recovery system, and boron thermal regeneration
system.
(3) Residual heat removal system (RH RS)
Principal functions of the
RHRS are:
(1) Removal of residual decay and sensible heat that continue after
the reactor has been shut down until the reactor cooling system is
able to realize the prescribed temperature and pressure (177C, 28
kgjcm 2).
(2) Functions as a component of the safety injection system. If there
is a loss of primary coolant, the reservoir of boric acid water used
during refuelings is injected into the core. After this water is injected, it is continuously withdrawn by the containment recirculation sump and recirculated in the core.
(3) Fills the fuel cavity with water during refueling. The cavity water is then pumped to a special tank designated for refueling water.
In a typical configuration, there are two trains, each containing
one cooler and one pump.
(4) Safety injection system (SIS)
In the event of an accidental loss of
primary coolant, the SIS injects boric acid water into the core to cool it
to prevent any major damage to the fuel or fuel cladding. Boric acid is
also injected if the main steam line is ruptured or obstructed for some
reason. By adding negative reactivity, the injected boric acid effectively quenches reactivity in the reactor.
Components of the safety injection system are the accumulator injection system, the high-pressure injection system, and the low-pressure injection system.
(5) Component cooling water system (CCWS)
This system cools a diverse range of auxiliary equipment that is instrumental in the running
of a nuclear reactor. For example, the pump and generator for the primary coolant, the non-regenerative cooler, and the seal-water cooler
are all cooled by this system. The system also cools numerous other
equipment during power operation, the residual heat removal system
during shutdown, and this same system and the containment vessel
spray system in the event of an accident.
679
The sea water system is comprised of a

(6) Sea water system (SWS)
sea water pump plus connecting pipeline and system valves. Auxiliary
cooling equipment, diesel generators, and coolers for air conditioning
purposes are commonly cooled by sea water.
The main components of this sys(7) Containment spray system (CSS)
tem are the containment spray pump, containment spray cooler, iodine removal drug tank, and refueling water storage tank. Primary functions
are to reduce the heat within the containment vessel in the event of an
accident involving the loss of coolant, and at the same time, remove
iodine that is given off in the vapor part of the containment vessel.
The containment spray pump is activated by a signal that the containment spray has been triggered. It pumps water from the refueling
water storage tank to the upper part of the containment vessel where
the spray heads are mounted. Caustic soda stored in the iodine removal
drug tank is added to the spray water. In recirculation mode, spray water is collected by the containment recirculation sump and returned to
the spray system.
Pumps, coolers, and piping for the spray system are implemented
in two independent trains.
The main steam sys(8) Main steam & feedwater system (MSFWS)
generator as far as
steam
the
from
line
steam
main
tem consists of the
on the main steam
Mounted
valve.
shut-off
steam
the turbine's intake
valve, main
safety
steam
line are a number of valves including main
and turbine
valve,
check
steam relief valve, main steam isolation valve,
bypass valve.
The turbine bypass system makes it possible for the operation to
continue when there is a sudden decline from the rated load, without
having to trip the reactor or emit a large quantity of steam, by diverting the surplus steam to the condenser.
Feedwater to the steam generator is generally supplied by the
feedwater pump. The volume of feedwater supplied is regulated by the
SG water-level control (the three elements governed are: water level,
feedwater volume, and steam volume).
In addition to the systems already covered, some
(9) Other systems
of the other systems found in a typical PWR plant are:
Waste disposal system (WDS)
Sampling system (SS)
Spent fuel pit cooling & clean-up system (SFPCS)
Refueling water system (RWS)
Primary makeup water system (PMWS)
SG blowdown system (BDS)
Instrument air system (lAS)
Auxiliary steam system (ASS)
680
Chap.lO
Heating ventilation and air conditioning system (HV AC)

(b) PWR instrumentation equipment
(1) Reactor instrumentation
Here we will be particularly concerned
with instrumentation used to monitor and collect information regarding conditions in the reactor core.
The out-core nuclear instrumentation system is designed to supervise the level of neutron flux, which is proportional to the output of
the reactor core. The system consists of neutron detectors mounted
around the reactor vessel. Three ranges are detected: source, intermediate, and power. The system provides reliable supervision of core neutron flux levels from shutdown up to 120% of the reactor's rated output.
The in-core nuclear instrumentation system is designed to monitor the output distribution of the core, and consists of equipment to
measure the temperature and neutron flux within the core. To supervise the output distribution of the core, the in-core temperature detection system uses chromel-alumel thermocouples to detect the temperature of the primary coolant at the outlet of the fuel assembles. The
in-core neutron flux detection system uses flexible, small-scale neutron detectors that are inserted into the center of the fuel assemblies.
This enables measurement of the neutron flux distribution parallel to
the fuel assemblies.
The control-rod position indicator system is based on position detectors that are mounted on each control-rod cluster. The position detectors are configured out of 42 coils mounted on the pressure housing
on the outside of the control-rod driving mechanism. Information indicating the positions of the control-rod clusters is converted to digital
signals and displayed on the central control panel.
(2) Process instrumentation
It is necessary to measure the volume
flows of all essential processes. This applies not only to the primary
coolant system, but to the processes of the various other support systems as well.
Process instrumentation consists of sensors and miscellaneous instruments installed in an instrument rack. Output from the instrument
rack is sent to the (a) reactor protection system, which automatically
shuts the reactor down in an emergency, (b) the engineered safety features operation system used to activate equipment essential in the
event of an accident, and (c) the reactor control equipment for controlling reactor output. In addition, important parameters from the instrument rack output are displayed, recorded and used to activate alarms
on the central control panel. To illustrate this type of instrumentation,
we will examine the process instrumentation of the safety protection
system, which provides essential parameters for the reactor protection
681
system and the engineered safety features operation system.

The process instrumentation for the safety protection system is designed to be redundant. The channels are all implemented separately
with their own sensors. The instrument rack is divided into four channels. In keeping with the functional segregation of the design, the power sources supplying the instrumentation of each channel are also independent. A portion of the safety protection system process instrumentation produces a control signal that is used by the reactor control system (described below). In this case, however, an isolation amplifier is used so that even if the reactor control system were to fail, no
adverse effects could spread back to the safety protection system. Signals to the indicators and recorders are also mediated by isolation amplifiers. Finally, it should be noted that all process instrumentation can
be tested to verify it is in good working order even while the reactor
is being operated at full power.
(3) Reactor control system
The control system of a nuclear reactor
has two main functions, namely, to compensate for changes in operating conditions that occur while PWRs are operating at a steady power
level and to regulate reactor power in response to load changes. Plant
output control during operation is effected by regulating the volume of
turbine steam and by controlling the reactivity of the core. Core reactivity is controlled through a combination of two basic methods; adjusting the positions of control-rod clusters and by controlling the concentration of boron in the coolant.
Control-rod clusters are adjusted according to need to compensate
for short-term changes in plant operating conditions such as power
and temperature, and to absorb excess reactivity in a high-temperature shutdown. Boron concentration control, on the other hand, is used
to compensate for long-term reactivity changes involving fuel burnup
and consumption of fission-product poisons, and to absorb excess reactivity in a low-temperature shutdown. Power control by adjusting control-rod clusters is done manually up to about 15% of the rated load of
the reactor; beyond 15% of the rated load, control-rod adjustment is
under automatic control. Within the automatic control range, load
changes can be ramped at a rate of 5%/min or stepped at a 10%.
Moreover, by making use of the turbine bypass system, extreme load
changes equivalent to about 50% (or even 95%) of the rated load can be
accommodated without have to shut down the reactor. In the rest of
this section, we will discuss the main control systems involved in controlling nuclear reactors. A schematic overview of the main control systems is shown in Fig. 10.61.
The control-rod control system is under automatic control and consists of the control-rod clusters in the control group. As was noted ear682
CD
Control rod control

Trip signal
(!) Neutron nux
Reactor protection system
Engineered safety features operation system
(j) Operation signal
Reactor instrumentation system
Other process instrumentation system

Pressurizer pressure control
@ Pressurizer water level control
@Average
@ Main steam relief valve control
Feed water control
@ Turbine bypass control
Boron concentration control
@Mixer
@Turbine
Condensor
Generator
Control mode
Boric acid water
@ Pure water
Fig. 10 61 Systematic diagram of PWR Atomic reactor instrumentation equipment
lier, the system is generally used to compensate for shorter-term

changes in plant operating conditions such as power and temperature
changes. A primary coolant average temperature signal is compared
against a program average temperature, which is proportional to turbine load, and a control signal is generated. This control signal plus a
compensation signal that is generated when transients occur based on
the rate of differential change between turbine load and neutron flux
are used to govern the speed of the control-rod clusters of the control
group and to maintain the primary coolant average temperature at the
required value. The control-rod clusters of the control group are arranged in four banks, designated A, B, C, and D. When the rods are in10.3
683
serted or withdrawn, some overlap is deliberately provided to linearize

the effects of the control rods.
Other groups of control rods perform other functions. For example, the shutdown group is kept in the full-out position, providing the
means (together with the control group) for fast reactor shutdown
without using automatic control. Another control-rod cluster is used
to adjust the power distribution.
As was mentioned, boric acid dissolved in the primary coolant controls only the slowly changing reactivity effects. Adjustment of the
boric acid concentration is done by the chemical and volume control
system; concentrated boric acid is manually added to increase boron
concentrations, and pure makeup water free of boric acid is added to
reduce boron concentrations. The concentration level is supervised by
means of regular sampling.
As much as possible, control rods are held in the withdrawn position during power operation. This not only provides a means of rapid
shutdown, but also helps shape the power distribution within the core
to make it uniform. Early in the core life, only bank D control-rod clusters in the control group are partially inserted (the so-called bite position) during power operation. If these control rods are either inserted
or withdrawn beyond their position limits, the concentration of boron
is adjusted so the rods can move back to their designated positions.
The rod withdrawal limit is calculated to enable a return to full power.
The pressure of the primary coolant system is controlled by the
pressurizer (pressurizer pressure control system). When pressure in
the pressurizer increases, the coolant on the low-temperature side of
the primary coolant system becomes proportional to the pressurizer
pressure signal, steam in the pressurizer is condensed by a subcooled
pressurizer spray, and this causes the pressure to drop. When pressure
falls and the water level in the pressurizer raises above the setpoint,
then heat is applied by proportional and backup immersion heaters
causing the pressure to raise. Thus, the drop in temperature of the liquid vapor in the pressurizer is compensated. If excess pressure builds
in the pressurizer beyond the capacity of the spray to correct, thermal
equilibrium is restored by the relief valve.
Volume of the chemical and volume control system is maintained
under automatic control by the pressurizer water-level control system
based on a deviation signal reflecting the difference between programmed and actual water level in the pressurizer. If the water level drops
to an abnormally low level, the extraction line valve is closed, thus
halting any further decline in water level.
Feedwater control is configured separately for each steam generator. Three control elements determine the aperture of the feedwater
684
Chap.lO
control valve; feedwater volume, steam volume, and the steam-generator water-level signal. If the steam generator water level raises to an
abnormal level, the feedwater control valve and bypass feedwater control valve are closed. In the opposite situation where the water level is
too low, the auxiliary feedwater pump is automatically activated to ensure that the ability of the steam generator to eliminate heat is not impaired. If it is necessary to boost power, water-level control can be
maintained either manually or automatically using the bypass feedwater control valve.
The turbine bypass control system diverts steam generated by the
steam generator past the turbines, conveying it directly to the condenser. This makes it possible for the plant to adjust smoothly to large,
step-shaped drops in load by eliminating heat transients in the primary
coolant system. In the case of high-temperature standby or plant cooling, pressure control is maintained by main steam header pressure to
either eliminate remaining heater or eliminate heat for cooling, as the
case may be.
The main steam relief valve control system works as follows. In
case of a large increase in the main steam pressure, due for example,
to an abrupt drop in load, a main steam pressure signal is produced,
opening the main steam relief valve. Recourse to the main steam safety valve is avoided if possible.
Immediately prior to reactor shutdown, the following interlocks
are implemented to enable many of the automatic systems.
Interlock to prevent withdrawal of control-rod clusters is effected
if intermediate range neutron flux, power range neutron flux, over
temperature LJT, and over power LJT reach the excess high set point.
This interlock prevents both automatic and manual withdrawal of the
control rods.
Turbine runback interlock reduces turbine output in case over temperature L1 T and over power L1 T reach excessive high levels. Also, to
restrict automatic control when reactor power is at a low level, automatic withdrawal of control-rod clusters is prevented when turbine
output falls below the setpoint.
(4) Reactor safety protection system
The system is designed to protect the system and keep it from failing in the event of abnormal transients or an accident. The system maintains the fuel assemblies, maintains the temperature and pressure of the containment vessel within
permissible design parameters, and safeguards the coolant. Two
subsystems are included in the reactor safety protection system: these
are the reactor protection system, which activates the automatic reactor shutdown system, and the engineered safety features operation system, providing access to various equipment that would be necessary in
685
the event of a reactor failure.

The reactor protection system senses any abnormal condition within the reactor and begins shutdown before damage to the structure can
occur by means of the out-of-core instrumentation and safety protection system process instrumentation described earlier. The protection
system consists of two main elements; a logic circuit and a reactor trip
breaker. The logic circuit outputs reactor-trip and interlock signals in
response to a bistable signal (i.e ., alarm activation signal of the alarm
setter) from the out-of-core and safety protection system process instrumentation. The reactor trip breaker is automatically opened when a
reactor trip signal is received. For added safety, the reactor protection
system is dispersed in either two or four independent trains. The principle is illustrated in Fig. 10.62 for a two-train configuration. All functions of the safety protection system (including the trip breaker) can
be tested to verify they are in good working order without shutting
down the reactor.
The operating principle of the reactor trip breaker is shown in
Fig. 10.62. An M-G set, three- phase, alternating-current supply system is connected to the control-rod drive mechanism. If the trip
breaker is opened, power to the control-rod clusters is cut off thus
causing the control rods to drop into the core under the influence of
gravity.
Control signals
Fig. 10 62
686
'----:---:---:--::--
--' To engi neered

Control rod
safety features B
PWR Atomic reactor control a nd protection system
The engineered safety features operation system is the means of

activating the reactor protection system for automatic shutdown of the
reactor in the event of failure. A number of equipment units are involved; a bistable signal from the safety protection system process
instrumentation triggers the emergency core cooling system, the containment vessel separation valve, and the containment vessel spray
equipment. Signals are also generated activating the various subsystems associated with these units. Note in Fig. 10.62 that here too the
engineered safety features operation system is dispersed in either two
or four trains. As with the reactor protection system, all system functions can be tested while the reactor is in full operation.
10.4
Instrumentation Applications in the Food Processing lndustry17)
-lQl
10.4.1 Overview
Japan's food processing industry has evolved very rapidly as
life-styles have become increasingly moderized and Westernized. Most
food products are made in batches. Handling systems in the food-processing industry generally require automatic line cleaning systems and
extra stringent quality assurance. These unique requirements call for
unique instrumentation techniques compared with other process industries.
Since the food-processing industry by its very nature produces
such a diversity of proda~<:ts, it is characterized more than other industries by a complex structure of many companies of all different sizes,
some dealing with raw materials, others with semi-finished products,
and still other companies with finished products. Nationwide, for example, there are some 85 000 establishments (employing approximately
1.15 million people) involved in the food-processing industry in Japan.
Of these, 98% have fewer than 100 employees, leaving a scant 2%
(about 1 700 firms) employing more than 100 people (1984 data).
Considering the great diversity of products, it comes as no surprise that almost as many unique automated techniques have been devised for their production.
In the following, we will describe some of the unique aspects of
food-processing instrumentation and sanitary detectors that are an integral part of food-processing lines. For our specific applications, we
will focus on two products that have achieved the highest degree of automation in their production: whiskey distilling and sugar refining.
(a) Food-processing instrumentation
When we speak of the food-processing industry, we are primarily
referring to those processes shown in Fig. 10.63. In terms of instrumentation, the following attributes are unique:
10.4 Instrumentation Applications in the Food Processing Industry
687
Raw material
(farm products ; marine products)
Raw material entry process

(I) Arrival and conveyance
(2) Weighing
(3) Stock control
f--
Manufacturing process
(I)
(2)
(3)
(4)
(5)
(6)
(7)
Utility
Inspection, Screening
Washing, Crashing, Mixing
Reaction, Fermentation, Concentration
Decolorization, Deodorization
Sterilization, Filteration
Quality control
Production control
(I)
(2)
(3)
(4)
Electric receiving control

Heat control
Water delivery
Waste water treatment
Commercial sugar delivery process

(I) Forming, Packing
(2)
(3)
(4)
(5)
Bottling, Labelling
Quality control
Delivery, Conveyance
Warehouse control
Fig. 10 63 From ran materials entry to delivery
(1) There are many processes where productive technology has not
been adequately established; therefore, establishment of standardized operating methods for food-industry equipment is exceedingly difficult.
(2) Many situations call for sensory evaluation of flavor, color, and
odor to determine product quality. Thus special sensors and analyzers are required.
(3) There is often considerable seasonal fluctuation in the operation
of food-processing equipment. This tends to drag down the overall
operating rate.
(4) Characteristics of the controlled objects can change quite readily
with variations in temperature or over time. Also, since most
food-industry materials are solids, material handling methods are
required.
(5) The process scale is often small, and a diverse range of products
are produced in small-lot batches; thus, sequence control is essential.
(6) Since the instruments come in contact with food products that
are susceptible to spoilage, sanitation is a primary concern. To
withstand the rigors cleaning and disinfecting, such devices must
be heat resistant, chemical resistant, and of course capable of
being sanitized. To meet these stringent requirements, the
688
food-processing industry uses sensors that are designed specifically for sanitary environments.
(b) Sanitary detectors
In contrast to sensors for general industrial applications, sensors
for use in the food industry are subject to extra stringent sanitation requirements. This is only natural considering the direct link between
food product quality and health.
In order to maintain the quality of foodstuffs, the following sanitation gidelines are observed:
(1) prevention of bacterial contamination;
(2) prevention of deterioration of food-product constituents;
(3) prevention of mixing contaminated foods; and
(4) proper use of food additives.
There are two international standards that stipulate in general
terms how these are conditions are to be met. These are the 3-A Sanitary Standards and the international standards compiled by the International Dairy Federation (IDF). With respect to sensors used as instrumentation devices, these standards stipulate four requirements pertinent to sanitation that should be complied with; compatible with a
high level of product safety, good cleanability, easy to disassemble,
and easy to inspect. For cleaning, a particularly important concept is
cleaning in place (CIP). All pipes and devices should be capable of being
cleaned and disinfected using either base or acid cleansers and hot water (or, steam) without them being removed or disassembled. This CIP
requirement also applies to sensors.
10.4.2 Whiskey distillery instrumentation
Whiskey can be broadly categorized into three types based on the
mixture of spirits that goes to make it:
1. Single-malt whiskey: A single malt stock.
2. Pure-malt whiskey: A blend of malt stocks.
3. Blended whiskey: A blended product made from malt stock and
grain whiskey.
Traditionally, the most commonly consumed whiskey has been
blended whiskey, but the demand for pure-malt whiskey has increased
in recent years. Malt and grain whiskeys are produced in exclusive distilleries, where the immature sprits are produced, allowed to age, and
finally blended to make the final product. Here we will focus on the
instrumentation in a malt-whiskey distillery and the instrumentation
applied to lautering equipment.
(a) Overview of a malt-whiskey distillery
A malt-whiskey distillery consists of eleven basic process steps, as
shown in Fig. 10.64, all of which are carried out as batch processes.
10.4
Instrumentation Applications in the Food Processing Industry
689
Bottling, Product
process
Fig. 10 64 Processes in a malt whiskey distillery
(1) Malting: Clean barley, the main material from which whiskey is
made, is steeped, allowed to germinate, dried in a kiln, and screened to remove the culm, or dried sprouts.
(2) Raw material processing: The malt is stored, separated, and
ground into a meal, or grist. Separation is to remove waste parts
of the grain, and grinding stimulates the production of enzymes capable of saccharifying the raw grain. The grist is then stored as
dictated by the production plan of the distillery.
(3) Mashing: The grist is mixed with hot water in a mash tub, and allowed to soak for 3 to 7 hours until the water has liquefied all of
the starches, and the diastase has converted the starches into maltose and dextrin. The liquid now is known as wort.
(4) Lautering: The wort is transferred to a Iauter tub, a tank with a
perforated false bottom, where a filter press method is applied and
the insoluble grain residue is separated from the wort.
(5) Cooling: The wort is cooled to zooc using a plate coil or similar
cooler, and then pumped into fermenting vats.
(6) Fermentation: Pure cultured yeast is added to lautered wort and
fermentation continues for 3 to 4 days. When fermentation is completed, a liquid with an alcohol content of 5 to 8% has been produced known as mash.
(7) Distillation: The mash goes into a pot still for two distillations.
Water is added to the now immature malt whiskey to reduce the
proof to under 60%.
(8) Storage: The immature whiskey is then put up in oak barrels and
aged for 3 to 8 years.
(9) Blending: After maturing, malt whiskeys are blended first and
then grain whiskeys are blended in to achieve the proper aroma,
flavor, etc. The blended whiskey is then rebarreled and allowed to
marry for about a month before bottling.
(10) Bottling and product processing: After this final aging, the
blended whiskey is reduced to a standard proof and bottled for
market.
690
(b) Lautering equipment instrumentation
In this section, we will consider the instrumentation requirements

for gravity-filter-type lautering equipment, a type of filtration equipment designed to separate the mash into wort and spent grain. The
spent grain itself serves as a filter thus making the use of another filter unnecessary.
(1) Lauter filter sequence control
As is shown in Fig. 10.65, the process begins by charging a fixed amount of water into the equipment (a
process called spurgging). Then, after transferring the mash (i.e., wort
+ spent grain) from the mashing process, the mash and water are circulated at a constant rate of flow for a fixed period of time (in some
cases, the time is determined by the quantity of mash added). After
the spent grain has been extracted, the wort is cooled as it is transferred to the fermentation process. Next, water is added at both top
and bottom of the Iauter tub, and husks and other grain residue are extracted by means of the awn-hacking unit. The water added at the top
of the tub is called spurgging, and water added at the bottom is called
backwash. After a fixed amount of water has been added, it is circulated at a uniform rate for a fixed interval. The second time, the liquid is
extracted from the grain. The wort is then sent back to the fermentation process passing through the cooler on the way. This is called the
No. 2 lautering.
By repeating the spurgging (backwash), circulation, and lauteting
Mashing
(TPC)
Mashing pan process
a--
Spurgging
Lauter main-process
Circulation
Lauter sub-process
1--A-w-n--.,.-ha-c.,.-ki-ng-sec-t.,.-io-n-------1~
Hook;,g
Fig. 10 65 Block diagram of Lauter filteration process
10.4
691
a number of times, the spent grain is completely removed, and one

batch is completed.
(2) Lauter equipment instrumentation and control
The instrumentation control involved in the operation of Iauter equipment is shown schematically in Fig. 10.66.
(1) Mash flow rate control (FQC-11): Mash (wort + spent grain) from
the mashing process is supplied to the main equipment in fixed
amounts by means of batch-add control. Usually, an electromagnetic flowmeter is used to measure the flow rate.
(2) Spurge and backwash flow control (FQC-8): For this control, an
electromagnetic flowmeter mounted on the mashing water supply
$
I
qJ
I
I
I
i Ii
I
.L
next process
Lauter
vessel
L - - -- -+----t:k::E--0--
Mashing water
(addition)
~~~~~~------:TToo~spentgrain
tank
for disposal by sale
Wort level (Ll-1)

Lauter differential pressure (Pdl-2)
Axial height (Ll-3)
Axial oil pressure (PI-4)
Axial current (AI-5)
Wort temperature (TI-6)
Wort outlet flow rate (FQC-7)

Mashing water flow rate (FQC-8)
Lauter vessel level (LIC-9)
Spent-grain bin Screw conveyer
current (AI-10)
Mash flow rate (FQC-11)
Fig. 10 66 Overview of Lauter instrumentation
692
Chap .lO
line is used to maintain batch-add control.

(3) Circulation and Iauter flow control (FQC-7): Circulation flow rate
and Iauter flow rate to extract spent grain and Iauter are carried
out by fixed valve control using orifice, vortex, or electromagnetic
flowmeters. Moreover, the valve- opening program control is governed by a control valve as well. Note that in cases where the
wort extraction is based on Iauter flow control, Iauter differential
pressure (Pdl-2) is sometimes used to determine the timing.
(4) /n-lauter temperature (T/-6): To monitor the quality of the wort,
a recording device regularly records the temperature within the
Iauter.
(5) Other measurements: Other measurements besides those already
mentioned include : axial height (LI-3), oil pressure (PI-4), current
(AI-5), and spent-grain screw conveyer current (AI-2).
Another technique of regulating the circulation volume is to use
level control (LIC-9) applied on the circulation line attached to the Iauter vessel.
The smooth operation of the Iauter equipment depends on the integration and synchronization of all the measurement controls and sequence controls described.
10.4.3 Sugar refinery instrumentation
Considered a daily necessity by people around the globe, sugar is
an international commodity that is widely produced and widely consumed.
World-wide production is close to 100 million metric tons (in 1982),
roughly 60% of which comes from cane sugar and 40% from beet sugar.
There are three different kinds of sugar refineries specializing in the
processing of raw sugar, beet sugar, and refined sugar. Here, we will
focus on the typical manufacturing processes of a refined sugar processing plant. The various kinds of sugar products are shown in Fig.
10.67, and a glossary of technical terms used in the sugar industry is
Crystal ugar
Beet sugar
Sugar
Soft sugar
ured sugar
Raw ugar
Refined
sugar
Processed uga r
Cane sugar
Liquid ugar
Stroop ugar
Molasses
Fig. 10 67
Sugar products
693
Table 10.6 Technical sugar terms
Terms
Definition
Washed sugar or
affina ted sugar
Magma
Liquor
Sugar which is washed and separated by centrifugal machine
Mixture of sugar and water, syrup or molasses

Liquor consisting of sugar dissolved in water, so far not
treated by crystalation process
Affination syrup
Syrup which is separated from magma of raw sugar and
devided into two groups; green syrup and white syrup
Raw liquor
Liquor consisted of washed sugar dissolved in hot water or
sweet water
Final liquor of washing process, just before being supplied to
Fine liquor
crystalization pan
Thin liquor which is obtained after washing equipments
Sweet water
Syrup
At raw sugar plant; liquor which is concentrated in
evaporator
At sugar refinary; liquor which is separated by centrifugal
machine
Liquor which is squeezed from raw materials (sugar cane,
Juice
sugar beet)
Molasses
Syrup which is separated by curing machine (term used in
raw sugar plant)
Visco
Thick invert sugar liquid, which is added in manufacturing
of soft sugar and includes a mixture of more than 90%
fructose and glucose
Mixture of crystal and mother liquor of sugar which is
Massecuite
acquired from boiling at pan
Operation which separates sugar cristals by heating in pan
Boiling
Concentration unit of sugar shown by weight percentage of
Brix degree, Bx
solid part
Ex. 60 Bx (60 Brix degree) Liquor contains 60% sugar
Nucleus of crystal (seed)
Seed
Two single crystals stuck on a specified plane
Twin
Crystal, based on agglomeration of multiple seeds other
Agglomerate
than twins
Crystal which grows spontaneously
Smear
Mean value of grain size in product
MA (mean aperture)
Dispersion of grain size in product
CV (coefficient of
variation)
Ratio of monocrystal in product
NCS (natural crystal
score)
Purity
Sugar ratio in solid content of liquor
Molasses, which has finished the recovery process
Final molasses
694
provided in Table 10.6.
(a) Overview of sugar refinery processes
Raw sugar transported by ship from overseas

(1) Raw sugar feeding
is unloaded and moved by conveyer belt to a raw warehouse. Then, in
measured quantities, the raw sugar is transported by conveyer to a
raw sugar storage bin.
Many impurities adhere to the surfaces of
(2) Washing and melting
sugar crystals. To remove these impurities, they are first loosened by
mixing with juice in a mixer-type scroll conveyer called a mingier.
The juice is then purged in a affination centrifugal, and dissolved in a
melter to become coarse sugar juice.
Lime milk is added to the coarse
(3) Carbonization and filtration
sugar juice, which reacts with carbon dioxide gas in a carbonator to
produce calcium carbonate. Impurities adhere to the calcium carbonate
which is then passed through a primary filter to remove these impurities, yielding a clear amber-colored sugar juice.
The filtered sugar juice is passed
(4) Decolorization and concentration
through a char cistern and an ion exchange resin tower. These processes decolorize and refine the juice. Then, after passing through a
check filter, the juice goes to a concentrator where it becomes more
concentrated fine liquor.
The fine liquor output from the concentrator is
(5) Crystallization
then crystallized in vacuum crystallization pans. After the concentration of the syrup is increased even further, product-specific seed crystals are added and sugar crystals grown to uniform standard size as
syrup continues to be added.
Granular sugar crystals are ex(6) Purging, drying, and finishing
tracted, or purged, from the mixture of crystals and surrounding syrup (massecuite) in a product separator. Remaining syrup is returned to
the crystallization process to reclaim further sugar. The final end product of the process is final molasses. The extracted sugar crystals are
passed through a dryer and cooler, then stored in a humidity- and temperature-controlled sugar bin. Next, the sugar is processed through a
multistage sieve to separate it into standard-grade products.
How sugar is packaged depends largely on the type of
(7) Packaging
sugar and how it is to be used. Sugar is usually either packed in bags
and warehoused or shipped directly in loose bulk.
The entire process from receipt of the raw sugar to final packaging is shown schematically in Fig. 10.68.
(b) Key sugar refinery controls
(1) Melter Brix (concentration) control in the sugar melting process
Raw sugar, the raw material for this process, is mixed with heavy
syrup in a mingier to form a magma. The magma is separated into wa10.4
695
Lime milk
Raw sugar
warehou e ~
d(~ g
J'
~ ~
C0 2
~~M~ingler
\]
Washed sugar
separator
Melter
Purging/ Product / Packing
<D
Raw sugar bin
Char cistern
@ lon exchange resin tower
@ Check filter
Fig. 10 68
Concentrator
Manufacturing process of sugar plant
shed sugar and green syrup in a centrifugal separator, and the washed
sugar is then dissolved in sweet waters in the melter to become raw
liquor, and sent to the carbonation process. An instrument called a
Brix meter is used to determine the concentration of the raw liquor.
Differential-pressure type Brix meters are especially widespread. the
configuration of the control loop involved is shown in Fig. 10.69.
Key points respecting the instrumentation are as follows.
(i) Since washed sugar is intermittently withdrawn from the centrifugal separator, the proper amount of sweet water charged to the
centrifugal is automatically adjusted and synchronized with the
centrifugal's discharge signal.
(ii) Sufficient residence time is necessary to produce high- concentration raw liquor with a stable Brix (concentration) value. This is accomplished by adding most of the sweet water and approximately
adjusting the Brix value in the pre-melter. The Brix value is then
fine- tuned with only a small quantity of sweet water in the main
696
Chap.lO
-.
I
~-r--~
I
I
~--$
I
r-::-:-'---. :
I
~i
I
I
I
I
Raw sugar
------~
I
I
I
I
:
I
L---- t----.
I
I
I
I
I
I
I
I
I
I
I
I
I
J
I
I
I
I
:
I
Steam
J-----4~--+--"\
Thermometer
L - - - - - - - - - - - - -- - _____ J
Fig. 10 69 Melter- Bri~ control system
melter.
(iii) By monitoring the Brix value of the sweet water, the quantity of
sweet water going into the pre- melter is automatically compensated. This control makes it possible to stabilize the Brix value and
obtain high-concentration raw liquor. The burden on downstream
processes is thus reduced and energy is conserved.
In the carbonator, lime
(2) pH control in the carbonization process
milk is added to raw liquor and it reacts with carbon dioxide gas to cohere impurities. Accurate pH control is extremely important in this
process. The control loop configuration is shown in Fig. 10. 70. The
key points respecting the instrumentation are as follows.
(i) The flow rate ratio between the raw liquor and the lime milk in
process must be controlled.
(ii) The lime milk flow ratio is corrected based on the concentration
of the lime milk and the pH of the carbonator. The control not
only stabilizes the pH but also improves the effectiveness of the
697
L--~)- To
Fig. 10 70
filter
PH Control in the carbonater process
following filtration process.

The primary filter
(3) Flow control and sequence control for the filters
removes impurities that cohere to calcium carbonate to yield a clear
amber-colored sugar juice.
Sequence control is applied to this operation mainly to coordinate
standby, precoat, filtration, residual blowing, desugaring, and sluicing processes. This filtration is carried out as a batch process, but sequence control is required because multiple filter lines are employed.
Sequence control coordinates the different processes, and also synchronizes filter-line rotation.
The control loop configuration is shown in Fig. 10. 71, and a time
chart showing how the operation of the different filter lines are
staggered is shown in Fig. 10. 72. Key points respecting the instrumentation are as follows.
(i) Flow control cascaded by the carbonation tank level and pressure
selector control at the filter are carried out.
(ii) The standby, sluicing, and multi-line filter system are under sequence control.
(iii) An operation should be implemented so as to maximize the efficiency of the filter cloth (i.e., the quantity of liquid passed
through the filters).
With these control measures, automatic operation can proceed on
the basis of combined capacity of multiple filter lines with minimal
waste, to achieve improved operational efficiency.
698
High-pressure hot water

Carbonation tank
LIC output
~
I
From
carbonation tank 1-.c:"o-,-~)-----cO:l..._--.J
Smearing pump
Mud
Fig. 10. 71
1 st
Liquor
Sweet water
Flow and pressure control for filter

One batch process
Precoat
Standby
\
I I
'\
o. I filter
o. 2 filter
o. 3 filter
o. 4 filter
'
Desugaring
Filtration Residu al! Sluicing
I
I l..t
;ol
/
Operation
~//~~~~ration
Operation
Opera tion
Operation
Operation
Operation
Since the precoat line is imprimented as a common line to

th e filtration operation , operation of the four filter lines
should be staggered so that the precoat process is not
duplicated
Fig. 10 72 Operation system of filters
(4) concentrator control

In the concentrator process, decolorized, refined liquor is boiled down to the desired concentration, then sent to
the crystallization process. The control loop configuration is shown in
Fig. 10.73.
(i) Feedforward control is implemented with an operational element
to calculate changes in feed concentration, the largest external disturbance. This measure reduces the adverse effects of lags in the
outlet concentration control loop to a minimum.
(ii) Using (a) the difference between the desired outlet concentration
and the measured concentration, and (b) the steam flow, the following calculation can be used to stabilize the set point of the desired
inlet flow concentration without altering the steam flow :
10.4
699
;--------ra51---,
I
~
I
!
-,
:d r:;:,
iI
1
. ~...J
re ~ :-Tv
I
I
1
@@
I
L - - - - - -,
iI :I
L ---
I
I
r-~ ,---~
~~ I
11
lv-}-,
~~
o
:I
I
I
I
I
I
I
I
I
I
I
I
CD
I
I
I
I
I
I
I
I
I
1 Vent
alcul ation
Flow rate
oncentration
@ Feed
Balance tank
Steam
Flow rate
oncentra tor
Concentration
When the difference between desired
a nd measured concentration is large.
This line is connected to the balance
tank
Fig.. 10 73
F=
I
Control system for the concentratio n proce
BqFs(I+k2)
(Bq- B;)(l + k,B;)
where Ft is the inlet flow (m 3/h), Fs is the steam flow (t/h), Bq is

the desired outlet concentration (0 to 1..0), B 1 is the inlet concentration (0 to 1..0), k1 is the concentration conversion factor, and k2 is
the evaporation multiplying factor..
Homogenization of the syrup concentration is essential to deliver a
stable volume of syrup to the crystallization pans in the next process .
The downstream
(5) Control between tanks in the purification process
crystallization process requires a stable supply of fine liquor from the
purification process. Thus, the buffer tanks between equipment units
must be monitored at all times by follow-up flow control of the
amount of product (melt) making sure the tanks neither overflow nor
run dry. The control loop configuration is shown in Fig. 10.74.
700
Chap.lO
L p
@~
LV I~
f
r----, FSP
CD : Y
LIC
---~
-1LVI'i r- Q
i@ r ' I
IMS '
:
I
Ml
le------L--:>--r I I
I
i1
I
I
'7' L _
\.!.)
1
1
I
I
I
I
I
I
L __ .J
I
_______ ...J
LL
_g
LC- t-~---r------------ ~
I1
I
L V2 '
SEQ
ILV2
1
I
HH
I
I
:
I
PVA 6 1I
I
.._ ____
I
I
I
I
: :r
IPvJ_~
rsEQI --t- --"----------1r
-r-(L-- -----I
t--
CD
on liner calculator
Adder
Signal selector
Fig. 10 74
@ High-low limiter
Function calculator
I
I
CD
1 I
I 1
1
Jl
Post
process
High -low alarm unit
@ Con tant va lue generator
High- low a larm unit
Loop configration of now control between ta nk s
(i) The most basic control loop implements a cascade-control of the

flow loop linked with the level. The flow and level setpoints for
each process are calculated from the production (melt) volume.
(ii) The volume of liquor passed by each process is determined by
supervising the level of buffer tanks both before and after the respective processes. The level signals are then used for the flow
loop set points after correction.
(iii) The level high-low values of each buffer tank are monitored on
a continuous basis. If the level deviates from the set range, a return valve is opened to redistribute liquor as required.
A number of points can be made with respect to the control loop
configuration shown in Fig. 10.74. We will assume that the setpoint
(LSP) of the level controller (LIC) for tank Vl has been pre-set to maximum capacity to make the best use of the tank. A nonlinear calculation of the deviation between the level setpoint (LSP) and the actual
level (L Vl) of the tank is calculated by calculator CD and input to calculator @.
In the meantime, the processing volume setpoint (which includes
the production volume) is output by the constant value generator (MS),
and added in to the calculation of adder @ such that
10.4
701
F~r.=Ms'
+Lc
where F~r. is the addition value, Ms' is the volume of processed liquor
(determined by the production volume), and Lc is the level correction
value.
Signal selector (the minimum value) then selects the lesser value between the result obtained by adder @ (i.e., the level correction
value plus the processed liquor volume) and the level of tank V2
(L V2'). This value then becomes the setpoint (FSP) of flow loop controller FIC.
Two other calculators should be mentioned at this point; loops are
configured for function calculator used to correct the level of tank
V2, and high-low limiter @) whose output is used to assure minimum
flow from the tank.
Also, in order to activate solenoid valves SVI and SV2, the combination of high-low level alarm units @ and (j) are used to detect the
high-low limits of tanks VI (sending) and V2 (receiving). Thus, if a situation develops where not enough flow is coming from tank VI, the
overflow from downstream tank V2 is shut off.
(6) Crystallizer control
Syrup close to saturation is delivered from
the previous process and charged to the crystallizer (intake). Since the
sucrose component would be decomposed at high temperature, evaporation is carried out in vacuum pans at low temperature (60oC) until supersaturation is reached (concentration). At the proper moment, seed
crystals are added to form the grain (seeding). Then, as the boiling progresses, syrup and feedwater supply valves are intermittently controlled in response to hardness and level signals until the crystals
reach their proper predetermined size (crystal-growing). The finished
product is then discharged and sent to the next process (sugar discharge).
(i) Sensors (hardness meters, more properly, rheometers) are required to precisely monitor the progress of crystal growth in the
crystallizer.
(ii) Crystallization is carried out based on a program control system
that has been implemented based on the knowhow of a seasoned
operator with extensive experience.
This control system, implemented with hardness meters and a
microprocessor, is called an automatic intermittent syrup boiling system.
The control system configuration for the crystallizer is shown in
Fig. IO. 75, and a schematic overview of the control operation is shown
in Fig. I0.76. A hardness meter, or rheometer, is shown in Fig. I0.77.
702
Chap.lO Jnstrumention to Manufacturing Industries
r - J .-----J
I
I
I
I
I
I
r-}--- ----- .J
l - -,
C ool ing
water
pump
<D Wash ing hot water

Seed
@ Level
Fig. 10 75
Concentra tio n
Process
Crystallizer
Pan pressure
Crystallizer control loop configration
Seed ing
Harden ing
C rysta II izi ng
Rheometer
va lue
(Hard ness)
Feedwater valve
Syrup valve
Seeding va lve
Pan level
Fig. 10 76 Operation of intermiuent syrup boiling ystem
10.4
703
f.
Fig. 10 77
Hardness meter (Rheometer)
In the intermittent syrup boiling system, fixed amounts of feedwater or syrup are alternately added to the pans as predetermined
rheometer (hardness) setpoints are reached, until the sugar crystals
achieve their prescribed size.
Briefly, the operation works as follows.
Syrup is charged into the pan and boiled rapidly until supersaturation is reached at setpoint (!). Seed crystals are added at setpoint @ ,
and feedwater is added at setpoint @ . Then, at setpoints from @) to
@ , feedwater is intermittently added (feedwater boiling). This is
followed by syrup boiling in which syrup is intermittently added at setpoints from (j) to @. Finally, the hardening is finished at setpoint @ .
It can be seen that various control actions are triggered in the
crystallizer as rheometer values reach setpoints from CD to @. Next,
each addition of syrup during the syrup boiling phase is controlled by
detecting a series of level in the pan from to (f).
(c) Total production control system at a sugar refinery
Today, sugar refineries, in common with many other kinds of
plants, must find new ways to conserve energy, resources, and labor.
704
More than ever before, advances will be won by improving product

quality and by catering to the more diversified tastes of consumers by
implementing production systems capable of turning out a great range
of products in small lots. Key to this development is the establishment
of integrated production control systems based on computers.
(1) Hierarchical production control system structure
A production control system is an information management structure for organizing the
information required to produce a product of consistent quality and
within in fixed time period by the most economical means possible.
The task is complicated by circumstances, both within and external to
the plant, that are easily subject to change.
Such a structure cannot be realized by merely mechanizing part of
a single process. Rather, the entire productive system must be optimized so it achieves economic balance.
Integrated production control systems commonly assume a hierarchical configuration with purview over every step from receipt of
the raw material to shipping the finished product.
At a sugar refinery, since the sugar process and the packing and
delivery process have different functional requirements in terms of control, different control systems are applied in each case. Also, because
the information used by the various systems making up the hierarchical structure is different, a centralized control system is necessary to
combine and integrate the disparate information from the other systems.
The structure of a hierarchical integrated production control sys-
Fig. 10 78
10.4
Hierarchical system structure of sugar refinery
705
tern for a sugar refinery is shown in Fig. 10.78.

(2) Production control functions
Production control functoins can be
separated into the optimal schedule level and the operation control
level.
The optimal schedule level, starting with the production schedule
based on the business plan, also includes the boiling schedule and the
packing schedule. Once these basic schedules are in place, the production volume (melt volume) can be established for the process control
system. This makes it possible to determine the operating schedules
for the crystallizer, the product separator, and the packer.
The operation control level encompasses a great number of control
items including boiling centrifuge control, product inventory control,
and delivery control. Through interface with the various control systems which monitor conditions throughout the plant, it has become possible to implement on-line operation control in real time.
The hierarchical structure classified according to functions is
shown in Fig. 10. 79.
10.5 Instrumentation Applications in the Paper Manufacturing
lndustry 20 ) - 23 ), 38)
10.5.1 Overview of an integrated paper mill
Paper manufacturing plants can be broadly classified as either
pulp plants or paper machine plants. Pulp is delivered to the latter
type of plant as the main raw material for the manufacture of paper.
Pulp plants produce pulp from wood or other fibrous raw material.
There are two general processes for producing wood pulps; mechanical
and chemical. In this section we will be primarily concerned with kraft
pulp processes, that fall under chemical processing.
A schematic overview of the kraft pump manufacturing process is
shown in Fig. 10.80. The wood for raw material is first put through a
chipping process where it is cut up into small, uniform-sized chips.
The chips are then conveyed to the cooking process, where they are
charged to digesters. A chemical solution, i.e., cooking liquor, is added
to the chips and the mixture is cooked with steam at a temperature of
about 170C at a pressure of several tens of kg/cm 2 After reacting
over several hours, lignin and other interfiber bonding materials in
chips are dissolved. When the cooked chips are discharged from the
digester to the blow tank, they are reduced to a suspended fibrous
slurry called brown stock that is already termed pulp. After washing
and screening, the pulp is then sent to the bleaching process. Here the
pulp is treated with the bleaching agents chlorine, sodium hypochlorite
(hypo), chlorine dioxide solution, and sodium hydroxide. Oxygen is
sometime also applied at this stage to remove residual lignin still adher706
'l
s"'
~-
,
1:)
"'
~
:::!
..
'"1::1
2;:::
o.
Be>.>-u
1;!
]=
"';;
:::!
<:;
~r
-....
~-
....~
~
:::!
~
"'....
c.,
o;:::,
......
Maintenance_
management
I
Boiling and melt
instructions/Result
Centrifuge
instruction/Result
Boiling centrifuge
control
Boiling schedule
I
I
Storage and
delivery
instruction
Delivery
instruction
Analysis line Packing line Automated

(Card leader
control system control system warehouse
\.'
control device
Packing
Automated
Laboratory
Delivery
process
warehouse
Analysis
daily report
Silo/Bin storage
and
Delivery control
---------------
Product
inventory control
~---
Packi~[l~
schedule
Packing process
submaterial control
Fig. 10 79 Hierarchical structure classified by functions
Raw sugar acceptance, Washing and melting

processes-Purification process-Crystallization
and centrifuge processes-Silo and bin utility
Process control system
L-----....1
Work totalization
control
Purification process
submaterial control
Melt capacity
Boiling schedule
for each sugar grade
a usticizing tank
<D
White liq uor tank
Lime slaker
@ D issolvi ng tank
@ Thick black liquor torage tank

@ High-concentration evaporator
(jJ) Evaporator
@ Ch lorina tion tower

@ Ex traction tower
Unbleached pulp chest
Hypo tower
White liquor clarifier
Th in
Lime mud filter
@ Di gester
@ Thickener
Lime kiln
@ Blow ta nk
Green liquo r clarifier
@ Knotter
@ Bleaching chest
@ leaner
Screen
<V
Recovery bo iler
Econo mi zer
b lack liquor storage tank
Fi ltrate tank
hl orine dioxide tower
Thickener
Bleached pu lp tower
Fig. 10 0 Outline of pulp plant
ing to the fibers. After passing through the bleaching process, the
pulp is washed and screened again and sent as raw material to the paper machine plant. In the meantime, the spent cooking liquor, now
called black liquor, is put through an elaborate recovery system. After
being concentrated through a condensation process, the organic
black-liquor solids are burned in recovery boilers and the inorganic
compounds are subjected to reduction recovery. The inorganic compounds (green liquor) are put through an additional process where it is
causticized with lime. The resulting clarified white liquor is returned
to the digester as cooking liquor.
A typical paper plant is schematically shown in Fig. 10.81. In the
initial stock preparation process, various kinds of pulps are mixed according to the type of paper that is to be produced. Also, fillers, dyestuffs, and other materials are added at this point depending on the surface smoothness, strength, color, and opacity desired for the final product. In the machine process, the prepared stock is delivered as a slurry
(web) onto a moving endless finemesh wire screen (wire), where water
is drained from the web. More water is forced out in the following
press section by press rolls. Then, the web is dried in steam driers to
708
Chap.lO
/nstrumention to Manufacturing Industries
From
pulp
( Stock preparation proces )

<Blending >
Other pulp/ hemicai/ Filler <J)
( Machine proces )
~--1 (!)
p~lant ~~
<D
Refine
r
Receiving chest
Refi~er chest
@ M1xmg chest
(!) Mixing box
~ . _@ .
...
@ Head box
Wire part
Press
Q)l Pre-dryer
Machine chest
@ Size press
@ After dryer
@ Calender
"
Cutter
.c:::ac. ~~
Fan pump
Chemical
Screen
<Dry pan >
<Wet pan >
=~
@ 0) ~ p-j 8 tP
.
Rew1ndcr
~~
e::::J
R 11
( Finishing process )
o
Fig. 10 81 Outline of paper machine plant
become paper. Throughout this processing, such qualitative factors as

basis weight (weight per unit area), moisture content, and thickness
are carefully controlled. The finished paper may then be sent on to a
coating machine for application of sizing or coating to one or both surfaces.
Finally, the paper is sent to the finishing process where it is
wound onto winders, cut, and/or trimmed to standard sizes, and
wrapped for shipment.
10.5.2 Pulp plant instrumentation
[ I ] Cooking-process instrumentation
The disgestion process is carried out in a digester, of which there
are two basic kinds; batch and continuous.
To illustrate instrument applications, we will describe a Kamyrtype continuous digester, the most common type of digester used in
Japan.
(a) Key instruments for continuous digesting
A typical continuous digester is a cylindrical vessel 20 to 80 meters in height and 3 to 8 meters in diameter. As can be seen in Fig.
10.82, metered chips pass through steaming vessel where they are preheated and air is expelled. The steamed chips are then continuously
fed concurrent with cooking liquor into the upper portion of the digester by a high- pressure rotary feeder. As the chips continuously move
downward in the digester, the cooking reaction proceeds with additions of heat from the cooking liquor. More heat is provided by recirculating liquor that has been withdrawn and heated. By the time the
chips have spent about half their retention time in the digester, the
cooking reaction is more or less completed. As the cooked chips reach
10.5 Instrumentation Applications in the Paper Manufacturing Industry
709
the bottom zone of the digester, they are plowed to a central well in
the bottom of the digester while being counterflow-washed with washing liquor. The chips are then discharged from the bottom of the digester by means of a bottom scraper.
In the operation of a continuous digester, the ideal is to optimize
two sets of values; the objective is to stay as close to the optimum
chip-cooking values as possible while at the same time maintaining the
various process parameters, which become primary factors of operation cost, at their optimum values. It is particularly important that the
following are controlled:
(1) Chip/pulp balance,
(2) Cooking liquor addition ratio,
(3) Temperature at different zones in the digester, and
(4) Degree of washing.
(b) Instrumentation methods
(1) Chipfpulp balance
(i) Chip feed control: The amount of chips fed into the digester is determined by the rotational speed of the chip-meter. The chip-meter speed
control is set based on the desired value of rotation, which is calculated from the target pulp production volume, the chip moisture content,
the filling factor of the chip-meter, and the basic pulp yield.
(ii) Chip-bin level: To assure a uniform and stable chip-meter filling factor (i.e., the quantity of chips delivered by one full turn of the chip-meter), the level of the chip bin must be controlled. The depth of chips in
the bin is detected by means of either a ultrasonic level detector or a
load cell, and control is maintained by regulating the feed volume from
the chip silo.
(iii) Chip level control in the digester: To maintain the chip level inside
the digester within a desired range, feedback data is used to adjust the
chip feed volume. Chip level is detected with a gamma-ray level indicator.
(2) Cooking liquor addition ratio
Referring to Fig. 10.82, it can be
seen that the desired flow rate of cooking liquor is obtained by multiplying the ratio (addition ratio) times the chip feed volume. The basic
addition ratio value is determined by the degree of cooking desired and
the type of tree the chips are derived from. This basic value is then
fine-tuned by factoring in cooking liquor alkalinity, moisture content
of the chips, and so on.
(3) Temperature at different zones in the digester
(i) Impregnation zone temperature control: The upper portion of the
digester above and around the cooking circulation strainer is called the
impregnation zone, because it is here that the chips are impregnated
with cooking liquor. The temperature in the impregnation zone is de710
.......
....
"'
'"I
;lt
~
.;,
'"I
"';;;
;lt
<:;
"
1'::.
~;lt
--
~
~
Extract strainer
Cooking circulation strainer
'
L-
L ----
o\LJ
i.... -~-,
:aonom 1 :
I scraper I I
: speed
: :
1 control ~ 1
r--..,
SIC
,-e
~---- ---~
I
I
Fig. 10 82 Continuous digesting process and control
Chip moisture content _._r----,,-,--,

Alkali concemration
Addition ratio __,
,
11
(j) Dilution water now control
Cooking circulation temperature

comrol
@ Ca lculation for washing liquor now Hi h
@) Impregnating circu lation strainer
ro~ary
CD
- Chip moisture contenl

Pulp yield
Filling factor of chip meter
termined by the temperature of the cooking liquor that is circulated in

the zone, which is controlled at about 145 to 150C.
(ii) Cooking circulation temperature control: The section of the digester
from the cooking circulation strainer to near the extract strainer is
called the cooking zone. The cooking reaction, in which most of the delignification occurs, takes place here. Temperature as well as retention
time in the cooling zone have a major influence on the cooking reaction; in fact , these are ultimate controlling determinants of the degree
of cooking. The cooking zone temperature is regulated by the temperature for the cooking liquor circulating in the zone, and is controlled at
about 165 to 170C. In some cases the temperature of the cooking zone
is monitored directly by inserting a thermal sensor into the digester.
(4) Degree of washing (Fig. 10.82)
Washing flow control: The flow rate of washing liquor (thin black
liquor) depends on the flow rate of bone-dry weight pulp descending in
the digester and the extent of washing desired. The ratio of the flow
rate of bone-dry weight pulp and flow rate of washing liquor is called
the dilution factor (DF). By multiplying the flow rate of bone-dry
weight pulp times the dilution factor, the desired washing flow rate
can be obtained. The dilution factor is set according to the degree of
washing desired.
[ H] Bleaching-process instrumentation
The bleaching process in pulp plants has undergone a true technological revolution. The traditional single-stage hypochlorite bleaching
method or batch bleaching method using diverse bleaching agents has
given way to a modern multistage bleaching process.
(a) Multistage bleaching process overview
In a typical multistage bleaching operation, separate reaction towers (bleaching towers) are implemented for each bleaching agent. The
pulp passes through the towers one at a time where the successive
CD
Chlorination tower
Fig. 10 83
712
@ Steam
Seal tank
Extraction tower
Washer
Hot water
(j) Chlorine dioxide tower
Bleached pulp chest

Bleached pulp
Four-stage bleaching proces flow
Chap.lO
bleaching operations take place. Fig. 10.83 shows the flow of a typical
four-stage bleaching process. The various bleaching stages have different functions as follows:
(1) Chlorination stage
The unbleached pulp is subjected to chlorination treatment which facilitates extraction of the chlorinated lignin in
the following process.
(2) Caustic soda stage
In this stage, lignin and impurities that impart color are extracted.
(3) Following the hypo stage
Remaining lignin is oxidized, extraction
is repeated, and the desired degree of whiteness is obtained.
(b) Instrumentation methods for each stage
(1) Chlorination stage
Figure 10.84 shows typical instrumentation
for the chlorination stage. The final objective of this stage is to
achieve a consistent kappa number (a measure of residual lignin) by the
end of the caustic soda extraction. The amount of chlorine supplied is
proportionate to the kappa number of the unbleached pulp.
(i) Chlorine feed control: The amount of chlorine injected is determined
by first deriving the bone-dry weight which equals the volume flow of
the pulp times pulp consistency. This product is then multiplied by the
----- ....
Cl 2 Gas - - - - ---'
I
I
Eject water
o
addition 1rau%
'--
----~Chlori ne now
MV
lI~s
~ M~-
Pulp consistency
$--~----I
1
~
I
I
I
I
--E:51
C1
:I
I,
~=:ql---"
c ---,
Dilution water
ra!_e comrolr-/
SV-,
'----=--
FIC
I
I
resid ual chlorine
I
I
.-----oo.
To
' ]1l
..,._.-.J
~Indication of
PV
_j__
---1
Ch lorination
tower
extraction stage
whit.enes meter
CD
Flow rate of bone dry pulp
@) Chlorine addit ion ralio
In -line mixer
CompensaJion for reaction velocity

Flow rate of bone dry pulp
@ ORP control or whiteness control
Fig. 10 84
Control of chlorination stage
713
chlorine addition ratio to derive the desired value for flow control.
Once the chlorine has been added, residual chlorine is detected
with a ORP meter for the purposes of feedback compensation. One
problem with this method is that since the injected chlorine reacts
very quickly after injection. The level of residual chlorine is indicated
by the ORP meter, but flow speed and temperature fluctuations can alter the reaction speed so that errors are introduced. Thus, calculations
to compensate for the time lag resulting from the detection position of
the residual chlorine, as well as pulp liquor temperature, have to be factored in.
(ii) Introduction of a whiteness gage: The ORP meter in the chlorine
feed control method indicates the inverse response of the residual chlorine to changes in the kappa number of the unbleached pulp. It is thus
fairly difficult to get the right amount of chlorine to be injected. Recently, a feedback control technique has been developed to overcome
this problem. It uses an optical sensor to detect pulp whiteness (whiteness gage).
(2) Caustic soda stage
A flow diagram illustrating control of the caustic soda stage is shown in Fig. 10.85. The control objective for caustic
soda feed is to maintain a consistent extraction kappa number after extraction. Because there is a high correlation between the extraction
kappa number and pH value at the outlet of the caustic soda stage, pH
control can be used to manipulate caustic soda feed volume.
(3) Following the hypo stage
Control after the hypo stage is based on
maintaining a constant ratio between the bone-dry weight flow of the
pulp that has come through the previous two stages and the volume of
<D
(j) Flow rate of bone dry pulp
Dead time compen ati on

Alkali add ition ratio
@) pH Control (Sample-a nd -hold PI control)
Extract ion lower
Seal lank
Fig. 10 5
714
Chap.JO
Jnstrumention to Manufacturing Industries
added chemicals.
As necessary, a whiteness gage is mounted at the outlet of either
the hypo stage or the final stage. The detected whiteness is used to
optimize the chemical flows of all the bleaching stages.
[ m] Recovery processes instrumentation
Chemical recovery processes include black-liquor concentration, recovery boiler, and causticizing and lime recovery processes. Here we
will focus on
(1) Black liquor concentration process, and
(2) Recovery boiler process.
(a) Black liquor concentration process
(1) Overview of black liquor concentration equipment (evaporator)
The
liquid discharge from the digester to be recovered is called thin black
liquor. The main constituents are cooking chemicals and lignin dissolved out of the chips. It has a soluble-solids content of 12 to 16%.
Since the black liquor cannot be burned in the recovery boilers as is, it
is first sent to evaporators where the liquor is concentrated under pressure to a solids content in excess of 60%.
Today, the evaporation is generally carried out in a multiple-effect evaporator with superior heat efficiency. The process flow of a
multiple-effect evaporator is shown in Fig. 10.86.
(2) Instrumentation methods
(i) Physical distribution control through evaporators: As can be seen from
Fig. 10.87, flow rate to the first evaporator must be controlled at a constant value coresponding to the volume of thin black liquor to be processed. Extraction of concentrated black liquor from each evaporator is
H igh concentrat ion
evaporator
Cooling water
Foul condensate
Thick black
liq uor storage
tank
To thin black liquor storage

tank
Th in black liquor
Q) To thin black liquor storage tank

intermediate tank
To
To
recovery boi ler
Fig. 10 86 Multi pleefTect evaporator process
715
MV~
r----Yo CD
IPV
High -concentration
evaporator
No. I
Thick
black
liquor
I
I
I
I
__ JI
L--- liquor
CD
hick black liquor concentration control

Bl ack liquor concentration meter
A Chamber
CV
B Chamber
C
hamber
<V
Feed liquor flow control
Level control
Ote : Black liquor pa es through the system in the rollwing

Sequence : o.5 B No.4
o.5A No.3
o .2
o . IC
Fig. 10 87
Intermediate tank
o. IB
Physical distribution control and thick black liquor concentration control
determined by feedback control from the liquid leveL An intermediate

tank positioned before the high-concentration evaporator serves as a
buffer ; it absorbs fluctuations in the flow distribution balance. The volume drawn off from the high concentration evaporator-i.e., the volume of finished thick black liquor produced-is determined by the tick
black liqure concentration control (described below). The liquid level of
each chamber is kept constant by feedback control of the feed volume.
(ii) Washer-vat control: Within the high-concentration evaporator,
washing is performed in a succession of chambers, A, B, and C.
Cut- over control of the various black liquor feed/extraction lines of
each chamber (steam/drain lines, thin black liquor feed for washing/extraction lines) is automated using sequence controL A wash cycle takes
anywhere from 3 to 5 hours.
(iii) Thick black liquor control: Concentration of the thick black liquor is
continuously monitored with a gamma-ray density meter or optical- type consistency meter mounted at the outlet of the high-concentration evaporator. The concentration of the thick black liquor is
controlled by feedback control, with the flow of drawoff extracted
from the evaporator.
Immediately after cut-over between chambers, since the concentration in the wash pan falls off substantially, valve opening is controlled
(as shown in Fig. 10.88) so as to minimize the duration of the transient
falloff in extraction concentration. Moreover, while steady-state thick
716
Chap .JO
CD
Steady state
@ Washing of thick black liquor

. l
Mv ( v ave
I openmg
Waiting for concentration to rise

Functional control
Steady state
Tt to Tz: Washing time for blow line

T3 to T, : Ri e time for concentrati on
control
TJ
Time
Fig. 10 88 Action of concentration control in transient phase
black liquor is sent to the thick black liquor tank, if the concentration
falls below a standard value, the line automatically cuts over to either
the intermediate tank or the thin black liquor tank as appropriate so
the liquor can be recycled.
(iv) Steam-line control: Steam, the evaporator heat source, is supplied
to the high-concentration evaporator as shown in Fig. 10.89. Steam
generated from the black liquor by heat exchange is used as a heat
source for the previous-stage evaporator. Then, in all succeeding downstream stages, generated steam is reused to maximize heat efficiency.
The volume of steam supplied depends on the volume of black liquor
fed to the evaporators. Also, where scaling is advanced (of course, every effort is taken to inhibit scaling, by such means as optimizing the
wash cycle), steam pressure may increase excessively at inlets. Thus,
steam pressure and steam flow override controls are sometimes implemented.
(b) Recovery boiler process
While the basic structure of a recovery boiler
for burning black liquor is more or less the same as a typical heavy-oil
boiler, in terms of volume of steam generated (approximately 50 to 250
Steam n ow
control
SV
Steam pres ure

control
SV
~~-Y_-,Ml~ r-M----, CD
,PV
<D
~ PV,
Vaporized steam
Flushing steam
Heating element
Fig. 10 89
10.5
Steam flow and control
Instrumentation Applications in the Paper Manufacturing Industry
717
H : Super heater
BB : Boiler bank
EP : Electric precipitator
Air
Chimney
<D
Salt cake
Black liquor feed nozzle

Weak liquor
Green liquor
Fig. 10 90 Schematic of recovery boi ler
t/h), it is much larger in scale. The basic configuration of a recovery

boiler is shown in Fig. 10.90.
The concentrated black liquor from the evaporator is first heated
and then injected into the furnace by black liquor spray-guns. As the
particles of black liquor descend in the furance, water content is vaporized, they are heated, dried, thermally decomposed, and the volatile
component (i.e ., organic) is burned. The burning operation leaves an accumulation of smelt called a char bed at the bottom of the furnace,
which is continuously removed from the furnace.
The main objectives of the recovery boiler operation are to obtain
the maximum energy (maximum amount of generated steam) from stable combustion, and to obtain the maximum rate of chemical recovery.
The main instruments to realize these objectives are discussed below.
(2) Instrumentation methods
(i) Combustion control
i) Air distribution ratio : The distribution ratio of combustion air is
important not only in terms of combustion efficiency, but also as an
extremely important parameter influencing. the reduction rate (described below). The distribution ratio is determined as a proportion
of all distribution flows making up the total air flow. An effective
method of finding the optimum ratio, however, is to monitor the
temperature distribution inside the boiler.
ii) Combustion control: It is a basic premise that the role of a recov718
Chap.lO
ery boiler is to burn all thick black liquor produced. It is therefore

not possible to adopt steady control of fuel supply as a standard for
generated steam pressure (boiler master), as is usual with ordinary
boilers. Moreover, because the heating value obtainable from fuels
varies depending on the solid content of the black liquor, it is difficult to apply conventional air/fuel ratio control techniques. Thus, in
order to manage and control the combustion, 0 2 and CO concentrations in the exhaust are detected and total air flow adjusted to obtain the optimum values for those concentrations. In cases where
feedback control is applied, since much dead time in the system can
be attributed to boiler capacity, a control algorithm designed to
cope with dead time has become necessary.
(ii) Char bed monitoring: The function of the char bed is recover the inorganic component in the black liquor reduced through chemical reaction at high temperature. To enhance reduction efficiency, the upper
part of the char bed should be kept at the optimum temperature (about
1050 to 1150C), and the optimum shape, or profile, of the char bed
should be maintained.
(iii) Soot-blower control: Every furnace unit of a recovery boiler generates a surprising accumulation of dust, which consists mostly of salt
cake. Soot blowing (sweeping with forced steam) is necessary much
more frequently than in the case of ordinary boilers. Applied to a recovery boiler, soot blower are usually implemented as follows. The boiler is divided into a number of sections; boiler bank, super heater, economizer, and so on. These sections are then further subdivided into
blocks, which are sequentially cleaned according to a prearranged
schedule.
10.5.3 Instrumentation applied to the papermaking process

(a) Stock preparation-process instrumentation
The object of stock preparation is to improve the suitability of the
pulp delivered from the previous process with chemical additives and
pigments depending on the quality and end use of the paper.
Another important treatment of the stock is mechanical beating in
a refiner. This improves the felting qualities of the fibers by fibrillation, causing them to bond together TUore readily in the papermaking
process to follow.
An overview of a continuous stock preparation process is shown in
Fig. 10.91.
Reflecting the objectives of preparation processing, there are
three main instrumentation applications:
(1) stock mixing ratio control
(2) chemical additive/pigment mixing ratio control
719
(3) beating degree control (freeness control)

These controls will now be taken up for closer consideration.
(1) Stock mixing ratio control
The purpose of this control is to deliver various stock pulps to the mixing chest in the proper predetermined proportional mix. The loop configuration for the control is
Pu lp torage
tank
1
Broke recovery
process
Fig. 10 91 Continuous tock preparation p rocess
a,
100
Q., = Qr X toO X c;-
Q.,: Set point of now rate for each materia l

Qr : Demand of now rate for total material
a; : Blending ra tio for each material l:a; 100 [%)
c, : Pulp consistency of each material
To machine chest
p, : Blending
ra tio for each chemica l and pigment

a, : Consistency of solid particles in each chemical and pigment
CD
Blending rat io calculation

Raw materials I
@ Raw materials 2
@ Raw materials 3
Chemica ls
(i) Pigments
Chemical 2
Fig. 10 . 92 Blending ratio control for materials as well as chemiea l a nd pigment
720
Chap.lO
The total stock flow into the mixing chest depends on the consumption of stock by the papermaking process. However, for the general control system, to keep the level of the mixing chest constant, the
manipulated variable of the level controller is taken as the demand signal for total stock flow.
Quantities of the various pulp stocks are simply determined by
multiplying the proportional factor by total required stock flow.
(2) Chemical additive pigment mixing ratio control
The quantity of
chemical additives and pigments added to the stock are figured as a
proportion of total stock sent to the mixing chest (not including stock
recovered from the paper-making process).
In the case of chemical additives and pigments, just as in the case
of the pulp stock, consistency of solids in the stock must be taken into
consideration in determining the flow setpoints.
(3) Beating degree control
The purpose of beating degree control is
to assure a uniform degree of beating (in terms of freeness, a measure
of pulp fiber cutting and fibrillation) of the pulp in the refiner.
In a typical configuration, a refiner has two disks, of which one or
both rotate and which are grooved with sharp-edged ridges. Watered
pulp passes between the two disks which cuts and fibrillates the fibers.
The degree of beating is varied by adjusting the clearance between
disks using hydraulic pressure.
For measuring the degree of beating, the conventional method has
been to use a batch-type sampling measuring device (a freeness tester,
which measures the filtered water content of the watered pulp, can
also be used to indirectly measure the degree of beating). This technique has drawbacks, however; the sampling time to carry out the
measurement is overly long, and it is impossible to eliminate short-cycle disturbance on the basis of feedback control by means of this signal
alone. A better alternative is to make use of the close correlation between motor load power and degree of beating. With this approach,
disk cleanrance is adjusted by increasing the motor load power until
the desired degree of beating is reached. It should be noted that when
pulp is fed into the refiner, pulp flow rate and consistency changes affect the load on the motor, and hence the beating degree changes.
Thus, feedforward control based on these variables is also be applied
to the motor to adjust the load as appropriate. The control loop configuration is shown in Fig. 10.93.
(b) Papermaking process instrumentation
In the paper-making process, prepared stock is delivered to a paper machine where paper is formed by draining water from the stock.
An overview of the process is shown in Fig. 10.94.
The stock is first diluted to around 1% in consistency, and it flows
721
r - - Freeness et
~.-----------------------:
I
: .d Electric power
Electric
5:.; \/
r-t-0>~~~ 1!8.
~
w Flow
rate
of bone
dry pulp
r-
l
adJUStment I
I
- - -,
Pulp
~onsistency
~
I
~
I
I
I
- -,
Load power
Pulp flow rate
Fig. 10 93
]
:
IMoving average I
1..-M-ea---'
n ~v"""'alue-,1
~ v
l~ara nce
~--+~--~
I
I
I
I
I
I
Freeness
Loop diagram of beating degree control
8 / M Sensor
Fig. 10 94
Paper making process
from a headbox to the top of a moving endless belt of wire screen

where it is uniformly dispersed. Drainage from the stock through the
wire screen is induced both by gravity and vacuum. In the press part,
more water is extracted by multiple press rolls. In the dry part, the paper is dried by steam-heated cylinders, and is finally wound into rolls
with a reel device.
Here we will focus on instrumentation applied to (1) the headbox,
and (2) the dry part.
(1) Headbox instrumentation
When the headbox delivers the diluted
stock on the wire, the delivery speed uj (jet velocity) must be controlled so it is in fixed relation to the speed of the wire screen uw in or722
der to obtain the best formation of paper (disposition of the fibers).

This ratio is called the jet/wire ratio a. The three operable parameters
are related as follows
The pressure (total head) yielding this jet velocity must be applied on
to the headbox. This value is given by
1
TH= 29 u]
where TH is total head (mAq), and g is gravitational acceleration ( =
9.8 x 60 2 m/min 2).
Using the pressure derived by this equation to control the headbox
is called total head control, and represents the single most important
value for headbox instrumentation.
In total head control, pressure in the headbox is measured and controlled by the stock flow rate, or the air pressure applied. Since adjustment of the stock flow rate can be varied over a wide range and high
precision is also required, a combination of rough-adjustment and
fine-adjustment valves is provided. In recent years high-precision motor rpm control has become a reality. Extending this approach, and to
conserve energy, the trend is toward using pump speed to control the
flow rate.
(2) Dry part instrumentation
After extraction of water in the press
part, the paper web still has a water content of approximately 60%.
The task of the dry part is to remove most of the remaining water
down to about 6%.
In the dry part, the paper web is dried by being held in contact
with cylinders that are internally heated by steam.
The steam flow configuration for a typical system is shown in Fig.
10.95.
Most paper machines are the so-called multiple-drum type, operating with anywhere from 40 to 100 cylinders. These cylinders are divided into 3 to 4 groups: a high-pressure, a middle-pressure, and a
low-pressure group (facing from the reel back toward the headbox).
Appropriate pressure differential is maintained between the groups (inter-group differential pressure). Steam is first supplied to the
high-pressure group. After the steam condenses, the water is removed, reconverted to steam in a flashtank, and successivley supplied
to the lower pressure groups.
Another common heat-recovery measure is to install a hood over
the drier section. This makes it possible to effectively capture for
reuse a major portion of the steam coming off the paper in the drier
723
FR
) -- _ J
CD
Main steam
Low -pressure dryers

Middle pres ure dryers
@ High-pressure dryers
@
ondenser
Fig. 10 . 95 Typical dryer steam now systew
section.
(c) Web quality management and control
There are various aspects that must be considered in web quality

control:
(1) basic quality considerations such as basis weight, thickness, and
dimensions related to terms of trade ;
(2) Structural quality relating to such attributes of the paper fibers
as dispersion, smoothness, tear strength, and tensile strength;
(3) optical quality considerations such as hue, whiteness, opacity;
and
(4) absence of such problems as shrinkage, wrinkling, curling that
are strongly influenced by moisture content.
Particularly important parameters in this regard are basis weight
(weight per unit area 1 m2 ) and moisture content. These two parameters are important not only in their own right, but also because they
have such a major influence on other quality attributes. We will focus
here, therefore, on the system for controlling basis weight and moisture content.
Basis weight and moisture content are measured by B/M sensors
(also see Chap. 3, Sec. 3. 9).
Basis weight-moisture content control is implemented in both machine and cross directions.
(1) Machine directional basis weight control
The control object is to
find the bone-dry basis weight from the basis weight and the moisture
content. The final control element is a high- accuracy stock control
valve (an adjustment resolution over 1/1 000) which is mounted below
the stock box . The control loop has the following characteristics:
724
(1) dead time is long (40 to 200 sec.), varying with the speed of the
machine;
(2) time constant is comparatively short (10 to 50 sec.); and
(3) measured value is obtained discretely (the average measured value is used in the cross direction).
Given these characteristics, since good results cannot be obtained
using a simple PID algorithm, sampled data control based on the finite
time settling response method is applied, with optimization of the dead
time by adjusting the speed of the machine.
Manipulated and process variable movements for the finite time
settling response control method are shown in Fig. 10.96.
Set
poim__,~l
_ _,_,- - ---+- -> ---.:J.L:_ _
r,
r,
r,
Tz
r.
Increment in
manupulated-+--'- >----t---l-...,.._--+----f.- T c -l
variable
: ontrol period
I
Process
, ............ -dR- -(~-:._-e
In )
!["--------+
1- a
variable-+r---+-~4--1-+--1--+----+-___..:_L__
I
1-- L ----1-r c-l
K t -LS
Process model 1 + ST
a = exp (- Tc/ T)
Fig. 10 96
Finite time settling r pon e control
This control is main(2) Machine directional moisture content control

tained by adjusting the pressure of the steam supplied to the driers.
Characteristics of the control loop are as follows:
(1) dead time is relatively short (10 to 30 sec.);
(2) time constant is long (50 to 500 sec.);
(3) measured variables are obtained discretely; and
(4) interference is received from the basis weight (every 1% change
in bone- dry basis weight produces a 0.1 to 1.0% change in moisture content).
Therefore, taking these characteristics into consideration, sampled
725
<D
Basis weight control algorithm

Non-interference algorithm
Moisture control algorithm
@ Setpoint of bone dry basis weight
Fig. 10 97
Measured value of bone dry basis weight

Measured value of moisture content
(J) Set point of moisture content
Basis weight/Moisture content control
data control based on bone-dry basis weight non-interferen ce and the

finite time settling response method is used. The basis weight/moisture content control is shown schematically in Fig. 10.97.
(3) Cross directional basis weight control
The headbox delivers a jet
of stock through a slice across the full width of the machine. The slice
openings are spaced about 150 mm apart and are adjusted by means of
slice bolts. The slice bolts are adjusted by matching transverse data detected by sensors (profile data) to a desired profile.
(4) Cross directional moisture content control
This control is implemented with auxiliary driers arranged so the entire width of the machine is divided into zones. This control loop is simple compared to the
cross directional basis weight control loop.
10.6
Waterworks Instrumentation Applications 24 l
- 32 )
10.6.1 Overview of waterworks facilities

In broad terms, waterworks encompasses facilities for the intake
and conveyance, purification, and distribution of water. Basic conditions of intake and conveyance facilities are that the planned intake
volume can be regularly assured and that the quality is good. The volume of water intake is controlled by intake gates and by controlling
the speed and number of pumps in operation.
Since water intake points are often located at some distance from
the water purification plants, supervision and control is generally exer726
Chap.lO
cised from the central control room at the purification plant by means
of telemeterftelecontrol (TM/TC) systems. Purification facilities include both purification processing and drainage/waste sludge processing. The former consists of arrival, settling, and filtering; while the
later encompasses facilities for washing drainage, settling waste
sludge, and treating sludge. The object of purification processing is to
remove the various particles, microscopic organisms, and bacteria
found in raw water thus making it fit for consumption as drinking water. The most common method of removing these contaminants is the
chemical sedimentation-rapid filter method. Typical water treatment
processing, from intake to purification to distribution, is schematically
10.6.2 Water treatment-related detectors
Detection devices commonly employed at waterworks facilities include general measuring devices for flow rate, level, and pressure; and
water-quality measuring devices. This latter class of devices is used
not only to supervise the quality of the water, but also to gage the effects of chemicals injected into the water. Since it would be impossible
to cover the whole range of waterworks-related devices in the few
pages allotted, only the most important devices and their uses will be
described.
(a) Water quality detectors
The most widely used water quality devices are the turbidimeter,
pH meter, residual chlorine meter, alkalinity meter, and electric conductivity meter. The points where these devices are applied and what
they are designed to detect are shown in Fig. 10.99.
(b) Common measurements
Waterworks, of course, deals with water that is stored and water
that is flowing. Level gages and flowmeters are thus essential to quantitatively measure water in these two basic states.
(1) Level gages
Among the various devices available to detect water
level, the most widely used are the float-type, differential-type,
throw-in-type, ultrasonic-type, and electrical capacitance-type level
gages. The features and uses of these various types of level gages are
summarized in Table 10.7.
(2) Flowmeters
Flowmeters are employed in processes for a variety
of different reasons. The most common uses are: for the purpose of
making estimates, for control (e.g., to determine the quantity of chemicals to be added to the water), to create a record of operations (i.e., by
measuring the volume of inflow, outlfow), to keep track of transaction
volume for the water-supply industry (on either the supply or delivery
side). In selecting the most appropriate type of device, degree of accu10.6 Waterworks Instrumentation Applications
727
CD
Sand basin
Pumping- up pump
Q) Water arrival well
Low-rate mixing basin

Waste Judge basi n
(j) C hemical sedimentation basin
(!) Flash mixing ba in
Fig. 10 98 Outline of
Table 10.7 Features and uses for various applications of miscellaneous liquid level
meters
Liquid level meters Features and uses
Float type
Simple principle and simple construction. This type is used in

open tanks such as sand basins or water arrival wells. If there
is any fluctuation on the liquid surface, however, a buffle tube
must be installed.
Differential type
This type is used for liquid level measurement of chemicals and

other storage tanks. The meter is installed at the low part of
the vessel and there must be enough space to access the meter.
Throw- in type
Diaphragm is used to detect liquid pressure. Simple installation

and simple maintenance.
Ultrasonic type
Certain amount of error will be introduced if there are any obstacles between the sensor and the surface of riquid or if dew
and drops are produced on the sensor. This type can also be implemented to measure in a flow .
Electrical
capacitance
type
Offers the widest measuring range and is therefore suitable for

measurement not only in tanks and basins on water purification
plant , but also in deep wells. Considerable error will result if
anything adheres to the electrodes.
718
Chap.JO
@ Rapid filler
Washing drainage ba in
Chlorine mixing chamber
Reservo ir for clean water
@ Water su pply a nd di stribution pump

@ Pump tation of pressurizing fo r d istribution reservoir
~ D istribution piping network
water works facilities
racy required and economy must be taken into consideration. Flow

measurement methods can be broadly divided into closed-conduit and
open-channel methods. Flowmeters representative of the closed-conduit method include electromagnetic, differential pressure, and ultrasonic flowmeters. For the open- channel method, there is the weir
flowmeter. For waterworks applications, electromagnetic flowmeters
are most widely used. Not only are electromagnetic flowmeters extremely accurate and subject to little pressure loss, they can also be applied to a wide range of pipe diameters from 2.5 to 2 600 mm. On the
negative side, this type of device is generally more costly than other
kinds of flowmeters; where the object is simply to obtain a rough estimate of flow , another type device would probably suffice. Typical
differential pressure flowmeters include the orifice plate and venturi
tube. For waterworks applications, venturi tubes are frequently used
because of their very small pressure loss. For measuring filtered flow
rate, ultrasonic flowmeters are sometimes used. While this type of
device is probably the easiest to mount on a pipe or conduit, they cannot be applied to wastewater applications or where the flow is accompanied by foaming. Also, where current velocity fluctuations are present, say after a pump outlet has been closed or around a pipe corner, a
sufficiently long section of straight pipe is necessary to prevent disturbances from being introduced.
10.6
Waterworks Instrumentation Applications
729
j Waler
dislriburion
Jnnow and ou tflow
warer al ch lorine
mixing chamber
:
I
-LI
I
:
I Fihered
wa rer
+
:
___,~$ -}>- 0-$- 4- <b--~-- -!--- f-- ------1

1
a w_a_r_e_r__
.___Ie_n_
l -------1
: : ! -- :~ ---~-: : -------'---.{: : : -~---{
l-$--$-$I
'
- ~ - - -
"'i-'" \..;..}
r-
3:
c:
- - - l..- - -1- ---~- '-- .. ~---,- ..

I
...
;;;
"
...
-----o ---f----'--- --- ---l- ----~

~4'> - ~
~
'+' - ~
I
I
I
l..- - - - - - - - - - ' \ ; . ]
I
I
I
I
~::>
0.
I
I
c:
.5:!
;;;
"E"
-~
-;;
"
<:r
...
3:
"'
"c:
;:
"'c:
c:
.2
;;;
c.,
E
~c:
;:
...E
~
""'E
-.::>
c:
"'"'
;:
"i5
Cl.
......
..
~
"
E
...c:
ij
-.;
::>
-o 'L:
-~ .
o::..c:
730
...
ij
?;>
:~
-~
u
.E
-.; -~" .,
"
c:
"""
<
Uj
"
il
..E
0
..c:
1-
L.-
"
ij
"
tliJ
ii:
c: -.::>
- c:
..9 E
..c: "
U -.::>
10.6.3 Filtration equipment instrumentation

Filtration equipment is used to screen fine suspended solids that
are not removed by chemical and sedimentation treatment by passing
the waste stream through a bed of sand. Key instrumentation for the
filter are filter rate control to maintain uniform velocity thorugh the
filter, and filter washing control applied to the backwash-storage compartment that holds water used to wash the filter bed. A flow sheet of
the instrumentation associated with the filter is shown in Fig. 10.100.
(a) Filter rate control
Standard filtration rates range from 120 to 150 meters/day, and
must be controlled to maintain a uniform rate. The total filter rate is
set by the main set station, which maintains common control over the
'
'I
'
~~J. 9
''
;--.J
r -- - ~
''
'
Rapid fil ter

From settling basin
To reservoir
for clean water
...__-c:>-'""!::~r-,-
L - --
--- w ;;hi~i -
---:
drainage basi n:
water~
To
arrival well
<D
Level indicating alarm meter on washing drainage basin

Level indicating alarm meter on washing water tank
@ Back washing water flow integrating controller

@ Surface washing water flow meter
Filter washing control equipment

(Sequence control equipment)
<V
Filtration head loss indicating alarm meter
@ Washing water tank

@ Surface washing valve
@ Back washing valve
Back washing water flow

control valve
Filter rate indicating controller
10.6
Filtered-water turbidimeter with alarm
@ From surface washing pump
Main set station
Fig. 10 100
Total filter rate integrator
Inflow valve
Instrumentation flow sheet of rapid filter
731
entire group of filters in operation, and the filter rate for each controller is set automatically. The number of basins in operation and total filter rate can be set either manually according to the operator's estimate
or automatically. The automatic setting method estimates water-delivery demand automatically (refer to Subsec. 10.6.8 below), then calculates balanced levels for the basins. Abruptly changing the filter rate
has an adverse impact on the sand bed, which can increase the turbidity of the effluent. For this reason, rate changes are carried out after the slow start procedure.
(b) Back-washing water flow control
To remove accumulated dirt clogging the sand bed, water is injected back up through the filter bed (back washing); the resulting vigorous agitation of the sand effectively washes the bed. This backwash
water is maintained at a uniform rate by flow control.
(c) Filtration head-loss meter
An important measurement item for supervising loss of filtering
capacity indicates the extent that the sand bed is clogged (head loss)
by filter resistance. Usually, when the value reaches 2 to 2.5 meters,
filter washing is required. Filter washing is also required when the upper limit of the filter run time is reached. Two other factors have an
important bearing on the wash demand: there is no basin during washing, and the drainage water in the washing drainage basin is below the
prescribed level.
(d) Wash sequence control
When the above wash conditions are met, sequence control for automatic washing can be implemented. The timing chart for such an
operation is shown in Fig. 10.101.
This type of filter wash sequence control and filter flow control
can be automated using a microprocessor-based control device.
10.6.4 Chemical injection equipment instrumentation
(a) Purposes of injections and places to be injected
Water is treated with chemicals is to clarify the water in compliance with water quality standards prescribed in the the water works
law so it is safe and fit for human consumption. Injection equipment
thus has an extremely important function at waterworks facilities. The
most common injection agents and the reasons they are added to the
water supply are summarized in Table 10.8. The points where these
agents are typically added to the purification process are shown in Fig.
10.102.
(b) Injection methods
There are three main methods of adding the chemical agents depending on the properties of the specific agent. These are: natural
732
c::.
......
"""
;:j
"'
;:;:
<:;
~
.....
'"
>
>
'"
"
;:;:
::t.
c:
.g
<:;
~
.....
"';:;:
;;l
.....
"'
"'~
;.,.
"'
~<:::>
0)
CD
Confirmati on of alarm
Fig. 10 101
Washing
Under washing
Preparatio n for
filteration
OD
Halt
@ Preparatio n for filteration
'\I
1\
Time chart of washing sequence
[!
/:
Cd[J
Stand-by
for
washing
Preparatio n for washing
:\
Normal water level :
I
I
I
I
I
I
!\
1\
JOD
Preparatio n for
washing
Washing
drainage valve
Back
washing valve
Surface
washing valve
Outflow
control valve
Level of
rapid filter
CD
........
Under filteration
Demand for
washing
Outflow valve
Inflow valve
Status of filter
Normal water Ieveli
I
I
I
Under filteration
Table 10.8 Injection purpose of coagulant

Injection agent
Purpose
Coagulant
Produces floc from suspended matter in raw water and promotes sedimentation as well as reducing turbidity of raw water, removal at the filter.
Coagulant aid
This is used to enhance the effect of coagulant depending on

the quality and condition of raw water.
Pre-alkali
Prevents reduction of alkalinity due to injection of coagulant

or pre-chlorine and raw water of low alkalinity. This aid is
used to keep alkalinity or pH high to acheive good coagulation and flocculation.
Middle alkali
Maintains pH at specified value or higher to prevent corrosion of pipes.
Post-alkali
Maintains pH at specified value or higher to prevent corrosion of pipe and to maintain water quality.
Pre-chlorine
Middle chlorine
Sterilizes and disinfects raw water, and removes iron, manganese, ammonia nitrogen, nitrite nitrogen and organic substance through acidification.
Post-chlorine
Chlorine injected to maintain residual chlorine (free available

chlorine) at a specified value to sterilize, disinfect and
prevant breeding of micro-organisms.
Middle chlorine
Middle alkali
Fig. 10 102 Main injection agents and injection points
flow, pump force feed, and ejector.

(1) To guarantee the required pressure at the point of injection for
the natural flow method, the injection tank must be placed in an elevated position. The injection dose is controlled by a flowmeter/control valve combination.
(2) With the pump force feed method, one method is to use a constant-rate pump to assure the required pressure at the injection
point. Another arrangement uses a pressure pump plus a flowmeter/control valve combination.
(3) With the ejector method, chemicals are injected in powder form,
734
and injection pressure is provided by the ejector itself.

(c) Injection dosage control
Although the injection dose will vary to some extent depending on

the quality of the water, ratio control derived by multiplying (volume
of water under treatment) x (injection ratio) is the most frequent method used.
(1) Injection control using flowmeter (electromagnetic)+ control valve
The measurement range of an electro-magnetic flowmeter is 0.3
to 10 metersjs. In cases where the injection dose exceeds either the
maximum or minimum values of this range, a combination of large and
small systems configured with a flowmeter and a control valve is required. Modern flowmeters that incorporate a microprocessor are capable of cutting over between different flow ranges. The loop configuration for such an injection method is shown in Fig. 10.103.
(2) Injection control with a constant-rate pump
With this method, the
injection dose is controlled by varying the speed and stroke of a constant-rate pump. The injection dose is simply calculated by multiplying (speed) x (stroke). Control is effected by either increasing or decreasing the speed until the set injection dose is achieved. If the upper
or lower limits of the speed are reached, the desired dosage can be
Storage
tank
AS
Control valve (Small)
Fig. 10 103 Loop diagram of flow meter and control valve (Wide range control type)
10.6 Waterworks Instrumentation Applications
735
Injectio.n ratio
r----------1
Stroke
I
Discharge rate
To injection point
Storage
tank
Fig. 10 104 Loop diagram of constant-rate pump (Speed and stroke control)
achieved by cutting over to the stroke and adjusting the speed. The
control loop for this method is shown in Fig. 10.104.
(d) Calculating the injection rate
The injection dosage for the water under treatment is determined
by the injection ratio. However, a number of methods have been developed or are being studied to enable automatic determination of the injection ratio. Here, we introduce a typical coagulant injection ratio
method. The coagulant injection ratio varies depending on water quality factors and temperature of the stream being treated. To enhance
the effectiveness of the coagulation and flocculation, coagulant aids
and alkaline are usually added at the same time, and the pH value is
kept above a prescribed level. Today, the most common method for determining the injection ratio is to conduct a jar test* (off-line). Two
methods of automating this testing are as follows:
(I) Store a correlation curve of turbidity (the main turbidity elements, or parameters) and the injection ratio in the memory of an
* An apparatus for measuring the appropriate injection rate for a coagulant. The
method involves gradually added a coagulant to a fixed amount of raw water while
measuring the turbidity, pH, and alkalinity, and observing the effects on flocculation and suspension. Based on an overall evaluation of the results, the optimum injection ratio is determined.
736
Chap.lO
automatic control device. Then, obtain the injection ratio based on

this data using an on-line turbidity meter.
(2) Compile a data file of jar test injection ratios and the water quality conditions that were under test (turbidity, pH, alkalinity, electric conductivity, and water temperature). Then, by applying statistical processing to the data (e.g., weighted regression analysis),
a correlation is obtained between the injection ratio and the various water quality factors (elements). The correlation can be expressed as a formula or organized as a table of injection rate correlation data for use as an injection method, and stored in the
memory of an automatic controller. Then, according to the reading of an on-line water quality indicator, the injection ratio is determined. One drawback of this method, however, is that the storage of the sample data needed for the statistical processing takes
considerable time. This is because the quantification of floc formation, one of the measures of coagulant effectiveness, is quite difficult. This problem must be addressed before this method can be
adopted on a widespread basis.
10.6.5 Instrumentation for water-supply and distribution facilities
After purification at a purification plant, water is supplied to
users maintaining appropriate quality standards and water pressure.
The link from the purification plant to the distribution reservoir is
called the supply system, and from the reservoir to end-users is called
the distribution system. A schematic of the two systems combined is
(a) Water supply flow control
Since distribution reservoirs are generally built on the highest
ground available, the water supply is regulated to maintain the reservoir at a certain level by controlling the number of supply pumps that
are in operation. Recently, the use of computers has had a great impact on the operation of waterworks. Computers have made it possible
to analyze demand fluctuation elements in terms of time units through..---------------,
Highlands and
J--:-:-:----'--- middle lands
Pump number control

Speed control
reservoir
~f---- Low
lands
Fig. 10 105 Water supply and distribution system diagram
10.6
737
out the course of a day. This data can be used to estimate water supply 24 hours ahead, which in turn, can be used to estimate the water
level fluctuation in the distribution reservoir. This advanced control
technique has been introduced at a number of municipal purification
plants, and makes it possible to maintain the reservoir level within upper and lower permissible limits.
(b) Distribution volume
Because gravitation flow can be used to distribute water to lowland areas, no special controls are generally needed. If water pressure
builds up to excess levels, presure is relieved with pressure-reducing
valves. In the case of high-lying areas, water is lifted to the higher
elevation by means of booster pumps. Control is effected either by adjusting pump speed or the number of pumps in operation to maintain
appropriate pipe-end pressure (above 1.5 kg/cm 2). This is called end
pressure control. An example of end pressure control, that includes
flow resistance within the pipe, is shown in Fig. 10.106.
H=kQ+Pe
Pe : Presumptive end pressure

kQ : Resistance of pipe
reservoir
Pe (Required end pressure)
-Flow rate Q
Fig. 10 106 An example of presumptive end pressure control in highlands
10.6.6 An integrated control system for large-scale, wide-area water

works facilities
In recent years, the trend has been toward wide-area waterworks
that supply water over an surprisingly dispersed geographical region.
Large-scale purification plants have been constructed that sell water
to multiple commuties and municipalities based on contracts individually arranged with the communities. Accordingly, more stringent man738
agement over both quantity and quality of water have come to be required. Moreover, since distances from purification plant to the local
municipal reservoir on the consumer end can range from several kilometers to several tens of kilometers, unmanned facilities such as relay
pumping stations and equalizing storage reservoirs must be installed at
intermediate points along the way. In order to supervise and control
these unmanned facilities, systems must be designed to permit centralized control from the control room at the purification plant. In short,
integrated system management has become indispensable in order to
maintain good quantitative balance in the supply system and to assure
the maintenance of contracted supplies of water at all times. Two approaches to implementing such an integrated system are to use telemeter/telecontrol (TM/TC) over a leased circuit or to use computer control via modem. The configuration of an integrated system is shown in
Fig. 10.107, and a typical control room is shown in Fig. 10.108.
10.6. 7 Water distribution information management system
So far, we have focused on the segment from intake (centered on
the purification plant) to the distribution reservoir-the so-called water production process. By contrast, the segment from distribution reservoir to each end-user over the distribution network can be likened to a
water distribution process. The distribution takes place over a network
of conduits and pipes that branches out to directly connect with
end-users. Topology is an important factor that must be taken into account when laying the network. The ideal of maintaining exactly the
right amount of water pressure (neither more nor less than needed) to
each end-user is extremely difficult to realize. In fact, the great difficulty of network analysis is one of the main reasons computer control
systems have not been widely introduced in the area of water distribution. This is in marked contrast to purification plants, where such systems have been extensively introduced for the purpose of automating
processes. Here, we will focus on the network (including the distribution reservoir), and describe network analysis using a piping network
calculation. Based on the network analysis, the entire water distribution process can be optimized (i.e., leakage minimized by optimal adjustment of the water pressure). The analysis also makes it possible to
produce a water-delivery forecast which can be used on the production-process side (i.e., the purification plant) to calculate a production
plan (the water supply), the most important objective for implementing
the system. Briefly, the piping network calculation is carried out as follows. The actual network of pipes is replaced by a logical diagram of
nodes linked by lines. The energy level (i.e., water pressure) at each
node, and the flow over segments between the nodes is then found.
10.6
739
"'
;;
't
't
-"'""
s.,
~
::!
""::!
o::!
"'~
<:::>
~
......
;:,:-
('")
<::)
>.
1:econdary
rE 'G --
1
5
l=~5:0:: -
~C RT
Washing drainage basin
Fig. 10 107 Overview of wide-area water works
Data way
Central
Information processi ng station
K.::l
Fig. 10 108 Typical control room
Generally, the Hazen Williams formula is used, which relates a number

of physical parameters of the pipe : D =pipe diameter, L =distance between nodes, Cis the flow coefficient of each segment of pipe, and q is
the demand volume at a node. The formula for each segment of pipe is
given by
QtJ = 1 002.7 x Cu x Du 2 63 x Lu-0 54 x (E 1-E1 ) 0 54
and the formula for the continuous flow rate at a node is given by
fQiJ+qj=O
as simultaneous equations for the unknown energy levels E 1 , E 1 of
each node. From the energy level of each node, the water pressure at
those points, and the flow between nodes can be found. To apply this
piping network calculation to actual networks, first the physical parameters at the pipe and node parameters are entered as a data base.
Then, based on facility data (data base information) and on- line data
(current water level in the reservoir, flow rate, etc.), the following can
be calculated:
(1) The current state of the piping network can be estimated. That
is, by knowing the water pressure of all nodes, any areas that are
out of balance are apparent.
(2) Simulation can be carried out using a computer model of the piping network (i.e., based on pressurizing and depressurizing elements). Simulation is useful for preevaluation of planned facilities,
and is also useful for providing information on renovation and/or
expansion of the piping network.
(3) Water pressure can be adjusted when there are points that need
to be depressurized; the degree of opening for valves to achieve
the desired water pressure can be calculated.
741
I Piping network calculation I
Piping network diagnosis

Water pressure distribution
Flow rate distribution
Flow direction
Simulated piping
network
Search of out-of-balance
water pressure areas
Pre-evaluation of
facilities at
planning stage
Search of pressurizing
or depressurizing
points for uniformity
Renewal and
expansion of
piping network
Simulation
Water pressure
adjustment
Appropriate
valve opening
Fig. 10 109 Application of piping network calculation
Piping network applications and a typical output result of the calculation are shown as Figs. 10.109 and 10.110, respectively.
Next, we consider the method of estimating water-delivery
demand. The purpose of the estimation is to forecast water-delivery in
daily and time units given the great fluctuations that occur due to
weather conditions and also depending on the day of the week. From
the perspective of ideal plant operations, on the other hand, better efficiency can be achieved if the facility is operated under uniform load
conditions. This makes it possible to operate the pumps at constant
load and to stabilize the amount of chemicals injected into the water.
In order to operate under constant load conditions at the plant, distribution reservoirs are constructed between the purification plant and
the distribution network. The reservoir serves to absorb such fluctuations as occur, and thus performs a buffer function. To make the most
effective use of this buffer function (based on upper and lower level
limits of the reservoir), estimations must be made of the total volume
supplied each day and the water-delivery for different times. From
that data, the water-level fluctuation curve can be estimated, which
makes it possible to calculate a water-delivery that will not exceed either the upper or lower level limits of the reservoir. One method of estimating the water-delivery is to estimate the load using a Kalman filter. One advantage of this method is that it is amenable to computer
processing, and in fact the method has been applied at a number of municipal purification plants. An outline of plant operations based on
demand forecast is shown as Fig. 10.111, and a procedure for calculating the necessary water supply is shown in Fig. 10.112. Finally, Figure
10.113 shows a method of forecasting water level, where Fout(m3/h) is
the estimated water-delivery, Fin(m3/h) is the water supply, and L(m)
742
Chap.lO
<::)
.......
~
.....
"''"
c;
~-
'"
c;
......
"';::;-'"
;:;!
"'......
C)
0~-~~~
:4.2 4.2\
4.2
~35
I/,
~I
cii5f
b.~/-
11// /
\D
,4.3
,,
,/
co.
\.
4.2
Fig. 10 110
,. ,. "
.... _ .... -
3.1
'"@
.._~ ,/'
/ ,"
.,"
,/' 2.8
1:
.....
..... I
-~.0
I
I
ti
Example of piping network display
r,"\ ......
/,'
,.~~, IS
.," 3.8
3414.0
+I
/:"/
,/
I
I
,,
4
value is written in red)

Arrows indicate direction of__. sign
beside pipe line accompanying
flow, and numbers indicate flow
rate (m3 /h)
=Nodos
~=Values
2 Numbers along pipe lines indicate
diameter in mm (detail is omitted)
3 Numbers beside nodes= node water
pressure (kgf/cm 2 ) (The normal
Note
Water delivery forecast

Daily water delivery
Time water delivery
I Water delivery
I fluctuation
Kalman filter
. ,.
factor
Water level estimate in

distribution reservoir
I Best use of water from
Application of the group control technique and

distributing the load to spare reservoirs
rEffective use of rl
distribution reservoir
rSharing
the load l
between reservoirs
upper level to lower level
Load fluctuation cannot be absorbed

by group of distribution reservoirs.
Load fluctuation can be

absorbed by group of
distribution reservoirs
Guide line to change water delivery
Constant load operation

of water purification plant
Normalization of
water supply
Normalization of
water intake
~
Pump constant Minimization
Pump running schedule
load running for start/Stop
Crequency
Fig. 10 111 Outline of plant operation based on demand forecast
Actual water delivery

Time
Weather
Date
Result of forecast
CRT display
Estimate of water-delivery F ,
c:::::> Result
of forecast
CTR display
Current water level

Water supply F;.
Calculation for
necessary
amount of
water supply( F5 )
Fig. 10 112
744
Procedure for calculation of necessary amount of water supply
Chap.lO
Current water supply

F;.
(Time i)
~
-t=:==l--
Distributing
reservoir
Estimate of water delivery

F
""'
A : Equivalent area of distribution reservoir
L Water level in distribution reservoir
General formula of water level at arbitrary time t
dL
A~=
F;.- F ,
0
L (t) =-tU,:F;.dt- J::F.,,dt)+Lo

Where F;. is constant, and at time i+ n (n= 1-24), estimated
water delivery is Qi+n and estimated water level is Li+n
Lt+t
I
=A (F;.-
Q,_tJ + L,
[L, is the current water level at time i)
I
L,~2 =A (F;.- Q;- 2 )
+ L,,. 1
.
I
L;+2< =A (F;.- Q,-2,)
+ L;- 23
Fig. 10 113 Method of estimating water level
is water level. Once the upper and lower level limits of the reservoir
are calculated, then the necessary amount of water supply Fs can be determined.
10.6.8 Wastewater system overview
Wastewater systems are maintained to enhance our living environment and also to safeguard the quality of public water bodies. The wastewater system consists of a network of sewer drains, pumping stations, and a treatment plant. Pumping stations become necessary when
drain pipes have to be buried beyond a certain depth.
For the past 80 or 90 years, the prevailing method of purifying
large volumes of low-density waste water has been the activated-sluge
method. The activated-sludge method is basically the same as the natural purification mechanism one observes taking place in rivers, only
much more intensified and concentrated. Activated-sludge processing
continues, even today, to be the fundamental method used in purification plants.
10.6.9 Overview of activated-sludge processes
The activated-sludge method is a biological treatment in ~hich
aeration and aerobic microbial action are used to eliminate organic pollutants. The processes involves, shown in Fig. 10.115, will be briefly
745
To estimate the water level at arbitrary time (t), the necessary amount of
water supply is determined using upper and lower alarm forecast.
(1)
_:::LHb---~---
~LL~
When LL<L(I)<LH is estimated within time I,
f--t--i
(2)
~ LH~---r------~
'
ILL---:------ ;:;,
~~-1
( 3)
be adopted.
:.
-::
I
LL ----:-------1
I
When LL> L(l) is estimated within time I,

Compensation factor AV1 =(L 1 -Lm;,) A should
F,=F;,+AV/1
When L(t) < LH is estimated within time I,.

Compensation factor AV2 = (Lm.,- LH) A should
be taken.
:.
F,=F;.- AV2 /I
(4)
When both LL<L(I) and L(t)>LH are estimated

within time I, it is considered an abnormal
situation and left to operator discretion.
Fig. 10 114 Method of calculating water supply

Final
effiuent
Waste activated
sludge
Fig. 10 115 Activated sludge process flow
described.
(a) Sand basin, screen, and wastewater pump
Gross suspended solids and trash are removed from the sewage inflow, and the stream is then pumped to the primary sedimentation
tank. The sewage stream then flows under the force of gravity
through the various downstream units where it is purified.
(b) Primary sedimentation tank
746
Chap.lO
Suspended solid particles are removed from the liquid stream by

simple gravitational settling to reduce shock loading at the following
aeration tank. The sludge that accumulates in the sedimentation tank
is periodically removed by pump.
(c) Aeration tank, final sedimentation tank
The sewage is churned and aerated by bubbling air through it in
the aeration tank to encourage microbial action. Aerobic microorganisms in the sewage are caused to grow under favorable conditions of
nutrients in the form of suspended organic matter in the waste being
treated. As the organisms grow, a microbial sludge is formed having
the consistency of gelatin that is known as activated sludge. Activated
sludge has excellent waste-water-purifying and settling characteristics.
The clarified liquid is separated from the activated sludge in the final sedimentation tank by simple gravity settling. The sludge is then
reaerated and returned to the aeration tank to contact fresh sewage.
Because the activated sludge gradually accumulates as it performs its
purifying role, excess sludge (waste sludge) has to be periodically removed.
(d) Chlorination equipment
The overflow from the final sedimentation tank is treated with
chlorine disinfectant for safety and hygienic reasons before the effluent is discharged.
10.6.10 Wastewater treatment instrumentation
An overview of the main instruments involved in waste-water
treatment is shown in Fig. 10.116. A brief description of each unit follows.
(a) Sand-basin and wastewater lift-pump instrumentation
Pump control is the main means of regulating the volume of
sewage inflow. The speed of the pumps and/or the number of pumps in
operation are adjusted to maintain the level of the sewage in the pump
basin at a certain level. The amount of sewage inflow is measured either by an electromagnetic flowmeter or a Parshall flume mounted on
the discharge side of the pump.
(b) Primary sedimentation tank instrumentation
The sludge settling on the bottom of the sedimentation tank is
scrapped into sludge hoppers from which it can be removed. There are
multiple hopper compartments, so the sludge removal valves is rotated
from one hopper port to the next. The amount of sludge removed is
generally determined using either a timer or a quantity setting device.
(c) Aeration tank and blower instrumentation
The amount of air supplied to the aeration unit is calculated to
747
;:
"'a.
"'
c
0
~
:,:
;;;
-.;
>
-.;
>
..!!
;:;
.!!
"'
a.
"a.
;;;
"
?;
.0
..!!
._
0
;q
3:
;;
-;;
..!!
;;;"
i5
..,
,.,
... "'c
:.0
3:
"
;;;
:;;
c
,...... ,..."
;;;
"
:.
e...
()
~
;;;
:!
:.0
;;
~
~
;:
"
""~""
"
.g
-.;
~
;:
>
..!!
3:
.."
""
~
;;;
" 3:
""""" o.
u;
!:,
""tii"
?;
)(
~
;:
.,0
c
0
'
:;
e"
3:
;:
"
E"
u
-;;
c
~
I
I
I
I
I
'
CD
Abnormal water level

Pump control
Excess sludge pulling out sequence

Chlorinator
IJ) Blower
Sewage
inflow
I
I
r .. -.J
I
I
'
'
Inflow gate
Sand basin
@Pump basin
@ Waste water pump
@ To sludge thickner
@ Final sedimentation
@ Raw sludge pump
@ Raw sludge pulling out sequence

@ Blower control
I
I
I
I
I
Primary sedimentation tank

Aeration tank
@ Return sludge
tank
Sludge pit
To sludge thicker
Chlorine contact tank
@ Final effluent
Fig. 10 116 Instrumentation diagram of waste water treatment
maintain the dissolved oxygen concentration in the sewage at specified

level. Power is also conserved by controlling this parameter. Since precision is not required in this instance, the most common approach is to
supply air as a proportion of the raw sewage inflow.
Almost all of sludge that settles in the final sedimentation tank is
recycled to the aeration tank. The quantity of return activated sludge
is usually expressed as a percentage of the sewage waste flow, and the
general practice is to use 20 to 40% of the average sewage inflow. Ratio control is used to keep this value.
(d) Final sedimentation tank instrumentaton
Settled sledge has to be scrapped and removed from the sedimentation tank on a continuous basis. Moreover, since this sludge is made up
of microorganisms, it must be returned to the aeration tank as quickly
as possible. Excess sludge, the incremental increase of organisms in
748
Chap.lO
the activated sludge, must be removed. The amount of excess to be removed is estimated by analyzing the density of the aertion tank suspended sludge. The removal of excess sludge is intermittently carried
out using a quantity setting device.
(e) Chlorination equipment instrumentation
Chlorine and sodium hypochlorite are used as disinfectants. The desired dose is that maintains the coliform group bacteria count in the final effluent of less than 3 000 groupsfcm3 Ratio control is used to determine the specific chlorine dose of 2 to 4 mg/1 of final effluent. Final
effluent is generally measured with a weir-type flowmeter mounted at
the outlet of the chlorination tank.
10.6.11 Sludge treatment instrumentation
Sludge produced by waste-treatment processes is first dewatered,
and then hauled to a land-fill site or incinerated. Here we will briefly
elaborate on the instrumentation needed to support the sludge disposal
processes shown in Fig. 10.117; thickening, digestion, sludge conditioning, mechanical dewatering, and removal by truck.
(a) Sludge thickener instrumentation
Sludge produced by waste-water treatment is first thickened and
then fed to the primary digestion tank. Thickener tanks are generally
circular in shape, and the sludge is removed in much the same way as
it is removed from the primary sedimentation tank, discussed previously.
(b) Sludge digestion tank instrumentation
In the digestion tank, the sludge gradually stabilizes and separates
from the liquid while digestion gas (primarily methane) is given off in
the process. This separation takes considerable time, and a number of
parameters have to be supervised while it is in progress; these include
the tank level, temperataure, pressure, and quantity of gas produced.
The sludge and separation liquid are periodically removed from the
tank. The removal, required to draw down the level of the tank when
it becomes full, is under sequence control and is triggered by a timer.
(c) Sludge elutriation tank instrumentation
The sludge discharged from the digestion tank has a slimy or viscous consistency, and thus is difficult to dewater as is. Washing the
sludge with water effectively removes the viscous component. Thickening is carried out in a two-stage counter-flow thickening process with
the detention in each thickening unit determined by timing control. After thickening, the sludge is stored in a storage tank.
(d) Coagulant mixing tank instrumentation
Prior to dewatering, coagulants and coagulant aids are added to
the sludge to facilitate the dehydration process. The chemical doses
749
'"
-;:;
>
..!!
0
c""
s0
00
:;
c.
>.
v
c
-~c
0
::, 00
"
-o
-o
"0
:;
.:><
s
c
.2
" ""' -ii~
-;;;
-o
1>1
"0
-:;
.2"
.:><
!:!
.
c
.2
ii
:0
~c
-o
to
..
-o
.5
0::
-o
"
-o
"
'"
-;:;
>
..!!
to
"'c
V)
3:
0
c
-;:;
>
.!!
s0
00
:;
c.
"00
-o
'"
3:
0
"'"'
.2
00
ii011
0 0
~
"0
-:;
.2"
.....
-"'
~
u
00
'
3:
0
c
u
011
.2
'
!lc
v
c
-~
"C'
u
u
00
00
E
u
'
00
"i)
-o
"
-~
3:
;;;
:?
>.
8.c.
.c
0.
c.
e.
u
.."
-"'
.c
;;;
-0>.
.c
~<-7
~'
' :
:
.-'
.<:
::,
Ep
<?!
'
'
'
'
'
''
L.,
''
<D
Sludge pulling out sequence
(f) Sludge thickner
@ Primary digestion tank
Desulfurizer
@Gas holder
(!) Boiler
Sludge pulling out sequence

Raw sludge and excess sludge
Fig. 10 117
@ Sludge storage tank

@ Ferrous and ferric
Secondary digestion tank

l.st stage sludge elutriation tank
@ 2nd stage sludge elutriation tank

@ Water for washing
chloride
@ Coagulant mixing tank
Chemical
@ Dewatering equipment
@ Cake derivery
Instrumentation diagram of sludge treatment
must be in proportion to the amount of solids in the sludge. By multiplying sludge flow rate times density of the sludge, the sludge solids
flow rate can be calculated. Ratio control can then be applied using
this value to determine the doses of the various coagulants.
Constant flow control regulates sludge feed to the dehydrator,
which is set according to the capacity of the dehydrator unit. The
method of flow control that is selected depends to a great extent on
the type of dehydrator. After the sludge has been dewatered, it is
moved by belt conveyer to the sludge cake hopper. Accumulated cake
is then transported by truck to another location, say a land-fill site, or
is incinerated.
750
10.7 Instrumentation Applications in the Automobile Industry

10.7.1 Overview of automobile industry instrumentation
The production of automobiles, a typical labor-intensive industry,
is shown schematically in Fig.10.118. First, the engine and other heavy
parts are cast at casting and forging factories. The pieces are machined into final shape at the machine factory, and then assembled at a
engine factory . In the meantime, vehicle bodies are progressing
through a succession of factories that constitutes the main line-the
press factory, the body factory, the painting factory, and the assembling factory . Still other components and parts (seats, windshields, and
so on) are delivered to the assembling factory by subcontractors. For
the production of an automobile, all these various kinds of factories
have to be systematically interrelated. Note that the production line in
the automobile industry is fundamentally different from continuous
processing as can be observed, for example, in the oil industry . The
automobile industry is an example of non-continuous processing; each
component, each body is handled as a discrete piece. Given the
non-continuous nature of the automobile industry, new production
technologies are constantly being introduced to realize the industry slogan "one more unit, one less yen." Here, we are referring to production technologies such robots and automatic machines which have been
extensively introduced in pursuit of automation, and computers which
have been introduced to aid in distributon and production management.
In this section, we will look at two instrumentation areas in the
auto industry : producton management of robots and automatic machines-probably better regarded as an integrated whole-at an automobile painting factory , and instrumentation to realize storage control.
Delivery
Fig. 10 118
Proces ing sequence at on automobile factory
751
10.7.2 Production management at a painting factory

Automobile production management seeks to manage all data pertaining to production of vehicles for the purpose of flexibly coping
with modifications to the production plan. Specifically, production management monitors the following five items; (a) production instructions
based on the production plan, (b) production speed (cycle time), (c) the
number of car bodies in each process, (d) the location of each body
(management maintained by assigning a control number to each body),
(e) progress management and actual results based on tracing data (control number) updated as the body passes through processes.
(a) Painting factory process
At the painting factory, bodies received from the body factory are
first put through a number of pre- treatment processes. Then, the
bodies are treated for corrosion protection and multiple coats of paint
are applied to provide a good visual appearance. This process flow is illustrated in Fig. 10.119.
Baking
Ba king
Under-coating
Baki ng
Baki ng
Fi na l top coati ng
Baki ng
Fig. 10 119
Pai nt in g process a nd layout
After degreasing and washing in the pre-treatment process, the

bodies are sent to the priming process. Next, any unevenness in the
prim coat and seams in the steel sheet are smoothed out, a sealant is
applied, and the bodies are sent on to the middle-coat process. After
the middle coat has been applied, the bodies pass by way of the middle-coat polishing process to the final top-coat process. Here, the
bodies are finished one unit at a time in accordance with purchasers'
752
specifications to produce a durable and visually attractive external appearance.

Car bodies are moved through the processes described by chain
conveyer that continues unbroken through the entire plant. There are
floor-type conveyers and overhead- type conveyers in use, as shown in
Fig. 10.120.
ms
( a ) Floor conveyer
( b ) Overhead conveyer
Fig. 10 120 Transfer by chain conveyer
(b) Production management instrumentation

We briefly touched on the contents of production management,
but neglected to mention how the data is kept. A data storage system
must be set up so data on the car bodies can be recorded and manipulated. In production management systems, a control number corresponding to each body is assigned. Then a file is prepared in a computer corresponding to the specifications of the specific vehicles represented by
omputer ror
production
management
Fig. 10 121
In trumentation diagram of production management
753
the control numbers. For the control media itself, token cards or magnetic cards are generally used.
(1) Production instructions based on the production plan
No matter
what kind of vehicle, the production sequence starts with the fabrication of the car prior to the painting processes. Once a body reaches the
painting factory, the specifications file for the body is looked up using
the token card control number. Then, the pertinent specifications are
sent to the workers by printer, and instructions are input to the paint
robots and automatic machines by contact signals. The token card is
read by a token card reader.
J_,
Token card
(B D ode)
Magnetic card
Fig. 10 122 Data torage med ia
(2) Production speed (cycle time)

Speed management in auto production is exacting, and is measured in units of seconds. For example, assuming a body can be processed in 90 seconds, if the cycle time slows
only slightly to 91 seconds, that means that only 791 bodies can be processes during a 20- hour operaating day versus 800 that could be processed at 90 seconds. The difference in productoin, in other words, is 9
bodies a day. Cycle time is in inverse proportion to the speed of the
conveyer. Pushers (i.e., hooks for latching onto the body conveyer
sleds) installed at fixed intervals along the conveyer trigger limit
switches. The limitswitch signals are used to measure time and monitor the cycle time.
I
ON
OFF
Cycle time
_jl
Fig. 10 . 123
ycle time
n n
n_
Limit witch signal
(3) Number of car bodies in each process

The car bodies trigger reset
switches that are mounted at the entrances and exits of each process.
By counting (incrementing or decrementing) the signals, the car bodies
in each process can be accounted for.
(4) Tracking car bodies, progress management and actual results
Buy754
Chap.l 0 lnstrumention to Manufacturing Industries
ers for the cars have already been lined up even before they are produced. Thus, the production proceeds in accordance with the customers' specifications. It is necessary, therefore, to monitor the progress of the bodies so it is always known exactly which process the
bodies are in. This task is managed with token card reader setpoints
and limitswitch installation points, whose setpoints demarcate the individual processes. Within a given process, bodies are precisely tracked
using the order in which the control numbers entered the process.
This data is also used for progress management and to monitor actual
results. To support such a production management system, high-performance computers are required that are capable of performing the
following functions within one time cycle: (a) reading the pass signal
from each setpoint limit switch, (b) outputting work instructions based
on control numbers from card readers, and (c) grasping the actual
results. All these requirements could not be satisfied without real-time
capability and multitask processing. With the development of computers specifically designed for factory automation (FA) applications, however, it has become fairly easy to construct extremely exacting production management systems.
Between processes ( I)
Between proces es (2)
Vehicle control
~~~hicle control
0-
cCS
Token card
reader
A 2 3 5 1
A 2 3 3 1
A 5 2 1 1
B 2 1 4 4
cC:b
P.
Limit
A 2 5 3 3
A 2 7 4 1
switch
A 3 5 5 9
Fig. 10 124
0.
A 4 4 1 7
Stock control
B 1 5 4 3
Cars are p resent in order

of veh icle control o.
o. between processes
10. 7.3 Storage control

In automotive manufacturing processes, conveyer lanes are used
to store car bodies between processing equipment units as well as between the painting factory and the assembling factory. This storage arrangement also functions as a buffer so that if the flow of vehicle
bodies from the previous process is cut off for some reason, no adverse
impact is felt at the succeeding process.
(a) Color- selection line control
The color selection line is situated between the middle-coating process and the top-coating process. The basic control function, implemented with an algorithm, is to load the vehicle into the proper col10.7 Instrumentation Applications in the Automobile Industry
755
To assembly
factory
PBS : Painted Body Storage

Color selection line
PBS Line
Fig. 10 125 Storage line
or-selection line from the middle-coating process and then unload the
bodies out of the right lane for delivery to the top-coating process.
The final surface coat and color in applied to the bodies in the
top-coat process. If a body out of the color selection line is inadvertently sprayed the wrong color, this creates a major inconvenience.
That batch of paint has to be discarded, the wrong color paint has to
be washed off the body with thinner, and the correct color then has to
be loaded into the sprayer. Not only is paint wasted, but considerable
time is lost. As a result, the production unit count falls, and along
with it, production efficiency. To minimize the chances of this happening, the control algorithm separates vehicle bodies into separate lanes
on the basis of color specifications. Then, every effort is made to send
all the bodies to be painted the same color to the top-coating process
in a continuous stream.
A schematic of the control functions for the color-selection line is
shown in Fig. 10.126; processing flow for the system is shown in Fig.
10.127.
(1) Production control for color-selection line
This control is activated
when a vehicle body is in the stop position at the entrance of the color-selection line. The control selects which lane the body is to be sent
and verifies its transfer to the selected lane.
CD
IMiddle
..
:.
. . ------ ----, r---- -----, r-------- -,

coating
:
I
:.I
0
I
0
1. . 1
:
L--- -----
CD
I
..JL.---------.J
I
I
L--------
I
I
Top coating!
I
I
I
I
I
I
I
I
I
I
I
y:
Stop sign al
I . Pass sgnal
_J
Production control for color-selection line

Inner control for color-selection line
Output production control for color-selection line
Fig. 10 126 Management functional diagram

756
Computer for color

selection control
Token card
Lane
Color idintification
idintification
(Booth control board)
(Conveyer control board)
Fig. 10 127 Processing flow for color selection
When a vehicle body is present in the stop positon, the operator inserts the vehicle's token card in the card reader to retrieve the vehicle's control number. The control number reveals the color specifications of the body. Then, if there's a lane available where the last unit
in the lane has the same color specifications as the body being processed, a pulse signal is sent identifying the lane to the conveyer control
board. The stop gate then opens, and the tracking of the body to the
designated lane is monitored by pass signals.
(2) Inner control for color-selection line
This control verifies the loading and unloading of vehicle bodies to the color-selection line, and also
tracks the bodies while they are in the line.
(3) Output production control for color-selection line
This control
sends bodies with the same color specifications to the top-coating process in a continuous stream. The control checks the bodies at the front
of the various lanes to see if there are any to be painted the same color
as the piece just sent; if there is, an unload signal is sent to the conveyer control board. When the stop barrier is opened and the workpiece transferred to the top-coat process, a pass signal detected by a
sensor mounted right in front of the top-coating process sends the color identification of the body ahead to the top-coating unit.
(b) PBS line control
The Painted Body Storage (PBS) line, capable of accommodating
anywhere from 100 to 200 vehicle bodies, is situated between the painting factory and the assembling factory. With the object of leveling or
evening out the work on the assembly line, the PBS line control has
two key functions: mixing delivery of vehicle bodies to the assembly
line to achieve a good load balance on the line, and sidetracking vehicle
(i.e., not sending them to the assembly line) which have component(s)
needed that are out of stock.
(1) Mixed delivery to achieve load balance
Auto manufacturing involves a great deal of assembly work. In order to meet the specifications
757
/Vehicle
l&..
/
control
No.
Pass signal
Token card
Lane idintification
(Conveyer control board)
Fig. 10 128 Process flow for load balance in assembly line
for a typical vehicle, many miscellaneous parts and components are assembled in numerous processes (anywhere from 80 to 100 processes) on
an assembly line. For each process, the time allowed to attach a component is strictly controlled at close to the minimum time possible for
the assembly work. Under these exacting conditions, when big-load vehicle bodies come down the line, if the work isn't completed within the
specified time allowed, the problem is passed on to the following downstream process. Start of assembly work on the next unit is delayed, until eventually, the line has to be shut down to catch up. To prevent
this situation from developing, bodies are taken off the PBS line and
delivered to the assembly line in such a way to facilite completion of
the assembly work within the designated time: a heavy-load body, for
example, is followed by a string of light-load bodies to even out the
line load. To illustrate the mixing principle, consider five types of vehicles (A, B, C, D, and E), with the following mixing conditions:
A: can be put on the line in succession
B: 2 units in a row prohibited
C: separate by at least 2 units
D: separate by at least 3 units
E: separate by at lesst 5 units
Given these conditions, the vehicle bodies might be put on the assembly line in the following order:
~--------,----1
r-L- ___ L 1
I 1
1 I
ABACDAABCEDACBDEAA
I
I I
'I
L-t-----.l...+---..!.-1
[__ ____ ..!_ ___ ~
(2) Countermeasures for missing componeuts

Components and parts
manufactured by outside subcontractors (seats, tires, engines, etc.) can
758
Chap.IO Jnstrumention to Manufacturing Industries
normally be delivered on a just-in-time basis within two hours for a

given vehicle. However, if a vehicle should come down the line and a
component is not available, say because it has been held up in traffic
or because the wrong part was delivered, this can create havoc on the
assembly line. This situation is avoided by registering missing parts in
an out-of-stock table. Then, by checking the table, bodies are held in
the PBS line until the missing component is back in stock.
(3) Processing flow
The PBS line control is implemented by means
of a card reader installed at the entrance to the line; the control number read in the card reader gives access to the specifications of the vehicle. Then, transferring bodies into and out of the PBS line is done in
compliance with the control conditions (mixing conditions and component availability). Bodies are tracked within the PBS line by means of
pass signals.
In the auto industry, the introduction of new equipment and the
revamping of manufacturing lines is practically an everyday occurrence. If lines aren't constructed around computer systems capable of
flexibly responding to line conditions, then they are locked in a rigid
mold, and quickly become out of date.
One final point to note is that the production management system
and storage control system described both rely on the same vehicle
specifications file. It would thus make sense from a systems point of
view to integrate the two systems in the same computer system.
10.8 Product Control in Batch Processing3 3 l- 37 l
The great number of new materials, bio-products, high molecular

compounds that have made their appearance in recent years is testimony to the steady stream of new products that are constantly being developed based on advanced technologies. Of course, for such new products to be created, new methods of production also must be devised. In
the case of large-scale chemical processes, say for the manufacture of
products such as ethylene and ammonia, large-volume, continuous processing methods can be applied. The profusion of new products, however, are almost always produced in small batches, which are considerably more difficult to automate. If maximum production is the only concern, technological effort is bent toward achieving economies of scale,
and continuous processing; whatever is necessary to boost volume.
Once we enter the age where quality is at least as important as quantity, however, a totally different approach is called for: Inventories are
reduced to the bare minimum, and only products where there is a
demand are produced, and only in the quantities required. Production
equipment is limited and quick modification of operating conditions is
10.8 Product Control in Batch Processing
759
emphasized in order to expand or contract producton in respnse to

changing demand. In short, this is the technology for producing small
quantities of diverse products, and the methodology is flexible automation. Basic to the new production technology is the desirability of
batch processing rather than continuous processing and more exacting
production plans. Implementing such an approach has greatly stimulated the development of automation technology.
On the other hand, automation itself, in the form of digital control
systems (DCSs), has had a profound impact. Not only can advanced control algorithms be implemented on DCSs, but even more signficant is
the ease of switching back and forth between control loops and sequence operations that is possible with DCSs. Such capability has accelerated the development of DCSs much more than would have occurred otherwise. Closely adhering to the production plan, it is a simple matter with DCSs to alter control loop parameters and change the
order of sequences as required for specific products. This, in essence,
is the basis of flexible automation for producing diverse products in
small batches.
10.8.1 Batch process recipe management

A common requirement of most small-volume batch processes is
the ability to change the operating conditions for different recipes or
products. However, there is great variation in detail depending on the
type of batch process and other conditions. Simplified prototypical
batch process configurations are shown in Fig. 10.129. Note that actual
batch processes can assume many different forms depending on the
product being produced; specific configurations consist of mafly pieces
of equipment linked together in complex ways. Moreover, there are
just about as many different recipe management methods as there are
different processes.
Recipe management for a single equipment unit is the most basic
configuration (Fig. 10.129(a)). The recipe management procedure for
this case is shown in Fig. 10.130. The recipe data base is stored in an
external file. When a batch is started, a command from the keyboard
fetches the appropriate recipe data, and the recipe-specific control loop
parameters and order of sequence operations are loaded. By specifying
a recipe name, corresponding data is read from an external file and displayed on a CRT, the amounts of raw materials needed to produce the
set production volume is calculated, and control loop parameters and sequence order changes are loaded.
When multiple units are configured in a parallel-line (Fig
10.129(b)), mix and charging systems such as shown in Fig.10.131 are
the most common approach. Generally, the mix and charging system re760
Chap.JO Instrumention to Manufacturing Industries
Product
( b) Multi units (Parallel line)

Raw
materials
Product
~~~-~
Utility
(c) Multi units (Series line)
Fig. 10 129 Batch process configurations
External file
CRT Console
Recipe name
Set data
Set data
Process sequence
Process sequence
r---,r'~--~1-------.,
e
Sequence tables
7~~; ~Kl
'
llj
~
Key boM
Key input for revision
of set data and
process sequences
Fig. 10 130 Recipe management procedure
fers to the line from the raw materials intake to the mixing tank. The
sequence that selects this line is often quite complex. When the recipe
is selected from the mix instruction screen, the proportions of the raw
materials (included in the recipe data) are brought up on the screen. If
10.8
Product Control in Batch Processing
761
<D
Reactor
Fig. 10 131 Mix and charging system
the entire batch volume is input for the setting, proportions for all of
the raw materials are calculated. Next, by entering the instruction for
the mixing to start, charging from the raw materials tank begins
through the line indicated by the data. In order to increase the operating efficiency of the mixing system, it is common to carry out multiple
operations in parallel. When this is done, it is necessary to implement
interlock between the processes.
When processing takes place over a number of stages in multiple
equipment units that are configured in a series, the recipe data must
encompass movement from unit to unit. Two things are required; recipe scheduling for the entire group of reactors involved, and reaction
cut- over conditions for each unit. Thus, the reaction conditions for
each unit must be entered into memory beforehand for each recipe.
Then, when a recipe is specified from the CRT, the control loop parameters and order of sequence operations are successively changed for
each unit following the path indicated by the recipe. Generally, the
762
Chap.lO Jnstrumention to Manufacturing Industries
product flow or unit path is fixed. However, this is not necessarily the
case in processes where multipurpose units are concerned ; that is,
where more than one reaction or operation takes place in a single unit .
Combining single-function units (where only a single operation takes
place) and multipurpose units, complicates the configuration. Moreover, since paths change with each recipe, sometimes the processing sequence path, or batch pass assignment from unit to unit is included in
the recipe itself. Two examples illustrating this kind of arrangement
are shown in Fig. 10.132.
Raw material
and addirives - - - ,
Barch parh of recipe A
Barch parh of recipe B
Fig. 10 132
Balch pass assignmenl of recipes
It is thus apparent that actual processes are considerably more

complex than our simplified configurations in Fig. 10.129. Recipe management methods are equally complex, and show considerable variation
depending on the product and the process. Nevertheless, using a basic
approach to recipe management (such as illustrated in Fig. 10.130) to
change control loop parameters, and the order and conditions of sequence operations, diverse products can be efficiently produced in
small batches.
10.8.2 Batch process control

It would be exceedingly difficlut to extend flexible automation to
diverse processes without first establishing control loop and sequence
operation methods. To facilitiate changing control parameters and the
order of sequence operations, a control loop and sequence operation
database must be constructed for the digital control system.
First, with respect to control loops, most commonly the same ap10.8
Product Control in Batch Processing
763
proach is adopted as applied to the construction of analog modules:

data is collected that is expressive of a particular modual function, and
designed so it can be constructed according to some method of coupling. For example, all data pertaining to the PID module, that is, PID
parameters, setpoint value, process variable, manipulated variable,
high and low alarm setpoint values, loop status, etc. and input/output
connection information would be lumped together in one module. In
the case of cascade control, two such modules could be interconnected
to form the loop. Since functions other than PID functions such as ratio, calculation, lag, and dead time are generally also implemented, all
control loops can be constructed out of these modules.
Consider, for example, the reactor control shown in Fig. 10.133.
For the reaction to proceed efficiently and safely, temperature control
is a primary consideration. The reaction is started by applying heat,
followed, at a critical moment of time just before the reaction becomes
critical, by removal of the heat. As a safety precaution, the temperature is always maintained at or below some prescribed temperature.
Or, there are situations where the reaction is divided into muiltiple
stages, and an additive has to be introduced over several or all of the
stages. For reaction temperature control in this kind of case, a control
loop can be configured out of standard modules such as a program setting module and a PID module that is capable of maintaining the desired temperature pattern. A screen depicting the operation of the temperature control loop is shown in Fig. 10.134. Note that the screen also
shows the program setting module and the temperature pattern in addition to the two temperature control modules. The advantage of DCSs
is clearly demonstrated when its time to change the temperature pattern for a different recipe; the data in the program setting module can
simply be written over to make the change.
Also, since there is some delay in the temperature response at the
beginning of the reaction, a compensating element is sometimes added
Raw material A
Raw material B
Add itive
Fig. 10 133 Reactor control
764
Chap.l 0 /nstrumention to Manufacturing Industries
Fig. 10 134 Display of reactor operation
\ \
\1ooc1
80
--~
6v
Witout t ompj"sati/n
20
lo I
fl--
ltemperatu
nterna~~
re
Pm.
h
Fig. 10 135 Control response of reactor temperature
765
to assure that the raise in temperature follows the prescribed pattern.

Furthermore, in the latter half of the reaction, reactant is depleted and
process sensitivity declines. When this happens, it is easy for temperature deviations to occur. An appropriate means to counter this effect is
conversion of temperature control sensitivity. Because the extent of
compensation needed and the amount of sensitivity cut-over for temperature control are recipe dependent, these data must be prepared in
advance for each recipe. Here, as we saw before, the data settings can
be easily changed whenever necessary by means of a DCS. Figure
10.135 shows an example of precise temperature ramping for a particular recipe using a DCS.
Turning now to sequence operations, Figure 10.136 shows how sequence phases can be changed for different recipes. In the case of
multipurpose reactors, not only operating conditions such as raw mate-
Fig. 10 136
766
Recipe-dependent sequence phases
Chap.l 0 /nstrumention to M anujacturing Industries
rial and charging volumes, temperature, and time, but also sequence
operation phases must be changed depending on the type of recipe.
Thus, the sequence control method must be configured in such a way
that the operation phases can also be rearranged in different ways according to the specific parameters involved. Moreover, each phase contains a series of detailed sequence steps: What valve is to be operated,
what pump is to be activated, how control loops are to be configured,
and so on. It is essential, therefore, that the detailed sequence steps
within the phases are also capable of being rearranged to coordinate
with the rearranged phases. In other words, the series of detailed sequence steps expressing, say, pump and valve actions, is typical of a
specific phase, and the sequence of phases is stipulated in the product-specific recipe data. How finelly the detailed sequence steps are
divided, or the scale of the units, will vary. Generally, however, it can
be said that if the degree of freedom is large, the detailed sequence
step units can be made small, and the corresponding phases will increase. Conversely, if the degree of freedom is reduced, the detailed sequence step units can be increased in size, and the corresponding number of phases can be reduced.
The relation between phases and detailed sequence steps is summarized in Fig. 10.137. So-called decision tables are shown in the same figure. When input conditions are entered in the upper tiers of the tables,
then the corresponding output operations shown in the lower tiers are
executed according to rules. This type of table is a useful device for describing the logical relations between input conditions and output
operations. Moreover, since operatoin instructions of one decision table can be executed from another decision table, this is also an appropriate description for sequence control methods with a hierarchical
structure. The step execution table is used to describe the detailed sePhase control table
Step
A
c D
Input Conditions of phase
transition
Output Instruction for each
phase execution
~
......
......
Step execution
table
......
Step
A Phase
Input
Conditions
A
Output Step
operations
Fig. 10 137
B Phase
Conditions
B
Step
operations
C Phose D Phose
Conditions Conditions
D
c
Step
Step
operations operations
Sequence phase and decision table structures for batch sequence control
767
quence steps for a single phase. These could include, for example, opening and closing valves, activating and stopping the temperature program setting module, implementing and severing control loop cascades,
changing settings, changing outputs, and so on. The phase control table, on the other hand, determines what step execution table is to be
executed and in what order according to the parameter conditions
designated for the particular recipe.
Since indicator lamps and push buttons are generally required for
batch process sequence operations, an example of what these can look
like is shown in Fig. 10.138. The selector switch is set to full-auto
mode when one wants the processing to proceed from phase to phase
automatically. Semi-auto mode allows automatic processing of only the
operations within a phase; when the phase is completed, the process
stops and waits for a instruction to proceed to the next phase. Setting
the selector switch to manual mode discontinues the automatic sequencing up to that point. This means that, as soon as a point is reached
where it is appropriate to stop, the process does so, and subsequent
operations are controlled manually. Start, stop, stepping, and reset are
activated by the push buttons labeled with those functions shown in
the same figure. Since this status transfer diagram can be easily modified using the decision table, the contents of the phase control table
can be designed. In other words, the system has been designed so that
the phases are controlled by the movement conditions of the status
transfer, and the order of the phases can be rearranged by referencing
the parameters.
Selector switch
Full-auto
Semi-auto
Manual
Push button
Lamps
Stand-by
Additive
charge
Raw material
charge
Reaction
Conveyance
Completion
CD
Not going ahead
Circles indicate sequence states;

arrows indicate transitions.
Fig. 10 138 Status transfer diagram of sequence and selector switches,

push buttons and lamps
768
Chap.JO /nstrumention to Manufacturing Industries
10.8.3 Recipe management and operation methods

In order to produce a diverse range of products by changing the
batch process operating conditions, two things are required; a great
volume of recipe data for each product, and a recipe data mangement
function. The recipe management function is used to select the product line, display it on screen, and view the operating status to see if
operating conditions need to be changed. In this section we will consider the recipe management function, and selection operation method.
It has already been mentioned that the recipe management function will differ depending on the configuration of the batch process.
The first configuration we will consider is a single unit process.
Through a sequence of panel operations such as shown in Fig. 10.139,
the recipe to be executed is selected, data respecting the selected recipe is brought up on the screen, and instructions altering operating conditions are entered. The first panel that appears is a menu of the recipes available. A selection is made from among the available choices,
and this brings the corresponding recipe data up on the screen. Since
the software permits altering the data on this panel, the batch isn't
started until after the recipe data has been verified. Once the batch
has been started, process operating conditions proceed in accordance
with the recipe data shown on the screen. Another point is that the recipe data is expressed as ratios of standard production units. In some
cases the ratios are calculated when the production volume is set. In
other cases, when the desired production volume cannot be ascertained
from the volume of the raw materials, the ratios are obtained indirectly either by calculating the residual volume or through a volume conversion based on the concentration or purity of a weight standard
chemical additive. Sometimes this recipe data is also shown on the
screen.
In configurations involving multiple equipment and reactor units,
a first requirement is to designate in which reactor a recipe is to be
produced. This is generally handled by creating lot numbers for all of
the reactors involved indicating the type of recipe and the reactor.
These numbers are then displayed along with the production conditions. In some cases, the kind of recipe that can be produced in a given
reactor will depend on the lines for charging raw material or additives,
or on the type of reactor. In such cases, the recipe selection to be carried out in a certain reactor is made only from among those recipes
that are appropriate for the reactor in question. When the product under production requires the use of a multi-unit configuration, the following information is shown: the unit in which the process starts, the
sequence of succeeding units, and the recipe.
769
85. 1&.13
23: 23
.:II:ICI:IIII:I:I
0
Recipe menu
o.
o. Name
a me
-+UI ~AIITJnl
Ill mLEi
Ill KARY
121 KAHHHTH
122 FASHIOn
123 JOHH
121m
m mson
125 HOPPEi
m JLm
127 noon
t2S DRm
m RAITAI
131 CHICMI
131 !LUE
132 SCORPIOn
ot~
c;~ ~uLE
lt5
117 Juct
119 iOHIC
In COLL!HS
Ill SWG
Ill JLOSSO"
112 SAIGUIRI
m REMaY
114 ZORIIE
115 liJRE
116 Rm~m
117 SUnRISE
113 sunm
m sunm
134 mK
135 Pm
136 SOUR
137 WE
m sm
m mum
1u sm
1(1
:mm
HI G~OIIYE
COPY 101 TO Ill
--------- ---------- - ----------------
lUll
TAG
fEr ~E1 1 COPY
1
Recipe menu panel
85. U.l4
3:41
11.
Recipe da1a base
001
MARTINI
tl T.C.TERP. I-+
12 PRG. TIRE Z
13 PRG. TEMP.Z
14 PRG. TlftE 3
ISPRG . TEMP.l
16 PRG. TIRE 4
17PR6. TERP.4
18 PR6. mE 5
nPRG. tm.s
II BASE POUR
II :1 POUR
POUR
IZ
13 :3 POUR
14 :4 POUR
:z
TAG
oc
OSEC
3 o oc
30SEC
4S. oc
6 om
4 s. oc
9 0 SEC
3S. oc
3 o.
1 0 Ol
SOl
SOL
SOl
SOl
1 0 SEC
6 SSEC
5 OSEC
7 0 SEC
70SEC
70SEC
7 OSEC
- 1 SEC
II I.C. START
12 PG. T.C. SIAI

13 J.POURUG
14 :s POURinG
IS :6 POURinG
16 :7 POURUG
17 :s POURinG
18 "An.POURinG
"
II
II
12
13
14 3-PROCESS
170SEC
------- ----- ---------- - -------------- - ------------------UJI

Recipe dala base panel
Fig. 10 139
770
Recipe menu a nd data base pa nel
Chap.JO
Recipe C
Fig. 10 140 Recipe ma nagement of multistage proce s
In cases such as illustrated in Fig. 10.140, where the equipment is

arranged in a multistage configuratoin, operating conditions cannot
simply cut over all at once in accordance with the recipe data. Rather,
cut over must follow the sequence of the equipment units. In other
words, the recipe data must be broken into segments corresponding to
each unit, so that when one unit is started, only the operating conditions for that unit will be modified. Figure 10.141 illustrates how the
unit sequence and recipe selection are shown on the screen. In the recipe shown in the figure, the maximum number of stages that can be
designated for each sequence, eight, has been designated. When one wishes to display that recipe data, first an overview of the data for each
recipe is displayed. In cases where the production involves a multistage unit configuration, the volume of recipe data generally increases.
Since it is not possible to display all of the data at the same time, the
data is broken up into groups. The groups, first displayed altogether
on a menu screen, are segmented so as to keep all data pertaining to
each stage of each unit together in one group. When a data group is indicated on the recipe overview screen, this brings up the recipe data
base for a single equipment unit, such as shown for example in Fig.
10.139.
If the unit start instruction is given on the multi-unit batch selection panel, only the data group for the relevant unit is selected for
modification from among the previous stage recipe data. In this case
the units' path is fixed, and represents a simple, series configuration.
Of course, there are possibilities: When the units are set up to serve diverse purposes, for example, the paths themselves can be changed according to the recipes. In situations such as this, the unit path is included as data in the recipe database. Then, depending on the path, the
recipe name of the stage prior to that unit and the recipe data group
for that unit must be searched. Figure 10.142 illustrates a recipe product cut-over method that contains a path table.
Just as process instrumentation varies widely depending on a multitude of different factors, recipe management methods also vary tre10.8 Product Control in Batch Processing
771
eS.IU3
~~111111
tS: 43
Prod uCI Iinc SlaiUS

Assign men I
Rocipe
name
In opera lion
1-+ Ul
~A~IIH!
Lo1
9HHI
Unil
A-TAHK
o.
118
IOHIC
95-JH~
21-1m
103
mY
e~-JH3
016 KA;.mH 95-JH<
o.
Rocipe
name
Lo1
C-Tm
6
7
s
PAHH -------------------- -- --- ----------- --- --- -----------------I) i-t
I 1- ~ I
Balch as.signmenl and currenl balch s1age display
9S.IU4
.illlll!l:lllll
3:37
Rocipe dala base overview
I AG
001
MARTINI
---- ---- ------------ --------------------- ----------------
Rocipe dala base overview panel
Fig. 10 141
772
Batch assignment and recipe data base over view panel
Path table
Unit I
Unit 2
Unit3
~r
or recipe
process path
Unit I
data
l .I Set
Process sequence
....... .. X
X
~ Set
:
Recipe data
Unit2
Recipe data
Unit 3
Recipe data
Fig. 10 142
Set data
Process sequence
Unit 3
Recipe data table
Unit I
Unit 2
data
Process sequence
I
l
l
Recipe management with pass assignment
mendously depending on the specific process configuration. Our approach here has been to describe prototypical schemes, or formulas, to
indicate the range of different possibilities. By changing the configuration of equipment, any number of application examples could have
been described. In adopting a formulaic approach, our intent was to
clearly reveal the unique advantage of digital control systems-that is,
flexible automation. There are other topics relating to recipe management that are also important ; collection of actual results data, formation and testing of recipe data, comparison of reaction patterns, optimization of operation scheduling, and so on. With respect to digital
control systems, however, it is safe to say that as these systems continue to develop and become more sophisticated, increasingly easy-to-use
configurations will be realized.
REFERENCES
1) T. Ishii : Process Control Instrumentation Systems, Denki Shoin (1973) (in Japa-
nese).
2) M. Tanaka: "Topping unit," Petrotech , 7 , 4 (1984) (in Japanese).
3) S. Hiramatsu, et a/.: "Problems in instrumentation at oil industry," Automation,
28,8 (1983) (in Japanese).
4) H. Hirano : MRP in Medium and Small Enterprises, Nikkan Kogyo Shinbun Sha
5) T . Fujimura, et a/. : "Instrumentation and control system for refinary off-site
equipment," Yokogawa Tech. Rep. , 29,3 (1985) 31-38 (in Japanese).
6) Y. Kawase, et al. : Oil Refinery Technology Handbook, Sangyo Tosho (1981) (in
References
773
Japanese).
The Iron and Steel Institute of Japan: Iron and Steel Industries in Japan (1985)
(in Japanese).
8) Comprehensive Bibliography for Iron Manufacturing Machinery '80, Jukogyo
Shinbunsha (in Japanese).
9) Nippon Steel Corporation: Iron Science (Making Process of Iron) (in Japanese).
10) Seitetsu Kenkyu, No. 308, Nippon Steel Corporation (1982) (in Japanese).
11) Tekko Kaiho, The Japan Iron and Steel Federation (June 1984) (in Japanese).
12) Sumitomo Metals, Sumitomo Metal Industries, Ltd. (in Japanese).
13) Tetsu-to-Hagane, 71, 3, The Iron and Steel Institute of Japan (1985) (in Japanse).
14) Kawasaki Steel Giho; 14, Kawasaki Steel Corporation (1982) (in Japanese).
15) Seitetsu Kenkyu; No. 313, Nippon Steel Corporation (1984) (in Japanese).
16) The Thermal and Nuclear Power, 29, 6/8, Thermal and Nuclear Power Engineering Society (1978) (in Japanese).
17) K. Shirano: Zymurgy, Kodansha (1982) (in Japanese).
18) Shokuryo Kogyo: Agricultural Food-Sugar, Koseisha Kosei Kaku (1985) (in Japanese).
19) R. Machida: "Production control system on YEWCOM at sugar refinery," Factory
Automation, 3, 7 (1985) (in Japanese).
20) Shigyo Times, The Latest Pulp and Paper Technology '80 (in Japanese).
21) Y. Murakami: Japan Journal of Paper Technology (Aug. 1983) (in Japanese).
22) A. Nomoto: Japan Journal of Paper Technology (Mar. 1982) (in Japanase).
23) T. Shibata: Data Systems for Automated Production and Material Hand.Jing,
Ryutsu Kenkyu Sha (Mar. 1984) (in Japanse).
24) Japan Water Works Association: Recommendation for Design of Water Works Facilities (1977) and Recommendation for Maintenance of Water Works (1982) (in
Japanese).
25) S. Nagase, et at.: "Data base management system on city water distribution
plants," Yokogawa Tech. Rep., 28,1 (1984) 8-13 (in Japanese).
26) S. Nagase: "Prediction of demand and total control systems for water supply
plants," Yokogawa Tech. Rep., 24, 1, (1980) 17-22 (in Japanese).
27) H. Kamei, eta/.: "On-line water demand predictions with Kalman filter," Yokogawa Tech. Rep., 25, 4, (1981) 36-40 (in Japanese).
28) K. Minamimura, et a!.: "Simulation of water distribution networks," Yokogawa
Tech. Rep. 28,1 (1984) 14-20 (in Japanese).
29) T. Yamamoto: "Hierarchical control system for wide-area water supply network," Yokogawa Tech. Rep., 28, 1, (1984) 21-25 (in Japanese).
30) Yokogawa Electric Corporation: Application Engineering Data Chemical Injection
Control and Water Quality Monitoring at Purification Plant (in Japanese).
31) Yokogawa Electric Corporation: Application Engineering Data Instrumentation
on Filter Basin and Washing Control (in Japanese).
32) Japan Sewage Works Association: Recommendation and Explanation for Design
of Wastwater Facilities (1984) (in Japanese).
33) K. Matsunaga: "The .latest instrumentation for batch process by distributed digital control system "CENTUM"," Yokogawa Tech. Rep., 23, 3 (1979) 26-31 (in Japanese).
34) T. Hirano, et at.: "Distributed control system applications in batch processes and
product grating," Yokogawa Tech. Rep., 25, 1, (1981) 44-52 (in Japanese).
35) K. Matsunaga: "Documenting process control sequence by decision tables," Yokogawa Tech. Rep., 26, 3, (1982) 44-50 (in Japanese).
36) T. Hirano, et al.: "Expediting the design of batch process control systems," Yok7)
774
37)
38)
ogawa Tech. Rep., 29, 3 (1985) 25-30 (in Japanese).

I. Miyazaki, et al.: "Functions and usage of CRT con sol on batch process," Instrumentation & Control Eng., 29, 6 (1986) 30-36 (in Japanese).
T. Kadoya: Paper Science, Chugai Sangyo Chosakai (1977) (in Japanese).
References
775
APPENDIXES
Appendixes
777
APPENDIX 1 REFERENCE THERMOELECTROMOTIVE FORCE TABLES

(Extracted from JIS C 1602-1981 which conforms to
IEC 584-1977, BS 4937-1973, DINIEC 584-1984,
ASTM E230-1983, ANSI MC96.1-1982)
Unit : pV
Type: R
Temp.
Temp.
(c)
-10
-20
-30
-40
-50
-51
-100
-145
-188
-226
10
20
30
40
50
60
70
80
90
0
647
1468
2 400
3 407
4 471
5 582
6 741
7 949
9 203
10 503
11846
13 224
14 624
16 035
17 445
18 842
20 215
54
723
1557
2 498
3 511
4 580
5 696
6 860
8 072
9 331
10 636
11983
13 363
14 765
16176
17 585
18 981
20 350
Ill
800
1647
2 596
3 616
4 689
5 810
6979
8196
9 460
10 768
12119
13 502
14 906
16 317
17 726
19119
20 483
171
879
1738
2 695
3 721
4 799
5 925
7 098
8 320
9 589
10 902
12 257
13 642
15 047
16 458
17 866
19 257
20 616
232
959
1830
2 795
3826
4 910
6 040
7 218
8 445
9 718
11035
12 394
13 782
15188
16 599
18 006
19 395
20 748
296
1 041
1923
2 896
3 933
5 021
6155
7 339
8 570
9 848
11170
12 532
13 922
15 329
16 741
18146
19533
20878
363
1124
2 017
2 997
4 039
5132
6272
7 460
8 696
9978
11304
12 669
14 062
15 470
16 882
18 286
19 670
21006
431
1208
2lll
3 099
4146
5 244
6388
7 582
8 822
10109
11439
12 808
14 202
15 611
17 022
18 425
19 807
501
1294
2 207
3 201
4 254
5 356
6 505
7 703
8 949
10 240
11574
12 946
14 343
15 752
17163
18 564
19 944
573
1380
2 303
3 304
4 362
5 469
6 623
7 826
9 076
10 371
11710
13 085
14 483
15 893
17 304
18 703
20 080
Temp.
(c)
0
100
200
300
400
500
600
700
800
goo
1000
1100
1200
1300
1400
1500
1600
1700
(c>
0
Temp.
(c)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
Remark: Temperature of reference junction is oc.

When temperature of reference junction is 20"C, subtract 111 JlV
from the value given in the above table.
778
Appendixes
Type: S
Temp.
("C)
0
Temp.
("C)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1 700
Unit: ;N
Temp.
-10
-20
-30
-40
-50
-53
-103
-150
-194
-236
10
20
30
40
50
60
70
80
90
0
645
1440
2 323
3 260
4 234
5 237
6 274
7 345
8 448
9 585
10 754
11947
13 155
14 368
15 576
16 771
17 942
55
719
1 525
2 414
3 356
4 333
5 339
6 380
7 454
8 560
9 700
10 872
12 067
13 276
14 489
15 697
16 890
18 056
113
795
1611
2 506
3 452
4 432
5 442
6 486
7 563
8 673
9 816
10 991
12 188
13 397
14 610
15 817
17 008
18 170
173
872
1698
2 599
3 549
4 532
5 544
6 592
7 672
8 786
9 932
11110
12 308
13 519
14 731
15 937
17 125
18 282
235
950
1785
2 692
3 645
4 632
5 648
6 699
7 782
8 899
10 048
11229
12 429
13 640
14 852
16 057
17 243
18 394
299
1029
1873
2 786
3 743
4 732
5 751
6 805
7 892
9 012
10 165
11348
12 550
13 761
14 973
16 176
17 360
18 504
365
1109
1962
2 880
3 840
4 832
5 855
6 913
8 003
9126
10 282
11467
12 671
13 883
15 094
16 296
17 477
18 612
432
1190
2 051
2 974
3 938
4 933
5 960
7 020
8114
9 240
10 400
11587
12 792
14 004
15 215
16 415
17 594
502
1273
2141
3 069
4 036
5 034
6 064
7128
8 225
9 355
10 517
11707
12 913
14 125
15 336
16 534
17 711
573
1356
2 232
3164
4 135
5136
6169
7 236
8 336
9 470
10 635
11827
13 034
14 247
15 456
16 653
17 826
("C)
0
Temp.
("C)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
Remark : Temperature of reference junction is 0 OC.

When temperature of reference junction is 20 OC, subtract 113 J1 V
from the value in the above table.
App. 1
Reference Thermoelectromotive Force Table
779
Type: K
Unit: pV
Temp. 'C) 0
-200
-100
0
Temp.
('C)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
-10
-20
-30
-40
-50
-60
-70
-80
-90
-5891 -6035 -6158 -6262 -6 344 -6404 -6 441 -6458

-3553 -3852 -4138 -4 410 -4 669 -4 912 -5141 -5354 -5550 -5 730
0 -392 -777 -1156 -1527 -1889 -2243 -2586 -2920 -3242
0
10
0
397
4 095 4 508
8137 8 537
12 207 12 623
16 395 16 818
20 640 21066
24 902 25 327
29128 29 547
33 277 33 686
37 325 37 724
41269 41657
45108 45 486
48828 49192
52 398 52 747
20
30
798
1203
4 919 5 327
9 341
8 938
13 039 13 456
17 241 17 664
21493 21919
25 751 26176
29 965 30 383
34 095 34 502
38122 38 519
42 045 42 432
45 863 46238
49 555 49 916
53 093 53 439
40
1611
5 733
9 745
13 874
18 088
22 346
26 599
30 799
34 909
38 915
42 817
46 612
50 276
53 782
50
2 022
6137
10 151
14 292
18 513
22 772
27 022
31214
35 314
39 310
43 202
46985
50 633
54125
60
2 436
6 539
10 560
14 712
18 938
23198
27 445
31629
35 718
39 703
43 585
47 356
50 990
54 466
70
80
2 850 3 266
6 939
7 338
10 969 11381
15132 15 552
19 363 19 788
23 624 24 050
27 867 28 288
32 042 32 455
36121 36 524
40 096 40 488
43 968 44 349
47 726 48 095
51344 51697
54 807
90
3 681
7 737
11793
15 974
20 214
24 476
28 709
32 866
36 925
40 879
44 729
48 462
52 049
Temp.
("C)-
-200
-100
0
Temp.
("C)0
100
200
300
400
500
600
700
800
900
I 000
1100
1200
1300
Remark : Temperature of reference junction is O'C.

When temperature of reference junction is 20 'C, subtract 798 pV
Type: E
Unit: pV
Temp.
('C)
-10
-20
-30
-40
-50
-60
-70
-80
-90
Temp.
('C)
-200 -8824 -9063 -9 274 -9455 -9 604 -9 719 -9 797 -9835

-200
-100 -5237 -5680 -6107 -6 516 -6 907 -7 279 -7631 -7963 -8 273 -8 561 -100
0
0 - 581 -1151 -1709 -2254 -2 787 -3 306 -3811 -4 301 -4 777
0
Temp.
0
10
20
30
40
50
60
70
80
90 Temp.
('t)
0
100
200
300
400
500
600
700
800
900
1000
0
6 317
13 419
21 033
28 943
36 999
45 085
53110
61022
68 783
76 358
591
6 996
14 161
21814
29 744
37 808
45 891
53 907
61806
69 549
1192
7 683
14 909
22 597
30 546
38 617
46 697
54 703
62 588
70 313
1801
8 377
15 661
23 383
31350
39 426
47 502
55 498
63 368
71075
2 419
9 078
16 417
24171
32155
40 236
48 306
56 291
64147
71835
3 047
9 787
17178
24 961
32 960
41045
49109
57 083
64 924
72 593
3 683
10 501
17 942
25 754
33 767
41853
49 911
57 873
65 700
73 350
4 329
11222
18 710
26 549
34 574
42 662
50 713
58 663
66 473
74104
4 983
11949
19 481
27 345
35 382
43 470
51513
59 451
67 245
74 857
5 646
12 681
20 256
28143
36190
44 278
52 312
60 237
68 015
75 608
('C)-
0
100
200
300
400
500
600
700
800
900
1000
Remark : Temperature of reference junction is 0 'C.

When temperature of reference junction is 20 'C, subtract 1192 pV
780
Appendixes
Type: J
Temp.
('C)
Unit: pV
0
-10
-20
-30
-40
-50
-60
-70
-80
-90
-200 -7 890 -8096

-100 -4 632 -5036 -5426 -5801 -6159 -6499 -6821 -7122 -7 402 -7659
0
0 -501 -995 -1481 -1960 -2 431 -2892 -3344 -3 785 -4 215
Temp.
<c)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
10
20
0
5 268
10 777
16 325
21846
27 388
33 096
39130
45 498
51875
57 942
63 777
69 536
507
5812
11332
16879
22 397
27949
33 683
39 754
46144
52 496
58 533
64 355
1 019
6 359
11887
17 432
22 949
28 511
34 273
40 382
46 790
53115
59121
64 933
30
40
1536 2 058
6 907
7 457
12 442 12 998
17 984 18 537
23 501 24 054
29 075 29 642
34 867 35 464
41013 41647
47 434 48 076
53 729 54 341
59 708 60 293
65 510 66 087
50
60
2 585
8 008
13 553
19 089
24 607
30 210
36 066
42 283
48 716
54 948
60 876
66664
3115
8 560
14 108
19 640
25 161
30 782
36 671
42 922
49 354
55 553
61459
67 240
70
80
3 649 4 186
9113 9 667
14 663 15 217
20192 20 743
25 716 26 272
31356 31933
37280 37 893
43 563 44 207
49 989 50 621
56155 56 753
62 039 62 619
67 815 68 390
90
4 725
10 222
15 771
21295
26 829
32 513
38 510
44 852
51249
57 349
63199
68 964
Temp.
('C)
-200
-100
0
Temp.
("C)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
Remark : Temperature of reference junction is 0 c.

When temperature of reference junction is 20 c, subtract 1 019 pV
Unit : pV
Type: T
Temp.
("C)
-10
-20
-30
-40
-50
-60
-70
-80
-90
-200 -5603 -5753 -5889 -6007 -6105 -6181 -6232 -6258

-100 -3 378 -3656 -3923 -4177 -4 419 -4 648 -4 865 -5 069 -5261 -5 439
0 -383 -757 -1121 -1475 -1819 -2152 -2475 -2788 -3089
0
Temp.
('C)
0
0
100
4 277
200 9 286
300 14 860
400 20 869
10
391
4 749
9 820
15 443
20
30
40
50
789
5 227
10 360
16 030
1196
5 712
10 905
16 621
1 611
6 204
11456
17 217
2 035
6 702
12 011
17 816
60
2 467
7 207
12 572
18 420
70
80
90
2 908
7 718
13137
19 027
3 357
8 235
13 707
19 638
3 813
8 757
14 281
20 252
Temp.
<c>
-200
-100
0
Temp.
("C)
0
100
200
300
400
Remark : Temperature of reference junction is 0 c.

When temperature of reference junction is 20 c, subtract 789 pV
App. 1 Reference Thermoelectromotive Force Table
781
APPENDIX 2 REFERENCE RESISTANCE VALUE OF Pt 100

(Extracted from JIS C 1604-1989 which conforms to
IEC 751-1983, BS 1904-1984, DINIEC 751-1985)
Temperature "C
-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
Temperature "C
-200
18.49
-190
-!80
-170
-160
-!50
22.80
27.08
31.32
35.53
39.71
22.37
26.65
30.90
35.11
39.30
21.94
26.23
30.47
34.69
38.88
21.51
25.80
30.05
34.27
38.46
21.08
25.37
29.63
33.85
38.04
20.65
24.94
29.20
33.43
37.63
20.22
24.52
28.78
33.01
37.21
19.79
24.09
28.35
32.59
36.79
19.36
23.66
27.93
32.16
36.37
18.93
23.23
27.50
31.74
35.95
18.49
22.80
27.08
31.32
35.53
-!90
-!80
-170
-160
-150
-140
-130
-120
-110
-100
43.87
48.00
52.11
56.19
60.25
43.45
47.59
51.70
55.78
59.85
43.04
47.18
51.29
55.38
59.44
42.63
46.76
50.88
54.97
59.04
42.21
46.35
50.47
54.56
58.63
41.79
45.94
50.06
54.15
58.22
41.38
45.52
49.64
53.74
57.82
40.96
45.11
49.23
53.33
57.41
40.55
44.70
48.82
52.92
57.00
40.13
44.28
48.41
52.52
56.60
39.71
43.87
48.00
52.11
56.19
-140
-!30
-!20
-110
-100
90
80
70
60
50
64.30
68.33
72.33
76.33
80.31
63.90
67.92
71.93
75.93
79.91
63.49
67.52
71.53
75.53
79.51
63.09
67.12
71.13
75.13
79.11
62.68
66.72
70.73
74.73
78.72
62.28
66.31
70.33
74.33
78.32
61.87
65.91
69.93
73.93
77.92
61.47
65.51
69.53
73.53
77.52
61.06
65.11
69.13
73.13
77.13
60.66
64.70
68.73
72.73
76.73
60.25
64.30
68.33
72.33
76.33
90
80
70
60
50
- 40
- 30
- 20
- 10
0
84.27
88.22
92.16
96.09
100.00
83.88
87.83
91.77
95.69
99.61
83.48
87.43
91.37
95.30
99.22
83.08
87.04
90.98
94.91
98.83
82.69
86.64
90.59
94.52
98.44
82.29
86.25
90.19
94.12
98.04
81.89
85.85
89.80
93.73
97.65
81.50
85.46
89.40
93.34
97.26
81.10
85.06
89.01
92.95
96.87
80.70
84.67
88.62
92.55
96.48
80.31
84.27
88.22
92.16
96.09
40
30
20
10
0
10
0
10
20
30
40
100.00
103.90
107.79
111.67
115.54
100.39
104.29
108.18
112.06
115.93
100.78
104.68
108.57
112.45
116.31
101.17
105.07
108.96
112.83
116.70
101.56
105.46
109.35
113.22
117.08
101.95
105.85
109.73
113.61
117.47
102.34
106.24
110.12
113.99
117.85
102.73
106.63
110.51
114.38
118.24
103.12
107.02
110.90
114.77
118.62
103.51
107.40
111.28
115.15
119.01
103.90
107.79
111.67
115.54
119.40
0
10
20
30
40
50
60
70
80
90
119.40
123.24
127.07
130.89
134.70
119.78
123.62
127.45
131.27
135.08
120.16
124.01
127.84
131.66
135.46
120.55
124.39
128.22
132.04
135.84
120.93
124.77
128.60
132.42
136.22
121.32
125.16
128.98
132.80
136.60
121.70
125.54
129.37
133.18
136.98
122.09
125.92
129.75
133.56
137.36
122.47
126.31
130.13
133.94
137.74
122.86
126.69
130.51
134.32
138.12
123.24
127.07
130.89
134.70
138.50
50
60
70
80
90
100
110
120
130
140
138.50
142.29
146.06
149.82
153.58
138.88
142.66
146.44
150.20
153.95
139.26
143.04
146.81
150.57
154.32
139.64
143.42
147.19
150.95
154.70
140.02
143.80
147.57
151.33
155.07
140.39
144.17
147.94
151.70
155.45
140.77
144.55
148.32
152.08
155.82
141.15
144.93
148.70
152.45
156.19
141.53
145.31
149.07
152.83
156.57
141.91
145.68
149.45
153.20
156.94
142.29
146.06
149.82
153.58
157.31
100
110
120
130
140
150
160
170
180
190
157.31
161.04
164.76
168.46
172.16
!57. 69
161.42
165.13
168.83
172.53
158.06
161.79
165.50
169.20
172.90
158.43
162.16
165.87
169.57
173.26
158.81
162.53
166.24
169.94
173.63
159.18
162.90
166.61
170.31
174.00
159.55
163.27
166.98
170.68
174.37
159.93
163.65
167.35
171.05
174.74
160.30
164.02
167.72
171.42
175.10
160.67
164.39
168.09
171.79
175.47
161.04
164.76
168.46
172.16
175.84
150
160
170
180
190
200
210
220
230
240
175.84
179.51
183.17
186.82
190.45
176.21
179.88
183.53
187.18
190.81
176.57
180.24
183.90
187.54
191.18
176.94
180.61
184.26
187.91
191.54
177.31
180.97
184.63
188.27
191.90
177.68
181.34
184.99
188.63
192.26
178.04
181.71
185.36
189.00
192.63
178.41
182.07
185.72
189.36
192.99
178.78
182.44
186.09
189.72
193.35
179.14
182.80
186.45
190.09
193.71
179.51
183.17
186.82
190.45
194.07
200
210
220
230
240
250
260
270
280
290
194.07
197.69
201.29
204.88
208.45
194.44
198.05
201.65
205.23
208.81
194.80
198.41
202.01
205.59
209.17
195.16
198.77
202.36
205.95
209.52
195.52
199.13
202.72
206.31
209.88
195.88
199.49
203.08
206.67
210.24
196.24
199.85
203.44
207.02
210.59
196.60
200.21
203.80
207.38
210.95
196.96
200.57
204.16
207.74
211.31
197.33
200.93
204.52
208.10
211.66
197.69
201.29
204.88
208.45
212.02
250
260
270
280
290
300
310
320
212.02 212.37
215.57 215.93
219.12 219.47
212.73 213.09 213.44 213.80 214.15 214.51 214.86 215.22 215.57

216.28 216.64 216.99 217.35 217.70 218.05 218.41 218.76 219.12
219.82 220.18 220.53 220.88 221.24 221.59 221.94 222.29 222.65
300
310
320
Temper
ature "C
782
10
-200
Temperature "C
Appendixes
Temperature t
Temperature t
10
330
340
222.65
226.17
223.00
226.52
223.35
226.87
223.70
227.22
224.06
227.57
224.41
227.92
224.76
228.27
225.11
228.62
225.46
228.97
225.81
229.32
226.17
229.67
330
340
350
360
370
380
390
229.67
233.17
236.65
240.13
243.59
230.02
233.52
237.00
240.47
243.93
230.37
233.87
237.35
240.82
244.28
230.72
234.22
237.70
241.17
244.62
231.07
234.56
238.04
241.51
244.97
231.42
234.91
238.39
241.86
245.31
231.77
235.26
238.74
242.20
245.66
232.12 232.47
235.61 235.96
239.09 239.43
242.55 242.90
246.00 246.35
232.82
236.31
239.78
243.24
246.69
233.17
236.65
240.13
243.59
247.04
350
360
370
380
390
400
410
420
430
440
247.04
250.48
253.90
257.32
260.72
247.38
250.82
254.24
257.66
261.06
247.73
251.16
254.59
258.00
261.40
248.07 248.41
251.50 251.85
254.93 255.27
258.34 258.68
261.74 262.08
248.76
252.19
255.61
259.02
262.42
249.10
252.53
255.95
259.36
262.76
249.45
252.88
256.29
259.70
263.10
249.79
253.22
256.64
260.04
263.43
250.13
253.56
256.98
260.38
263.77
250.48
253.90
257.32
260.72
264.11
400
410
420
430
440
450
460
470
480
490
264.11
267.49
270.86
274.22
277.56
264.45
267.83
271.20
274.55
277.90
264.79
268.17
271.53
274.89
278.23
265.13
268.50
271.87
275.22
278.56
265.47 265.80
268.84 269.18
272.20 272.54
275.56 275.89
278.90 279.23
266.14
269.51
272.88
276.23
279.56
266.48
269.85
273.21
276.56
279.90
266.82
270.19
273.55
276.89
280.23
267.15
270.52
273.88
277.23
280.56
267.49
270.86
274.22
277.56
280.90
450
460
470
480
490
500
510
520
530
540
280.90
284.22
287.53
290.83
294.11
281.23
284.55
287.86
291.16
294.44
281.56
284.88
288.19
291.49
294.77
281.89 282.23
285.21 285.54
288.52 288.85
291.81 292.14
295.10 295.43
550
560
570
580
590
297.39
300.65
303.91
307.15
310.38
297.72 298.04 298.37

300.98 301.31 301.63
304.23 304.56 304.88
307.47 307.79 308.12
310.70 311.02 311.34
600
610
620
630
640
313.59 313.92
316.80 317.12
319.99 320.31
323.18 323.49
326.35 326.66
650
329.51
Temperature "C
314.24
317.44
320.63
323.81
326.98
329.82 330.14
282.56 282.89
285.87 286.21
289.18 289.51
292.47 292.80
295.75 296.08
283.22 283.55
286.54 286.87
289.84 290.17
293.13 293.46
296.41 296.74
283.89 284.22
287.20 287.53
290.50 290.83
293.79 294.11
297.06 297.39
500
510
520
530
540
298.70
301.96
305.20
308.44
311.67
299.02 299.35
302.28 302.61
305.53 305.85
308.76 309.09
311.99 312.31
299.68 300.00
302.93 303.26
306.18 306.50
309.41 309.73
312.63 312.95
300.33 300.65
303.58 303.91
306.82 307.15
310.05 310.38
313.27 313.59
550
560
570
580
590
314.56 314.88
317.76 318.08
320.95 321.27
324.13 324.45
327.30 327.61
315.20 315.52
318.40 318.72
321.59 321.91
324.76 325.08
327.93 328.25
315.84 316.16
319.04 319.36
322.22 322.54
325.40 325.72
328.56 328.88
316.48 316.80
319.68 319.99
322.86 323.18
326.03 326.35
329.19 329.51
600
610
620
630
640
330.45
330.77
331.08
331.40
331.71
332.03
332.34
332.66
10
650
Temper
ature "C
The reference resistance value of Attached Table shall be calculated from the following
formulas:
Range of -200'C to 0 'C: R =R. [ 1 +At+ Bt2 +C(t-100) t 3 ]
Range of 0 'C to 650'C :
Rr =R. ( 1 +At+ Bt2 )
where,
A= 3.908 02X1()3 'C- 1
B = -5.802 X1() 7 c-2
c=
-4.273
sxw-12 c-
Remarks: 1 . R. means a resistance value at 100 Q and R at t'C

2 . The relationship formulas given above are for the calculation of
the reference resistance values for this Standard, and not
intended for obtaining the characteristics of the individual
resistance thermometer sensors.
App. 2
Reference Resistance Value of Pt 100
783
APPENDIX3
TABLES OF LAPLACE TRANSFORM
Table of Laplace transform pairs

No.
/(t)
F(s)
No.
/(/)
F(s)
o(t)
12
1,.(1-cos at)
a
s~s 2 +a 2 )
u(l)
1...
13
- 1-(e-- e-)
b-a
~s+a)~s+ b)
14
- 1 -(ae-- be-)
a-b
s
(s+a)(s+b)
e-at
1
s+a
15
te-
1
(s+a)
lo-e-')
16
e- 01 (1- at)
-h-<e- 01 +at -1)

a
17
~{1-(1- at)e- 01 )
sin at
18
t sin at
s
2a (s+a')
cos at
19
t cos at
Ts'+li')'"
sinh t
--;r=az
20
e- sin bt
b
(s+a) 2 +b 2
10
cosh t
s
s-a
21
e- cos bt
11
~at-sin at)
1
s 2(s 2 +a 2)
s~s+a)
1
s2(s+a)
~
~
1
1
~s+a)
s~s+a)
sz-az
s+a
~s+a) 2 +b 2
Laplace transform theorems

No.
/(I)
F(s)
No.
f(t)
F(s)
!!!iJl
sF(s)- /(0)
/(I -a)
e-F(s)
j;<t)dt
l..[F(s)+ r'(O))
s
e'J(t)
F(s-a)
c,J,(t)+c./.(1)
c,F,(s)+czF.(s)
e- 01/(t)
F(s+a)
/(at)
1zF(~)
10
lim/(!)
limF(s)
/(t/a)
aF(as)
11
lim/(!)
limF(s)
/(t +a)
eF(s)
-
--
12
f/.(r)/z(t- r)dr
F,(s)F.(s)
784
dt
-~
Appendixes
INDEX
A/D conversion
351
ABMS
659
accuracy
67
accuracy rating
68
adaptive control
512
advanced control
483
advanced thermal reactor
672
aeration tank
747
AFC
669
air reservoir
444
alarm check function
267
analog/digital conversion
351
analyzer
492
angle valve
369
anti-surge control for compressor
536
application
471
ARMA model method
516
ash content
188
ASTM
239
ATR
672
automatic burner management system
659
automatic combustion control
651
ACC
651
automatic control
12
automatic frequency control
669
automatic intermittent syrup boiling system
702
automatic lineup
617
B
band-pass filter
164
base metal thermocouple
73
BASIC
304
basic design review
473
basis weight
188,724
Index
batch process
249,760
batch reactor
514
battery backup
447
bellows
139
Bernoulli's equation
100
billet
635
bit-slice microprocessor
341
black body radiation
93
black liquor concentration process
715
blast furnace
624
bleaching-process
712
block diagram
13
bloom
635
board operation
257
Bode diagram
19
boiler time constant
652
boiler/turbine coordinate mode
668
boiler-follow mode
668
boiling water reactor
671
Bourdon tube
140
bridge circuit
64
brushless DC motor
215
builder/mainte nance function
289
bumpless switching
226,270
burnout function
198
butterfly valve
371
BWR
671
c
CAD/CAM system
427
cage valve
369
calibration
69
capacitance liquid-level meter
147
capacity control of a reciprocating compressor
538
cascade control
39
central processing unit
307
ceramic-sealed element
88
CGWU
311
785
char bed
718
characteristics equation
22
chemical and volume control system
678
chemical injection equipment instrumentation
732
CIM
247
CIP
689
cleaning in place
689
closed loop
506
cold junction
72,198
color-selection line control
755
colorimetry
183
column system
160
combustion control
649
communication gateway unit
311
communication interface
299
compensating lead wire
80
component cooling water system
679
compressor characteristic
535
compressor control
533
computer- integrated manufacturing
247
concentration control in distillation column train
568
concentrator control
699
concentric orifice
102
conditional control (monitor control)
51
configuration of furnace control system
547
consecutive reaction
583
constant-rate drying
596
containment spray system
680
continuous casting
635
continuous digester
709
continuous process
248
control of a batch fluidized-bed dryer
595
control of a batch polymerization reactor
578
control of a continuous fluidized-bed dryer
598
control of a recycle reaction system
588
control of distillation column pressure
560
control of fluid-to-fluid heat exchanger
542
control of heat exchanger
540
control of reactor
573
786
control of refrigerant compression

590
control panel
453
control station
256
control system for binary-component continuous distillation
556
control-rod control system
682
controlled variable
11
controller with external feedback
566
cooking-process
709
CPU
307
cross-limit
653
CRT
347
CRT operation
257
crude oil feedpump control
609
current-to-current positioner
386
current-to-pneumatic positioner
384
CVCF
438
cycle time
752
D
data-base management system
321
DDC
2,249
dead time
27,492
dead-time element
26
decision table
52,272,293
deflection method
63
degrees of freedom of process
23
delay in conveying
492
derivative action
36
design review
471
desired value
11
detailed design review
473
development support software
320
deviation
11
diaphragm
139
diaphragm valve
369
differential pressure flowmeter
99
differential transformer method
149
diffusion current
184
digital counting tachometer
156
digital signal processor
341
dipping thermocouple
84
direct digital control
249
direct measurement
62
discrete process
5,247,251
displacer liquid-level detector
144
distillation column
490
distillation equipment constraints and control
563
Index
distributed control system

247,256
distributed DDC
407,414
distributed direct digital control
249
disturbance
13
disturbance compensator
500
DNC
251
Doppler method
134
368
double-seated valve
144
dry leg method
dry part
723
drying speed
597
dual-port disk
315
duplex system
314
E
eccentric rotating plug valve

371
economical load dispatching
669
eddy current displacement gage
151
effective wavelength
93
ELD
669
electric pressure transmitter
138
electrolytic cleaning
644
electrolytic galvanizing
642
electromagnetic induction method
148
emergency operation for furnace protection

551
emissivity
93,96,97
end pressure control
738
engineered safety features operation sys682,685
tem
engineering interface
257
equal-percentage
361
equivalent dead time
30
equivalent time constant
30
error
66
estimation error
525
estimation water-delivery demand
742
evaporator control
592
event recording
212
652
excess air ratio
expansion correction factor
100
517
expert method
expert system
527
445
explosion-proof construction
F
5,251
FA
FA computer
Index
factory automation
5
failsafe
448
fan rule
535
fast breeder reactor
673
FBR
673
FDD
344
feasibility study
392
feedback control
11
feedback control function
267
feedforward control
486
feedwater control
655
feedwater supply control
649
FIF
289
file management
319
fill-in-the-form
232,289
731
filter rate control
filtration equipment instrumentation
731
first-order-lag
29,493
first-order-lag element
28
flash point
613
flexible automation
760
flexible disk drive
344
float liquid-level meter
141
float-type area flowmeter
105
100
flow coefficient
flow sheet
403
flowchart
271
fluidized-bed reactor control
585
flying capacitor multiplexer
202
444
FMEA
frequency response
18
FTA
444
fuel/air ratio control
652
fuel-air ratio control system
549
furnace
490
furnace blower
634
furnace purge system
659
fuzziness
527
fuzzy control
527
FWC
655
G
gain margin
23
gain-scheduling control
514
gate valve
369
Gaussian white noise
524
glass membrane
169
glass sealed element
88
globe valve
367
251,295
787
hard disk drive

346
HDD
346
headbox
722
heat control of topping unit
610
heat exchanger
486
heater control by drain valve adjustment
541
heating furnace control
546
296
hierarchical distributed system
high top-pressure operation
631
higher-order lag system
30
hookup drawing
466
hot junction
72
hot stove
632
human interface package
322
hybrid recorder
218
hydrogen ion activity
169
hyperbolic
361
inter-computer communication package

324
interaction
228
interaction coefficient
508
interface list
409
interlock
451
ionization chamber
180
ironmaking
622
J
JOB control
job summary
618
436
Kalman filter
524
Kalman vortex street
125
knowledge base
517
kraft pump manufacturing process
706
L
IEEE-488 instrument bus

329
IEEE-802 .4 method
335
incomplete derivative
37
incomplete differentiation
226
indirect measurement
62
industrial instrument
2
industrial measurement
61
inherent flow characteristic
361
inherent rangeability
363
injection dosage control
735
injection method
732
input noise
524
input scanner
220
installed flow characteristics
362
installed rangeability
363
instrumentation
1
instrumentation air supply
441
instrumentation for water-supply and dis737
tribution facilities
instrumentation work
460
integral action
34
integral element
27
integral optimal regulator
522
213
integration type AID converter
integrative process
496
intelligent input/output equipment
309
intelligent terminal
312
788
ladder diagram
293
ladle
635
LAN
252,334
language
304
Laplace transform
15
Jaw of intermediate conductors
73
law of successive temperatures
73
LD
353
learning control
527
LED
353
light-emitting diode
353
limit cycle method
516
linear
361
linearizing
195
liquid crystal
350
local area network
252
logic circuit
271
low frequency square wave excitationmethod
121
luminance temperature
93
M
machining center
251
magnetic balance method
152
magnetic flowmeter
117
magnetic storage device
344
magnetic strain method
153
Index
magnetic wind method

166
main steam & feedwater system
680
main steam relief valve control system
685
man-machine interface
257,276,301
11
manipulated variable
manual control
12
manufacturing automation protocol
327
MAP
327
maritime blending
614
mass absorption coefficient
179
melter Brix (concentration) control
695
membership function
527
MFT
660
mica-insulated element
88
mismatch
495
mix and charging system
760
mixture preparation control for a reactant
gas
587
model reference adaptive control
513
moisture content
724
mold
635
mold molten steel level control
638
molten steel level meter
638
MTSF
446
multi-component distillation column con566
trol
multi-stage fixed-bed reactor control
580
multi-tubular reactor control
582
multiple-effect evaporator
592
N
251
Nernst equation
170
noble metal thermocouple
73
nominal resistance
86
non-continuous process
5
non-interacting control
502,509
non-linear characteristic
514
nozzle
103
numerical control
251
Nyquist stability determination method
22
NC
observation noise
Index
524
off-site
614
33
offset
on-line identification device
513
on-line maintenance
262
on-off control
32
on-site
614
open loop
506
operating system
303,316
operation
403,453
operator station
256
optical fiber
353
optical pyrometer
96
optimal adjustment
48
521
optimal control
orifice plate
99
output equation
522
oval gear flowmeter
111
overhead method
462
overhead-type conveyer
753
override control
44
overshoot
45,519
oxidation-reduction potential meter
171
p
P&I
393
PA
5
paper plant
708
paper thickness
192
papermaking process
721
parameter change
501
parameter estimation
525
PBS line control
757
PC
251,409
PCI
634
peripheral integrated circuit
342
petroleum industry
604
pH control in the carbonization process
697
phase
493
phase control table
768
phase margin
23
pickling
644
PID control
249
PID control algorithm
269
piping network calculation
739
plant operation engineering
530
platinum resistance temperature detector
85,199
Pneumatic pressure transmitter
136
pneumatic pressure positioner
384
789
polarography
183
potentiometer
63
pr~paration of reactant-gas mixture
585
pressure detector
138
pressure differential liquid-level meter
141
pressurized water reactor
671
pressurizer pressure control system
684
pressurizer water-level control system
684
priority processing
318
process automation
5,247
process control
247,248
process control system
247
process data
407
process data acquisition package
322
process data highway
262
process dataway
328
process gain change of fluid-fluid heater
544
process interface
256,298
product specifications switchover control
578
production line control system
247,291
production management computer
247,306
production management computer system
247
program control (process control)
51,693
programmable controller
251,292,409
proportional band
33
proportional control
32
proportional element
26
proposal
427
proposal final review
473
proposal review
473
protective tube
91
PROWAY
328
pulse flow signal transmitter
238
pulverized coal injection
634
pump characteristic
531
pump control
530
purge-type liquid-level meter
145
purged gas
589
PV derivative
38
PWR
671
790
Q
quadrant edge orifice
103
quality control in continuous polymerization
575
quick-opening
361
R
93
radiation thermometer
RAM
343
rangeability
363
raster scan recorder
209
ratio control
41
reaction temperature control
764
reactor control equipment
681
reactor control system
682
reactor coolant system
678
reactor protection system
681,685
reactor safety protection system
685
real-time operating system
317
reboiler steam
503
recipe management
760
recovery boiler process
717
redundant system
446
reference resistance element
87
reflux flow
503
relative gain
504
relative volatility
554
reproducibility
68
reset windup
35,270,579
residual heat removal system
679
resistance potentiometer method
148
resistance ratio
87
resistance temperature detector with protective tube
89
resistance thermometer
84
rheometer
704
Riccati equation
523
rolling
623
ROM
343
Routh/Hurwitz stability determinationmethod
22
RS-232 C interface
331
s
safety injection system
679
safety protection system
682
sampling controller
494
Index
sampling PI controller
500
sanitary detector
689
Saunders valve
369
scaling
212
scheduling
762
sea water system
680
second-order lag element
29
selectivity
584
selector control
42
self-balancing method
63
self-balancing recorder
209
290
self-documentation function
self-powered valve
388
self-regulation
25
self-tuning controller
511
semiconductor detector
164
semiconductor laser diode
353
semiconductor memory
343
sensitivity
69
separation of distillation
556
sequence control
249
sequence control function
267
sheathed resistance bulb
90
sheathed thermocouple
78
short-circuit ring method
149
side-stream concentration control
571
silicon radiation thermometer
98
single loop controller
249
single-chip microprocessor
338
single-seated valve
368
sizing
409
slab
635
sliding pressure operation
667
sludge treatment instrumentation
749
Smith controller
494
software package
305
soot blowing
719
SPC
3
specific gravity-volume conversion table
239
spectral radiant emittance
93
speed regulation
664
spray cooling water control
638
square root
361
standard current
86
stand.ard thermocouple
78
startup
392,460,467
startup preparation review
473
state
522
state equation
522
Index
state feedback
521
state transition diagram
272
STC
656
steam temperature control
650,656
steelmaking
623
step response
17,495
step writing method
232
stock preparation-process
719
storage control
755
Strouhal number
126
successive identification method
516
sugar product
693
system configuration
403
system engineering
391
system generation
427
T
tachometer
153
tank-inventory control
616
target response curve
519
task management
317
TCD
160
techonology for high reliability
264
telemeter/telecontrol
727
temperature control by heat exchanger
bypass
544
temperature control of heater by threeway
valve
545
thermal conductivity detector
160
thermistor thermometer
92
thermocouple
72
thermocouple for high temperature use
84
thermocouple with protective tube
78
thermoelectric thermometer
72
thermoelectromotive force
72
three-way valve
369
time chart
271
time delay
489
time slice
318
token card
754
top pressure control
631
top-charging
626
topping unit
606
total FA
252
total FA network
252
total head
723
total production control system at a sugarrefinery
704
total reflux operation for distillation
791
554
traceability
69
trade-off
395,477
transfer function
13
transient response
16
transient response method
50
transit time differential method
131
transmission-scattering method
174
triangular coordinate
586
trim
364
trip
451
tundish
635
turbine bypass control system
685
turbine flowmeter
113
turbine-follow mode
668
tuyere
628
two-wire signal transmission
200
u
ultimate-sensitivity method
ultrasonic flowmeter
130
ultrasonic liquid-level meter
ultrasonic position detector
underground method
462
uninterruptible power supply
UPS
438
utility
393,407
792
w
wash sequence control
732
wastewater treatment instrumentation
747
water content
188,595
water distribution information management system
739
water supply flow control
737
water-run
394
wavelength spectrum
194
wet leg method
142
whiteness
714
y
48,516
146
215
YEWMAC
296
yield
583
437
zero method
Ziegler-Nichols
zirconia method
v
vapor-liquid equilibrium
variable coupling method
vector locus
19
vector transposition
522
venturi tube
103
volumetric flowmeter
110
vortex flowmeter
125
553
150
2-out-of-3
3-wire type
63
516
166
447
87
Index

Instrumentation System Fundamentals

Uploaded by

Copyright:

Available Formats

Instrumentation System Fundamentals

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Instrumentation System Fundamentals

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some of the topics covered in the book?

What are some of the topics covered in the book?

Instrumentation Systems

-Fundamentals and Applications-

Tasuku Senbon, Futoshi Hanabuchi (Eds.)

Instrumentation Systems-Fundamentals and Applications.

All rights reserved. No part of this publication may be

1991 Springer-Verlag Berlin Heidelberg

Hisashi Tamura, Senior Vice President

ABOUT THE ENGLISH-LANGUAGE EDITION

About the English-Language Edition

Chapter 1 INDUSTRY AND INSTRUMENTATION

Chapter 3 DETECTION AND CONVERSION OF

3. 8. 5 Moisture/humidity meters .................................... 172

5 .1. 2 Configuration of a total FA system ......................... 252

6 02 03 Flow characteristics ........................................... 361

Chapter 7 SYSTEM ENGINEERING

7. 4. 1 Overview ........................................................ 460

Chapter 9 CONTROL OF PROCESS UNITS (Application I )

Chapter 10 INSTRUMENTATION TO MANUFACTURING INDUSTRIES

10.3. 7 Pressurized water reactor control system ................. 675

INDUSTRY AND INSTRUMENTATION

The Word "Instrumentation"

The Word 'Instrumentation'

Definition of Industrial Instrument:

1.2 The Development of Instrumentation

Chap.l Industry and Instrumentation

Table 1.1 Development of instrumentation technology in Japan

Recovery period after World War ll

Construction of petroleum industrial

Practical use of electronic control

Construction of blast furnace

Computer control systems

Second oil crisis

Distributed control systems

Low economical growth period

Single loop DDC

lers. This is known as Supervisory Process Control or Setpoint Control

The 1970's to the present

Trend toward Total FA

Chap.l Industry and Instrumentation

FA(Factory automa tion)

Fig. 1 1 "Upstream" and "Downstream" processes

Instrumentation system hierarchy

facturing instructions received and produce products according to the

As part of what is generally termed "instruments," this section

Chap.l Industry and Instrumentation

Figure 1.3 shows instrument classification. This particular example

Measuring equipment (calipers, micrometers, water-meters, analytical

Fig. 1 3 Classification of measuring instruments

Process use industrial instruments

Fig. 1 4 Classification of industrial instruments

1.4 Classification and Use of Instruments

are classified under electronic application equipment in Fig. 1.3. In

Electric measuring instrui~/

Fig. 1 5 Trends in production of measuring instruments

Table 1.2 Production of industrial instruments (1985)

Total industrial instruments