Visit https://ebookultra.

com to download the full version and

explore more ebooks

Process Modelling and Model Analysis Process Systems

Engineering 1st Edition Ian T. Cameron

_____ Click the link below to download _____

https://ebookultra.com/download/process-modelling-and-
model-analysis-process-systems-engineering-1st-edition-
ian-t-cameron/

Explore and download more ebooks at ebookultra.com

Here are some recommended products that might interest you.
You can download now and explore!

Model Driven Business Process Engineering 1st Edition

Kevin Lano

https://ebookultra.com/download/model-driven-business-process-
engineering-1st-edition-kevin-lano/

ebookultra.com

Model Based Control Case Studies in Process Engineering

1st Edition Paul Serban Agachi

https://ebookultra.com/download/model-based-control-case-studies-in-
process-engineering-1st-edition-paul-serban-agachi/

ebookultra.com

Exposing Fraud Skills Process and Practicalities 1st

Edition Ian Ross

https://ebookultra.com/download/exposing-fraud-skills-process-and-
practicalities-1st-edition-ian-ross/

ebookultra.com

Designing complex web information systems integrating

evolutionary process engineering 1st Edition Roberto
Paiano
https://ebookultra.com/download/designing-complex-web-information-
systems-integrating-evolutionary-process-engineering-1st-edition-
roberto-paiano/
ebookultra.com
Food Process Engineering and Technology 1st Edition Zeki
Berk

https://ebookultra.com/download/food-process-engineering-and-
technology-1st-edition-zeki-berk/

ebookultra.com

Stochastic Modelling in Process Technology First Edition

Herold G. Dehling

https://ebookultra.com/download/stochastic-modelling-in-process-
technology-first-edition-herold-g-dehling/

ebookultra.com

Aspects of Modelling Road Vehicles For Model Railways 1st

Ed. Edition Ian Moreton

https://ebookultra.com/download/aspects-of-modelling-road-vehicles-
for-model-railways-1st-ed-edition-ian-moreton/

ebookultra.com

Process Engineering Equipment Handbook 1st Edition Claire

Marguerite Soares

https://ebookultra.com/download/process-engineering-equipment-
handbook-1st-edition-claire-marguerite-soares/

ebookultra.com

The Global Manufacturing Revolution Product Process

Business Integration and Reconfigurable Systems Wiley
Series in Systems Engineering and Management 1st Edition
Yoram Koren
https://ebookultra.com/download/the-global-manufacturing-revolution-
product-process-business-integration-and-reconfigurable-systems-wiley-
series-in-systems-engineering-and-management-1st-edition-yoram-koren/
ebookultra.com
Process Modelling and Model Analysis Process Systems
Engineering 1st Edition Ian T. Cameron Digital Instant
Download
Author(s): Ian T. Cameron, Katalin Hangos
ISBN(s): 9780121569310, 0121569314
Edition: 1
File Details: PDF, 21.75 MB
Year: 2001
Language: english
 PROCESS MODELLING
AND MODEL ANALYSIS
This is Volume 4 of
PROCESS SYSTEMS ENGINEERING
A Series edited by George Stephanopoulos and John Perkins
 PROCESS
MODELLING AND
MODEL ANALYSIS

K. M. Hangos
Systems and Control Laboratory
Computer and Automation Research Institute of the
Hungarian Academy of Sciences
Budapest, Hungary
L T. Cameron
Cape Centre Department of Chemical Engineering
The University of Queensland
Brisbane, Queensland

r^ ACADEMIC PRESS
\,,__^^ A Harcourt Science and Technology Company

San Diego San Francisco New York Boston London Sydney Tokyo
This book is printed on acid-free paper.

Copyright © 2001 by ACADEMIC PRESS

All Rights Reserved.

No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, recording, or any
information storage and retrieval system, without the prior permission in writing from the
publisher.
Academic Press
A Harcourt Science and Technology Company
Harcourt Place, 32 Jamestown Road, London NWl 7BY, UK
http://www.academicpress.com

Academic Press
A Harcourt Science and Technology Company
525 B Street, Suite 1900, San Diego, California 92101-4495, USA
http ://w w w. academ icpress. com

ISBN 0-12-156931-4

Library of Congress Catalog Number: 00-112073

A catalogue record of this book is available from the British Library

Typeset by Newgen Imaging Systems (P) Ltd., Chennai, India

01 02 03 04 05 06 BC 9 8 7 6 5 4 3 2 1

Transferred to digital printing in 2007.

 Dedicated to

Misi, Akos, Veronika

and
Lucille, James, Peter, Andrew
This Page Intentionally Left Blank
CONTENTS

INTRODUCTION xiii

I F U N D A M E N T A L PRINCIPLES A N D PROCESS
MODEL DEVELOPMENT

I The Role of Models in Process

Systems Engineering
1.1. The Idea of a Model 4
1.2. Model Application Areas in PSE 7
1.3. Model Classification 10
1.4. Model Characteristics 12
1.5. A Brief Historical Review of Modelling in PSE 13
1.6. Summary 17
1.7. Review Questions 17
1.8. Application Exercises 17

VII
VIII CONTENTS

2 A Systematic Approach to Model Building

2.1. The Process System and the Modelling Goal 20
2.2. Mathematical Models 22
2.3. A Systematic Modelling Procedure 24
2.4. Ingredients of Process Models 32
2.5. Summary 36
2.6. Review Questions 36
2.7. Application Exercises 37

3 Conservation Principles

3.1. Thermodynamic Principles of Process Systems 42

3.2. Principle of Conservation 51
3.3. Balance Volumes in Process System Applications 58
3.4. Summary 61
3.5. Review Questions 62
3.6. Application Exercises 62

4 Constitutive Relations
4.1. Transfer Rate Equations 65
4.2. Reaction Kinetics 70
4.3. Thermodynamical Relations 72
4.4. Balance Volume Relations 75
4.5. Equipment and Control Relations 75
4.6. Summary 79
4.7. Review Questions 79
4.8. Application Exercises 80

5 Dynamic Models—Lumped Parameter Systems

5.1. Characterizing Models and Model Equation Sets 83
5.2. Lumped Parameter Models—Initial Value
Problems (IVPs) 84
5.3. Conservation Balances for Mass 86
5.4. Conservation Balances for Energy 89
5.5. Conservation Balances for Momentum 95
5.6. The Set of Conservation Balances for Lumped Systems 98
5.7. Conservation Balances in Intensive Variable Form 99
5.8. Dimensionless Variables 101
5.9. Normalization of Balance Equations 102
5.10. Steady-State Lumped Parameter Systems 103
CONTENTS IX

5.11. Analysis of Lumped Parameter Models 104

5.12. Stability of the Mathematical Problem 114
5.13. Summary 117
5.14. Review Questions 118
5.15. Application Exercises 118

6 Solution Strategies for

Lumped Parameter Models
6.1. Process Engineering Example Problems 124
6.2. Ordinary Differential Equations 125
6.3. Basic Concepts in Numerical Methods 126
6.4. Local Truncation Error and Stability 129
6.5. Stability of the Numerical Method 133
6.6. Key Numerical Methods 137
6.7. Differential-Algebraic Equation Solution Techniques 149
6.8. Summary 155
6.9. Review Questions 156
6.10. Application Exercises 156

7 Dynamic Models^—Distributed
Parameter Systems

7.1. Development of DPS Models 163

7.2. Examples of Distributed Parameter Modelling 174
7.3. Classification of DPS Models 182
7.4. Lumped Parameter Models for Representing DPSs 185
7.5. Summary 186
7.6. Review Questions 187
7.7. Application Exercises 187

8 Solution Strategies for Distributed

Parameter Models
8.1. Areas of Interest 191
8.2. Finite Difference Methods 192
8.3. Method of Lines 201
8.4. Method of Weighted Residuals 203
8.5. Orthogonal Collocation 206
8.6. Orthogonal Collocation for Partial Differential Equations 216
8.7. Summary 218
8.8. Review Questions 218
8.9. Application Exercises 219
 CONTENTS

9 Process Model Hierarchies

9.1. Hierarchy Driven by the Level of Detail 225
9.2. Hierarchy Driven by Characteristic Sizes 233
9.3. Hierarchy Driven by Characteristic Times 239
9.4. Summary 245
9.5. Further Reading 246
9.6. Review Questions 246
9.7. Application Exercises 246

II ADVANCED PROCESS MODELLING A N D

MODEL ANALYSIS

10 Basic Tools for Process Model Analysis

10.1. Problem Statements and Solutions 251
10.2. Basic Notions in Systems and Control Theory 253
10.3. Lumped Dynamic Models as Dynamic System Models 264
10.4. State Space Models and Model Linearization 269
10.5. Structural Graphs of Lumped Dynamic Models 277
10.6. Summary 281
10.7. Review Questions 281
10.8. Application Exercises 282

11 Data Acquisition and Analysis

11.1. Sampling of Continuous Time Dynamic Models 286
11.2. Data Screening 289
11.3. Experiment Design for Parameter Estimation of
Static Models 294
11.4. Experiment Design for Parameter Estimation of
Dynamic Models 295
11.5. Summary 296
11.6. Further Reading 296
11.7. Review Questions 296
11.8. Application Exercises 297

12 Statistical Model Calibration and Validation

12.1. Grey-Box Models and Model Calibration 300
12.2. Model Parameter and Structure Estimation 302
12.3. Model Parameter Estimation for Static Models 314
12.4. Identification: Model Parameter and Structure Estimation of
Dynamic Models 318
12.5. CSTR: A Case Study of Model Parameter Estimation 323
12.6. Statistical Model Validation via Parameter Estimation 330
CONTENTS XI

12.7. Summary 331

12.8. Further Reading 331
12.9. Review Questions 331
12.10. Application Exercises 332

13 Analysis of Dynamic Process Models

13.1. Analysis of Basic Dynamical Properties 336
13.2. Analysis of Structural Dynamical Properties 341
13.3. Model Simplification and Reduction 350
13.4. Summary 359
13.5. Further Reading 359
13.6. Review Questions 360
13.7. Application Exercises 361

14 Process Modelling for Control and

Diagnostic Purposes
14.1. Model-Based Process Control 364
14.2. Model-Based Process Diagnosis 370
14.3. Qualitative, Logical and AI Models 372
14.4. Summary 384
14.5. Further Reading 384
14.6. Review Questions 385
14.7. Application Exercises 385

15 Modelling Discrete Event Systems

15.1. Characteristics and Issues 388
15.2. Approaches to Model Representation 388
15.3. Solution of Discrete Event Dynamic System Models 404
15.4. Analysis of Discrete Event Systems 408
15.5. Summary 410
15.6. Further Reading 411
15.7. Review Questions 412
15.8. Application Exercises 412

16 Modelling Hybrid Systems

16.1. Hybrid Systems Basics 415
16.2. Approaches to Model Representation 420
16.3. Analysis of Hybrid Systems 430
16.4. Solution of Hybrid System Models 431
16.5. Summary 434
16.6. Further Reading 434
XII CONTENTS

16.7. Review Questions 435

16.8. Application Exercises 436

17 Modelling Applications in Process Systems

17.1. Copper Converter Dynamics 438
17.2. Destruction of Phenol in Wastewater by
Photochemical Reaction 445
17.3. Prefermenter System for Wastewater Treatment 451
17.4. Granulation Circuit Modelling 456
17.5. Industrial Depropanizer using Structural Packing 462
17.6. Summary 469

18 Computer Aided Process Modelling

18.1. Introduction 472
18.2. Industrial Demands on Computer Aided Modelling Tools 472
18.3. Basic Issues in CAPM Tools 474
18.4. Approaches to CAPM Tool Development 483
18.5. Summary 492

19 Empirical Model Building

19.1. Introduction 493
19.2. The Modelling Procedure Revisited 494
19.3. Black-Box Modelling 497
19.4. Traps and Pitfalls in Empirical Model Building 511
19.5. Summary 515
19.6. Further Reading 515
19.7. Review Questions 516
19.8. Application Exercises 516

Appendix: Basic Mathematic Tools 517

A. 1. Random Variables and Their Properties 517
A.2. Hypothesis Testing 521
A.3. Vector and Signal Norms 522
A.4. Matrix and Operator Norms 523
A.5. Graphs 524

BIBLIOGRAPHY 527

INDEX 535
INTRODUCTION

Process modelling is one of the key activities in process systems engineering. Its
importance is reflected in various ways. It is a significant activity in most major com-
panies around the world, driven by such application areas as process optimization,
design and control. It is a vital part of risk management, particularly consequence
analysis of hazardous events such as loss of containment of process fluids. It is a
permanent subject of conferences and symposia in fields related to process systems
engineering. It is often the topic of various specialized courses offered at graduate,
postgraduate and continuing professional education levels. There are various text-
books available for courses in process modelling and model solution amongst which
are Himmelblau [1], Davis [2], Riggs [3] and Rice and Do [4]. These however are
mainly devoted to the solution techniques related to process models and not to the
problem on how to define, setup, analyse and test models. Several short monographs
or mathematical notes with deeper insights on modelling are available, most notably
by Aris [5] and Denn [6].
In most books on this subject there is a lack of a consistent modelling approach
applicable to process systems engineering as well as a recognition that modelling is
not just about producing a set of equations. There is far more to process modelling
than writing equations. This is the reason why we decided to write the current book in
order to give a more comprehensive treatment of process modelling useful to student,
researcher and industrial practitioner.
There is another important aspect which limits the scope of the present material
in the area of process modelling. It originates from the well-known fact that a par-
ticular process model depends not only on the process to be described but also on the
modelling goal. It involves the intended use of the model and the user of that model.
Moreover, the actual form of the model is also determined by the education, skills
and taste of the modeller and that of the user. Due to the above reasons, the main
emphasis has been on process models for dynamic simulation and process control
purposes. These are principally lumped dynamic process models in the form of sets
of differential—algebraic equations. Other approaches such as distributed parameter
modelling and the description of discrete event and hybrid systems are also treated.
Finally the use of empirical modelling is also covered, recognizing that our knowledge

XIII
XIV INTRODUCTION

of many systems is extremely shallow and that input-output descriptions generated

by analysing plant data are needed to complement a mechanistic approach.
Process modelling is an engineering activity with a relatively mature technol-
ogy. The basic principles in model building are based on other disciplines in process
engineering such as mathematics, chemistry and physics. Therefore, a good back-
ground in these areas is essential for a modeller. Thermodynamics, unit operations,
reaction kinetics, catalysis, process flowsheeting and process control are the helpful
prerequisites for a course in process modelling. A mathematical background is also
helpful for the understanding and application of analysis and numerical methods in
the area of linear algebra, algebraic and differential equations.

Structure of the book

The book consists of two parts. The first part is devoted to the building of process
models whilst the second part is directed towards analysing models from the view-
point of their intended use. The methodology is presented in a top-down systematic
way following the steps of a modelling procedure. This often starts from the most
general case. Emphasis is given in this book to identifying the key ingredients, devel-
oping conservation and constitutive equations then analysing and solving the resultant
model. These concepts are introduced and discussed in separate chapters. Static and
dynamic process models and their solution methods are treated in an integrated man-
ner and then followed by a discussion on hierarchical process models which are related
by scales of time or degree of detail. This is the field of multi-scale modelling.
The second part of the book is devoted to the problem of how to analyse process
models for a given modelling goal. Three dominant application areas are discussed:
• control and diagnosis where mostly lumped dynamic process models are used,
• static flowsheeting with lumped static process models,
• dynamic flowsheeting where again mostly lumped dynamic process models
are in use.
Special emphasis is given to the different but related process models and their
properties which are important for the above application areas.
Various supplementary material is available in the appendices. This includes:
• Background material from mathematics covering linear algebra and math-
ematical statistics.
• Computer science concepts such as graphs and algorithms.
Each chapter has sections on review questions and application examples which
help reinforce the content of each chapter. Many of the application exercises are
suitable for group work by students.
The methods and procedures presented are illustrated by examples throughout
the book augmented with MATLAB subroutines where appropriate. The examples
are drawn from as wide a range of process engineering disciplines as possible. They
include chemical processing, minerals process engineering, environmental engineer-
ing and food engineering in order to give a true process system's appeal. The model
analysis methods in part two are applied to many of the same process systems used
in part one. This method of presentation makes the book easy to use for both higher
year undergraduate, postgraduate courses or for self study.
INTRODUCTION XV

Making use of the book

This book is intended for a wide audience. The authors are convinced that it will be
useful from the undergraduate to professional engineering level. The content of the
book has been presented to groups at all levels with adaptation of the material for
the particular audience. Because of the modular nature of the book, it is possible to
concentrate on a number of chapters depending on the need of the reader.
For the undergraduate, we suggest that a sensible approach will be to consider
Chapters 1-6, parts of 10-12 as a full 14-week semester course on basic process
modelling. For advanced modelling at undergraduate level and also for postgraduate
level. Chapters 7-9 can be considered. Chapter 17 on Modelling Applications should
be viewed by all readers to see how the principles work out in practice.
Some industrial professionals with a particular interest in certain application areas
could review modelling principles in the first half of the book before considering the
specific application and analysis areas covered by Chapters 11-19. The options are
illustrated in the following table:

Chapter Undergraduate Postgraduate Professional Professional

introduction advanced introduction advanced

1. Role of Modelling in Process / /

Systems Engineering
2. A Modellling Methodology / / / /
3. Conservation Principles in / /
Process Modelling
4. Constitutive Relations for / /
Modelling
5. Lumped Parameter Model / /
Development
6. Solution of Lumped Parameter
Models
>r /
7. Distributed Parameter Model / /
Development
8. Solution of Distributed / /
Parameter Models
9. Incremental Modelling and / /
Model Hierarchies
10. Basic Tools for Model / / / /
Analysis
11. Data Acquisition and Analysis / ^ / /
12. Statistical Model Calibration / / / /
and Validation
13. Analysis of Dynamic Process / /
Models
14. Process Modelling for Control / /
and Diagnosis
15. Modelling of Discrete Event / /
Systems
16. Modelling of Hybrid Systems / /
17. Modelling Applications / / / /
18. Computer Aided Process / /
Modelling
19. Empirical Model Building / /
XVI INTRODUCTION

For instructors there is also access to PowerPoint presentations on all chapters

through the website: h t t p : / / d a i s y . c h e q u e . u q . e d u . a u / c a p e / m o d e l l i n g /
index.html.

Acknowledgements

The authors are indebted to many people who contributed to the present book in
various ways. The lively atmosphere of the Department of Chemical Engineering of
The University of Queensland and the possibility for both of us to teach thefinalyear
"Process Modelling and Solutions" course to several classes of chemical engineering
students has made the writing of the book possible. Presenting sections to professional
engineers within Australian industry and also to Ph.D. and academic staff in Europe
through the Eurecha organization, has helped refine some of the content.
We would especially thank Christine Smith for the care and help in prepar-
ing different versions of the teaching materials and the manuscript. Also to Russell
Williams and Steven McGahey for help in reviewing some of the chapters and to
Gabor Szederkenyi for his kind help with many of the LaTeX and figure issues.
We are conscious of the support of our colleagues working in the field of mod-
elling in process systems engineering. These include Prof. John Perkins at Imperial
College, Prof. George Stephanopoulos at MIT, Professors Rafique Gani and Sten Bay
Jorgensen at the Danish Technical University, Prof. Heinz Preisig at Eindhoven and
Prof. Wolfgang Marquardt at Aachen.
Special thanks to Dr. Bob Newell for many years of fruitful discussion and
encouragement towards realism in modelling. We thank them all for the advice,
discussions and sources of material they have provided.
We readily acknowledge the contribution of many whose ideas and pub-
lished works are evident in this book. Any omissions and mistakes are solely the
responsibility of the authors.
PARTI
FUNDAMENTAL PRINCIPLES
AND PROCESS MODEL
DEVELOPMENT
This Page Intentionally Left Blank
I THE ROLE OF MODELS IN
PROCESS SYSTEMS
ENGINEERING

Models?—really nothing more than an imitation of reality! We can relate to models of

various types everywhere—some physical, others mathematical. They abound in all
areas of human activity, be it economics, warfare, leisure, environment, cosmology or
engineering. Why such an interest in this activity of model building and model use?
It is clearly a means of gaining insight into the behaviour of systems, probing them,
controlling them, optimizing them. One thing is certain: this is not a new activity,
and some famous seventeenth-century prose makes it clear that some modelling was
considered rather ambitious:
From man or angel the great Architect
Did wisely to conceal, and not divulge
His secrets to be scanned by them who ought
Rather admire; or if they list to try
Conjecture, he his fabric of the heavens
Hath left to their disputes, perhaps to move
His laughter at their quaint opinions wide
Hereafter, when they come to model heaven
And calculate the stars, how they will wield
The mighty frame, how build, unbuild, contrive
To save appearances, how gird the sphere
With centric and eccentric scribbled o 'er.
Cycle and epicycle, orb in orb,

[John Milton (1608-74), English poet. Paradise Lost, Book. 8: 72-84]

3
 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

Our task here is a little more modest, but nevertheless important for the field of
process engineering. In this chapter, we want to explore the breadth of model use in the
process industries—where models are used and how they are used. The list is almost
unlimited and the attempted modelling is driven largely by the availability of high
performance computing and the demands of an increasingly competitive marketplace.
However, we must admit that in many cases in process engineering much is
concealed by our limited understanding of the systems we seek to design or manage.
We can relate to the words of Milton that our efforts and opinions are sometimes
quaint and laughable and our conjectures short of the mark, but the effort may well
be worthwhile in terms of increased understanding and better management of the
systems we deal with.
The emphasis in this book is on mathematical modelling rather than physical
modelling, although the latter has an important place in process systems engineering
(PSE) through small scale pilot plants to three dimensional (3D) construction models.

I . I . THE IDEA OF A MODEL

A model is an imitation of reality and a mathematical model is a particular form of

representation. We should never forget this and get so distracted by the model that
we forget the real application which is driving the modelling. In the process of model
building we are translating our real world problem into an equivalent mathematical
problem which we solve and then attempt to interpret. We do this to gain insight
into the original real world situation or to use the model for control, optimization or
possibly safety studies.
In discussing the idea of a model, Aris [5] considers the well-known concept of
change of scale as being at the root of the word "model". Clearly, we can appreciate
this idea from the wealth of scale models, be it of process plant, toys or miniature
articles of real world items. However in the process engineering area the models we
deal with are fundamentally mathematical in nature. They attempt to capture, in the
form of equations, certain characteristics of a system for a specific use of that model.
Hence, the concept of purpose is very much a key issue in model building.
The modelling enterprise links together a purpose V with a subject or physical
system <S and the system of equations M which represent the model. A series of
experiments £ can be applied to M in order to answer questions about the system S.
Clearly, in building a model, we require that certain characteristics of the actual
system be represented by the model. Those characteristics could include:
• the correct response direction of the outputs as the inputs change;
• valid structure which correctly represents the connection between the inputs,
outputs and internal variables;
• the correct short and/or long term behaviour of the model.
In some cases, certain characteristics which are unnecessary to the use of the
model are also included. The resultant model has a specific region of applicability,
depending on the experiments used to test the model behaviour against reality. These
issues are more fully investigated in the following chapters.
THE IDEA OF A MODEL

Real world 1 Mathematical 2 Mathematical 3

problem problem solution
A

FIGURE I . I Real world modelling process.

The following definition derives from Minsky, who stated that:

DEFINITION 1.1.1. A model (M) for a system (S) and an experiment (£) is
anything to which £ can be applied in order to answer questions about S.
For those who wish to investigate the fuller implications of the etymology of the
word "model", we refer to informative discussion given by Aris [5] and Minsky [7].
This overall process is represented schematically in Fig. 1.1, which shows the
four key steps in the overall modelling process.

Each of the steps in Fig. 1.1 has very important issues attached to them, which
are covered in the subsequent chapters. At this point, it is worth mentioning a few of
the issues to be raised at each of the steps seen in Fig. 1.1. These are:

I . I . I . Reality to Mathematics (Step I)

Here we have to deal with the task of translating the real problem to one represented
in mathematical terms. Some of the key issues which have to be dealt with here are:
• What do we understand about the real world problem?
• What is the intended use of the mathematical model?
• What governing phenomena or mechanisms are there in the system?
• What form of model is required?
• How should the model be structured and documented?
• How accurate does the model have to be?
• What data on the system are available and what is the quality and accuracy of
the data?
• What are the system inputs, states, outputs and disturbances?

1. 1.2. Mathematical Solution (Step 2)

Having generated some mathematical description of the real world system, it is then
necessary to solve this for the unknown value of the variables representing that system.
Key issues here are:
• What variables must be chosen in the model to satisfy the degrees of freedom?
• Is the model solvable?
• What numerical (or analytic) solution technique should be used?
6 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

• Can the structure of the problem be exploited to improve the solution speed or
robustness?
• What form of representation should be used to display the results (2D graphs,
3D visualization)?
• How sensitive will the solution output be to variations in the system parameters
or inputs?

I.L3. Interpreting the Model Outputs (Step 3)

Here we need to have procedures and tests to check whether our model has been
correctly implemented and then ask whether it imitates the real world to a sufficient
accuracy to do the intended job. Key issues include:
• How is the model implementation to be verified?
• What type of model validation is appropriate and feasible for the problem?
• Is the resultant model identifiable?
• What needs to be changed, added or deleted in the model as a result of the
validation?
• What level of simplification is justified?
• What data quality and quantity is necessary for validation and parameter
estimates?
• What level of model validation is necessary? Should it be static or dynamic?
• What level of accuracy is appropriate?
• What system parameters, inputs or disturbances, need to be known accurately
to ensure model predictive quality?

LI.4. Using the Results in the Real World (Step 4)

Here we are faced with the implementation of the model or its results back into the
real world problem we originally addressed. Some issues that arise are:
• For online applications where speed might be essential, do I need to reduce
the model complexity?
• How can model updating be done and what data are needed to do it?
• Who will actually use the results and in what form should they appear?
• How is the model to be maintained?
• What level of documentation is necessary?
These issues are just some of the many which arise as models are conceptualized,
developed, solved, tested and implemented. What is clear from the above discussion is
the fact that modelling isfar more than simply the generation ofa set of equations. This
book emphasizes the need for a much broader view of process modelling including
the need for a model specification, a clear generation and statement of hypotheses
and assumptions, equation generation, subsequent model calibration, validation and
end use.
Many of the following chapters will deal directly with these issues. However,
before turning to those chapters we should consider two more introductory aspects to
help "set the scene". These are to do with model characterization and classification.
The important point about stating these upfront is to be aware of these issues early
MODEL APPLICATION AREAS IN PSE

TABLE I. I Model Application Areas

Application area Model use and aim

Process design Feasibility analysis of novel designs

Technical, economic, environmental assessment
Effects of process parameter changes on performance
Optimization using structural and parametric changes
Analysing process interactions
Waste minimization in design
Process control Examining regulatory control strategies
Analysing dynamics for setpoint changes or disturbances
Optimal control strategies for batch operations
Optimal control for multi-product operations
Optimal startup and shutdown policies
Trouble-shooting Identifying likely causes for quality problems
Identifying likely causes for process deviations
Process safety Detection of hazardous operating regimes
Estimation of accidental release events
Estimation of effects from release scenarios (fire etc.)
Operator training Startup and shutdown for normal operations
Emergency response training
Routine operations training
Environmental impact Quantifying emission rates for a specific design
Dispersion predictions for air and water releases
Characterizing social and economic impact
Estimating acute accident effects (fire, explosion)

on and thus retain them as guiding concepts for what follows. Before we deal with
these issues, we survey briefly where models are principally used in PSE and what is
gained from their use.

1.2. MODEL APPLICATION AREAS IN PSE

It was mentioned that the list of applications in process engineering is almost endless.
We can, however, categorize the use of models into several well-defined areas. These
are outlined in Table 1.1, which sets out the typical application area and the aim of
the modelling.
Clearly, the list can be extended in each application area and the individual
categories can be extended too. It does show the wide range of applications and
hence the importance of process modelling on the modern design, optimization and
operation of process systems. What is sometimes not obvious is that each of the
application areas may require quite different models to achieve the desired outcome.
In the next section, we consider the issue of model classification and show the great
diversity of model types which have been used.
However, we first give some examples of models and their application from
several process systems areas.
EXAMPLE 1.2.1 (Fire radiation modelling). Safe operations are vital in the
process industries and for land-use planning purposes. Liquefied petroleum gas can
 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

FIGURE 1.2 BLEVE fireball caused by the rupture of an LPG tank (by permission of A.M. BIrk,
Queens University, Canada).

200 250 300 350 500

distance(metres)

FIGURE 1.3 Radiation levels for a BLEVE incident.

be dangerous if released and ignited. One type of event which has occurred in several
places around the world is the boiling liquid expanding vapour explosion (BLEVE).
Figure 1.2 shows the form of a BLEVE fireball caused by the rupture of an LPG
tank. Of importance are radiation levels at key distances from the event as well as
projectiles from the rupture of the vessel.
Predictive mathematical models can be used to estimate the level of radiation at
nominated distances from the BLEVE, thus providing input to planning decisions.
Figure 1.3 shows predicted radiation levels (kW/m^) for a 50-tonne BLEVE out to a
distance of 500 m.

EXAMPLE 1.2.2 (Compressor dynamics and surge control). Compressors are

subject to surge conditions when inappropriately controlled. When surge occurs, it
MODEL APPLICATION AREAS IN PSE

F I G U R E 1.4 Multi-stage centrifugal compressor (by permission of Mannesmann Demag Delaval).

12
Surge line
10 I

Shutdown
0
0 5 10 15 20 25 30 35
flow (kg/s)
F I G U R E 1.5 Head-flow dynamics during compressor shutdown.

can lead to serious damage or destruction of the equipment. Effective control systems
are necessary to handle load changes. To test alternative control designs, accurate
modelling and simulation are useful approaches. Figure 1.4 shows a large multi-stage
compressor and Fig. 1.5 shows the predicted behaviour of the first stage head-flow
dynamics under controlled shutdown.
10 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

1.3. MODEL CLASSIFICATION

We can devise several ways of classifying models. Each leads to a variety of model
characteristics which have an impact on the solution techniques as well as the potential
application areas where they can be used. Some model types are inappropriate in
certain circumstances, such as a steady-state model for batch reactor start-up analysis.
Table 1.2 gives an overview of model types, their basic characteristics and the final
form of the models.
The next section explains some key aspects of these model types and later chapters
are concerned with the development of several model types.

H H TABLE 1.2 Model Classification

Type of model Criterion of classification

Mechanistic Based on mechanisms/underlying phenomena

Empirical Based on input-output data, trials or experiments
Stochastic Contains model elements that are probabilistic in nature
Deterministic Based on cause-effect analysis
Lumped parameter Dependent variables not a function of spatial position
Distributed parameter Dependent variables are a function of spatial position
Linear Superposition principle applies
Nonlinear Superposition principle does not apply
Continuous Dependent variables defined over continuous space-time
Discrete Only defined for discrete values of time and/or space
Hybrid Containing continuous and discrete behaviour

1.3.1. Characteristic Nature of Process Models

Mechanistic models are also referred to diS phenomenological models because of their
basic derivation from system phenomena or mechanisms such as mass, heat and
momentum transfer. Many commonplace models in process engineering applications
are derived from a knowledge of the underlying mechanisms. However, most mech-
anistic models also contain empirical parts such as rate expressions or heat transfer
relations. Mechanistic models often appear in design and optimization applications.
They can be termed "white box" models since the mechanisms are evident in the
model description.
Empirical models are the result of experiment and observation, usually not relying
on the knowledge of the basic principles and mechanisms which are present in the
system being studied. They employ essentially equation fitting where the parameters
have little or no physical meaning. Empirical models are widely used where the actual
underlying phenomena are not known or understood well. These models are often
termed "black box" models, reflecting the fact that little is known about the real
mechanisms of the process.
The most common form of model used in process engineering is a combination
of mechanistic and empirical parts and hence is termed "grey box".
MODEL CLASSIFICATION I I

• • • EXAMPLE 1.3.1 (Empirical BLEVE Model). In Example 1.2.1 the radiation

levels from a BLEVE were illustrated. The size and duration of a BLEVE fireball
have been estimated from the analysis of many incidents, most notably the major
disaster in Mexico City during 1984. The empirical model is given by TNO [8] as
r = 3.24m^•^^^ t = 0.852m^'^^,
where r is fireball radius (m), m the mass of fuel (kg) and t the duration of fireball (s).
• • •
Stochastic models arise when the description may contain elements which have natural
random variations typically described by probability distributions. This characteristic
is often associated with phenomena which are not describable in terms of cause and
effect but rather by probabilities or likelihoods.
Deterministic models are the final type of models characterized by clear cause-effect
relationships.
In most cases in process engineering the resultant model has elements from
several of these model classes. Thus we can have a mechanistic model with
some stochastic parts to it. A very common occurrence is a mechanistic model
which includes empirical aspects such as reaction rate expressions or heat transfer
relationships.
• • • EXAMPLE 1.3.2 (Mechanistic compressor model). Example 1.2.2 showed the
prediction of a compressor under rapidly controlled shutdown. The model used here
was derived from fundamental mass, energy and momentum balances over the com-
pressor plenum. By assuming ID axial flow the model was reduced to a set of ordinary
differential equations given by:

Mass: —— = m\ — m2\
at
dE
Energy: -— = m\h\ — m2h2\
at
dM
Momentum: — = AmiPti -Ptj)-^ ^net

where m\,m2 are inlet and outlet mass flows, h\,h2 the specific enthalpies, Am is the
inlet mean cross section, P/. are the inlet and outlet pressures and Fnet is net force on
• • • lumped gas volume.
Table 1.2 also includes other classifications dependent on assumptions about
spatial variations, the mathematical form and the nature of the underlying process
being modelled.

1.3.2. Equation form of Process Models

We can also consider the types of equations which result from such models when
we consider steady state and dynamic situations. These are shown in Table 1.3.
The forms can involve linear algebraic equations (LAEs), nonlinear algebraic equa-
tions (NLAEs), ordinary differential equations (ODEs), elliptic partial differential
equations (EPDEs) and parabolic partial differential equations (PPDEs).
Each of the equation forms requires special techniques for solution. This will be
covered in subsequent chapters.
12 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

TABLE 1.3 Model Equation Forms

Type of model Equation types

Steady-state problem Dynamic problem
Deterministic Nonlinear algebraic ODEs/PDEs
Stochastic Algebraic/difference equations Stochastic ODEs or difference equations
Lumped parameter Algebraic equations ODEs
Distributed parameter EPDEs PPDEs
Linear Linear algebraic equations Linear ODEs
Nonlinear Nonlinear algebraic equations Nonlinear ODEs
Continuous Algebraic equations ODEs
Discrete Difference equations Difference equations

1.3.3. Characteristics of the System Volumes

When we develop models, it is necessary to define regions in the system where we
apply conservation principles and basic physical and chemical laws in order to derive
the mathematical description. These are the balance volumes. A basic classification
relates to the nature of the material in those volumes. Where there are both tem-
poral and spatial variations in the properties of interest, such as concentration or
temperature, we call these systems "distributed". However, when there are no spa-
tial variations and the material is homogeneous, we have a "lumped" system. The
complexity of distributed parameter systems can be significant both in terms of the
resulting model description and the required solution techniques. Lumped parameter
models generally lead to simpler equation systems which are easier to solve.

1.3.4. Characteristics of the System Behaviour

When we consider system modelling, there are many situations where discrete events
occur, such as turning on a pump or shutting a valve. These lead to discontinuous
behaviour in the system either at a known time or at a particular level of one of the
states such as temperature or concentration. These are called "time" or "state" events.
A model which has both characteristics is termed a hybrid system. These are very
common in process systems modelling.
Not only do we need to consider the classification of the models that are used in
PSE applications but it is also helpful to look at some characteristics of those models.

1.4. MODEL CHARACTERISTICS

Here we consider some of the key characteristics which might affect our modelling
and analysis.

• Models can be developed in hierarchies, where we can have several models for
different tasks or models with varying complexity in terms of their structure
and application area.
A BRIEF HISTORICAL REVIEW OF MODELLING IN PSE 13

• Models exist with relative precision, which affect how and where we can use
them.
• Models cause us to think about our system and force us to consider the key
issues.
• Models can help direct further experiments and in-depth investigations.
• Models are developed at a cost in terms of money and effort. These need to be
considered in any application.
• Models are always imperfect. It was once said by George E. Box, a well-known
statistician, "All models are wrong, some are useful"!
• Models invariably require parameter estimation of constants within the model
such as kinetic rate constants, heat transfer and mass transfer coefficients.
• Models can often be transferred from one discipline to another.
• Models should display the principle of parsimony, displaying the simplest
form to achieve the desired modelling goal.
• Models should be identifiable in terms of their internal parameters.
• Models may often need simplification, or model order reduction to become
useful tools.
• Models may be difficult or impossible to adequately validate.
• Models can become intractable in terms of their numerical solution.

We can keep some of these in mind when we come to develop models of our
own for a particular application. It is clearly not a trivial issue in some cases. In other
situations the model development can be straightforward.

1.5. A BRIEF HISTORICAL REVIEW OF MODELLING IN PSE

As a distinct discipline, PSE is a child of the broader field of systems engineering as

applied to processing operations. As such, its appearance as a recognized discipline
dates back to the middle of the twentieth century. In this section, we trace briefly the
history of model building, analysis and model use in the field of PSE.

1.5.1. The Industrial Revolution

It was the industrial revolution which gave the impetus to systematic approaches for
the analysis of processing and manufacturing operations. Those processes were no
longer simple tasks but became increasingly complex in nature as the demand for
commodity products increased. In particular, the early chemical developments of
the late eighteenth century spurred on by the Franco-British wars led to industrial
scale processes for the manufacture of gun powder, sulphuric acid, alkali as well
as food products such as sugar from sugar beet. In these developments the French
and the British competed in the development of new production processes, aided by
the introduction of steam power in the early 1800s which greatly increased potential
production capacity
In dealing with these new processes, it was necessary for the engineer to bring
to bear on the problem techniques derived from many of the physical sciences and
engineering disciplines. These analysis techniques quickly recognized the complex
interacting behaviour of many activities. These ranged from manufacturing processes
I4 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

to communication systems. The complexities varied enormously but the approaches

took on a "systems" view of the problem which gave due regard to the components
in the process, the inputs, outputs of the system and the complex interactions which
could occur due to the connected nature of the process.
Sporadic examples of the use of systems engineering as a sub-discipline of indus-
trial engineering in the nineteenth and twentieth centuries found application in many
of the industrial processes developed in both Europe and the United States. This also
coincided with the emergence of chemical engineering as a distinct discipline at the
end of the nineteenth century and the development of the unit operations concept
which would dominate the view of chemical engineers for most of the twentieth cen-
tury. There was a growing realization that significant benefits would be gained in
the overall economics and performance of processes when a systems approach was
adopted. This covered the design, control and operation of the process.
In order to achieve this goal, there was a growing trend to reduce complex
behaviour to simple mathematical forms for easier process design—hence the use of
mathematical models. The early handbooks of chemical engineering, e.g. Davis [9],
were dominated by the equipment aspects with simple models for steam, fluid flow
and mechanical behaviour of equipment. They were mainly descriptive in content,
emphasizing the role of the chemical engineer, as expressed by Davis, as one who
ensured:
... Completeness of reactions, fewness of repairs and economy of hand labour
should be the creed of the Chemical Engineer.
Little existed in the area of process modelling aimed at reactor and separation
systems. In the period from 1900 to the mid-1920s there was a fast growing body
of literature on more detailed analysis of unit operations, which saw an increased
reliance on mathematical modelling. Heat exchange, drying, evaporation, centrifuga-
tion, solids processing and separation technologies such as distillation were subject
to the application of mass and energy balances for model development. Many papers
appeared in such English journals as Industrial & Engineering Chemistry, Chemistry
and Industry, and the Society of the Chemical Industry. Similar developments were
taking place in foreign language journals, notably those in France and Germany. Text-
books such as those by Walker and co-workers [10] at the Massachusetts Institute of
Technology, Olsen [11] and many monographs became increasingly analytic in their
content, this also being reflected in the education system.

1.5.2. The Mid-twentieth Century

After the end of the Second World War there was a growing interest in the application
of systems engineering approaches to industrial processes, especially in the chemical
industry. The mid-1950s saw many developments in the application of mathematical
modelling to process engineering unit operations, especially for the understanding
and prediction of the behaviour of individual units. This was especially true in the
area of chemical reactor analysis. Many prominent engineers, mathematicians and
scientists were involved. It was a period of applying rigorous mathematical analysis
to process systems which up until that time had not been analysed in such detail.
A BRIEF HISTORICAL REVIEW OF MODELLING IN PSE 15

However, the efforts were mainly restricted to specific unit operations and failed to
address the process as a "system".
This interest in mathematical analysis coincided with the early development and
growing availability of computers. This has been a major driving force in modelling
ever since. Some individuals, however, were more concerned with the overall process
rather than the details of individual unit operations.
One of the earliest monographs on PSE appeared in 1961 as a result of work
within the Monsanto Chemical Company in the USA. This was authored by T.J.
Williams [12], who wrote:
... systems engineering has a significant contribution to make to the practice and
development of chemical engineering. The crossing of barriers between chemical
engineering and other engineering disciplines and the use of advanced mathematics
to study fundamental process mechanisms cannot help but be fruitful.
He continued,
... the use of computers and the development of mathematical process simulation
techniques may result in completely new methods and approaches which will justify
themselves by economic and technological improvements.
It is interesting to note that Williams' application of systems engineering covered
all activities from process development through plant design to control and operations.
Much of the work at Monsanto centred on the use of advanced control techniques aided
by the development of computers capable of performing online control. Computer-
developed mathematical models were proposed as a basis for producing statistical
models generated from their more rigorous counterparts. The statistical models, which
were regression models, could then be used within a control scheme at relatively
low computational cost. It is evident that significant dependence was placed on the
development and use of mathematical models for the process units of the plant.
In concluding his remarks, Williams attempted to assess the future role and impact
of systems engineering in the process industries. He saw the possibility of some 150
large-scale computers being used in the chemical process industries within the USA
for repeated plant optimization studies, these computers being directly connected to
the plant operation by the end of the 1960s. He wrote:
... the next 10 years, then, may see most of today's problems in these fields
conquered.
He did, however, see some dangers not the least being
... the need for sympathetic persons in management and plant operations who
know and appreciate the power of the methods and devices involved, and who will
demand their use for study of their own particular plants.

1.5.3. The Modern Era

Clearly, the vision of T.J. Williams was not met within the 1960s but tremendous
strides were made in the area of process modelling and simulation. The seminal
I6 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

work on transport phenomena by Bird et al. [13] in 1960 gave further impetus to the
mathematical modelling of process systems through the use of fundamental principles
of conservation of mass, energy and momentum. It has remained the pre-eminent book
on this subject for over 40 years.
The same period saw the emergence of numerous digital computer simulators
for both steady state and dynamic simulation. These were both industrially and aca-
demically developed tools. Many of the systems were forerunners of the current
class of steady state and dynamic simulation packages. They were systems which
incorporated packaged models for many unit operations with little ability for the
user to model specific process operations not covered by the simulation system. The
numerical routines were crude by today's standards but simply reflected the stage of
development reached by numerical mathematics of the time. Efficiency and robustness
of solution were often poor and diagnostics as to what had happened were virtually
non-existent. Some things have not changed!
The development of mini-computers in the 1970s and the emergence of UNIX-
based computers followed by the personal computer (PC) in the early 1980s gave a
boost to the development of modelling and simulation tools. It became a reality that
every engineer could have a simulation tool on the desk which could address a wide
range of steady state and dynamic simulation problems. This development was also
reflected in the process industries where equipment vendors were beginning to supply
sophisticated distributed control systems (DCSs) based on mini- and microcomputers.
These often incorporated simulation systems based on simple block representations of
the process or in some cases incorporated real time higher level computing languages
such as FORTRAN or BASIC. The systems were capable of incorporating large scale
real time optimization and supervisory functions.
In this sense, the vision of T.J. Williams some 40 years ago is a reality in cer-
tain sectors of the process industries. Accompanying the development of the process
simulators was an attempt to provide computer aided modelling frameworks for the
generation of process models based on the application of fundamental conservation
principles related to mass, energy and momentum. These have been almost exclu-
sively in the academic domain with a slowly growing penetration into the industrial
arena. Systems such as ASCEND, Model.la, gPROMS or Modelica are among these
developments.
What continues to be of concern is the lack of comprehensive and reliable tools for
process modelling and the almost exclusive slant towards the petrochemical industries
of most commercial simulation systems. The effective and efficient development of
mathematical models for new and non-traditional processes still remains the biggest
hurdle to the exploitation of those models in PSE. The challenges voiced by T.J.
Williams in 1961 are still with us. This is especially the case in the non-petrochemical
sector such as minerals processing, food, agricultural products, pharmaceuticals,
wastewater and the integrated process and manufacturing industries where large scale
discrete-continuous operations are providing the current challenge.
Williams' final words in his 1961 volume are worth repeating,
... there are bright prospects ahead in the chemical process industries for systems
engineering.
One could add, . . . for the process industries in general and for modelling in
particular!
APPLICATION EXERCISES 17

1.6. SUMMARY

Mathematical models play a vital role in PSE. Nearly every area of application is
undergirded by some form of mathematical representation of the system behaviour.
The form and veracity of such models is a key issue in their use. Over the last 50
years, there has been a widespread use of models for predicting both steady state
and dynamic behaviour of processes. This was principally in the chemical process
industries but, more recently, these techniques have been applied into other areas
such as minerals, pharmaceuticals and bio-products. In most cases these have been
through the application of process simulators incorporating embedded models.
There is a maturity evident in traditional steady state simulators, less so in large
scale dynamic simulators and little of real value for large scale discrete-continuous
simulation. Behind each of these areas there is the need for effective model develop-
ment and documentation of the basis for the models which are developed. Systematic
approaches are essential if reliable model use is to be demanded.
The following chapters address a systematic approach to the mathematical devel-
opment of process models and the analysis of those models. The idea that one model
serves all purposes is fallacious. Models must be developed for a specific purpose,
and that purpose will direct the modelling task. It can be in any of the areas of model
application discussed in Section 1.2. This point is emphasized in the following chap-
ter, which develops a systematic approach to model building. The rest of the book
provides detailed information on underlying principles on which the models are built
and analysed.

1.7. REVIEW QUESTIONS

Ql.l. What are the major steps in building a model of a process system? (Section 1.1)
Q1.2. What key issues might arise for each of the overall modelling steps?
(Section 1.1)
Q1.3. What major areas of PSE often rely on the use of models? What are some of
the outcomes in the use of those models? (Section 1.2)
Q1.4. How can models be classified into generic types? Are these categories mutually
exclusive? If not, then explain why. (Section 1.3.1)
Q1.5. Explain the fundamental differences between stochastic, empirical and mech-
anistic models. What are some of the factors which make it easier or harder to develop
such models? (Section 1.3.1)
Q1.6. What are some of the advantages and disadvantages in developing and using
empirical versus mechanistic models for process applications? (Section 1.3.1)

1.8. APPLICATION EXERCISES

Al.l. Consider the model application areas mentioned in Section 1.2. Give some
specific examples of models being used in those applications? What were the benefits
I8 CHAPTER I THE ROLE OF MODELS IN PROCESS SYSTEMS ENGINEERING

derived from developing and using models? Was there any clear methodology used
to develop the model for its intended purpose?
A1.2. What could be the impediments to effective model building in process systems
applications? Discuss these and their significance and possible ways to overcome the
impediments.
A1.3. Consider a particular industry sector such as food, minerals, chemicals or
pharmaceuticals and review where and how models are used in those industry sectors.
What forms of models are typically used? Is there any indication of the effort expended
in developing these models against the potential benefits to be derived from their use?
A1.4. Consider the basic principles of mass, heat and momentum transfer and the
types of models which arise in these areas. What are the key characteristics of such
models describing, for example, heat radiation or heat conduction? How do you
classify them in terms of the classes mentioned in Table 1.3.
A SYSTEMATIC APPROACH TO
MODEL BUILDING

In the past, process modelling was often regarded as "more art than science" or
"more art than engineering". However, it is increasingly an engineering activity
with a growing maturity. A general systematic approach to modelling has emerged
from the numerous models which have been set up and used in process engineer-
ing. This systematic approach endeavours to provide good engineering practice
in process modelling. It can be regarded as providing some "golden rules" for
the task.
This chapter contains the basic principles of model building, outlining the
elements and procedures of a systematic approach. It addresses the following
issues:

The concept of a process system and the modelling goal, as well as the effect
of the goals on the process model (Section 2.1).
The general notion of a model, different types of models and mathematical
models (Section 2.2).
The description of the steps in a modelling procedure illustrated by simple
examples and an explanation of the iterative nature of the model building
process (Section 2.3).
The necessary modelling ingredients, namely the necessary elements of a pro-
cess model: the assumptions, the model equations and any initial and boundary
conditions which are put into a systematic format (Section 2.4).

19
20 CHAPTER 2 A SYSTEMATIC APPROACH TO MODEL BUILDING

u y
S
inputs outputs

system
X

FIGURE 2.1 General system schematic.

2.1. THE PROCESS SYSTEM AND THE MODELLING GOAL

2.1.1. The Notion of a Process System

If we want to understand the notion of a process system, we can start from the general
notion of a system. This can be defined in an abstract sense in system theory. A
system is a part of the real world with well-defined physical boundaries. A system is
influenced by its surroundings or environment via its inputs and generates influences
on its surroundings by its outputs which occur through its boundary. This is seen in
Fig. 2.1.
We are usually interested in the behaviour of the system in time t eT.A system
is by nature a dynamic object. The system inputs u and the system outputs y can be
single valued, giving a single input, single output (SISO) system. Alternately, the
system can be a multiple input, multiple output (MIMO) system.
Both inputs and outputs are assumed to be time dependent possibly vector-valued
functions which we call signals.
A system can be viewed as an operator in abstract mathematical sense transform-
ing its inputs u to its outputs y. The states of the system are represented by the vector
X and are usually associated with the mass, energy and momentum holdups in the
system. Note that the states are also signals, that is time-dependent functions. We can
express this mathematically as

M:T->WC7^^ J:T->3^c7^^ S:U-^y, y = S[u]. (2.1)

Here, the vector of inputs M is a function of time and this vector-valued signal is
taken from the set of all possible inputs W which are vector-valued signals of dimension
r. The elements of the input vector at any given time r, M/(0 can be integer, real or
symbolic values. Similarly, for the output vector}^ of dimension v.
The system S maps the inputs to the outputs as seen in Eq. (2.1). Internal to the
system are the states x which allow a description of the behaviour at any point in time.
Moreover x{t) serves as a memory compressing all the past input-output history of
the system up to a given time t.
A process system is then a system in which physical and chemical processes are
taking place, these being the main interest to the modeller.
The system to be modelled could be seen as the whole process plant, its envir-
onment, part of the plant, an operating unit or an item of equipment. Hence, to define
our system we need to specify its boundaries, its inputs and outputs and the physico-
chemical processes taking place within the system. Process systems are conventionally
THE PROCESS SYSTEM AND THE MODELLING GOAL 21

inputs

FIGURE 2.2 General model schematic.

specified in terms of aflowsheetwhich defines the boundaries together with inputs and
outputs. Information is normally available about the internal structure of the system
in terms of the operating units and their connections.

2.L2. The Modelling Goal

The modelling goal specifies the intended use of the model. The modelling goal has a
major impact on the level of detail and on the mathematical form of the model which
will be built. Analogous to the idea of a process system shown in Fig. 2.1, a model
acts in some way to mimic the behaviour of the real system it purports to represent.
Thus, Fig. 2.2 shows the model with certain inputs and outputs. The use of the model
can take various forms depending upon what is assumed to be known and what is to
be computed.
Amongst the most important and widely used modelling goals in process
engineering are the following:
Dynamic simulation
With the process model developed to represent changes in time, it is possible to predict
the outputs o given all inputs /, the model structure M and parameters p.
Static or steady-state simulation
Here, the process system is assumed to be at steady state, representing an operating
point of the system. Again the simulation problem computes the output values o given
specific inputs /, a model structure M and its parameters p. This is sometimes known
as a "rating" problem.
Design problem
Here, we are interested in calculating the values of certain parameters p from the set
of parameters /?, given known inputs / and desired outputs o and a fixed structure
M. This type of problem is normally solved using an optimization technique which
finds the parameter values which generates the desired outputs. It is also called a
"specification" problem.
Process control
The fundamental problems in process control are to consider a dynamic process model
together with measured inputs / and/or outputs o in order to:
• design an input for which the system responds in a prescribed way, which
gives a regulation or state driving control problem'.
22 CHAPTER 2 A SYSTEMATIC APPROACH TO MODEL BUILDING

• find the structure of the model M with its parameters p using the input and
output data, thus giving a system identification problem;
• find the internal states in M given a structure for the model, thus giving a state
estimation problem which is typically solved using a form of least squares
solution;
• find faulty modes and/or system parameters which correspond to measured
input and output data, leading to fault detection and diagnosis problems.
In the second part of the book which deals with the analysis of process models,
separate chapters are devoted to process models satisfying different modelling goals:
Chapter 13 deals specifically with models for control. Chapter 15 with models for
discrete event systems, while Chapter 16 deals with models for hybrid or discrete-
continuous systems.
It follows from the above thai a problem definition in process modelling should
contain at least two sections:
• the specification of the process system to be modelled,
• the modelling goal.
The following example gives a simple illustration of a problem definition.
wmm EXAMPLE 2.1.1 (Problem definition of a CSTR).
Process system
Consider a continuous stirred tank reactor (CSTR) with continuous flow in and out
and with a single first-order chemical reaction taking place. The feed contains the
reactant in an inert fluid.
Let us assume that the tank is adiabatic, such that its wall is perfectly insulated
from the surroundings. The flowsheet is shown in Fig. 2.3.
Modelling goal
Describe the dynamic behaviour of the CSTR if the inlet concentration changes. The
desired range of process variables will be between a lower value x^ and an upper
• • • value x^ with a desired accuracy of 10%.

2.2. MATHEMATICAL MODELS

In Chapter 1, we discussed briefly some key aspects of models and, in particular,

mathematical models. We saw that models of various sorts are constructed and used
in engineering especially where undertaking experiments would not be possible, fea-
sible or desirable for economic, safety or environmental reasons. A model can be
used to help design experiments which are costly or considered difficult or danger-
ous. Therefore, a model should be similar to the real process system in terms of its
important properties for the intended use. In other words, the model should describe
or reflect somehow the properties of the real system relevant to the modelling goal.
On the other hand, models are never identical to the real process system. They
should be substantially less complex and hence much cheaper and easy to handle so
that the analysis can be carried out in a convenient way. The reduced complexity of
the model relative to the process system is usually achieved by simplification and
MATHEMATICAL MODELS 23

/.C4
in-flow

O^

out-flow

FIGURE 2.3 Flowsheet of a continuously stirred tank.

elimination of certain controlling mechanisms during the modelling process. It also

relates to the number and type of inputs and outputs that are considered appropriate.
We use modelling assumptions to describe the specific knowledge we have about the
process system. This knowledge can help simplify and eliminate what we consider
to be unnecessary aspects of the model's description. Because of the less complex
nature of the model, the number of variables will be much smaller in dimension than
the dimension of the real system. Hence, we can write that

dim(jM) =PM <^ dim(ys) = ps, dim(MM) = rM <^ dim(M5) = rs,

where yM, UM denotes the input and output vectors of the model and ys, us are those
of the system being modelled.

2.2.1. White, Black and Grey-Box Models

In Chapter 1, we classified models using various criteria. In particular, models can
be classified according to their physical nature. There are analogue models where the
descriptive power of the model is based on physical, or physico-chemical analogues
between the real process system and the model. Laboratory scale equipment or pilot
plants use full physico-chemical analogy but there are also analogue models of process
systems based on mechanical or electronic principles, using springs and damper pots
or analogue computing components. In contrast, mathematical models are models
consisting of mathematical objects, such as equations, graphs, or rules.
There are at least two fundamentally different ways of obtaining process models.
First, we can use our process engineering knowledge to describe the physico-chemical
processes taking place in the system on the level required by the modelling goal. This
is a "first principles" engineering model. Such a model is termed a white-box model
(mechanistic) to indicate that the model is totally transparent or understandable to a
process engineer. In this case, we do not directly use any measured data about the
24 CHAPTER 2 A SYSTEMATIC APPROACH TO MODEL BUILDING

process. However, we do make use of measured data indirectly, through the form and
the value of the system parameters.
Second, the alternative means of obtaining process models is often dictated by a
lack of accumulated engineering knowledge of the system. In this case, we can use
measurement data of the inputs and outputs to build a process model. The need to
stimulate or excite the system to obtain useful information for model identification
is a major issue for any model builder. This includes the amount of data and its
frequency content which enables the key responses to be captured in the resulting
model. In this case, we can use structure and parameter estimation methods developed
mainly in the field of process identification. This kind of model is called a black-box
model (empirical) because knowledge of the process is only available through the
measurement data and not from the underlying mechanisms. A good coverage of
system identification methods is given by Ljung [14]
In process engineering practice, however, both purely white and purely black-
box models are rare. In most cases, we use a suitable combination of our a priori
process engineering knowledge to determine the structure and some of the parameter
values of the model. We then use measured data to build the model. This involves
the definition of kinetic and transport mechanisms, the estimation of key kinetic
and transport parameters as well as validation of the model against the performance
specifications. These combination models are termed grey-box models.
The terms used in this section are more recent inventions; however, the underlying
principles and practice have existed for decades.

2.3. A SYSTEMATIC MODELLING PROCEDURE

Like other engineering tasks, good practice requires models to be constructed fol-
lowing a well defined sequence of steps. These steps are arranged in a "Seven (7)
Step Modelling Procedure" which is introduced below and shown schematically in
Fig. 2.4.
However, it should be noted that model development is inherently iterative in its
nature. One must usually return to and repeat an earlier step in case of any problems,
unusual or unwanted developments later in the process. No one gets it rightfirsttime!
In fact, we never get a perfect model, just one that is usable.
Before starting to setup a process model the problem definition should be clearly
stated. This defines the process^ the modelling goal and the validation criteria. This
is part of the formal description in the SEVEN STEP MODELLING PROCEDURE, which
is given in the form of an algorithmic problem. An algorithm being a systematic
procedure for carrying out the modelling task, named after the Persian mathematician
Abu Ja'far Mohammed ibn Musa al Khowarizm of the ninth century [15].
In a formal description of an algorithmic problem, one should formally specify
the following items:
• the inputs to the problem in the Given section,
• the desired output of the procedure in the Find or Compute section,
• the method description in the Procedure or Solution section.
Problem statements and their use in analysing computational or algorithmic
complexity will be described later in Section 10.1 in Chapter 10.
A SYSTEMATIC MODELLING PROCEDURE 25

1. Problem
definition

2. Identify
controlling factors

3. Evaluate the
problem data

^'
4. Construct
the model

^'
5. Solve
the model

'r
6. Verify
the model solution

'r
7. Validate
themlodel

FIGURE 2.4 Systematic model building steps.

Applying the above principles to the general modelling procedure, the following
algorithmic problem statement can be constructed.

SEVEN STEP MODELLING PROCEDURE

Given:

• a process system
• a modelling goal
• validation criteria
26 CHAPTER 2 A SYSTEMATIC APPROACH TO MODEL BUILDING

Find:
• a mathematical model

Procedure:
The model is built following a systematic procedure given below with seven steps.
The steps of the modelling procedure are as follows:

1. Define the problem

This step refines the sections already present in the problem definition: the description
of the process system with the modelling goal. Moreover, it fixes the degree of detail
relevant to the modelling goal and specifies:
• inputs and outputs
• hierarchy level relevant to the model or hierarchy levels of the model in the
case of hierarchical models (see Chapter 9 for hierarchy levels)
• the type of spatial distribution (distributed or lumped model)
• the necessary range and accuracy of the model and
• the time characteristics (static versus dynamic) of the process model.

2. Identify the controlling factors or mechanisms

The next step is to investigate the physico-chemical processes and phenomena taking
place in the system relevant to the modelling goal. These are termed controlling
factors or mechanisms. The most important and common controlling factors include:
• chemical reaction
• diffusion of mass
• conduction of heat
• forced convection heat transfer
• free convection heat transfer
• radiation heat transfer
• evaporation
• turbulent mixing
• heat or mass transfer through a boundary layer
• fluid flow
Figure 2.5 shows schematically the situation which exists when attempting to model
a process system. As we consider the system under study, we recognize the following
issues:
• There is a set of all process characteristics which are never fully identified.
• In the set of all process characteristics for the system we often:
- only identify and include in the model a subset of the essential characteris-
tics. This means that some essential characteristics for the application can
be missing from our model description;
- include non-essential characteristics of the system in our models, which
lead to unnecessary complexity and/or model order (or size).
• We sometimes incorrectly identify process characteristics which are actually
not part of the system and include them in our model.
The previous issues are often difficult to resolve and very dependent on our
understanding of the system and our insight into what is important. Model validation
and the principle of parsimony, which seeks the simplest representation for the task
are the tools that need to be employed to address these issues.
A SYSTEMATIC MODELLING PROCEDURE 27
Identified essential
process characteristics

Identified
non-essential
process
characteristics

' All process

characteristics

All essential
Incorrectly process
identified characteristics
process
characteristics

Identified essential
process characteristics

Incorrectly identified Identified non-essential

process characteristics process characteristics

FIGURE 2.5 Characteristics of system and model.

It is very important to emphasize that one has to filter carefully the set of all
possible controlling mechanisms taking into account the following key elements in
the problem definition:
• the hierarchy level(s) relevant to the model,
• the type of spatial distribution,
• the necessary range and accuracy,
• the time characteristics.
The effect of the modelling goal on the selection of balance volumes are discussed
separately in Section 3.3.
EXAMPLE 2.3.1 (Identifying key controlling mechanisms). As an example,
consider the modelling of a jacketed tank which is well-stirred and heated using
a hot oil feed as shown in Fig. 2.6. If we were to model this system to predict the
dynamic behaviour of the liquid temperature, then some key controlling mechanisms
could be:
• fluid flow of liquid into and out of the tank,
• fluid flow of hot oil in and out of the jacket,
• convective heat transfer between the jacket and tank contents.
28 CHAPTER 2 A SYSTEMATIC APPROACH TO MODEL BUILDING

Cold feed in

Hot oil in

Hot oil out

Heated liquid out

F I G U R E 2.6 Heated tank.

• work input done through the agitator,

• convective heat losses from the jacket to the environment,
• convective heat losses from the surface of the heated liquid to the environment,
• evaporative losses of liquid from the tank contents to the environment.
We would need to assess the most important or essential characteristics required
in the model for the stated modelling goal. This initially requires an order of magnitude
assessment of mass and heat transfer rates to see what is of major importance.
The models for these controlling factors are developed in physical chemistry,
reaction engineering, mass and heat transfer and unit operations. They form partial
models within the overall process model. These partial models are generally black
box in their nature.
3. Evaluate the data for the problem
As it has been already noted, models of real process systems are of the grey-box
type, therefore, we almost always need to use either measured process data directly
or estimated parameter values in our models.
It is very important to consider both measured process data and parameter val-
ues together with their uncertainties or precision. Some default precision values
might be:
• industrial measured data is ±10 to 30%,
• estimated parameters from laboratory or pilot-plant data is ± 5 to 20%,
• reaction kinetic data is ±10 to 500%,
if nothing else is specified.
At this step, we may find out that there are neither suitable parameter values
found in the literature nor measured data to estimate them. This situation may force
us to reconsider our decisions in steps 1 and 2 and to return there to change them.
4. Develop a set of model equations
The model equations in a process model are either differential (both partial and
ordinary differential equations may appear) or algebraic equations. The differential
Discovering Diverse Content Through
Random Scribd Documents
 CHAPTER XIX.
RICH MEN’S QUARRELS.

M y father was himself again. He was clothed in his right mind

once more. He even appeared to have forgotten that I had
emptied the bottle the day before, and treated me as kindly as
though nothing had occurred to mar the unity which had always
subsisted between us. My mother seemed to be quite happy, too;
and, while I was at work in the garden, she told me she had talked till
daylight with him, after his return from Colonel Wimpleton’s. He had
bitterly bewailed his error, and solemnly promised not to taste
another drop of liquor. He was conscious that he had lost his twenty-
four hundred dollars by getting intoxicated, and he had very little
hope of ever seeing it again.
More than this, my mother had explained my conduct to him, and
he was satisfied with it. The night visit of Waddie, and the colonel’s
unreasonable harshness to him, had probably done more to
convince him than any words of my mother. He had lost his situation,
and had been treated with gross injustice, for the great man would
not accept his explanation of the blow he had given his son.
“Wolf,” said my father, after he had granted me permission to
accept Major Toppleton’s offer, “I am afraid we shall soon be in
trouble all round.”
“I hope not.”
“If I had the money to pay off the mortgage on the house, I should
not care so much. As it is, I may lose even the thousand dollars I
have paid on it. The colonel will foreclose on me at once, and people
here will not dare to bid when it is put up at auction, if he tells them
not to do so.”
“I heard you say you had an offer of thirty-five hundred dollars for
the place.”
“So I had; Bingham offered that for it.”
“I would go to him, and take the offer at once.”
“What, sell the place?”
 “Yes; you can pay off the mortgage, and then have fifteen
hundred left.”
“That’s a good idea,” replied my father. “But I don’t know that
Bingham will give thirty-five hundred now.”
“I would try him, at any rate. I think we had better move out of
Centreport.”
“Perhaps it would be as well, after what has happened,” said he,
in deep thought. “I will see what can be done.”
My father hastened to the village to see Bingham, and soon after
I pulled across the lake to report for duty to Major Toppleton. I was
shown into his elegant library; but I found the magnate of Middleport
in violent wrath.
“I have called, sir, to say that I will accept the offer you were so
kind as to make to me yesterday,” I began, with the utmost
deference.
“Very well, boy, I am a man of honor, which cannot be said of
every man who lives on the other side of the lake,”—by which, of
course, he meant Colonel Wimpleton,—“and I will keep my
agreement; but if the business were to be done over again, I
wouldn’t have anything to do with a person from Centreport.”
“I’m sorry you think so hard of us, sir,” I ventured to reply. “I will
do the very best I can for you; and I hope we shall not live in
Centreport much longer.”
“Well, I don’t know that I need to blame you for what Wimpleton
does. He is a mean man, and his soul is smaller than a mosquito’s.
This morning the old rascal sent his agent over here to offer the
engineer of my flour mills twenty dollars a month more than he is
getting now. The villain was paid up to last night, and left without
giving me any notice, and my mills are all stopped.”
Major Toppleton walked the library in a violent rage, and I waited
for further developments before I dared to speak.
“He hired my engineer away from me, I’m told, because I
employed you,” added the magnate, pausing before me.
“I’m very sorry I made any trouble,” I answered, diffidently.
“You didn’t make it. I only wonder how Wimpleton was my friend
for so many years. He omits no opportunity to stab me when he gets
a chance. I suppose he is gloating over it now because no smoke
rises from my mills.”
“Do you want an engineer, sir?” I had the audacity to ask at this
opportune moment.
“Of course I do. Wimpleton sent over for mine solely to vex me,
and I would give a thousand dollars to be even with him this
moment.”
“I can run the engine of your mills,” I replied.
“You?”
“Yes, sir; I have run the Centreport mills for a week at a time.”
“But I want you on the dummy.”
“I will bring you an engineer, then, in an hour. What wages will
you give, sir?”
“I will give the same that Wimpleton pays the man he stole from
me—eighty dollars a month, and engage him for a year.”
“I will have him here in one hour, sir.”
“But who is he?”
“My father, sir.”
“Oh, ho!”
“Colonel Wimpleton discharged him before daylight this morning.”
“Then I am to take a man whom Wimpleton has discharged, and
pay him twenty dollars more than he was having before.”
“He discharged my father in order to punish him,” I replied; and
then I told him the sequel to the story I had related the day before.
“Very good! Excellent! I will help Wimpleton punish your father by
giving him eighty dollars a month, which is twenty dollars more than
any engineer ought to have. Go for him at once.”
I never pulled across the lake so quickly before as I did then. I
found my father at home; he had just returned from his visit to
Bingham.
“Back so soon, Wolf?” said he; and he looked quite sad.
“Yes, sir. Did you see Bingham?”
“I did; but it’s too late. He has heard of the quarrel, and won’t buy
the house at any price. It will go hard with me, I’m afraid.”
“Never mind, father. It will come out right in the end, I know.”
“What did you come back for?”
 “Major Toppleton wants you, and will engage you for a year, at
eighty dollars a month,” I replied, with proper enthusiasm.
“Eighty dollars!”
I explained what Colonel Wimpleton had done, and what Major
Toppleton wished to do.
“He wants you right off, this minute,” I added.
“I’m all ready.”
“When rich men quarrel, poor men ought to profit by it, if they can
do so honestly,” I suggested.
“The colonel will be the maddest man this side of the north pole,
when he hears of my good fortune,” said my father.
“I dare say he will, for it appears that he has only discharged you
to open the way to a much better position.”
“Exactly so!” exclaimed my father, delighted with the situation. “If
rich men will be fools, we cannot help it, as you say, Wolf.”
My father took the bundle of old clothes he had just brought from
the mill; and we went down to the wharf, where we embarked in the
skiff for the other side of the lake.
“If you could only sell the place, father, we might move over to
Middleport at once,” said I, pulling with all my might at the oars.
“I don’t think I can do it. By this time everybody knows that the
colonel has quarrelled with me, and no one will run the risk of
offending him by buying it,” replied my father. “I hope Mortimer will
catch Christy, and get back part of my money, if not the whole of it.”
We landed in Middleport, and hastened to the mansion of the
major. He was ever so much better humored than when I had called
upon him before. He had evidently considered the nature of the
victory he had won over his powerful rival, for he had certainly
cheated the colonel out of his revenge upon my father, and
practically nullified his punishment. He appeared to be duly
comforted.
“I am glad to see you, Mr. Penniman,” said he, graciously, as my
father bowed low to him.
“I am very grateful to you for your kind offer, sir, and I accept it
thankfully,” replied my father.
“I wish to see the smoke rising from the chimney of the mill at
once,” added the major, briskly. “I want Wimpleton to understand that
he can’t shut me up. Go to the mill, and get up steam as fast as you
can; and the more smoke you make, the better, for that will be my
sign of triumph.”
“I’ll fire up at once,” replied my father, leaving the room, and
hastening to his work.
Young and inexperienced as I was, I could not help feeling sad at
this exhibition of malignity on the part of the rich man of Middleport.
The colonel had taken the opportunity afforded by the dismissal of
my father to strike his rival in a tender place. It was mean; but such
was the character of the dealings between them, when they had any.
The major rubbed his hands with delight, and paced the library under
the exhilaration of the moment. It was a pity that these men, with
such vast means of doing good within their grasp, should quarrel
with each other, and debase and demoralize a whole neighborhood
by their actions.
“Well, Mr. Penniman, I suppose you are ready to go to work,” said
the major, pausing before me after a time.
Mr. Penniman! I felt an inch taller to have a handle applied to my
name by such a magnificent man.
“Yes, sir; I am waiting for orders.”
“I suppose you think that dummy isn’t much of an engine,” he
added, with a very pleasant smile upon his face.
“I think it works very well, sir.”
“I dare say you do; but I want to say, a thing like that is not the
height of our ambition,” he continued, rubbing his hands under the
influence of some undeveloped idea.
“I’m sure I shouldn’t wish for anything better than the dummy.”
“It answers very well to begin with; but I have a regular
locomotive and two cars in process of building, and I shall have them
on the track this fall.”
“Is it a big locomotive?” I asked, curiously.
“No, it’s a small one; and it will be the prettiest plaything you ever
saw. I’m determined that the Toppleton Institute shall be the most
popular one in the country.”
“I suppose Colonel Wimpleton will do something to offset this
movement on your part,” I suggested.
 “What can he do?” asked the major, anxiously. “Have you heard
of anything?”
“No, sir. I only know they feel very bad about the Lake Shore
Railroad over there.”
“They will feel worse before we get through with it,” replied the
magnate, shaking his head. “What can they do? They can’t build a
railroad, the country is so rough. We can keep ahead of them now.
But I want that dummy in motion. You must run it every half hour for
the rest of the day between Middleport and Spangleport. Carry
everybody who wishes to ride. I want the Centreport people to see it,
and to know that we are alive on this side.”
“Will the students be with me?” I inquired.
“This afternoon, when they are dismissed from the school-room,
they will be. I will send you a conductor. Let me see; Higgins is too
sick to study, and just sick enough to play. He shall run with you.
Now keep her going, as though you meant business.”
“I will, sir; I will put her through by daylight,” I replied, as I left the
library.
 CHAPTER XX.
THE BEAUTIFUL PASSENGER.

I found the dummy just as I had left it on the preceding day. I

kindled a fire in the furnace, rubbed down the machinery, filled up
the water tank, and took on a supply of coal, which was the kind of
fuel intended to be used under the boiler. I assure my sympathetic
reader that I felt a real pleasure in the discharge of these duties, and
in the consciousness that I was actually the master of the machine.
Though my taste was rather inclined towards the engine of a
steamer, I was more than satisfied with my present position, and
deemed myself the luckiest dog in the world.
Higgins, the invalid student, who was to officiate as conductor,
stood by and watched all my movements with the most intense
interest. He looked like a clever fellow, and I proceeded to make
friends with him in due form, by declaring that I was sorry he was
sick.
“I’m not sorry,” said he with a grin. “I’m rather glad I’m sick. In
fact, I’m not very sick.”
“Well, I thought you were; the major said so; at least he said you
were too sick to study, and just sick enough to play.”
“Did he say that?”
“He did.”
“Well, he knows a thing or two,” laughed Higgins. “My mother
thinks it makes my head ache to study; and in fact it does when the
lessons are hard.”
“I dare say. Are they hard to-day?” I asked.
“Not so very hard; but, to tell the truth, I thought there was to be
some fun going on here and I wanted to be on hand. My mother
wrote to the principal that she did not wish me to study very hard, for
something ailed my head.”
“I’m afraid the jar of the dummy will hurt your head,” I suggested.
“Oh, no, it won’t,” protested the candid Higgins. “It feels better
now than it did this morning; in fact, it always feels better after school
begins.”
“But I’m really afraid it will injure you to ride on the dummy, with
all the excitement of the highly responsible position of conductor,” I
added, gravely. “I think I had better mention the matter to Major
Toppleton when I see him.”
“Oh, no; don’t do that,” pleaded Higgins, plaintively. “Between you
and me and the smoke-stack of the dummy, I am as well as you are.”
“Precisely so; and I think the major understands your malady, if
the principal does not.”
“Don’t say a word this time, and I won’t ‘soger’ any more.”
“It’s none of my business, Higgins, but you are sawing off your
own nose, and playing the trick upon yourself. I would be a man and
face the music like one, if I were you.”
“I will face the music if you won’t say anything.”
“All ready, Mr. Conductor!” I shouted, when I had steam enough.
“All aboard!” yelled Higgins, rather glad to change the subject
when he found that I did not appreciate his deception.
I ran the dummy out of the house, and stopped her near the head
of the steamboat wharf. The car was still so great a novelty that
many people gathered around to examine it. The cushions were now
well dried, and though the cloth had suffered somewhat from the
effects of the bath, it looked very nice inside of her.
“Have you a watch, Higgins?” I asked of the gentlemanly
conductor.
“Yes,” replied he, producing a small gold one.
“We will leave Middleport on the hour, and Spangleport on the
half hour,” I added. “I wish I had a watch.”
“Why don’t you have one?”
“I’m not a rich man’s son, and I can’t afford to have such
playthings. But I suppose I must get one, if I run on this dummy.”
“I’ll lend you mine for to-day, Wolf.”
“Thank you. I want to time the running, so as to know where we
are,” I answered, taking the watch, and attaching the chain to my
vest. “It is nearly twelve o’clock, and we will start soon.”
“All aboard for Spangleport!” screamed Higgins, as though the
announcement was intended for the people on the other side of the
lake.
 “Folks will understand that nothing ails your lungs, Higgins,
whatever is the matter with your head,” I added, gently, to the
zealous conductor. “I wouldn’t yell so. Boys always make fools of
themselves by hallooing when there isn’t the least need of it.”
Higgins, in a milder tone, invited the ladies and gentlemen who
were inspecting the car to step in and make the excursion to
Spangleport, promising that they should return in just fifty minutes.
Quite a number of them accepted the invitation; and I was about to
start, when I saw a very beautiful young lady hastening towards us.
She was elegantly dressed, and her movements were as graceful as
those of a fawn. The “gentlemanly conductor” rang the bell for the
engine to start, and the young lady, hearing it, made a motion with
her sunshade for us to wait for her. I was too happy to find she was
to be a passenger in the car to start without her, in spite of my
laudable ambition to be “on time.”
The moment Higgins saw her, he jumped off the platform, took off
his cap, bowed and scraped like a French dancing-master, and
helped her up the steps. There was a glass window in the partition
between the engine-room and the passenger compartment, for
which at that moment I felt extremely grateful to the builder, for it
enabled me to obtain an occasional glance at the beautiful young
lady. I beg leave to say that this unwonted enthusiasm on my part
was as surprising to myself as it will be to my readers, for I had
hardly ever looked at any person of the feminine persuasion before,
except my mother and sisters. I had certainly never seen any lady
who attracted me so strongly, or for whom I felt so great an
admiration. She was not more than fifteen or sixteen years of age;
but she wore a long dress, and had a mature bearing.
Higgins conducted her to a seat, and she took possession of it as
gracefully as though she had been schooled in the polite art for a
whole lifetime. I could not help gazing at her, and I envied Higgins
the rapture of being permitted to speak to her. She looked around,
and bowed to several persons in the car, with the sweetest smile that
ever lighted up a young lady’s face. I was wholly absorbed in gazing
at her, and actually forgot that I was the young engineer of the Lake
Shore Railroad, till the sharp snap of the bell brought me to my
senses, and assured me that Higgins was not so fascinated as I
was.
I was a minute behind time, and I let on the steam to make it up. I
was obliged to turn my back on the beautiful being in the car, and
look out for “breakers ahead” through the door and windows in the
end of the engine-room; but I had the pleasing satisfaction of
thinking that in running backwards from Spangleport I should face
the other way.
What a fool I was! Of course I was. A young man always has a
time to be a fool, just as he has to take the measles, though he
seldom has it so young as I did. I did not know who the young lady
was, and I did not crave any other privilege than that of simply
looking at her, just as I should at a pretty picture. If she had fallen
overboard, I should certainly have jumped in after her. If she had
been in the claws of a lion, I should certainly have smitten the lion. If
she had been in the upper story of a house on fire, I should certainly
have run the risk of being singed for her sake. But she did not fall
overboard, or into the claws of a lion, and she was not in a burning
house; and, provoking as it was, I could not do anything for her,
except turn my back to her,—and I was not sure that this was not the
most agreeable service I could render her,—and run the dummy at
its highest speed.
I could not help seeing the beautiful young lady even through the
back of my head; and I am sorry to say that I forgot to look at my
watch, when we passed Ruggles’s barn and the Grass Brook bridge,
as I had intended; and at a quarter past twelve the dummy sizzled
into Spangleport, shivering like an over-driven horse. I had the self-
possession, however, to stop her when she got there; but I have
since wondered that, under the circumstances, I did not run her into
the lake, or over the hill to Grass Springs. I had made the distance in
just fourteen minutes.
The passengers got out of the car, and for a time I lost sight of
the elegant young lady. Higgins came round to me, and declared that
we had made a “bully trip.” I was entirely of his opinion; but I was not
willing to confess that a certain absentmindedness had induced me
to run the machine so as to gain five minutes, and make up one. The
conductor left me, and I fancied that he had gone to find the
interesting person who had fascinated me, and with whom he
seemed to be acquainted.
“Will you allow me to get in there and see the machinery?” said a
silvery voice, while I was rubbing up the works.
I turned, and my face felt as though all the steam in the boiler had
been discharged upon it when I discovered that the speaker was no
other than the bewitching being who was uppermost in my thoughts.
“Certainly,” I replied, leaping to the ground, and endeavoring to
imitate the polite gyrations of the gentlemanly conductor.
“Thank you, Mr. Wolf,” added she, with the sweetest of smiles.
Mr. Wolf! Involuntarily my head went up, and I felt prouder of the
handle to my name then when the mighty major himself had applied
it.
“I’m afraid you will find the engine-room a very dirty and greasy
place,” I had the courage to suggest, flustered as I was by having the
beautiful girl speak to me—actually speak to me!
“Oh, never mind! I have on my old clothes.”
If these were her old clothes, I wondered what her best were.
“I suppose you don’t know me, Mr. Wolf; but I have heard a great
deal about the young engineer, and I assure you I am delighted to
see you,” she added, with a kind of roguish look, which made me
feel just as though I was “going up.” “I am Grace Toppleton.”
The daughter of the major! I had heard what a pretty, gentle,
amiable girl she was, and I was positively sure that the reports did
not belie her.
“I have often heard of you, though I never had the pleasure of
seeing you before,” I replied, as gallantly as my flustered state of
mind would permit.
Still imitating the gentlemanly conductor, I took her gloved hand,
and attempted to help her up the high step of the engine-room. I felt,
at this particular moment, just as though I was in the seventh
heaven. As the elegant young lady was about to step up, a rude
grasp was laid on my shoulder; so rude that Miss Grace lost her
foothold on the step, and was thrown back upon the ground.
Turning round, I discovered that my rough assailant was Captain
Synders, the constable of Centreport. He was attended by Colonel
Wimpleton and the skipper of the canal boat which had been blown
up. To my astonishment, Miss Grace leaped upon the dummy
without my help, and I was held back by the savage grasp of the
officer. My blissful dream had suddenly been disturbed, and I was
mad. The envious Centreporters had chosen the moment of my
greatest joy to pounce upon me.

THE VISITORS FROM CENTREPORT.—Page 226.

 CHAPTER XXI.
SOME TALK WITH COLONEL WIMPLETON.

I was very intent upon explaining to Miss Grace Toppleton the

mysteries of the dummy engine, and I was not pleased to have
the agreeable interview broken off. I was vexed, annoyed, and
disconcerted. The beautiful young lady looked at me, and I thought I
could see the indications of sympathy upon her face.
“If you will excuse me a moment, Miss Toppleton, I will show you
the engine,” I said to her, with all the politeness of which my nature
was capable.
“I guess not,” added Captain Synders, with a coarse grin, as
though he had evil intentions in regard to me.
“If you will let me, Mr. Wolf, I want to ride back in the engine-
room, and see the machinery work,” replied she, in her silvery tones.
“I guess not,” repeated Captain Synders; and I turned my
attention from her to him.
I could not conceive why Colonel Wimpleton and his odious
associate had chosen to come down upon me at Spangleport, rather
than Middleport, unless it was because their appearance would
make less excitement. The boat in which they had come lay at the
wharf, and they must have started long before the dummy left
Middleport. Possibly they expected to interrupt the trips of the
engine, and have it left five miles from its headquarters without an
engineer.
Colonel Wimpleton had with him Captain Synders, the constable.
It had not yet occurred to me that I should actually be arrested, and
held to answer for the destruction of the honest skipper’s canal boat,
though the appearance of the officer had suggested the idea to me.
They could not arrest me without including Waddie in the warrant, for
he had confessed his agency in the mischief. I did not know of any
way by which I could be punished without involving the scion of the
great house on the other side.
 “What do you wish with me?” I asked, in a very ill-natured tone;
for I beg to remind the reader that I am human, and that Miss
Toppleton occupied the engine-room of the car.
Captain Synders glanced at the colonel, as though he expected
him to do the talking, and that distinguished gentleman looked down
upon me with unutterable severity. The honest skipper did not
appear to have much sympathy with his companions, and looked
very pleasant for a man who had experienced so heavy a loss as
that of his canal boat.
“Wolf!” said the colonel, in stern and lofty accents.
“Sir!” I replied, with a dignity becoming the engineer of the Lake
Shore Railroad.
“We have been looking for you,” he added, glancing at the
constable, as if to direct my attention to him.
“You have been lucky enough to find me, sir. I wish to say, sir,
that the car starts for Middleport at half-past twelve, and therefore I
have only ten minutes to spare,” I replied, consulting Higgins’s gold
watch, the appearance of which, I think, produced a sensation in the
minds of my visitors.
“Humph! I think you will wait my pleasure.”
“That will depend somewhat upon the length of time your
pleasure demands my presence. Will you please to tell me what you
want of me?”
“Where is your father, Wolf?”
“At Middleport, sir.”
“What is he doing there?”
“He is at work, sir.”
Colonel Wimpleton looked as though he wanted to swear; for I
am sorry to say this influential man sometimes indulged in the
wicked habit of using profane language. It did not seem quite proper
that the menial, whom he had discharged as a punishment, should
find work so soon.
“What is he doing?” demanded the magnate of Centreport, biting
his lips to conceal his vexation.
“You were so kind as to make an opening for him, sir, by hiring
away Major Toppleton’s engineer, and my father has taken his place,
at the same wages—eighty dollars a month—as you pay your new
engineer.”
The great man stamped his foot with rage, and uttered an
expression with which I cannot soil my paper. As wicked, tyrannical,
overbearing men often do, he had overreached himself in his anxiety
to strike my father. If it was unchristian for me to rejoice in his
discomfiture, I could not help it, and I did so most heartily.
“I have been to see him about your conduct,” continued the
colonel, when his wrath would let him speak again. “I want to know
what he is going to do about paying his share of the loss of the canal
boat which you and Waddie blew up?”
“I can speak for him, sir, if that is all you want. He is not going to
pay the first cent of it,” I replied.
“Here is the captain of the boat, and he wants to know what you
are going to do about it,” added the colonel, trying to enjoy the
confusion which he thought I ought to feel in view of such a demand.
“Yes, I want to know who is going to pay for the mischief,” said
the honest skipper; but as he already knew, he did not put much
heart into the words, and actually chuckled as he uttered them.
“Captain,” I continued, turning to the master of the canal boat, “I
say to you, as I have said to others, that I had nothing whatever to
do with blowing up your boat, and I did not know anything about it till
the explosion took place. That is all I have to say.”
“I don’t know anything about it,” replied the skipper.
“I do,” interposed the colonel. “He has confessed that he had hold
of the string when the boat blew up.”
I took the trouble to explain to the honest skipper that Waddie
had asked me to pull in his kite line; that I had picked it up, but,
fearing some trick, had done nothing with it; and that Waddie had
pulled the string himself.
“All aboard for Middleport!” shouted Higgins, as moderately this
time as a gentlemanly conductor should speak.
“My time is nearly up, sir,” I added to the colonel. “If you have any
further business with me, please to state it as quickly as possible.”
“You must go over to Centreport with me, and arrange this
business,” replied the magnate, gruffly.
“No, sir; I cannot do that.”
 “Then Captain Synders must arrest you.”
“Very well, sir; let him do so. I am willing to go to jail and stand
trial on the blowing up. Have you made a complaint against me?” I
asked of the honest skipper.
He was too candid to tell a lie, and he made me no answer.
“Have you a warrant for my arrest?” I demanded of Captain
Synders.
“I can take you without a warrant,” growled the constable.
“Do so, then. If you wish to arrest me, I will submit.”
My friends may think I was putting a very bold face upon the
matter, but I candidly admit that I should have been glad to have the
charge against me investigated; though I was very certain no steps
would be taken in that direction. It is possible Colonel Wimpleton
believed that I had been concerned with his hopeful in the blowing
up of the canal boat; yet the guilt of his son was settled, and, if
convicted, some stupid judge might sentence us both to the
penitentiary, for the case would have to go to the shire town of the
county, out of the reach of the great man’s influence, for trial.
My father had told me that, at the interview with the colonel in the
night, the latter had threatened him with prosecution for abusing his
son; but when my father suggested that Waddie had broken into his
house in the night time, it was plain enough that the young
gentleman was liable to a turn in the state prison. Waddie’s crimes
and mistakes continually stood in the way of his taking his revenge. I
considered myself fully protected in the same manner.
“Captain Synders, if you are going to arrest me, please to do it at
once,” I added, as the lady passengers began to get into the car, and
some of the gentlemen came up to the spot where I stood.
“What’s the row?” asked Higgins.
“These gentlemen from the other side talk of taking me up for the
mischief to that canal boat. If they do so, Higgins, I want you to go to
my father, and tell him about it. If I mistake not, Waddie Wimpleton
will be arrested before night for breaking into our house.”
“I’ll do it!” exclaimed the enthusiastic conductor.
“Arrest Waddie!” ejaculated the colonel, gnashing his teeth with
rage.
 It was mortifying to the great man to find that he had come to the
end of his rope; that even his power to annoy and persecute his
inferiors had a limit.
“All aboard!” repeated Higgins.
“If you are going to arrest me, Captain Synders, now is your
time,” I added.
The people who had gathered around us began to laugh and
enjoy the scene, and, being mostly Middleporters, they had no
particular sympathy for the colonel.
“Wolf, we shall meet again,” said the great man, sullenly, as he
turned upon his heel, and, followed by his companions, walked
rapidly down to the wharf, where his boat lay.
Even while I knew his power, and felt that he could annoy our
family, and perhaps ruin us, I was quite ready to meet him again.
Waddie’s indiscretions stood between me and his wrath for the
present, but his time might come. I leaped into the engine-room of
the dummy, where Miss Toppleton had stood listening to our
conversation.
“Mr. Wolf, I think you are real smart,” said she, with a sweet smile
of approbation.
“I thank you, Miss Toppleton, for your good opinion. Colonel
Wimpleton is very hard upon me just now.”
“I heard father tell about it. I think that Colonel Wimpleton is a real
wicked man; and I only wonder that he and father were good friends
for so many years.”
“I am all ready to start now,” I added. “I wish I had better
accommodations for you.”
“Oh, this is very nice!” exclaimed she.
I opened the valve, and let on the steam.
“What did you do then?” she asked, pointing to the handle of the
valve.
“I let on the steam;” and then I gave her a full description of the
engine, which was hardly finished when we came in sight of
Middleport.
I found it a delightful task to expatiate on my favorite theme to
such a beautiful and interested listener, and I pointed out the cylinder
wherein the piston worked, the connecting rod which moved the
crank, and showed her how the valves which admitted the steam to
the cylinder were worked. I flattered myself, after the lesson I had
given her, that she was almost competent to run the dummy herself.
“I think it is real nice to ride in here, and see the machinery
move,” said she.
“So do I; and I enjoy it very much; more just now, I fear, than I
ever shall again.”
“Why so?” she asked, innocently.
I was not quite bold enough to explain the principal reason, and
so I replied that it was a new thing to me.
“I hope you will let me ride with you again, some time,” she
added.
“With the greatest pleasure,” I replied. “Whenever you please.”
We ran into Middleport, and Miss Toppleton thanked me very
prettily for my kindness in showing her about the engine; and I really
wished I had it to do over again. By this time the students were
turned out of school, and all of them gathered around the dummy,
anxious to begin the afternoon’s fun. I had brought over my dinner,
and I ate it before the next trip. At one o’clock I was ready to start for
Spangleport again.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com