The Industrial Practice

of
Chemical Process Engineering
Samuel W. Bodman
The MJ.T. Press
Massachusetts I nstitute of Technology
Cambridge, Massachusetts, and London, England
Copyrigt 1968 by
The Massachusetts Institute of Technology
Printed and bound i the United States of America by
The Maple Press Company, York, Pensylvania.
All rights resered. No part of this book may be
reproduced or utilized in any form or by any meas,
electronic or mechanical, including photocopying,
recordig, or by any iformation storage ad retrieval
system, without permission in writig from the publi
s
her
Library of Congress catalog card number: 68-18232
Preface
With study of the engineering sciences now a dominant factor in the
curricula of many academic engineering departments, only limited time is
available for the creative application of thcoretical fudamentals to practical
chemical processing problems. I particular, the study of process design
and financial evaluation is often either disregarded or given only cursory
attention. As a result, the young engineering graduate frequently encouters
dificulties in becoming acclimated to the environent of a commcrcial
organization and in ma.ximizing the benefits which can be gcnerated from his
technical background.
In order to prepare its studcnts more fuly for the challenges of
industrial work, the M.I. T. Chemical Engineerng Department has developed
two specific programs. First, tht department's School of Chcmical Engineer
ing Practice exoses graduate students to actual problems i an industrial
environment uder the direct and close supervision of a faculty member.
This program has been in operation for over fifty years and has proven to be
an effective contributor to a student's total academic experience. To com
plemcnt the Practice School program, a senior-ycar synthesis course in
process design has been developed. Professor Thomas K. Sherwood
summarizcd his original work in the development of the M.I. T. process desig
course in his teli "A Course in Process Design". The course, as conceived
by him, is based upon a series of design cases; for each case the student is
reuired to devise and analyze process schemes that might lead to the
solution of the design problem.
The present text seeks to combine some aspects of te Practice School
program with elements of Professor Sherwood's case-study approach to in
strction i engineering design. It is hoped that the result of this combiation
will not only be useful as a text for academic process design instrction but
will also serve as a reference book for te youg engieer embarkig on a
career i idustr.
iii
iv Preface
An introductory chapter briefly describes te interaction and inter
dependence of market researh, process desig, and financial evaluation
functions in commercializing a chemical prduct. Methods of planing and
analyzig laboratory experiments, of utiizig market and financia infor
mation, and of preparig and presenting a chemical process desig are
discussed. The introduction is followed by a series of si case studies in
engineerig desig, tyical of those encountered by a youg engineer i his
initial idustrial assigments. The need for accurate aaysis and correla
tion of laboratory data is given great emphasis. Secondly, the need for
devisig creative and practical solutions to prcessig problems is given
some discussion. Finally, each case illustrates methods of combining tech
nical and finacial inforation to provide a realistic evaluation of a proposed
process. Each case is concluded wit a recommended design as well as
suggestions for furer work that would be required in subsequent, more
detailed, design efforts.
A great deal of emphasis has been placed on the use of the digital
computer i the aalysis and presentation of desig problems; computer
programs are presented for three of the cases discussed. These progras
and others lie tem have proven to be particularly stimulatig when used in
a "computerized classroom, " wherei the student ca communicate directly
wit the machine. In preparing porions of the text, it has been assumed that
the reader has at least some kowledge of te FORTRN codig language.
The present volume was written while the author sered as Director of
the Boud Brook Station of M. I.T.'s School of Chemical Engieerig Practice.
This station is located wth the Organic Chemical Division of the American
Cyanamid Company, Boud Brok, New Jersey. The data used in the prep
aration of Chapter 3 were gathered by a student group as a part of a project
at the Bund Brook Station. The author is grateful to American Cyanamid for
its permission to use these data as well as for the company's hospitality during
the 1965-66 and 1966-67 academic years.
Many individuals have contributed substatially to the case studes
sumaried i this book. First of all, thas are due to Professor Thomas
K. Sherwood, who has continued his gidance in the instrction of chemical
process desig at M.I.T. It was he wo origially suggested the preparation
of the present text, ad Chapters 2 and 4 are based almost entirely on desig
cases developed by him for his process design courses. Sincere appreciation
is el'ressed to Professor Robert York of Corell University for his gidance
i the economic evaluation of chemica projects ad for his instrctive re
view of portions of the present text. Professors C. J. King and Scott Lynn of
the University of Califoria i Berkeley were kind enough to review the entire
text of tis book. Their comments were particularly significant i givig
additional breadth and meaig to several of te case studies. Dr. Howard
Kehde and his colleages at Dow Chemical Compay, Midland, Michigan, very
kidly reviewed the cases concered with styrene production; their comments
provided an invaluable dimension to those cases by bringng to bear the knowl
edge ad sophistication of a major producer of styrene monomer. Professor
Giles R. Cokelet of Califora Institute of Technology provided a valuable
commentary on te sulfur transporation case. Professor R. G. Thorpe of
Corell University originally acquainted the author with some of the difficulties
Preface
v
assoiated wth te vacuu fractionation of styrene-ethylbenzene mitures;
this inforation was most usefl in th preparation of Chapter 6. The data
used as a basis for Chapter 3 were gathered uder te direction of
Professor Michael Modell of M.I. T.; his cooperation in making this in
foration avaiable is greatly appreciated. Several forer teaching assist
ants and students at M.I. T. contributed substantialy to the development of
computer progras for Chapters 5, 6, and 7. I this respect, specia
thaks are due to Robert L. Blumberg, Bruce Crocker, Avelino R. Rodrigez,
and Robert L. Sandel.
Machie computations for this book were perfored at the M.I. T.
Computation Center. The author gratefully ackowledges the financial suppor
of a research initiation grat from the Rohm ad Haas Compay which was
instrmental in the preparation of tis text.
The autor wshes to express his tanks to Messrs. Georges F. Doriot,
Henry W. Hoaglad, William H. Congleton, John A. Shane, ad Miss Droty
E. Rowe of American Research and Development Cororation. Their gidace
durig te past four years in various aspects of finacial aalysis ad business
j udgment contributed substantially to the present work.
The greatest source of encouragement ad assistace in the preparation
of tis mauscript has come from the author's wife Betsy. Without her many
very thoughtfu ad skillful contributions, this work could not have been com
pleted.
S. W. Bodma
Cambridge, Massachusetts
March, 1968
Contents
Preface iii
1. Introduction 1
2. Reactor Desig, Optiization, and Control i the Production
of Monochlorobcnzene 18
3. Process Improvement for Liquid-Liquid Extraction of Fcnway
AcW 51
4. Catalytic Reactor Desig for Benzene Hydrogenation 75
5. Evaluation of New Methods for Sufur Trasporation 100
6. Vacuu Fractionator Desig for the Purification of Styrene
Monomer 125
7. Process Desig and Evaluation for te Production of Styrene
Monomer
151
Appendi 199
Index 229

1. Introduction
The industrial practice of chemical engineering design requires the
application of many talents and skills. By the definition of his pro
fession, the process design engineer in the chemical industry imple
ments the work of the chemist and development engineer in providing
mechanical and structural specialists with a realistic description of
the required equipment. Sherwood (44) has recently described the
following functions which must be performed to span the gap between
bench-scale chemistry and the operating plant:
1. Recognition of an economic opportunity
2. Conception of a plan or design
3. Preliminary analysis of the design
4. Completion of a final design
5. Implementation of construction and operation
These functions are useful indicators of the stages through which
a design must pass and of the various scales of thinking which are re
quired of the design engineer. The foregoing list is quite helpful in
placing each individual job in the perspective of the over-all effort
which is required. I the analysis of a design project, it is to be re

membered that an economic evaluation for the entire proj ect must be
completed after each step in the design; in the absence of a favorable
evaluation after each step, the time and effort required by the next
step cannot be justified.
To complete the transition from laboratory conception to operating
plant successfully, the design engineer in industry must call upon a
wide background not only of technical fundamentals but also of finan
cial and social understanding. For example, knowledge of geographi
cal factors may be important in selecting a proper plant site, while
a grasp of corporate finance and economics may be a prerequisite to
a proper evaluation and presentation of the economic benefits to be
gained from a speciic chemical project. Clearly an understanding of
the operation of modern computation eqUipment is now an essential
part of the training of a design specialist. In addition to having a
1
2 Chapter 1
broad technical background, the successful design engineer must con
tinually develop talents in analyzing his own technical performace
and in evaluating the efforts and contributions of other people.
The present text describes the type of technical and economic
background necessary for the successful completion of a chemical
process design. These fundamentals will be summarized briefly in
the present chapter. This introduction is followed by a series of de
sign case studies that serve to illustrate important aspects of indus
trial chemical engineering design.
The cases selected for presentation illustrate not only different
types of engineering projects but also vriations in the requ;.red de
gree of completion for a design. Cases requiring chemical reactor
design, separation equipment design, and pipeline sizing and optimi
zation are included. One of the most dificult aspects of these design
cases, or for that matter of any realistic technical problem, is that
of defining the real nature of the problem, i. e. , deciding what is re
quired. One of the more common failings of a technical program re
sults from a tendency to answer a question that has not been asked or
to complete work that has not actually been requested. This dificulty
is particularly prevalent in design work, where many different types
and degrees of effort may be required. The case studies here pre
sented illustrate industrial problems and are summarized in the form
of memoranda. In some instances the data were obtained from the
literature, while other cases represent actual industrial problems
where the data originated in a company laboratory and where the in
dicated result constitutes the actual solution presented to manage
ment.
In each of three cases cited, the design analysis is summarized in
the form of a computer program which is reproduced in the text. Re
sults generated by each program are employed to evaluate the econo
mics of a design as a function of the important process variables.
However, results for all possible combinations of these variables have
not been obtained, and the computer programs are presented in order
that they may be further exloited to refine the economic evaluation
of the projects to which they apply. These programs have been found
particularly useful when applied in a computer classroom-a situation
in which a class can "communicate" directly with digital computation
equipment.
The use of a computer to evaluate in detail a chemical process
represents one extreme of the situations that might be encountered
by a typical industrial design group. At the other extreme is the
situation where no detailed design or economic evaluation is required
and only a modest number of calculations is needed to establish the
most liely coniguration for the ultimate design. In the cases of the
latter type which are presented in this volume a consideration of limit
ing cases and the use of shortcut methods in preparing the calculations
prove to be most helpful.
By presenting a series of cases having not only a vried technical
content but also a varied degree of required sophistication, this text
attempts to illustrate some of the concepts associated with the suc-
II/troductiol/ 3
cessful completion of an industrial process desig. By definition,
process design is involved with the application of technical principles
to the available experimental information in order to produce a work
able manufacturing process. As such, design cases have traditionally
been examined individually with relatively little emphasis on a consis
tent set of principles necessary for the proper understanding and
successful execution of general c1.asses of problems. To some degree,
such a set of prinCiples can be established; as a gide in maing the
ensuing case studies meaningful beyond their own particular boun
daries, the following set of principles may be considered:
1. Determination of design requirements
2. Comprehension of market conditions
3. Evaluation of experimental data
4. Establihment of critical design parameters-simulation
and optimization
5. Evaluation of process economics
6. Presentation of design results
The present introduction does not purport to be a thorough review
of each subject listed. Only a cursory discussion of each point is of
fered, together with a review of a few pertinent references. Neither
should the reference citations mentioned be considered as a complete
literature review; they are merely those that have proved useful in
working with young chemical engineers encountering their first in
dustrial design problems.
DETERMNATION OF DESIGN REQUIREMENTS
The sophistication of a process design must be tailored to meet the
requirements of the individual situation. As mentioned before, the use
of data-processing equipment allows the designer more freedom than
ever before to investigate various combinations of system parameters.
Indeed, one of the major functions of the present text is to demon
strate the utility of machine computations in studying various aspects
of a process design. Nevertheless, the advent of modern computers
makes it quite easy for the user to pass through the point of diminish
ing returns. A great deal of objective thining is required to avoid
solving problems merely for the intellectual satisfaction gained from
the solution. As in all aspects of engineering for industry, if the value
of the programming and computer time used is not exceeded by the
value of the design improvement gained, then both engineering and
computer time have been misspent.
To determine the possible need for a detailed design calculation,
it is most useful to analyze limiting aspects of a desig situation by
means of Simple hand calculations. Computations for limiting cases
4 Cliapter 1
are often quite straightforward, since simplifying assumptions can
usually be made. For exmple, if a laboratory reactor has been oper
ated adiabatically between two temperature levels, the results of such
an experiment can be scaled directly to a commercial-scale adiabatic
unit operating between the same two temperature levels. Compari
son of this result with that obtained by assuming an isothermal opera
tion at the lower temperature level sets the extremes between which
nonisothermal designs must fall. By noting the variations in the im
portant design parameters as the design is shifted from one limit to
aother, one can assess the need for maing more detailed and often
more time-consuming calculations at intermediate conditions. Such a
calculation procedure is illustrated by Chapter 4, in which a reactor
for the hydrogenation of benzene is designed. The need to make a
comparison of limiting conditions would seem obvious; nevertheless,
this simple technique is often overlooked, resulting in an unnecessary
exenditure of engineering and/or computer time.
One of the most valuable talents that can be developed by the de
sign engineer is an ability to perform the Simple calculations neces
sary to establish the limiting cases of a design, where the most dif
ficult technique is often that of making the appropriate assumptions
in order to simplify the calculations. This ability usually must be
developed through may years of exerience, and the novce often
finds it quite dificult to achieve. It is to be appreciated, however,
that the ability to perform a simple but meaningful analysis of a prob
lem does not develop automatically with experience. A conscious ef

fort must be made to compare the results evidenced in the final opera
ting plant with the assumptions made in the early stages of the deSign.
Only by such a feedback and by comparison can the quality of sub
sequent estimates be upgraded.
The need for limiting-case calculations cannot be overemphasized;
such calculations should be applied as early as possible in the con
Sideration of any chemical project. More and more effort is being
made today and will be made in the future to provide the research
manager with a quantitative estimate of the probability for the tech
nical and economic success of a particular research project. Clearly
a major ingredient in such an estimate must be a preliminary fore
cast of the capital and operating costs for the project. This type of
estimate is necessarily based on little or no data, and the computa
tions must result from some sort of limiting-case analysis. Thus the
ability to perform such an analysis is valuable not only i establishing
preliminary limits on the process variables but also i determining
whether the probability for financial success justifies the expenses
involved in the bench-scale and pilot-scale experimental work.
In this light, the deSign engineer should enter into consideration of
a chemical project at the bench stage. I his initial calculations do
not show a high probability for financial success, the very existence
of the bench work should probably be reconsidered unless external
Circumstances, such as a raw-material position, are more important
than economic factors. Similarly, as work proceeds through bench
and pilot-scale development, discussions between development groups
and a process design engineer may be very important, particularly in
Introductioll 5
coordinating technical progress with the efforts of market research
groups.
COMPREHENSION OF MKT CONITIONS
Cooperation between the desig engineer and the market research
and development groups is of critical importance. One of the most
essential pieces of information required for the completion of pro
cess design calculations is an estimate of both present and future
market demands for the product under consideration. In many cases
the ability of a design engineer to analyze technical information and
to provide an accurate scale-up to commercial equipment will have
only a modest influence on the economics of the final operating plant.
O the other had, an accurate sales forecast for a product is usually
quite critical to a realistic prediction of the ultimate financial per
formace of the operating unit.
For example, a 20 per-cent error in a kinetic constant or heat
transfer coefficient may be damped out at that stage in the calcula
tions in which the over-all process economics are considered. How
ever, a similar percentage error in a market forecast may well be
amplified as it is transmitted through the calculations leading to an
economic evaluation of the project.
Sources of market information run all the way from government
reports to the annual reviews published by the various trade journals.
Most important, however, are the personal dscussions of salesmen
with customers, and it is this type of interaction which forms the best
basis for sales forecasts. Typically, the sales forecasts will be pre
pared in the very early stages of process development, but they are
subject to rapid and substantial deviations as the market research
work proceeds. It is essential that the process design group be con
tinually kept inormed on the status of the maket estimates. Only in
this way can a final desig be produced which will be justified by
present and future market estimates. Finally, it should be remember
ed that plant construction is usually finished two years or more after
the desig plans are completed. If economic conditions are favorable,
the need often arises to expand the plant facilities even before con
struction is complete. This factor further emphasizes the need for
accurate market forecasting procedures.
Because of the difficulty i gathering and processing meaningful
raw data, the chemical engineering literature has historically given
only scant attention to the subject of marketing. More recently, the
availability of the electronic computer has made feasible the collec
tion and assessment of market information sufficiently broad and ac
curate to allow the development of useful marketing theories as ap
plied to the chemical industry. A corresponding increase in research
and publication activity in this area has been evidenced. Of particular
note is a series of papers presented at a 1965 Arerican Institute of
Chemical Engineers symposium in which the interaction of research,
marketing, and design efforts was discussed (9,1 1,13,21,12). These
papers were prepared by men familiar with all aspects of product
6 Chapter 1
commercialization in the chemical industry, and the series provides
an excellent exposition of the advantages to be gained and the prob
lems encountered by efforts to coordinate marketing and research
programs. The papers are particularly valuable in illustrating the
various methods by which an engineer can assure that an adequate
market picture will be obtained and that a correspondingly accurate
financial evaluation will be achieved. Another compendium of papers
dealing with chemical marketing has been published by the American
Chemical Society (l). This book, which contains twenty contributions,
serves as an excellent background source for the specialized areas
of marketing. For example, the roles of product advertising, applica
tions research, and product delivery methods are given detailed treat
ment.
There is a very definite need for a complete review of recent
chemical marketing literature. Such a review, preferably carried out
by someone with a strong background of industrial marketing expe
rience, would not ony clarify the situation for the student but would
also hopefully lead to better market analysis techniques for the in
dustry as a whole.
EVALUATON OF EXPEIMENAL DATA
I the manufacture of a particular chemical, the required process
steps generally follow the sequence shown below:
1. Preparation of reactants
2. Carrying out of reaction(s)
3. Heating or cooling of reaction products
4. Separation of reactants from products and puriication of
products
Almost without exception, the design engineer is required to base
his analysis of each step upon laboratory data generated by other in
vestigators or obtained from the literature. For those having only a
modest exosure to the chemical literature, Mellon (32) has provided
a very useful guide to the proper methods to be used in searching the
literature. In using literature data for the engineering analysis of a
process operation, it is critical to develop an appreciation for the
fuality of the information to be used. For example, data reported
may years ago may have been obtained before sufficient theory had
been developed to allow a proper analysis and presentation of the ex
perimental information. In the absence of such a theory, early ex
perimenters sometimes failed to measure a variable necessary for
proper analysis. Obviously, the exerimental equipment available in
the early engineering laboratories was not as sophisticated as that
currently available; it is therefore important to develop an apprecia
tion for the strengths and failings of various types of laboratory ap
paratus.
Difficulties with the proper interpretation of published data are
frequently compounded by industrial censorship of process informa
tion. The suppression of technical information is obviously necessary
IntroducLion
7
to protect the commercial value of a process. However, from a tech
nical viewpoint, censorship often requires the engineer to make a
"reasonable" assumption in order to be able to proceed with his
analysis.
A good example of the censorship of industrial information is pro
vided by MacMullin (30), who discusses the distribution of reaction
products for the chlorination of benzene. He presents data that estab
lish the distribution of the various chlorinated compounds as a func
tion of the total amount of chlorine reacted. This information is of
course not sufficient for the design and evaluation of a manufacturing
process, since the kinetic parameters for the reactions are not dis
closed. In order to complete a design, reasonable values of the chemi
cal kinetic constants must be assumed; such a procedure was follow
ed in preparing Chapter 2, in which various processes for the chlorin
ation of benzene are discussed. When it is necessary to proceed in
this manner, it is most desirable to obtain literature information from
as many different sources as possible. By comparing and combining
all available information, one is often more likely to establish a realis
tic basis for a design. I fact, the technique of gathering and compar
ing information from a number of sources is frequently useful in
many aspects of a process design.
Before investing the time and effort required even by a prelimin
ary design calculation, it is prudent to assess the validity and con
sistency of the laboratory findings upon which the design is to be
based. This assessment is most easily accomplished by comparing
the data directly with appropriate literature information. For example,
the general accuracy of a set of vapor-liquid equilibrium data for a
mixture of two components may be checked most directly by com
paring them with those for the same two com)ounds but for other con
ditions of temperature and pressure. I such data are not available
for the desired compounds, the relative volatility computed from the
laboratory result might be compared with that calculated for an ideal
mixture by using Raoult's law. It may also be inform'tive to compare
the relative volatility with that for other compounds having Similar
chemical structures. Finally, the thermodynamic consistency of the
data should be assessed by invoking one form of the Gibbs-Duhem
equation.
Similarly, by plotting the observed solubility of a solid in a liquid
versus the reciprocal of absolute temperature on semilog paper, one
should obtain a straight line from whose slope the heat of solution
can be computed. A comparison of this heat of solution with heats of
solution or heats of fusion for chemically similar compounds yields
a check on the validity of the experimental data. An analogous tech
nique applied to chemical kinetic data or chemical equilibrium data
would yield an activation energy or a chemical enthalpy change that
could then be compared wit literature values.
Table 1-1 has been prepared to summarize the methods for assess
ing the validity of those types of data most often encountered in com
pleting design projects for the chemical process industries. It is to
be emphasized that this table is not a summary of design methods
but merely a set of criteria by which to judge the quality of technical
8 Chapter 1
data to be used in carrying out a design. The references shown are
not mant to be comprehensive; moreover it is clear that the table
vastly oversimplifies the types of operations carried out in the
chemical industry as well as the theoretical and technical background
necessary for the completion of even a simple design problem.
Nevertheless. the information summarized has proved very useful in
applying the results of theoretical considerations to engineering prob
lems of practical significance. Naturally, in many instances it is
both desirable and necessary to supplement the elementary methods
described in Table 1-1 by using some of the more advanced the ore -
tical developments.
The utility of the information summarized in Table 1-1 naturally
varies significantly from one segment of the chemical process in
dustrv to another. For example, the organic chemical industry makes
great use of extraction and leaching processes, and the simple tech
nique of plotting solubility data on semilog paper to obtain a heat of
solution can prove to be of great and frequent utility. Once confidence
in the exerimental data is developed, the design calculations to op
timize the number of extraction or leaching stages can proceed quite
smoothly.
Wen theoretical correlation of process inormation is impossible,
the use of a factorially designed exerimental technique may be of
great value. The use of statistically designed experiments is parti
cularly valuable in reducing the required amount of exerimental
and analytical effort to solve a problem for which there is little or no
theoretical basis. The following references, arranged by Koehler,
provde an excellent introduction to the application of statistical con
cepts to a variety of problems encountered in the chemical industry.
Besides a discussion of statistical designs in the analysis of labora
tory and pilot plant data (7,23), the series also effectively presents
the advantages to be gained by the application of statistical techniques
to in-plant exerimentation (22), to the use of computers in data re
duction (41), to the selection of appropriate production-line control
charts ( 17), and to the general improvement of quality-control
methods (26). When a sound theoretical basis for a process design is
limited, the use of statistical methods can provide a highly useful
foundation for the necessary design and evaluation calculations.
One last point frequently overlooked is the need for preparing an
adequate error analysiS. If the probable error as computed for the
experimental technique used is approximately equal to the random
deviation of the data about a correlating line, then it can be assumed
with a high degree of confidence that all sources of error have been
properly established and accounted for. Such an analYSis lends a
great deal of confidence to the use of the data in a design, particularly
when the exenditure of a large capital investment is required.
ESTABLISHMENT OF THE CRTICAL DESGN PARMETER
SIMULATON AN OPTIMZATION
In his text, Sherwood (41) states quite appropriately that the de
signer must be willing to make assumptions. Once sufficient informa-
T
a
b
l
e

1
-
1
.

U
s
e
f
u
l

M
e
t
h
o
s

o
f

E
v
l
u
a
t
i
n
g

a
d

C
o
r
r
e
l
a
t
i
n
g

C
h
e
m
i
c
a
l

E
n
g
i
n
e
e
r
i
n
g

E
x
e
r
i
m
e
n
t
a
l

D
t
a

P
r
o
c
e
s
s

O
p
e
r
a
t
i
o
n

R
e
q
u
i
r
e
d

D
e
s
i
g
n

P
a
r
a
m
e
t
e
r

F
l
u
i
d

t
r
a
n
s
p
o
r
t
a
t
i
o
n

F
r
i
c
t
i
o
n

f
a
c
t
o
r

F
l
u
i
d

h
e
a
t
i
n
g

o
r

c
o
o
l
i
n
g

H
e
a
t
-
t
r
a
n
s
f
e
r

c
o
e
f
f
i
c
i
e
n
t

C
h
e
m
i
c
a
l

r
e
a
c
t
i
o
n

C
h
e
m
i
c
a
l
-
k
i
n
e
t
i
c

c
o
n
s
t
a
n
t

C
h
e
m
i
c
a
l
-
e
q
u
i
l
i
b
r
i
u
m

c
o
n
s
t
a
n
t

C
a
t
a
l
y
t
i
c

c
h
e
m
i
c
a
l

E
f
f
e
c
t
i
v
e
n
e
s
s

f
a
c
t
o
r

r
e
a
c
t
i
o
n

M
a
s
s
-
t
r
a
n
s
f
e
r

c
o
e
f
f
i
c
i
e
n
t

f
r
o
m

f
l
u
i
d

t
o

c
a
t
a
l
y
s
t

p
e
l
l
e
t

s
u
r
f
a
c
e

S
l
u
r
r
y

r
e
a
c
t
i
o
n

M
a
s
s
-
t
r
a
n
s
f
e
r

c
o
e
f
f
i
c
i
e
n
t

(
r
o
m

l
i
q
u
i
d

t
o

c
a
t
a
l
y
s
t

p
e
l
l
e
t

s
u
r
f
a
c
e

S
p
a
r
g
e
d

r
e
a
c
t
i
o
n

M
a
s
s
-
t
r
a
n
s
f
e
r

c
o
e
f
f
i
c
i
e
n
t

a
t

t
h
e

b
u
l
k

l
i
q
u
i
d

s
u
r
f
a
c
e

M
a
s
s
-
t
r
a
n
s
f
e
r

c
o
e
f
f
i
c
i
e
n
t

f
r
o
m

a

g
a
s

b
u
b
b
l
e

t
o

b
u
l
k

l
i
q
u
i
d

A
b
s
o
r
p
t
i
o
n

S
o
l
u
b
i
l
i
t
y

o
f

g
a
s

i
n

a

l
i
q
u
i
d

B
o
i
l
i
n
g

o
r

V
a
p
o
r

p
r
e
s
s
u
r
e

c
o
n
d
e
n
s
a
t
i
o
n

D
i
s
t
i
l
l
a
t
i
o
n

V
a
p
o
r
-
l
i
q
u
i
d

e
q
u
i
l
i
b
r
i
u
m

(
r
e
l
a
t
i
v
e

v
o
l
a
t
i
l
i
t
y
)

T
r
a
y

e
f
f
i
c
i
e
n
c
y

C
r
y
s
t
a
l
l
i
z
a
t
i
o
n

S
o
l
u
b
i
l
i
t
y

a
n
d

s
u
p
e
r
s
o
l
u
b
i
l
i
t
y

E
x
t
r
a
c
t
i
o
n

D
i
s
t
r
i
b
u
t
i
o
n

(
p
a
r
t
i
t
i
o
n
)

c
o
e
f
f
i
C
i
e
n
t

B
a
s
e
s

f
o
r

E
v
a
l
u
a
t
i
o
n

o
r

C
o
r
r
e
l
a
t
i
o
n

o
f

D
a
t
a

1
/
2

=

F
(
N
H
e
)

j

/

=

F
(
N
H
e
)

A
r
r
h
e
n
i
u
s

c
o
r
r
e
l
a
t
i
o
n

v
a
n
'
t

H
o
f
f

c
o
r
r
e
l
a
t
i
o
n

T
h
i
e
l
e

m
o
d
u
l
u
s

j
o

=

F
(
N
l
e
)

k
L

=

F
(
D
.

N
.
D

p
.
l
.

A
p
)

k
L

=

F
(
D
,

N

k
L
a

=

F
(
D
,

N
,

V
s
)

H
e
n
r
y
'
s

l
a
w

o
r

R
o
u
l
t
'
s

l
a
w

C
l
a
u
s
i
u
s
-
C
l
a
p
e
y
r
o
n

e
q
u
a
t
i
o
n

G
i
b
b
s
-
D
u
h
e
m

e
q
u
a
t
i
o
n

M
u
r
p
h
r
e
e

e
f
f
i
c
i
e
n
c
y

r
e
l
a
t
i
o
n
s
h
i
p

C
l
a
p
e
y
r
o
n
-
t
y
p
e

c
o
r
r
e
l
a
t
i
o
n

N
e
r
n
s
t
'
s

l
a
w

L
i
t
e
r
a
t
u
r
e

R
e
f
e
r
e
n
c
e
s

(
3
-
1
)

(
3
1
)

(
6
)

(
.
5
2
)

(
1
0
)

(
1
0
)

(
1
9
)

(
2
7
)

(
5
)

(
2
1
)

(
5
2
)

(
3
7
)

(
3
7
)

(
U
)

(
1

1
)

-

c

<

- -
.

g

10 Chapter 1
tion is available to analyze a process completely and accurately,
typically the financial incentive for completing the design will be sub
stantially diminished. However, i order to mae appropriate pre
liminary assumptions, the designer must develop the capability of
isolating the variables that are critical to the determination of the
over-all economic performance of the process. Frequently a pre
liminary design calculation is required to establish the identity of
the most important design variables. Such a case is illustrated in
Chapter 4, in which the hydrogenation of benzene is considered.
For the more common process operations, the critical variable
frequently is well known; e.g. , in the design of a distillation column,
the reflux ratio usually serves as a m0st sensitive index of the pro
cess economics. I the first analysis of a unique or a highly complex
design, the determination of the most critical variables often is left
to the judgment of the engineer. Typically, when a multivariable de
sign is approached, a base case is selected by arbitrarily establish
ing "reasonable" values for many of the process variables that are
believed to be least critical in determining the process economics.
The variables thought to be most critical are allowed to vary, and a
preliminary economic evaluation and optimization of the design are
completed. Then variations from the base case are considered by
allowing variations in the parameters that had previously been fixed.
As show in Chapter 7, in which the economics of styrene production
is considered, the use of a digital computer can greatly facilitate the
evaluation of the base case and the variations from this case that are
significant. As illustrated in the same chapter, it is essential to re
turn to the original set of assumptions in order to establish the effect
of each assumption on the over-all economics of the process. It is
also important to realize that in dealing with a multi variable design
problem, several local minima may exist; a certain amount of judg
ment must be exercised in determining whether a local or an over
all minimum has been achieved.
The calculation procedure described suffers from the lack of an
organized approach to the problem, and a great many decisions must
be reserved for the judgfnt of the engineer. In certain types of de
sign problems, a more quantitative approach to the logic required for
a design calculation may be achieved by applying the technique of
linear programming. This method has application in design cases
that result in linear algebraic relationships. Happel's very useful text
on chemical process economics (18) includes a brief description and
an example of linear programming techniques. It is important to note
that many chemical economics problems are highly nonlinear and that
in such instances linear programming techniques are not directly
applicable. However, by linearizing the appropriate analytical expres
sions, linear programming techniques may be helpful in establishing
the general nature and relative importance of the cost functions under
study.
A significant portion of modern chemical engineering research
has focused on the development of advanced methods for process
optimization as applied to individual sections of a process as well as
to the over-all design result. Particular emphasis has been placed
Introduction 11
on the optimization of chemical reactor systems; much of this work is
well summarized by Aris (5, 6), Denbigh (15), and Kramers and
Westerterp (28). One of the more important of modern optimization
techniques is that of dynamic programming, i which the last of a
series of staged operations is first optimized; the last two stages are
then optimized as a single unit. One stage is added on in a stepwise
fashion, working backward until the optimum conditions for the entire
process are established. This method has obvious application in the
aalysis of continuous stirred-tank reactor systems (5 ). The cal
culus of variations has been shown to have particular value in the op
timization of tubular reactors; the method is especially useful in the
case of adiabatic reactions (15). In addition to the aforementioned, the
method of "steepest ascent" has also proved useful i the optimization
of reactor systems (20). This technique allows one to approach an
optimum as closely as desired by successive quantum changes in the
process variables until the design parameter in question (e.g. , yield)
is optimized.
Concurrent with the development of optimization methods has been
the establishment of new techniques to simplify and minimize design
computational effort. Rosen (38) has presented a machine computa
tion method for performing process material balances, and Rvicz
and Norman (36) have reported a more flexible program incorpora
ting both heat and material balances. Sargent and Westerberg (39)
have developed a general-purpose approach that is very helpful in
organizing the programming work required for the computer simula
tion of a chemical process.
More recently, Lee, Christensen, and Rudd (29) have made a very
significant contribution in the study of multi vriable design problems
where assumptions are required to proceed with the design. These
investigators have established a method that allows the preassign
ment of values to a selected set of design variables in such a way as
to minimize the computational effort. This work is one of the first
signs of a trend in the chemical engineering profession to make more
effective use of computers in the analysis of process design problems.
EVALUATON OF PROCESS ECONOMCS
When the chemist undertakes a fundamental research project
in the laboratory, he frequently has only a limited notion of the ulti
mate applications and possible financial success of his product. This
sort of freedom has proved invaluable in developing a creative atmos
phere for the development of new products. However, once a chemical
project emerges into applied research or development, each subsequent
step in the project should be evluated finanCially before the work for
that stage is undertaken. In this manner, projects having a low proba
bility for an acceptable finacial return can be " weeded out" at an
early stage, and more technical effort can be applied to developing
those products which are most likely to realize the greatest economic
return.
One of the critical fWCtions of the design engineer tor any other
person responsible for evaluating the process economics) is to keep
12 Chapter 1
abreast of the latest technical developments in a new product and to
interpret these developments in the light of their ultimate effect on
the profit potential . For example, if the production yield of a new
product is increased by carrying out the reaction step i a particular
solvent, the cost for separating the product from the solvent must be
estimated. It may be that the cost of recovering the product over
shadows the savings generated by the use of the solvent. I such is
the case, the course of the early process research and development
for the product will have to be altered accordingly.
The actual form of a financial evaluation varies widely within the
industry; and because of the obvious commercial Significance of evalu
ation techniques, little meaningful information is available i the
literature. However, some general references are available which
have proved very useful in orienting chemists and engineers to the
problems associated with a financial evaluation.
I first approaching an economic evaluation, one of the major dif
ficulties encountered by the technically trained person is that of
understanding financial terminology. The work of Beattie and Vivian
(10) greatly mitigates this problem; it provides a detailed compilation
of the definitions for most terms common to financial analysis as
applied in the chemical industry. In addition, the importance of using
consistent terminology is well illustrated.
In its simplest form, the financial evluation of a chemical project
provides an estimate of the capital required for the construction of
the manufacturing facility and a forecast of the costs required to
operate the proposed process. By deducting the total operating costs
and income taxes from the anticipated sales revenues, the net opera
ting income for the project can be established. In addition to the an
nual dollar income volume, profitability may also be exressed as
the fraction of the capital ivestment needed to establish the antici
pated return. The philosophy giding the analysis of corporate ven
tures in the chemical industry is well summarized in an A.I.Ch. E.
publication (2).
Each company has its own peculiar raw-material position to pro
tect, its own process know-how, its own accounting system, and its
ow plans for future exansion. As a result, the identical proposed
project may meet different fates, depending upon the company that
considers it. For example, consider an American company that is
domestically selling products that are facing competition from identi
cal products manufactured in Europe. The European concern might
be accounting for the cost of its goods on an incremental basis, Le., a
basis in which depreciation and overhead costs are allocated entirely
to the domestic portion of production while the production deSignated
for export is burdened only with direct expenses. Clearly in such a
situation the European firm might find itself in a very advantageous
competitive situation even after the transportation costs are taken
into account. Thus a difference in cost-accounting techniques can
cause a vast disparity in the commercial market place. Similar ex
amples are available where difference in raw-material position and
process know-how lead to substantial commercial implications.
Introduction 13
Some of the more standard methods which may be used in account
ing for chemical project costs are elucidated by several texts (ls, 35,
43,50). It should be noted that each company has its own required
minimum return on investment for a project; for obvious reasons
these figures are necessarily held in confidence. However, in the ab
sence of external circumstances (e.g. , raw-material position), an
estimate of 8 to 10 per cent is probably a realistic lower limit on the
return on invested capital required for project apPJoval. Much higher
return percentages are obviously desirable and are frequently ob
tained.
The mathematics involved in evaluating the economics of a chemi
cal project becomes quite involved when the current values of the
various cash flows into and out of the project are considered. In a
significant article, Souders (47) has produced an exemplary discus
sion of this issue together with an interesting comparison of the ef
fects of various profitability criteria on the ultimate investment de
cision. For the reader interested in a more complete background i
the fundamentals of engineering economics, this paper also includes a
short but useful bibliography of definitive works in this field.
One of the great difficulties in providing an accurate financial eval
uation lies in the fact that both the market price and the demand for
individual products as well as the general development of the national
economy are dynamic functions. The financial performance of a plant
should be examined as a function of both short-term, high-frequency
variations and of long-term, gradual growth or decline in the dollar
volume of product sales. Schenk (,12) presents a brief discussion and
a useful example of such a study. He pOints out that the economics of
a chemical project are most strongly affected by variations in the
selling price. Following after the selling price, and in order of im
portance, variations in the sales volume, sales exenses, and capital
investment have been found the most critical factors afecting the re
turn on investment. The need for adequate price and sales forecasting
is well illustrated by his discussion. An examination of the inluence
of variations in selling price, sometimes called a risk analysis, is
frequently a required component in the final presentation of the finan
cial evaluation of a new chemical project.
Twaddle and Malloy (19) provide a useful discussion of the effects
of long-term demand, selling price, and capacity variations on the
economic return to be antiCipated for a given chemical project.
Various methods of graphically illustrating the economic performance
of a plant are well illustrated in this reference. I particular, it is
effectively demonstrated that accounting for time variations in demand
and price has a dramatic effect on the optimum plant size and may
radically influence the decision whether the plant should be built at
all.
A major deterrent to the proper financial evaluation of preliminary
designs arises from the paucity of reliable capital-cost information
for process equipment. Chilton's excellent compilation (12) provides
a useful background to the problems of cost estimating and to the
proper methods of cost data correlation. The correlations contained
in his text are quite adequate for preparing preliminary cost esti-
14 Chapter 1
mates; however, it is recommended that other sources (-1,50) be con
sulted in order to verify cost estimates. In particular, the price
estimates for the larger pieces of equipment should be checked by
using two or three ditferent references. When the design reaches its
final stages, it is normal practice to contact suppliers in order to
verify the estimates for all Significant pieces of equipment.
PRESENTATION OF DESIGN RESULTS
Regardless of the quality of the technical work that has contributed
to the design and evaluation of a chemical project, the total effort may
be valueless unless it is properly presented to those who must make
the ultimate investment decision. I particular, the need for lucid
technical writing has long been recognized as an important issue, and
several excellent texts are available for this purpose (25,33,48).
Generally, a lack of quality in a technical report can be traced to a
lack of effort in preparing and polishing the report rather than to a
lack of knowledge of grammar, style, or proper report organization.
Often the results of a technical effort must be presented orally.
For the speaker who must summarize a mass of technical informa
tion in a short time, the American Institute of Chemical Engineers has
prepared an excellent booklet (3) which summarizes the important
points to remember in preparing the presentation. It is particularly
valuable in pOinting out the most effective m2thods of using slides
and other visual aids. The texts of Atwood (8), Flesch (16), and Weaver
(51) serve as valuable references to the more general aspects of pre
paring and executing oral presentations.
SUMMRY
The foregoing sections provide a brief introduction to some of the
more important issues that confront the industrial deSign engineer
in the practice of his profession. Some of the pOints discussed sug
gest various types of formal training necessary to the successful
execution of design work. Other issues suggest specific approaches
to design problems which have been found useful. In particular, the
appropriate application of electronic computation equipment to aid in
the design calculations requires more than ever before the use of
engineering talents to comprehend and analyze the computed results.
The following chapters provide a series of design case studies where
in the important aspects of industrial design practice are brought to
bear on realistic problem situations having commercial Significance.
NOTATION
a Interfacial surface area
D Diffusion coefficient
D p Catalyst pellet diameter
f
Friction factor
jD Mass-transfer j-factor
jJ Heat-transfer j-factor
IlL Liquid-phase mass-transfer coefficient
N Rotational speed of impeller
NRe Reynolds number
Vs Superficial gas velocity
J Dynamic viscosity
Ap Density difference between solid and liquid
REFERENCES
Introductioll 15
1. American Chemical Society, Chemical Marketing in the Competi
tive Sixties, Advances in Chemistry Series, No. 24, Washington,
D.C. (959).
2. American Institute of Chemical Engineers, Venture Analysis,
A.I.Ch.E. Publication Department, New York960).
3. American Institute of Chemical Engineers, Guide for Writers ad
Speakers, A.I.Ch.E. Publication Department, New York (196
4. Aries, R. S., and R. D. Newton, Chemical Engineering Cost Esti
mation,McGraw-Hill Book Co., Inc., New York (1955).
5. Aris, R., The Otimal Design of Chemical Reactors, Academic
Press, New York (1961).
6. Aris, R., Introduction to the Analysis of Chemial Reactors,
Prentice-Hall, Inc., Englewood Cliffs, N.J. (1965).
7. Atkinson, A. C., Cher. Eng. 73, No. 10, 149 (1966).
8. Atwood, R. L., When You Tal, Atwood Corporation, Melrose,
Mass. (1959).
9. Bare,B.M.,Chem. Eng.P61,No.10,26 (1965).
10. Beattie, R. D., and J. E. Vivian, Cher. Eng. 60, No.1 (1953), re
printed in reference 12, p. 24.
11. Bradley, J. W., Cher. Eng. Progr. 61, No. 10, 15 (1965).
12. Chilton, C. H., Cost Engineering in the Process Industries,
McGraw-Hill Book Company, Inc., New York (196).
13. Craver, J. K., Cher. Eng. Progr. 61, No. 10, 24 (1965).
14. Denbigh, K. G., The Principles of Chemical Equilibrium, Cam
bridge University Press, Cambridge (1961).
16 Chapter 1
15. Denbigh, K. G., Chemical Reactor Theory, Cambridge University
Press, Cambridge U5).
16. Flesch, R. The Art of Plain Talk, Harper Brothers, Inc., New
York (1946).
1 7. Freund, R. A., Cher. Eng. 73, No.3, 70 (1966).
18. Happel, J., Chemical Process Economics, John Wiley & Sons, Inc.,
New York (1958) .
19. Harriott, P.,A.I .Ch.E.J.8, 93 ( 1962).
20. Horn, F.,and U. Troltenier, Cher. Ing.-Tech. 32, 382 (1960).
21. Hougen, O. A., K.A. Watson, and R. A. Ragatz, Chemical Process
Principles, Part I, John Wiley & Sons, Inc., New York (1954).
22. Hunter, J. S., Cher. Eng. 73, No. 7, 111 (1966).
23. Hunter, W. G., and A. C. Atkinson, Cher. Eng. 73, No.12 , 159 (1966).
24. Kennel, W. E., Cher. Eng. Progr.61, No.10, 20 (1965).
25. Kobe, K.A.
,
Cemicl Engneering Reports: How to Search the
Literature and Prepare a Report, Inter science Pblishers, Inc.,
New York (1957).
26. Koehler, T. L., Cher. Eng. 73, No.1, 81 (1966).
2 7. Kozinski, A. A., and C. J. King, A.I.Ch.E.J. 12,109 (1966).
28. Kramers, H., and K. R. Westerterp, Elements of Chemical Reactor
Design and Oeration, Academic Press, New York (1963).
29. Lee, W., J. H. Christensen, and D. F. Rudd, A.I.Ch.E.J.12, 1104
(1966).
30. MacMullin, R. B., Cher. Eng. Progr.44, No. 3,183 (1948).
31. McAdams, W. H., Heat Transmission, McGraw-Hill Book Company,
Inc., New York (1954).
32. Mellon, M. G. Searching the Chemical Literature, American
Chemical Society Publications, Washington, D.C. (1964).
33. Nelson, J. R., Writing the Technical Report, McGraw-Hill Book
Company, Inc., New York (194 7).
34, Perry, J. H., Chemical Engineers' Handbook, McGraw-Hill Book
Company, Inc., New York (1950).
35. Peters, M. S., Plant Design and Economics for Chemical Engin
eers, McGraw-Hill Book Co., Ic., New York (1958).
36. Ravicz, A. E., and R. L. Norman, Cher. Eng. Progr. 60, No. 5, 71
(1964).
37. Robinson, C. S., and E. R.Gilliland, Elements of Fractional Dis
tillation, McGraw-Hill Book Company, Inc., New York m
38. Rosen, E. M., Cher. Eng. Progr. 58, No. 10, 69 (1962).
Introduction 17
39. Sargent, R. W. H., and A. W. Westerberg, Trans. Inst. Chem. Eng.
42, T190 ( 1964).
40. Satterfield, C. N., and T. K. Sherwood, The Role of Diffusion in
Catalysis, Addison-Wesley Publishing Co .. Inc . Reading. Mass.
(1963).
41. Savitzky, A., Chem. Eng. 73, No.5, 99 (1966).
42. Schenk, G., Chem. Eng. Progr. 61, No. 10, 16 ( 1965).
43. Schweyer, H. E., Process Engineering Economics, McGraw-Hill
Book Company, Inc., New York (1955).
44. Sherwood, T. K., A Course in Process Design, The M.I.T. Press,
Cambridge, Mass. (1963).
45. Sideman, S., O. Hortassu, and J. W. Fulton, Ind. Eng. Chem. 58,32
( 1966).
46. Smith, J. M., Chemical Engineering Kinetics, McGraw-Hill Book
Company, Inc., New York (1956).
47. Souders, M., Chem. Eng. Progr. 62, No.3, 79 (1966).
48. Souther, J. W., Technical Report Writing, John Wiley & Sons, Inc.,
New York (1957).
49. Twaddle, J. J., and J. B. Malloy, Chem. Eng. Progr. 62, No. 7, 90
(1966).
50. Vilbrandt, F. C., and C. E. Dryden, Chemical Engineering Plant
Design, McGraw-Hill Book Company, Inc., New York (1959).
51. Weaver, R. M., The Ethics of Rhetoric, Henry Regnery Co.,
Chicago, Ill. (1953).
52. Weber, H. C., and H. P. Meissner, Thermodynamics for Chemical
Engineers, John Wiley & Sons, Inc., New York (1957).

2. Reactor Design, Optimization, and
Control in the Production of
Monochlorobenzene
Ths cusc dccIs tIh Ihc dsgn oJ u j+occss /o+ Ihc chIo+icIoii o)
bizi. In!uI ciilmss s ]Iucd on I|ic +oc+ co++InIon o]
Iubo+uIo+ dnIc IhuI dcsc+bc Ihc lincIcs o] I|ic Ih+ s+qucnIuI
chIo+nc!oii +ncIions. 3c:+uI +occss schciii+s, iii c!udng boIh buIch
cnd coiiIiiitous oc+uIons, uc cxcmiicd um| cnch s oIiiizd ccono-
mcnII. Thc mosI u!I+ucIc +ocss s scIcIcd b| iicuiis oJ :umin-
iiig IIc +suII s o] I|ic /nuncuI c:ItuIoii. IiicII, Ih Ihc+mcI s!cbI-
I) oJ /lic oIiiitIi :occss is n:sIgcIcd ciid /hc Jsigii s iiod/d
u]+o+nIcI ii o+d+ Io cisu+c s!nbIc oc+c!on.
TIic c|nIc+ +o:idcs ui o]]o+Itn! ]o: I|ic n:iIon uiid cucIss
o/ t+oits ]+occss sc|cmcs. 3cic+cI cIIeItcI: dcsgns u+ u:uIcbI
tt|iic|i mgJI o//c+ cc+Iun ud:iiIugcs o:+ !Iosc cuscs coiisidc+cd in
I|c +cscii! IxI. Thcsc cIIc+iuIcs qui!c +ocI could /o+iii !|c
bcss /o+ JitI|ic+ dsigns cid ccoiioii!c c:/itnIoiis. T|ic ]]cIi:+ncss
oJ IIis c|m!c+ |iigcs oii II dc!c+:inu!oii o] Ihc c]]+o+iuIc dgn
n+ciiicI+ sucu /huI I|c ccoiioncs oJ t+oits Is o] ++ucIo+ dcsgiis
cnii bc coiiic+cd i u mcuiiug/uI un).
Aja Pharmaceutical Company
Secaucus, New Jersey
To: J. A. Smith, Engineer
!roH: W, Allen, Manager of Technical Operations
Ajax has often thought of maing its own monochlorobenzene in
stead of purchasing this intermediate from the outside. Since the
"Chlorpan-A" production is now shut down, there are three reactors
that might be made available for chlorinating benzene, You are re-
18
Reactor Desigl l , OjJtilllizatiOll, alld COl/trol 1
quested to evaluate the feasibility of this proj ect under the assump
tion that both benzene and chlorine would have to be purchased.
The three reactors are equipped with coils for heating and coolig
and with gas distributors to assure that the chlorine is well dispersed
so as to keep the liquid saturated with chlorine. The reactors are
necessarily operated at atmospheric pressure. Each has a reflux
condenser, a chlorine gas recycle system, an agitator, and 550 sq ft of
cooling surface in the form of a double helical coil of O. 5-in. o. d.
tubing. With the agitator in operation, over-all heat transfer coef
ficients of 1 50 Btu/(hr) (sq ft) (OF) have been obtained for benzene at
60C with cooling water in the coil . During operation, with chlorine
being fed, each reactor will hold 400 gallons of liquid. Adequate dis
tillation capacity is available to separate the unconverted benzene and
the chlorinated products: technical -grade monochlorobenzene and a
miture of di- and trichlorobenzenes. It is anticipated that this latter
miture cannot be sold and that it will have to be di scarded.
The kinetics of the liquid-phase chlorination of benzene have been
investigated by the Research Department; their results confirm the
literature reports that the chlorination of benzene, monochlorobenzene,
and dichlorobenzene is in all cases first -order and irreversible, based
on rate equations in which mole fractions (instead of moles per unit
volume) are employed. The laboratory tests were carried out in a 2-
liter flask equipped with a reflux condenser that returned everything
but excess chlorine. The liquid was well agitat ed by the bubbling
chlorine, which was fed through a fritted glass bubbler. Small amounts
of ferric chloride were used as a catalyst.
The results of the laboratory tests are summarized as follows:
Mole Fraction
Time Mole Fraction Benzene NGnDCh1C1CDCD2EDE
~
(hrs) 40"C "C T0"C 40"C 55C 70C
0 I I 1 0 0 0
0. 3 0.3 0. 35
0. 5 0.4 0. 49
1. 0 0.91 0. 0.2I 0.09 0. 32 0. 60
2. 0 0.3 0.44 0.0 0.I 0. 53 0. 51
4. 0 0. 70
6.0 0. 56 0. 09 0. 43 0. 75
12. 0 0. 3 1 0. 63 0. 59
Evidently, the formation of trichlorobenzene is so small that it may
be neglected.
The possibility of using the three existing reactors for batch
chlorinations has already been examined, and it seems clear that
some form of continuous operation would be preferable. Your study
Z Chapter Z
should be restricted to continuous operation for a maximum of 7000
hrs/year. The maximum operating temperature is 70C; higher tem
peratures lead to excessive vaporization and strain the capacity of
the condensers. We need 8 million lbs of monochlorobenzene per year;
we have no way of selling additional production at more than an un
acceptable distress price.
The attached cost figures are approximate but should be adequate
for the purpose of your analysis. Specifically, you are asked to evalu
ate the possible financial return to be achieved by producing mono
chlorobenzene. U addition, you are to specify the optimum manner in
which the process should be run.
Data 0Assumptions
Chemical costs and net sales and use values:
Chlorine
Monochlorobenzene (b. p. 1 32C)
Dichloro-trichlorobenzene mix
(b.p. 1 74-208C)
Benzene (b.p. 80C)
$0. 045/lb
$0. 1 05/lb
negligible
$0. 034/lb
Recovery costs (total, including both fixed and operating costs) :
Monochlorobenzene
Benzene (s.g. O. 88)
Dichloro- trichlorobenzene mix
Oerating costs, each reactor:
Reactor heating and/or cooling
Labor and supervision
Chlorine recycle
$0. 01 0/lb
$0. 01 5/lb
zero (waste)
$2 . 40/hr
$2. 90/hr
$4. 00/hr
Total fixed charges on each reactor, including piping, controls, and
auxiliaries: $14, OOO/year.
As an approximation, assume that the molal volume of each of the
products is the same as that of benzene and that there is no volume
change on mixing these liquids.
Physical Data
Product
Benzene (liquid)
Monochlorobenzene
(liquid)
Dichlorobenzene
(liquid)
Chlorine (gas)
HCI (gas)
L
|
Btu/{lb mole) (I) Al Formation at 2 5C
33. 4
3 5. 6
3 8. 4
8. 1
6. 96
1 1 , 700 gcal/mole
2, 500
-4, 900 (average)
-22, 063
Reactor Design, Optimization, and Control Z1
Sggested Nomenclatre:
/ Heat-transfer area available in each reactor, sq ft
a Cost for benzene converted to MCB, dollars/year
Cost for benzene converted to DCB, dollars/year
L Cost for chlorine converted to MCB, dollars/year
d Cost for chlorine converted to DCB, dollars/year
e Cost for separating and recycling benzene, dollars
/
year
J
Direct operating costs for the three reactors, dollars/year
)
Flow rate to a reactor, lb mol es/year

Cost of recovery of MCB, dollars/year

/ Fixed charges on the three reactors, dollars/year
k
1
Reaction- rate constant for monochlorination reaction, hour-1
k
2 Reaction- rate constant for dichlorination reaction, hour
-
1
I
g
_ k_.Y_/ ) with _ evaluated at temperature Tn' dimensionless
I
2n /2
N
R
/ )
,with /'2 evaluated at temperature T_,dimensionl ess
I_ R/
)
dimensionless
N
R
Moles in each reactor, lb mol es
0,
Sensible heat absorbed by the cold entering benzene feed stream,
Btu/hr
{
c
Rate of heat transfer to the cooling water in a reactor, Btu/hr

Total rate of heat generated by the chemical reactions in reactor

n, Btu
/
hr
Recycle rate of benzene, lb moles/hr
l Time, hr
T Temperature, oK
T
e
Temperature of cooling water, C
T
n
Temperature of reactor n, C
T
R
Temperature in a reactor, C
Over-all heat transfer coefficient, Btu/(hr) ( sq ft) (OF)
7 Mole fraction of benzene in reactor n, dimensionless

Mole fraction of monochlorobenzene i n reactor n, dimensionless

x__ Mole fraction of dichlorobenzene i n reactor n, dimensionless
tI Enthalpy of chlorination reaction, cal/ g mole of chlorine
tT T
R
-T
e
,oC
J Oerating time, hr/year
ZZ
Chapter Z
PRELIMNAY DESIGN ANALYSIS
The reactors are to be operated continuously; it is therefore clear
that the flow rate to be employed will be an important operating
variable. The reactors may be operated with series flow through the
three, or piped for two, or all three, in parallel. Recycled benzene can
be returned to the first reactor or distributed to all three. Each
reactor may be operated at a different temperature.
The reactor system will be desiged by first examining several
diferent process schemes where each reactor is operated at the
same speCified temperature. Once the proper process flow pattern
has been established, variation in the operating temperatures of each
reactor will be considered. Finally, the stability of the optimum
reactor system will be examined.
At first thought, one might conclude that the maximum permissible
temperature should be employed in order to obtain high reaction
rates. The best choice of temperatures will be considered after the
laboratory data have been analyzed and the best reactor flow pattern
has been selected.
ANALYSIS OF THE LABORTORY KNTC DATA
Before studying the different possible reactor flow plans, the
chemical kinetics for the chlorination reactions must be derived by
referring to pertinent literature references [b) and by examining the
available laboratory data. The chemical reactions to be considered
are
C6H6 ' Cl2 _ C6H5CI HCI
C6H5CI + Cl2 * C6H4Cl2 * HCl
C6H4Cl2 Cl2 _ C6H3Cl3 + HCI
The original memorandum indicates that the rate equations for these
reactions may be written as
GX_
l:_x_ ( 2. 1 )
J/
''
= k

k
;^
( 2. 2)
d/
Formation of trichlorobenzene is to be neglected, and the mole frac
tion of dichlorobenzene, __ will be assumed equal to A

The laboratory-batch data are to be compared with integrated

forms of the two rate equations, in order that the rate constants may
be evaluated for use in optimizing the operation of the three reactors.
Reactor Design, Opti mization, and Control 23
The desired relations between

,
@
_ and t are easily obtained:
CX

= k
_

"
<

(x_ = 1 . 0 at t = 0)
dl
Solving this last, with the conditions
@ =
0 at t = 0:
k

"
1
(<

c)
`1
^
4
(
2 . 3)
(2. 4)
Values of
^
_ are most easily obtaied from the laboratory data by
plotting x

vs. l on a semilogarithmic graph and drawing best straight

lines through the pOints for each of the three temperatures.
Even with k_ known, ther e is apparently no way to plot the data on
7

so as to evaluate k
4
by drawing a best straight line through the
exerimental points. Accordingly, Eq. 2.4 is solved by trial to obtain
values of R corresponding to each data poit, and these are simply
W
0
M C.I

C.|00l
30 80
8U8f0lUf8

L
J
O
W
Figure 2-1 . First-order kinetic con
stants for benzene chlorination.
(Symbol s L and. represent values
computed from exerimental data. )
Z4 Chu]Ic: 2
averaged a
t
each temperature. The rate constants so obtained are as
follows:
40C 5 5C 70C
_ hr-
1
0. 0965 0. 41 2 1. 55
k
2
, hr-
1
0. 0045 0. 055 0. 4 5
Figure 2 -1 shows these results graphically and may be used t o inter
palate between data pOints. Figre 2 -2 is a check on the calculations;
the data are compared with curves calculated by the use of Eqs. 2 . 3
and 2. 4 and the tabulated rate constat s.
l
1Im6 hOurS
|Z l9 lb
Figure 2 -2. Concentration profiles for chlorina
tion of benzene. (Solid lines are computed
from Eqs. 2. 3 and 2. 4. )
Legend
Symbol Mole Fraction Temperature, OC
L 7
A
40
7
A
5 5
A 7
A
70
7
8
40
^ 7
8
5 5
7
8
70
SELECTION OF A REACTOR FLOW
Five different flow plas will be studied, employing the three
reactors given in the problem statement to determine the optimum
Reactor Desigll, Optimizati oll , und Control 3
production costs. These process cases are:
I. Three reactors i series, recycle to the first
2. Three reactors i series, recycle to the second
3 . Three reactors i series, recycle to the third
4. Three reactors i n parallel
5. Batch operation
The batch process will be studied so as to compare its optimum pro
duction costs with the production costs for the four continuous opera
tions. The five process plans listed here are not the only possible
alternatives. However, by first investigating these several cases it
should be possible to determine those type s of process schemes which
merit further consideration.
The combination of sparged chlorine and mechanical agitation may
be assumed to maintai the liquid well stirred in each reactor and to
j ustiy the assumption that the composition of the effluent stream i s
the same as that of the reactor contents. Each reactor i s a CSTR
( continuous stirred tan reactor) , and the operation can be analyzed
by the methods outlined by Denbigh ) and described in several
chemical engineering texts. Since the volume change on reaction i s
very small, the total moles I_ present i n each reactor may be as
sumed to be the same and constant.
To permit the selection of an optimum flow pattern each reactor
will be assumed to operate at 55C. The effects of variation i tem
perature will be inve stigated at a later stage.
Case I: Tree Reactors i Series with Recycle to te First
The first scheme to be analyzed i s represented by the flowsheet
of Fig. 2-3 . In order to achieve the most useful result, the process
show i n thi s figure will be analyzed i n a generalized fashion. The
general results may then be applied easily to any specific situation.
W6ze recjcM
!t z
T

at az
mtm

sz ^0

f0

Cbw Cb|otm Cbkr

Figure 2 - 3 . Block flowsheet of continuous stirred
reactor system for the chlorination of benzene.
0cbkto- 0M
trcb|oo~
w6ze6e
Z Chapter Z
Let F represent the rate (lb moles/hour) of feed of benzene to the
first reactor, including recycle. The analysis then proceeds as fol
lows:
First Reactor

7 "

@
J
"
1

:1
where

]
"
~ !
g
W
8
/ !,with k
g
based on the temperature 1
"
in the nth
(in this case, the first) reactor.

1
~
t \
1
@_ ^;

|
11
~

(1 K

1
|
(1
7 ";'
Second Reactor

1
X

q
:2
-
1 +K1
2
(
1
+Ku) (
1 +K
1
2
)
\
;
~
|
~
^
8
\

,; ;

;
|
I

;
^;
=

(1 |
(1

;
|
(1

;;
|
Third Reactor
^
,
I

1
( 2. 5)
Reacto1' Dcsi, O]limzcl.`oii,and ConLml 27
all three reactors are operated at the same temperature
(T
1
7 T
2
T3) , these exressions for x
A
3 and
x
B
3 reduce to
and
1

-
[1 +K
1
)
3
Costs
(2. 6)
(2. 7)
( 2. 8)
The annual costs for the required production rate of 8 million
lbs/year of monochlorobenzene are expressed D terms of % and 7

400 K 62. 3 ^ 0. 88
_
37 6 lb

- . mo es
7. 48 K 78
J
8 Y 1 06 71 , 000
--
-
hr/year
1 1 2. 6 FXB3 F
XB3
where J must be less than or equal to 7000.
( a) Benzene for monochlorobenzene:
8 Y 1 06 Y 78
1 1 2 . 6
K 0. 034 = $1 89, 000/year
[b) Benzene for dichlorobenzene: $1 89, 000
X
C
3
!year
x
B
3
38 Cltc[Icr 3
(c) Chlorine for monochlorobenzene:
8 K 10
6
Y 71 Y 0. 045

7 $227, 400/year
1 1 2 . 6
[J) Chlorine for dichlorobenzene:
$227, 400 Y 2 (

s
) = $454, 800
C3
/year
^

s ^

s
(e) Benzene recycle costs:
8 K 10
6 Y 78 (x
s
)
+s
/ - 0. 015 = $83, 100 - year
1 1 2. 6
^

s ^

s
() Operating costs, direct:
( 2. 40 + 2 . 90 + 4. 00) Y 3 Y J = $27.90 e/year
(-) Monochlorobenzene recovery:
8 K 10
6
Y 0. 01 = $80, ODD/year
(/)
Fixed charges: 3 Y 14, 000 = $42, ODD/year
By summing items a through h the total annual operating cost for
this process is established; this parameter is a function of J 1 T
and T_ The total costs will evidently pass through a minimum as the
feed rate is varied. At low flow rates, with J approaching 7000, the
conversion is high and the cost of chlorine and benzene for the un
wanted dichlorobenzene becomes large. At very high flow rates the
conversion is low and the cost of benzene recycle becomes excessive.
For the particular case when all the reactors are operated at 55C,
Eqs. 2 . 7 and 2. 8 may be applied. U this instance:
(0. 412) ( 37. 6) 1 5. 5
[
1
= 7
! ! (2. 9)
( 0. 055) (3 7. 6) 2. 07

4
7
! ! ( 2. 1 0)
Using Eqs. 2. 7 to 2 . 1 0, calculations have been carried out in order to
optimize Case I when all the reactors are operated at 55C. The re
sults are summarized in Table 2-1 .
Table 2-1 indicates that a minimum annual cost is achieved for
Case I when the feed rate is about 55 Ib moles
/
hr and the reactor sys
tem is operated approximately 2660 hrs/year. It is to be noted that
the minimum in the total cost curve is a very broad one. Therefore,
T
a
b
l
e

Z

1
.

R
e
s
u
l
t
s

o
f

t
h
e

D
e
s
i
g

C
a
l
c
u
l
a
t
i
o
n
s

f
o
r

C
a
s
e

D
e
s
i
g

p
a
r
a
m
e
t
e
r
s

!

1
0

2
0

3
0

4
0

5
0

1
1

1
.
5
5

0
.
7
5
5

0
.
5
1
6

0
.
3
8
7

0
.
3
1
0

'
;

0
.
2
0
7

0
.
1
0
3

0
.
0
6
9

0
.
0
5
1
8

0
.
0
4
1
4

^
A
3

0
.
0
6
0
3

0
.
1
7
9

0
.
2
8
8

0
.
3
7
4

0
.
4
4
6

B
3

0
.
5
8
6

0
.
6
5
2

0
.
6
1
4

0
.
5
6
0

0
.
5
1
0

0
g

0
.
3
5
4

0
.
1
6
9

0
.
0
9
8

0
.
0
6
6

0
.
0
4
4

0

1
2
,
1
0
0

5
,
4
4
0

3
,
8
6
0

3
,
1
7
0

2
,
7
8
0

C
o
s
t
s
,
d
o
l
l
a
r
s
/
y
e
a
r

a
+
c
+
g
+
h

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

4
9
,
0
0
0

3
0
,
2
0
0

2
2
,
3
0
0

1
6
,
3
0
0

d

1
1
8
,
2
0
0

7
2
,
6
0
0

5
3
,
6
0
0

3
9
,
2
0
0

2
2
,
8
0
0

3
9
,
0
0
0

5
5
,
5
0
0

7
2
,
7
0
0

J

1
5
2
,
0
0
0

1
0
8
,
0
0
0

8
8
,
5
0
0

7
7
,
6
0
0

T
o
t
a
l

8
8
0
,
4
0
0

7
8
8
,
2
0
0

7
5
8
,
3
0
0

7
4
4
,
2
0
0

6
0

7
0

1
0
0

0
.
2
5
9

0
.
2
2
1

0
.
1
5
5

0
.
0
3
4
5

0
.
0
2
9
6

0
.
0
2
0
7

R

0
.
5
0
2

0
.
5
4
9

0
.
6
4
9

0
.
4
6
4

0
.
4
2
4

0
.
3
3
8

~

"

0
.
0
3
4

0
.
0
2
7

0
.
0
1
3

C

2
,
5
4
0

2
,
4
0
0

2
,
1
0
0

>
`

~
5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

.

~
.

1
3
,
8
0
0

1
2
,
1
0
0

7
,
3
0
0

m

~
3
3
,
4
0
0

2
9
,
0
0
0

1
7
,
5
0
0

~
.

g

9
0
,
0
0
0

1
0
7
,
3
0
0

1
5
9
,
3
0
0

C

7
1
,
0
0
0

6
7
,
0
0
0

5
8
,
6
0
0

7
4
6
,
6
0
0

7
5
3
,
8
0
0

7
8
1
,
1
0
0

g

~

o

<

J Chapter 2
wide variations in the operation of the three reactors for this pro
cess can be tolerated since the total conversion cost is only slightly
afected. This characteristic should mae it convenient to schedule
other products for manufacture in the same three reactors.
It i s now appropriate to proceed with the investigation of Case s II
and Il, so that the effect of other types of recycle can be ascertained.
Case . Three Reactors i Series, Recycle to the Second
This process case is the same as that shown in Fig. 2-3, except
that the recycle stream is now mixed with the product from the first
reactor. The combined stream is fed at a rate of [!+ R) lb moles/hr
to the second reactor. The analysis of this case proceeds as follows :
First Reactor
The material balances follow directly from Case I, whereby:
Second Reactor
A material balance on benzene around the junction of the recycle
and the line between reactors 1 and 2 gives
F
X
@
J
d R - (
F R
)x
Al
Solving for A
@
__ the composition of the stream entering the second
reactor, one obtains
where
g
= R/F. A benzene balance around the second reactor re
sults in
which may be rewritten as
ReactoY Design, O/J/iJllizatiol1, and Control J1
S1D1l01 material balances for monochlorobenzene yield
K_ K

X
;
-

-
[J !
;
!

}
;
JK_} [J
K
;
I

}
;
K_

[J _} [JK
;
} (1 K
;

}
Third Reactor
Material balances on benzene and monochlorobenzene around the
third reactor re sult in
1
q
K3 [J K

} K

<
3
- (1
q
K1) (1 + Kl
q
K3)
2
( 1 !_

}
;
K1(1
q
K3)

B
3
(1 K
2
q
K3
)
2
(1
q
Kl
K3) (1
q
K
1)
Kl (1 K3)
q
(1

K1) (1
q
K
2
) (1
K
2
* Ka
)
2
!_ [J!

}

[J _}J !

}
;
J _ !

}
!

J_

}
;
J
;

}
[2,JJ}
[ 2,J2)
One more relation may be obtained by noting that for the recycle
R [! R

Therefore,
and
[2,J}
3Z Chu[/cr3
Equation 2 . 1 3 can be solved by trial and error so that one may obtain
1 as a function of 1

. The only remaining equation needed is that

which relates the flow rate and conversion to J.
71, 000
J =

[! R}\

(2. 1 4)
The economic evaluation of Case II is now carried out directly,
since Eq. 2. 1 3 allows the proper value of R to be chosen for each feed
rate so that the required annual production can be achieved. The
method of computation and the results of the se calculations are sum
marized in Table 2 -2.
Case ! Three Reactors i Serie s_ Recycle to the Third
This case also may be visualized by using Fig. 2 -3. The same
proce ss scheme is used as H that figure, but here the recycle stream
i s combined with the product from the second reactor ad the com
bined stream is fed to the third reactor.
The analysis for this case is exactly analogous to that for Case II.
Its result is
and
1

-"~~~ ~~~~~~

(1 1

}
;
(1
"
+

} (1
;

+-- ----
(1
'

'
} (1
';

'

}
'
+-~ ~~ "~""
(1
'

}
;
(1 1
;
} (1
';
1

}
'1
( 2. 1 5)
( 2. 16)
where 1

R,!. A material balance on the recycle stream yields

1
~

(1 +}

(1 +
'

}
;
1 (2. 1 7)
In a manner similar to that of Case II, a value of _ is selected, and
the calculations for the economic evaluation of Case III follow direct
ly from Eqs. 2. 1 4 to 2 . 17. These computations are summarized in
Table 2-3 .
T
a
b
l
e

Z
.

Z
.

R
e
s
u
l
t
s

o
f

D
e
s
i
g

C
a
l
c
u
l
a
t
i
o
n
s

f
o
r

C
a
s
e

D
e
s
i
g

P
a
r
a
m
e
t
e
r
s

K
_

0
.
2

0
.
3

0
.
4

0
.
6

K
1
,

E
q
.

2
.
1
3

1
.

0
3

0
.
8
7

0
.
7
7

0
.
6
6

K
2

~

0
.
1
3
4
K
1

0
.
1
3
9

0
.
1
1
6

0
.
1
0
3

0
.
0
8
8
2

)

1
5
.

5
/
K
1

1
5
.
0

1
7
.
8

2
0
.
1

2
3
.
5

R

~

!
K
_

3
.
0

5
.
3
4

8
.
0

1
4
.
1

x
_
_
,

E
q
.

2
.

1
1

0
.
1
6
7

0
.
2
3
1

0
.
2
8
7

0
.
3
7
7

x

E
q
.

2
.

1
2

0
.
6
4
0

0
.
6
2
2

0
.
6
0
7

0
.
5
5
1

x
c
_

0
.
1
9
3

0
.
1
4
7

0
.
1
0
6

0
.
0
7
2

E
q
.

2
.
1
4

6
,
1
6
0

4
,
9
5
0

4
,
1
6
0

3
,
4
3
0

C
o
s
t
s
,

d
o
l
l
a
r
s
/
y
e
a
r

+

L

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
7
,
0
0
0

4
5
,
0
0
0

3
3
,
0
0
0

2
5
,
0
0
0

d

1
3
7
,
0
0
0

1
0
7
,
6
0
0

7
9
,
4
0
0

5
9
,
4
0
0

L

2
1
,
6
0
0

3
0
,
9
0
0

3
9
,
3
0
0

5
6
,
8
0
0

J

1
7
2
,
0
0
0

1
3
8
,
0
0
0

1
1
6
,
0
0
0

9
5
,
5
0
0

T
o
t
a
l

9
2
6
,
0
0
0

8
5
9
,
9
0
0

8
0
6
,
1
0
0

7
7
5
,

1
0
0

0
.
8

1
.
2

2
.
0

0
.
6
0

0
.
5
3

0
.
4
7

0
.
0
8
0
2

0
.
0
7
1

0
.
0
6
3

2
5
.
8

2
9
.
3

3
3
.
0

R

2
0
.
7

3
5
.
1

6
6
.
0

o
0
.
4
4
4

0
.
5
4
7

0
.
6
6
7

0
.
5
0
1

0
.
4
1
9

0
.
3
1
5

0
.
0
5
5

0
.
0
3
4

0
.
0
1
8

'

3
,
0
5
0

2
,
6
3
0

2
,
2
8
0

.

i
'

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

2
1
,
0
0
0

1
5
,
0
0
0

1
0
,
8
0
0

g
'

5
0
,
0
0
0

3
7
,
0
0
0

2
6
,
0
0
0

5

7
3
,
5
0
0

1
0
8
,
4
0
0

1
7
5
,
8
0
0

8
5
,
0
0
0

7
3
,
5
0
0

6
3
,
7
0
0

"

7
6
7
,
9
0
0

7
7
2
,
3
0
0

8
1
4
,

7
0
0

~

T
a
b
l
c

Z
~
3
.

L
8

L
8
C
U

0
B
8

0
I

C
a

D
e
s
i
g

p
a
r
a
m
e
t
e
r
s

K
l

K
3
,
E
q
.
2
.
1
7

F

=

1
5
.
5
/
K

_

=
K
3
F

K
2

=

2
.
0
7
/
F

x
B
3
'

E
q
.

2
.
1
6

X
A
3
'

E
q
.

2
.
1
5

x
C
3

0

E
q
.

2
.
1
4

0
.
5

0
.
6

8
.
3
3

1
.
8
9

3
1
.

0

2
5
.
8

2
5
8
.
0

4
8
.
8

0
.
0
6
6
8

0
.
0
8
0
3

0
.
1
0
1

0
.
3
1
3

0
.
8
9
3

0
.
6
5
5

0
.
0
0
6

0
.
0
3
2

2
,
4
3
0

3
,
0
4
0

C
o
s
t
s
,
d
o
l
l
a
r
s
/
y
e
a
r

u

5
3
8
,
4
0
0

5
3
8
,
4
0
0

1
1
,
5
0
0

1
9
,
5
0
0

d

2
7
,
0
0
0

4
6
,
4
0
0

7
3
5
,
0
0
0

1
7
4
,
0
0
0

J

6
7
,
8
0
0

8
4
,
8
0
0

T
o
t
a
l

1
,
3
7
9
,
7
0
0

8
6
3
,
1
0
0

0
.
7

0
.
8

0
.
9
8

0
.
6
3
3

2
2
.
2

1
9
.
4

2
1
.
8

1
2
.
3

0
.
0
9
3
3

0
.
1
0
7

0
.
4
4
0

0
.
5
1
4

0
.
4
9
5

0
.
3
8
7

0
.
0
6
5

0
.
0
9
9

3
,
6
6
0

4
,
3
6
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

2
8
,
0
0
0

3
7
,
0
0
0

6
7
,
2
0
0

8
7
,
6
0
0

9
3
,
5
0
0

6
2
,
5
0
0

1
0
2
,
2
0
0

1
2
1
,
7
0
0

8
2
9
,
3
0
0

8
4
7
,
2
0
0

~

0
.
9

1
.
0

1
.
2

1
.
5

~

0
.
4
4
7

0
.
3
3
3

0
.
2
0
9

0
.
1
2
0

|

1
7
.
2

1
5
.
5

1
2
.
9

1
0
.
3

7
.
7

5
.
1
7

2
.
7
0

1
.
2
4

0
.
1
2
0

0
.
1
3
3

0
.
1
6
0

0
.
2
0
1

0
.
5
5
9

0
.
5
8
9

0
.
6
0
3

0
.
5
8
6

0
.
3
0
8

0
.
2
5
0

0
.
1
3
3

0
.
1
6
6

5
,
0
9
0

5
,
8
8
0

7
,
5
5
0

1
0
,
5
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

4
5
,
0
0
0

5
4
,
0
0
0

1
0
8
,
4
0
0

1
2
9
,
2
0
0

4
5
,
7
0
0

3
5
,
5
0
0

1
4
2
,
0
0
0

1
6
4
,
2
0
0

8
7
9
,
5
0
0

9
2
1
,
3
0
0

Reactor Design, Optimization, uHd Conh'ol JJ
Case : Three Reactors in Prallel
the three reactors are operated in parallel with an equal flow to
each, the analysis follows directly from the conside ration of the first
reactor in Case I. The results are
and
1
7
~ m
1 +K
1
(2. 1 8)
(2 . 1 9)
Here, K
1
and K
2
are evaluated using a value for ! which corresponds
to the total rate of feed including recycle to each reactor. It may
therefore be written that
m
8

23, 1
J =
-
=
M
3( 1 1 2 . 6)Fx

FX
B
(2. 2)
From Eqs. 2. 1 8 to 2 . 20 the computations for the evaluation of Case IV
follow directly and are shown H Table 2 -4.
Case V: Batch Reaction
If the chlorination were carried out batchwise in three reactors
simultaneously, the same analysis used t o interpret the labratory
data would apply. Thus, Eqs . 2. 3 and 2 . 4 may be used to calculate the
extents of both reactions as functions of time. A pseudo feed rate to
the reactors may then be computed as
3 N
R
1 1 2. 8
F = *
=
=
l (2.21 )
where l represents the time each batch is allowed to react. the time
for loading and unloading is considered negligible, the annual reactor
usage follows directly as
o Y

63 0e
J= =
-
( 1 1 2.6)Fx
B
x
B
(2. 22)
From Eqs. 2 . 3 , 2 . 4, 2. 2 1 , and 2 .22 the ideal operating cost s for the
hat.rh rpart.nr havp hppn C.alr1I1at.pr anr arp !ummari7.pr in Tahlp Z~.
COMPARISON OF THE FVE REACTOR CASES
The results of the designs for the five reactor cases are com
pared in Fig. 2 -4, which shows the annual operating cost for the

8
D

4
.

0
8
0

D
s
i

L
8

C
0

D
D
8

D
I

L
8
8
0

D
e
s
i
g

p
a
r
a
m
e
t
e
r
s

F

1
0

1
5

3
0

3
0

K
l

1
.
5
5

1
,
0
3
3

0
.
T
T
5

0
.
5
1
6

K
2

0
.
3
0
T

0
.
1
3
8

0
.
1
0
3

0
.
0
6
9

x
A

0
.
3
9
3

0
.
4
9
3

0
.
5
6
3

0
.
6
5
9

x
B

0
.
5
0
3

0
.
4
4
8

0
.
3
9
5

0
.
3
1
8

X
c

0
.
1
0
6

0
.
0
5
9

0
.
0
4
3

0
.
0
3
3

4
,
T
3
0

3
,
5
3
0

3
,
0
0
0

3
,
4
8
0

C
o
s
t
s
!

d
o
l
l
a
r
s
/
y
e
a
r

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

3
9
,
8
0
0

3
4
,
8
0
0

3
0
,
1
0
0

1
3
,
T
0
0

9
6
,
0
0
0

5
9
,
8
0
0

4
8
,
4
0
0

3
3
,
0
0
0

c

6
4
,
6
0
0

9
1
,
5
0
0

1
0
8
,
3
0
0

1
T
3
,
0
0
0

1
3
3
,
0
0
0

9
6
,
3
0
0

8
3
,
9
0
0

6
9
,
3
0
0

T
o
t
a
l

8
T
0
,
8
0
0

8
1
3
,
8
0
0

T
9
9
,
1
0
0

8
3
6
,
3
0
0

%

~
4
0

6
0

"

0
.
3
8
T

0
.
3
5
9

0
.
0
5
1
8

0
.
0
3
4
5

0
,
T
3

0
.
T
9
5

0
.
3
6
4

0
,
1
9
9

0
.
0
1
6

0
,
0
0
6

3
,
3
4
0

1
,
9
T
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

1
1
,
5
0
0

5
,
T
0
0

3
T
,
6
0
0

1
3
,
8
0
0

3
3
6
,
5
0
0

3
3
3
,
0
0
0

6
3
,
5
0
0

5
5
,
0
0
0

8
6
6
,
5
0
0

9
4
4
,
9
0
0

T
a
b
l
e

2
-
5
.

R
e
s
u
l
t
s

o
f

D
e
s
i
g
n

C
a
l
c
u
l
a
t
i
o
n
s

f
o
r

C
a
s
e

N

D
e
s
i
g

p
a
r
a
m
e
t
e
r
s

l

h
r
s

1

2

2
.
5

3

4

5

6

)

E
q
.
2
.

2
1

1
1
2
.
8

5
6
.
4

4
5
.
1

3
7
.
6

2
7
.
2

2
2
.
5

1
8
.
8

E
q
.

2
.

3

0
.
6
6
3

0
.
4
3
4

0
.
3
5
6

0
.
2
9
1

0
.
1
9
2

0
.
1
2
8

0
.
0
8
3

R

E
q
.
2
.

4

0
.
3
2
7

0
.
5
3
4

0
.
5
9
5

0
.
6
4
5

0
.
7
0
6

0
.
7
3

0
.
7
3
5

-

o
^

0
.
0
1
0

0
.
0
3
2

0
.
0
4
9

0
.
0
6
4

0
.
1
0
2

0
.
1
4
2

0
.
1
8
2

.

E
q
.

2
.

2
2

1
,
9
2
5

2
,
3
6
0

2
,
6
5
0

2
,
9
3
0

3
,
5
7
0

4
,
3
1
0

5
,
1
5
0

C
o
s
t
s
,

d
o
l
l
a
r
s
/
y
e
a
r

<

a

+

l

+

I

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
3
8
,
4
0
0

5
,
8
0
0

1
1
,
3
0
0

1
5
,
5
0
0

1
9
,
7
0
0

2
7
,
2
0
0

3
6
,
9
0
0

4
6
,
7
0
0

.

~
.

1
3
,
9
0
0

2
7
,
3
0
0

3
7
,
3
0
0

4
4
,
7
0
0

6
5
,
4
0
0

8
9
,
8
0
0

1
1
2
,
3
0
0

~
.

c

1
6
8
,
0
0
0

6
7
,
6
0
0

4
9
,
8
0
0

3
7
,
5
0
0

2
2
,
6
0
0

1
4
,
5
0
0

9
,
5
0
0

g

J

5
3
,
7
0
0

6
5
,
9
0
0

7
3
,
8
0
0

8
1
,
6
0
0

9
9
,
5
0
0

1
2
0
,
0
0
0

1
4
4
,
0
0
0

T
o
t
a
l

7
7
9
,
8
0
0

7
1
0
,
5
0
0

7
1
4
,
8
0
0

7
2
1
,
9
0
0

7
5
3
,
1
0
0

7
9
9
,
6
0
0

8
5
0
,
9
0
0

Jo Cliu]lc+ 3
t

i90P

i
'
80

.^
!
Figure 2- 4. Economic evaluation of five design cases.
chlorination reaction plotted as a function of F. Cases I and V provide
substantial savings over the other suggested methods.
At low F vlues the operating costs for Cases I, II, and V are al
most equal because of the long residence times at low flow rates which
lead to low recycle costs for all processes. At these values the
method of recycle is therefore relatively unimportant. The dominating
variable charges at low flow rates are those for raw material s lost
in conversion to dichlorobenzene. and these are approximately the
same for all cases.
Several chemical engineering texts [1,5) point out that as the num
ber of reactors H a series of CSTR's is increased the resulting con
version-re sidence time relationship gradually approache s that for a
batch or plug-flow reactor. Thus, the maximum conversion at any
specified time will be achieved in a batch reactor for all but zero or
negative -order reactions) , because the concentration driving force for
the reaction is fully utilized in either a batch or a plug-flow reactor.
In a CSTR the reactant concentration in the feed is diluted when this
stream is mixed with material that has already partially reacted.
Thus, the rate of reaction would be maximized in the feed stream
where the concentration of reactants is highest; however, the full effect
of thi s concentration driving force i s lost when the stream is diluted.
U line with this reasoning, Fig. 2 -4 reveals that the batch reaction
achieves the least exensive operation. However, its economics are
closely approximated by those for Case I. the number of CSTR's
were increased to more than three, the cost-versus-flow- rate curve
would gradually drop and would approach that for Case V asympto
tically as the number of CSTR's is increased to ifinity. Even in the
case of three CSTR's the curves for Cases 1 and 5 show in Fig. 2-4
gradually approach each other at high feed rates, where the recycle
costs dominate.
Actually in the present economic analysiS the curve for Case V
should be considered only as a theoretical limit, since labor and
Reaclor Dfsigll , Opt imi:atioll, and Cont rol 39
time for loading, unloading, and operating the reactors have been
neglected. These factors undoubtedly would be very signiicant in
contributing to the total cost of operating a batch reaction system.
Thus, Case represents the optimum practical reactor flow plan for
the present analysis. It is convenient to thin of Case V as a limit
that might be approached if the number of reactors were increased
indefinitely. Figre 2-4 shows that there i s only a minor advantage
to be gained by using more tha three reactors.
As mentioned previously, the high raw-material costs cause all
five cases to be very uneconomical at low flow rates and high resi
dence time s. The curve shown for Case II nearly coincides with that
for Case V for a considerable distance as F is inc reased. Even
tually, however, the increasing recycle charges caused by incom
plete use of the recycle stream in Case II lead to significantly in
creased total costs and a re sulting sharp minimum in the curve. A
similar, and even more striing, effect is observed when Case III i s
compared with Cases V and .
Case IV corresponds exactly to the operation of a single CSTR
having three times the volume of the individual reactors Wder con
sideration in the present problem. For reasons mentioned previously,
the loss of the beneficial effects of staging the reactor volume would
be exected to lead to higher recycling costs. This effect is observed
i n Fig. 2-4, where Case IV i s show t o be one of the least desirable
systems studied.
The operation of three reactors in series with recycle to the first
is thus the most economical of the process schemes considered.
Figure 2-4 indicates that neither the operation of parallel reactors
nor the use of recycle to an intermediate point leads to improved re
sults. Thus it may be concluded that Case I, as shown in Fig. 2-3
represents the optimum arrangement that can be obtained with three
reactors.
At this point, it is well to emphasize that a great deal of unnec
essary calculation has been presented in order to provide a com
parison of various modes of reactor operation. Clearly, Cases II and
could have been eliminated at the outset on a qualitative basis;
such a decision would have been entirely in accordance with te sug
gestion of Chapter that computational effort be minimized. Never
teless, the comparative results summrized i n Fi g. 2-4 are valuable
in illustrating the effect of various types of recycle.
All the methods considered have been restricted to isothermal
operation of the reactors. It now becomes appropriate to investigate
the effect of temperature variations. This subj ect wll be treated
using the reactor system shown in Fig. 2-3 as a basiS for calculation.
EFFECT OF REACTOR TEMPERATUE ON PROCESS ECONOMCS
Calculations of the type shown in Table 2-1 have been carried out
for the case of three reactors in series with recycle to the first.
Figre 2 - 5 summarizes the results of these calculations and presents
9 Chapte 3
F ml0 f, m8/N

m %
M
CC 55
M M
ff
FF
U
_,x
% %
M M
55 55
M M

M M
Figure 2-5 . Effect of temperature on the economics
of chlorination.
the annual costs as functions of feed rate for five cases in which the
temperatures in all reactors are the same. The optimum feed rate
and the minimum annual costs are approximately as follows:
2_ = 2
;
2

!, lb moles/hr
60C
80
Costs, dollars/year 945, 000 773, 000 744, 000 743, 000 760, 000
It is found, however, that the required production rate of 8 million lb
per year cannot be achieved at 40C for any feed rate.
The cost curves are flat and the annual costs relatively insensitive
to temperature and feed rate. This is significant, since operation at
higher than optimum feed rate could make the equipment available
for some other operation a large part of the year without adding very
much to the cost of the monochlorobenzene.
OPTIMUM TEMPERTURE SEQUENCE
Figure 2-1 indicates that the activation energy for the unwanted
second reaction is far greater than that for monochlorobenzene for
mation. I t follows that for maximum yield of monochlorobenzene the
temperature should decrease as t he reaction proceeds. The tempera
ture can be relatively high in the first reactor, since there is little
monochlorobenzene present to be further chlorinated. It should then
be reduced as monochlorobenzene accumulates; yet the requi red pro
duction rate must be achieved in no more than 7000 hr /year.
It is evident that there is an infinite number of possible combina
tions of 7_ 7_,and 7 each 70C or less, which could be specified,
and it would seem probable that one of these might reduce dichloro
benzene formation appreciably and result in an annual cost less than
the $ 743, OOO/year calculated for operation of all these reactors at
60C. One possible combination is Tl * 70, 7
;
7 60, and 2

* 50C.
From Eqs. 2 . 5 and 2 . 6 a minimum annual cost of $ 750, 000 at a feed
Reactor Dsgu, Optimization, alld Control 4
rate of about 1 1 0 lb moles/hr can be calculated. Figure 2-5 compares
the results of this combination with those for isothe rmal operation.
Pursuit of further combinations of reactor temperatures is clearly
a j ob for a computer, since the possibilities are too numerous for
slide- rule calculation. Formulation of the problem for a digital com
puter is relatively simple, since the cost exressions are not com
plicated. The rate constant s must be expressed as functions of tem
perature, either by empirical polynomials or by Arrhenius-type
equations. The value s of III and k
2
shown in Fig. 2-1 are found to
be linear in l/T on a semilogarithmic graph, and may be exressed
by
6_ 7 5.1 Y 101
2
e-19
6
00fRT
1
2
2. 9 K 1
0
2
0
e
-
3
2
6
00/RT
where T is now in oK.
(2. 23)
(2 . 24)
An incomplete computer search of temperature sequence s in
dicates that there is little to be gained by temperature variations,
even though total anual costs in the vicinity of $73 5, 000 to $740, 000
can be attained. The principle that the temperature should be reduced
as the reaction proceeds i s sound [ see Denbigh (4) , Chapter 5], but
the obj ect is now minimum cost rather than maximum yield. The
cost data applicable to the present analysis are such as to give very
flat optima in the economic evaluation curves.
The selection of an optimum temperature sequence is a classic for
the more advanced optimization techniques mentioned in Chapter 1 .
A more sophisticated approach to the selection of reactor tempera
tures would provide a worthwhile exosure to some of the se modern
techniques .
SUMMAY OF THE REACTOR DESIGN
The minimum total annual cost for MCn production is approxi
mately $740,000; this includes purchased chemicals, recycle and pro
duct recove ry, operating labor, and fixed charges, and would result in
an annual saving of $ 1 00, 000 as compared with use of monochloro
benzene purchased at $0. 1 05/lb Otimum operation is at abut 60C
with a benzene feed rate (including recycle) of 80 lb moles/hr
( 14. 2 gallons/min) . Oerating at this temperature and flow rate
should make it possible to produce the requi red amount of mono

chlorobenzene in only 1770 hr /year.

A slight gain could be made by operating successive reactors at
lower temperatures, but the possible cost reduction would not appear
to j ustify the more complicated control system that would be needed.
The total annual cost is relatively insensitive to feed rate and
would increase to only about $775, ODD/year if the feed rate at 60C
were doubled and the yearly operating hours cut from 1 770 to 1 290.
This means that the eqUipment could be made available much of the
year for some other use .
42
Chapter 3
It should D pOinted out that the pi'oduction of dichlorobenzene
represents a substantial cost for chemicals. Its recovery and puri
fication would not be very difficult, and the proposed process would
look very much more attractive i a market could be found for a few
hundred thousand pounds per year.
HEAT-TRSFER A REACTOR-STAILITY CONSIERTIONS
With the basic nature of the process established, it is essential
that the reactor cooling system be examined in some detail. The
adequacy of the cooling medium and the available heat-transfer area
must be ascertained before the reaction equipment can be recommen
ded for MCB production. L equal importance, the possibility that
the reactor operation might be unstable under the proposed operating
conditions should be inve stigated.
To compute the rate of heat release in each reactor, the following
enthalpy changes of the chlorination reactions must be calculated
from the thermodynamic data on the reactants and products:
(
a) C6H6 C12= C6HsCI HCI
1 1.70 0 = 2 . 50 -- 22. 06 D
D= -31 . 26 Kcal
() C6HsCI Cl2 ~ C6H4Cl2 HCI
2 . 50 + 0 = -4 . 90 - 22 . 06 - D
D = -29 . 46 Kcal
The rate of heat evolution in a single reactor may now be computed
as
Q_ ~ 3 1 , 260 ( 1 . 8) (NJ1
x
A
) 29, 460 ( 1 . 8) (NRk1xB
)
= 2, 1 20, 000 k
1
xA
+ 2, 000, 000 /:_x_ Btu/hr (2. 25)
At the suggested f eed rate of 80 l b moles/hr with each reactor at
60C, the liquid compositions, found from Eqs. 2. 5 through 2. 8, are
as follows :
.

Mole fraction Mole fraction

First reactor 0.760 0. 228
Second reactor 0. 578 0. 389
Third reactor 0.440 0. 501
Evdently the total moles of chlorie reacting are greatest in the
first reactor ( 1 O. 760 O. 012 0. 252 in the first, O. 76 O. 578
+ 0. 033 - 0.01 2 0. 203 in the second, and 0. 578 - 0. 440 + 0. 059
- 0. 033 O. 1 64 in the third) . Sice the enthalpy decrease per mole
Reactor Design, Optimizatioll, and Control 43
of chlorine is about the same for each of the two steps of the chlorin
ation sequence, it is clear that the cooling load is greatest in the first
reactor.
At 60C, k
l
is 0. 65 hr-1 and k
2
is 0. 1 1 3 hr-
1
; therefore
Q_ ~ 2, 1 20, 000 (0. 65) (0. 76) + 2, 000, 000 (0. 1 1 3) (0. 228)
~ 1 , 1 00, 000 Btu/hr
The benzene, if fed at 20C, would absorb (60 - 20) ( 1 . 8) (80) (33 . 4) ~
1 93, 000 Btu/hr. The heat to be removed from the first reactor is then
1,1 00, 000 - 1 93,000 907, 000 BtU/hr.
The feed to the second reactor absorbs no heat, but the heat trans
fer to the cooling coil is still less than in the first reactor ( see Ap
pendix) . Excess chlorine passing to the condenser provides some re
flux cooling, but this will be neglected in order to make a conservative
estimate of the adequacy of the cooling coil s .
Under the assumption that cooling water i s available at 75F and
that a temperature rise of 40F may be allowed, the appropriate tem
perature driving force for heat transfer is computed as
A
T
ave
~ 60 ( 1 . 8) + 32 - 90 ~ 50F
The product required in the first reactor is then given by
907, 000
(
V
A)required
=
~ 1 8, 1 00 Btu/(hrWF)
50
but the avilable product was specified as
\'^ava);a),
~ (1 5 0) ( 550) ~ 82, 500 Btu/(hrW
F)
The coil is evidently more than adequate to hadle the cooling load
for cooling water averaging 90F.
There remains the question of reactor stability. Will a smal upset,
causing the temperature to rise slightly, so increase the reaction rate
(and hence the rate of heat evolution) that the cooling coil cannot re
turn the temperature to the control point of 60C? If the temperature
does not return, the rate of heat evolution will continue to icrease,
and the contents will boil over. There is a serious safety hazard due
to benzene and chlorine, and also the possibility of damage to the
equipment. the reactor is inherently unstable, it may still be pos
sible to provde controls so as to maintain it at 60C. Exenses for
control eqUipment will of course be much lower i the operation is
inherently stable.
If the rate of heat release goes up because the temperature rises,
the increased cooling due to the larger temperature difference for
heat transfer should cause the temperature to return to the desired
60C. To determine whether this will happen, the heat release in the
first reactor is calculated as a function of temperature. At 50C, for
example, _ and ! corresponding to a feed rate of 80 lb moles/hr
are 0.1 2 1 and 0.01 21 , respectively; thus, 7
@
_ i s 0. 892 and A
_
is 0. 1 07.
4J Chc[Icr2
The heat release Q, calculated by the procedure described earlier, i s
then 494, 000 Btu/hr.
Part of the heat generated by the chemical reaction is used to pre
heat cold benzene feed, and the rest must be transferred to the cool
ing medium. Let 2 represent the temperature of the liquid in the
reactor, and assume the benzene to be supplied at 20C. The heat
absorbed by the feed is then
g_ 7 (80) ( 33. 4) [1.B)
\
7_ 20) 4, 820 [T_ 20) Btu/hr
The diference Q1
g_ represents the heat to be transferred to the
cooling coil. At 50C, for example, this amouts to 494, 000 - 4, 820
( 50 - 20) ~ 3 88, 000 Btu/hr. Corresponding values at other tempera
tures versus 2_ are shown as line I on Fig. 2-6. The points used in
plotting this line have been calculated from values of

_ obtained from
Eq. 2. 25, with values of x__ and

_ for steady-state operation of the

first reactor at F 7 80.
o
s

L
x

Feed tce 8D!b moIes/hr
Qqn::y

pIoeo
| U - r
Vr r
||I
U
> r
n
c
B
c
C
c
0
c
|.0f
os~
s
r
r
r
4o so co
60C\0t l0U0l0\ut6
t
"L
ro
Figure 2-6. Stability diagram for the
reactor system shown in Fig. 2- 3.
o
The rate of heat transfer to the cooling coil is given by
{ = / [ 7_
- 2} 1 50 ( 550) ( 60 2_
|
[J. 8)
~ 1 48, 500 ( 60 ~ T) Btu/hr
Reactor Design, Optimization, and COll trol 4
where T is the mean temperature of the cooling water or brine.
{
is to be equal to (Ql { ) at 600e, T_ must be given by
907, 000
1 ~ 60
~ 60 - 6. 1 ~ 53. goe
1 48, 500
Line on Fig. 2 - 6 represents the cooling rate g_ for a T value of
53. goe and a varying value of Tl '
The reactor is stable at 600e (point P) by the Van Heerden criter
ion (7) with the specified U. product of 82, 500 and with a cooling
fluid temperature of 53. goe. A small temperature rise, say from 60
to 61 e, would increase the cooling rate more than it would the net
rate of heat generation. Accordingly, the temperature would drop back
to 600e. The opposite will happen H the temperature drops slightly
below 600e.
It may seem surprising that reactor instability can be prevented by
using warm water for cooling; intuitively one might thin that cold
brine would be needed. At 600e the net heat generation increases
roughly 8 per cent per degree increase in T
u
To maintain stability,
the cooling rate must increase at least 8 per cent per degree. With
U1 fixed, the cooling rate is proportional to 2_ T_ ,which must then
increase at least 8 per cent per degree. The T must therefore not
be greater than about 1 2e. For an elaboration of this point, see
Boynton, Nichols, and Spurlin (3) .
U case cold water were supplied for cooling e. g.,
*c
20C), the
cooling line g_ would pass through the 60 point U the U product
had a value of 22, 600 instead of the available 82, 500 Btu/hr OF. Line
J on Fig. 2 -6 represents the
0
cooling line in this situation. The
operation is now seen to be unstable at 600e, since an increase in
reactor temperature from 60 D 61e would cause the rate of heat
generation to increase more than the rate of cooling: The reaction
would " run away. "
At this point it seems clear that the stability question has been
disposed of: The criterion of stability discussed by Van Heerden [ )
and many other writers i s satisfied. At the operating point J the
cooling line is steeper than line I, the curve representing the net
rate of heat generation. The latter curve, however, represents
[Q_ { ) under steady-state conditions at T
8
In calculating Q_,the
compositions employed were obtained from the equations for steady
state operation at each T
u
As reaction proceeds isothermally, the
benzene disappears and Q_ decreases. It follows that Q_
0
,
) would
be greater at 61 e U the liquid composition remained at the steady
state value at 600e as the temperature climbed than if it changed to
the new steady-state value at 61 e. Since the composition change may
be exected to lag behind the transient temperature change, the steady
state criterion of stability that has been employed is not reliable. The
error, U any, in employing the steady-state values of reactant con
centrations in the calculations is in the unsafe direction.
This point is discussed at length by Bilous and Amundson (1), who
derive the relation between composition and temperature as T
u
changes
46 Chapter 2
slightly from a control point. For a simple irreversible first-order
reaction there is another criterion which must be met in addition to
that of Van Heerden U stability is to be assured. For the pair of
sequential first-order reactions encountered in benzene chlorinations
there are three stability criteria. The analysis is somewhat complex
and its use may not be necessary.
Since the benzene concentration is decaying and the second reaction
is slow, the worst that can happen is that the steady- state composi
tion at 60 remains unchanged as the temperature increases. Under
such conditions the reactor is stable U the slope of the cooling curve
at 60C is greater than that of the ( Ql ~ q) curve, when the latter
is found using the 60 steady- state values of
@_ = O. 760 and
x
B
1
= 0. 228.
From Eq. 2. 25 ,
Q
1
7 2, 1 20, 000 (0. 760 Il l) 2, 000, 000 (0. 228 k
2
)
and since
then,
g_ = 4820 ( Tn - 53. 9)
o (Q ) dkl dll
1
= 1 , 61 0, 000 456, 000
2
~ 4820
o TR dTR dTR
dink
1
dl nll
2
= 1 , 61 0, 000 k
1
- 456, 000 k
2
- 4820
dT
R
dTR
= 1 , 61 0, 000 (0. 65)
(
1 9, 600)
RTf
+ 456, 000 ( 0. 1 1 3)

(32 600)
RT
- 4820
o (Q ~ q ) _
= 93, 000 7, 650 4820 = 95, 830 Btu/(hr} ( OC) .
o T
R
This value is to be compared with the slope of the cooling curve,
which is U1 * 1 50 Y 550 Y 1 . 8 = 1 48, 000 Btu/hr) (OC) .
Evidently the reactor will respond t o temperature transients even
U composition does not change from its steady- state value at 60C.
Any change in composition with temperature will be in the direction
of greater thermal stability.
Reae/or Dcs/n, Opt i mizatiol l , uIlu COll t rol 47
Since U is the same in each reactor, the cooling lines for the
second and third reactors will have the same slopes as that for the
first reactor. These are shown as lines C and D on Fig 2 -6 . The
mean coolant temperatures in the three reactors are seen to be 53 . 9,
54.0, and 55. 1 C. The mean temperatures for reactors I and II are
approximately equal, because the lower reaction rate in the second
reactor is almost exactly balanced by the lack of a cooled feed stream
in the second reactor. In order to avoid confusion, curves and L
have been superimposed in Fig. 2 -6.
In designing the process, part of the cooling water leaving the first
reactor can be used in the second reactor and part of that from the
second can be used in the third. Part of that leavng the third can
then be mixed with cold water so as to provide warm water to cool
the first reactor.
the temperature rise of the cooling water i the first reactor is
1 0F, some ( 1 , 1 00, 000 - 1 93. 000)/1 0. or 90. 700 lb/hr will be needed.
all of this were to be put through a single O. 5 -in. coil with O. 4-in.
Ld., the mean water velocity would be 460 ft/sec. This is obviously
excessive, and it will be necessary to connect the coil so as to handle
a number of water streams in parallel short coil sections.
There are many questions to be exlored further the project is
pursued. Regarding stability, it i s evident that a factor of safety
could be introduced by operating at 55 instead of 60C. Figure 2 -6
shows that the slopes of the cooling line and the heat-generation
curve would then differ markedly at the control point. Operation of
all reactors at 55 instead of 60C would increase the annual costs by
only a very slight amount (see Fig. 2 -
5) .
The subject of control has not been discussed. What contingencies
must the control instruments handle? Power failure? Loss of cooling
water? Unexected increase in catalyst activty? Variation in feed
rate? Perhaps the quickest way to stop a runaway is to cut the chlo
rine supply. But H cooling water were cut off, would the di ssolved
chlorine continue the reaction far enough to bil over benzene? All of
these questions should be treated in some detail before the final de
sign is approved. Consideration of these types of problem during the
design phase of a project can save many hours of difficulty during
plant start -up and subsequent operation.
SUMARY OF HAT-TRANSFER A STAILITY CONSIERATIONS
It has been show that the cooling coils in the existing reactors
are adequate t o tae care of the heat of reaction under the conditions
of the proposed operation at 60C and a feed rate of 80 lb moles/hr of
benzene. However, stability considerations indicate that operation at
5 5C and at a somewhat reduced feed rate would be preferable, even
though the total annual costs would be slightly increased.
Water can be used as a cooling medium, but it is essential to em
ploy relatively hot water for this purpose. the temperature dif
ference between reacting liquid and water is small, the increased rate
of cooling due to a small increase in reactor temperature may be
4o C|c]I+ 3
made larger than the increase i rate of heat generation. This will
cause the temperature to fall and to stabiliz e at the control tempera
ture. The use of hot water as a cooling medium would undoubtedly
cause scaling problems. The possibility of using treated water or
some other heat-transfer medium should be evaluated.
Since the average temperature difference must be small [!1 OF) ,
the water temperature can ri se only perhaps l OF in passing through
the coil. This means that the water rate must be rather high, and it
will be necessary to modify the coil connections so that several coil
sections can be used in parallel .
Finally, if the proposed proj ect is to be undertaen, a careful study
to determine the best type of control equipment will be necessary.
Reactor Design, Optimization, and Cont rol 49
APPENI L L Z
Result s of Slide -Rule Calculations of Mole Fractions Benzene and
Monochlorobenzene (xA and
xB
) and of Heat of Reaction Q (Btu/hr)
(Each reactor ope rated at the same temperature)
)
( lb
noles
/
hr)
40 60 80 100 1 20
40C
xA1
0. 909 0. 943 0. 958 0. 965 0. 970
xB
l
0. 082 0. 057 0. 043 0. 035 0. 029
Q
1
1 87, 000 1 94, 000 1 96, 000 1 97, 000 1 98, 000
XA2
0. 832 0. 899 0. 91 6 0. 931 0. 942
xB2
0. 1 5 8 0. 1 1 0 0. 085 0. 069 0. 058
Q
2
1 71 , 000 1 85, 000 1 88, 000 1 91 , 000 1 93, 000
x
A3
0. 762 0. 841 0. 877 0. 900 0. 91 3
xB3
0. 226 0. 1 60 0. 1 24 0. 1 01 0. 085
Q
3
1 57, 000 1 73, 000 1 80, 000 1 85, 000 1 88, 000
50C
xA1
0. 804 0. 861 0. 892 0. 91 1 0. 925
xB
l
0. 1 90 0. 1 3 7 0. 1 07 0. 088 0. 072
Q
1
450, 000 478, 000 494, 000 502, 000 5 1 0, 000
x
A2
0. 646 0. 741 0. 795 0. 83 1 0. 856
xB2
0. 339 0. 2 53 0. 201 0. 1 67 0. 1 42
Q
2
3 70, 000 41 8, 000 445, 000 464, 000 475, 000
x
A3
0. 520 0. 63 8 0. 71 0 0. 757 0. 792
x
B3
0. 454 0. 3 5 1 0. 283 0. 238 0. 205
Q
3
307, 000 3 6 7, 000 402, 000 426, 000 443, 000
60C
XA1
0. 621 0. 71 1 0. 760 0. 804 0. 850
XB
1
0. 343 0. 2 71 0. 228 0. 1 88 0. 1 63
Q
1
933, 000 1 , 040, 000 1 , 1 00, 000 1 , 1 48, 000 1 , 1 92, 000
xA2
0. 385 0. 506 0. 578 0. 645 0. 689
xB2
0. 523 0. 445 0. 3 89 0. 332 0. 294
Q
2
648, 000 797, 000 885, 000 963, 000 1 , 01 6, 000
x
A3
0. 239 0. 360 0. 440 0. 5 1 9 0. 5 72
J
Chapt er 3
( lb
moles/
hr} 40
60 80
1 00 1 20

0. 61 6 0. 553 0. 501 0. 439 0. 396

V

468, 000 620, 000 71 9, 000 81 6, 000 876, 000

70C

0. 408 0. 504 0. 580 0. 632 0. 674

0. 41 6 0. 384 0. 345 0. 3 1 5 0. 286

VI
1 , 71 4, 000 2, 004, 000 2, 2 1 8, 000 2, 3 63, 000 2, 477, 000
;
0. 1 66 0. 25 8 0. 336 0. 406 0. 453

;
0. 462 0. 494 0. 487 0. 468 0. 444
;
963, 000 1 , 293, 000 1 , 543, 000 1 , 755, 000 1 , 888, 000
x
~

0. 068 0. 1 31 0. 1 95 0. 252 0. 305

x

0. 3 93 0. 484 0. 5 1 8 0. 527 0. 5 1 9
V
5 77, 000 866, 000 1 , 1 06, 000 1 , 3 03, 000 1 , 469, 000
REFERENCES
1 . Bilous, L. and N. R. Amundson, A. I . Ch. E.J. l, 5 12 ( 1 95 5) .
2. Boynton, D. E. , W. B. Nichols, and H. M. Spurlin, Ind. Eng. Cher.
51, 489 ( 1 959) .
3. Chermin, H. A. G. , and D. W. Van Krevelen, Chem. Eng. Sci. 9 58
(1 961 ) .
4. Denbigh, K. G. , Chemical Reactor Theory, Cambridge University
Press, Cambridge {1965} .
5. Kramers, H. , and K. R. Westerterp, Elements of Chemical Reactor
Design and Operation, Academic Press, New York {193} .
6. MacMullin, R. B. , Chem. Eng. Progr. 99 No. 3, 1 83 ( 1 948) .
7. Van Heerden, C. , Ind. Eng. Cher. 4 5, 1242 ( 1 9 53) .

3. Process Improvement for the Liquid
Liquid Extraction of Fenway acid
Tlds case is based uj)OIl [abomtoJ)' data obtained ill (II industrial
laboratory />riOl" to lite design of a contillllOus extractiol! /n'ocess. Tile
names of tile actual c/lemicals being cOllsidered lI(u'e been c/I(ed for
/mr/JOses of comme1'cial security. Some discussioll is directed to
lW1"d IIle use of faclorially designed eX/.erimellts ill gatherin and
analyzing cltemical engineering data. The im/J01"t(IIICe of utilizing tile
/Jroper /01'111 of inde/Jelldent variables ilZ tile fOl'lnulalioll of designed
eX/Je1"imelll s is also em/J/Ulsized
Tile design computations in lids clla/Jte1' lead /0 a classical/ill(l
cial emluatioll in ll'hicll ca/.ital and operating costs are balallced in
the determination of all optimum design. III tills case only a prelimi
nary desigl/ is required, al/d liberal use of appmximations greatly
simplies the calculation effm'l. An ol'er-all underslandillg of lite
process economics is obtained by all allalysis of a limiting case in
zcliell all illfillite lIumber of ext1'aclion slages is used.
Fenway Chemical Company
Bston, M:!!lrhu!ptt!
To: Michael Higgins, Group Leader
From: William Sullivan, PreSident
As you know, the company is currently operating a batch process
for the production of Fenway acid salt in our Back Bay plant. First,
Fenway acid is produced by sulfonating Fen-aromol-A with mono
chlorosulfonic acid. This reaction is carried out in a stirred-tan
reactor containing an aromatic solvent. Sodium hydroxide is then
51
52 C/wple y 3
added to the mixture to produce the stable sodium salt of Fenway aCid;
this neutralization is necessary to prevent hydrolysis of the sulfonate
group. The salt is extracted into water and then amidated to the
sodium salt of Brighton acid, the desired end product.
Since the aromatic solvent lowers the yield of the amidation step,
it must be removed before the reaction can proceed. Moreover, a
highly pure aromatic solvent is necessary for the recycle to the sul
fonator, because water reacts with the monochlorosulfonic acid. This
separation is at present achieved by a cumbersome and costly series
of washing and stripping operations carried out on each batch of pro
duct.
Since the batch extraction process is currently operated at pea
capacity and increases in product demand are anticipated, it is de
sired to investigate the possibility of a continuous extraction opera
tion. Your design group is requested to complete a preliminary pro
cess design that will allow the extraction of Fenway acid to be carried
out on a continuous basis. The present batch operation is carried out
at 70C, with a PH in the aqueous phase of 8. Since only a preliminary
design is requested, you may wish to restrict your attention to these
same conditions. However, the manuacturing technical group has ob
tained sufficient data to allow consideration of temperatures between
20 and 70C and of pH values from 8 to 12.
The plant has available several agitated kettles that could easily
be adapted to a continuous mixer-settler operation for the proposed
extraction. Since this eqUipment is currently idle, you are requested
to direct at least some attention to a continuous mixer-settler opera
tion; however, any ideas you may have about other types of processes
would also be appropriate.
The design requirements for other sections of the process specify
the following:
1. Feed to the continuous extraction and purification system
(after addition of NaOH to form the acid salt):
Aromatic solvent 3, 000 lb/hr
Fenway acid salt 680 lb/hr
(Plus negligible water which accompanies the NaOH)
2. Product of the continuous extraction and purification system
(to be fed to the amidation reactor) :
Aromatic solvent 5. 6 lb/hr
Fenway acid solvent 666. 0 lb/hr
Water I, 995. 0 lb/hr
The following sections summarize the data relevant to your design,
which have been gathered by the plant technical group.
Molecular Weights
Although both the Fenway acid and the aromatic solvent are com
posed of mixtures of several components, the following average mole-
Process im/JYocemellt for Liquid-Liquid Extractioll 53
cular weights may be ascribed to these compounds:
Compound
Sodium salt of Fenway acid
Aromatic solvent
Water
Suggested Nomenclature
Molecular
weight
246
123
18
A Refers to the sodium salt of Fenway acid
B Flow rate of the bottoms or product stream, lb moles/hr
C Capacity of a process unit, cu ft
D Diameter of a fractionating column, in.
f Fugacity, atm
F Feed rate to a fractionating column, lb moles/hr
J Distribution or partition coefficient, dimensionless
L Liquid flow rate in a fractionating column, lb moles/hr
nz Mass flow rate of a process stream, lb/hr; or quantity of
material present, lb.
n Molal flow rate of a process stream, lb moles/hr; or quantity
of material present, lb moles
S Refers to the aromatic solvent
T Temperature, C
V Vapor flow rate, lb moles/hr or lb/hr
W Refers to water
x Mole fraction in a liquid phase
y
Mole fraction in a vapor phase
(
Per cent of the total Fenway acid salt appearing i the aqueous
phase
" Activity coefficient
Superscripts
o
Refers to a component in its standard state
54
Chapter 3
Subscripts
AS
Refers to component A in the solvent phase
A
w
Refers to component A i the aqueous phase
D
Refers to the solvent content of the bottoms stream from a frac-
tionating column
s
Refers to the aromatic solvent
sw Refers to the aromatic solvent in the aqueous phase
s
s
Refers to the aromatic solvent in the organic phase
v Refers to the solvent content of the overhead product from a
fractionating column
w
Refers to water
ws
Refers to water in the organic phase
ww
Refers to water i the aqueous phase
Mutual Solubility Data
The mutual solubilities at 20C of the pure aromatic solvent and
water were obtained from the International Critical Tables. They are
Water in solvent: 0. 24 g/100 g
Solvent in water: 0. 19 g/100 g
Liquid-Liquid Equilibrium Data
An exerimental program was carried out to establish the relative
solubility of Fenway acid in the water and the solvent phases. A 23
factorial program was carried out; the factors and setting levels for
these exeriments are as follows:
Factor Levels
Temperature 20 and 70C
pH 8 and 12
W:S weight ratio 1. 0 and 2. 15
The data obtained from these exeriments are summarized in Table
3-1.
Process IlIlprOl'elllelll for Liquid-Liquid Extradioll 55
Table 3-1. Results of the Liquid-Liquid Exraction Studies
Experiment I: 20C;PH = 8; 1l1
w
/m
s
= 2. 15
Aqueous phase Organic phase
Component
K
g mole mole fro
K
g mole mole fro
A 35. 90 0. 146 0. 00818 3. 66 0. 0149 0. 0125
W 318. 6 17. 70 0. 991 0. 585 0. 0325 0. 0272
S 1. 94 0. 0158 0. 00089 141. 2 1. 148 0. 960
356. 4 17. 86 1. 00 145. 4 1. 195 1. 00
Experiment II: 20C;PH = 8; m
w
/
m
s
= 1. 0
Aqueous phase Orgaic phase
Component

g mole mole fro
.I
g mole mole fro
A 13. 10 0. 0533 0. 00405 3. 10 0. 0126 0. 00617
W 236. 0 13.10 0. 995 0. 991 0. 0550 0. 0269
S 0.873 0. 0071 0. 00054 242. 8 1. 975 0. 967
m
250. 0 13. 16 1. 0 246. 9 2. 043 1. 00
Exeriment III: 20C;PH = 12;
m
w
/m
s
= 2. 15
Aqueous phase Organic phase
Component

g mole mole fro
K
g mole mole fro
A 33. 4 0. 136 0. 00798 3. 42 0. 0139 0. 0114
W 304. 5 16.90 0. 991 0. 546 0. 0304 0. 0249
S 1. 21 0. 00985 0. 00058 144. 4 1. 175 0. 964
.
339. 1 17. 05 1. 00 148. 4 1. 219 1. 00
Experiment IV: 20C;PH = 12; m
w
/m
s
= 1. 0
Aqueous phase Organic phase
Component

g mole mole fro
,
g mole mole fro
A 18. 39 0. 0748 0. 00565 4.33 0. 0176 0. 00874
W 237. 0 13. 15 0. 993 0. 844 0. 0468 0. 0233
S 0. 563 0. 0046 0. 00035 240. 0 1. 950 0. 969
256. 0 13. 23 1. 00 245. 2 2. 014 1. 00
56 Chapter 3
Table 3 -1. Results of te Liquid-Liquid Extrction Studies (cont.)
Exeriment V: 70C;PH = 8; ml
V
/ms ~ 2. 15
Aqueous phase Organic phase
Component

g mole mole fro
I
g mole mole fro
A 42. 0 0. 171 0. 00972 4. 33 0. 0176 0. 0147
W 313. 0 17. 40 0. 989 0. 537 0. 0298 0. 0250
S 3. 09 0. 025 0. 00142 141. 0 1. 145 0. 960
--
358. 1 17. 60 1. 00 145. 9 1. 192 1. 00
Exeriment V: 70C;PH = 8; 111w/n1
S
= 1. 0
Aqueous phase Organic phase
Component

g mole mole fro
I
g mole mole fro
A 14. 10 0. 0573 0. 00440 5. 50 0. 0224 0. 0107
TV 234. 0 13. 00 0. 995 1. 150 0. 0640 0. 0306
S 0. 825 0. 0067 0. 00051 246. 0 2. 00 0. 958
248. 9 13. 06 1. 00 252. 6 2. 086 1. 00
Experiment VII: 70C;PH = 12; ml
V
/ms = 2. 15
Aqueous phase Organic phase
Component

g mole mole fro
g-
g mole mole fro
A 39. 2 0. 159 0. 00845 3. 20 0. 0130 0. 0108
IV 336. 0 18. 65 0. 990 0. 480 0. 0267 0. 0222
S 2. 35 0. 0191 0. 00101 143. 0 1. 163 0. 967
--
377. 5 18. 83 1. 00 146. 7 1. 203 1. 00
Exeriment VI: 70C;PH = 12; n1
w
/m
s
= 1. 0
Aqueous phase Organic phase
Component

g mole mole fro
K
g mole mole fro
A 14. 65 0. 0596 0. 00474 4. 76 0. 0193 0. 00943
W 225. 0 12. 50 0. 995 1. 19 0. 0661 0. 0323
S 1. 18 0. 00960 0. 00077 241. 0 1. 960 0. 958
240. 8 12. 57 1. 00 246. 9 2. 045 1. 00
Process 11ll/)rOl'emellt for Liquid-Liquid Extraction 57
The Statistics Department has analyzed these data uSing a standard
computer program and has obtained the following correlating equation:
{ = 86. 11 + 9. 82 (:
:
- 1. 56) - 0. 425 (pH - 10)
+ 0. 0215 (T - 45) + O. 478 (
m
W
- 1. 56) (pH - 10)
's
+ 9. 25
X 10-3 (:
s
w 1. 56
)
(T - 45)
- 8 X 10-3 (pH 10) (T 45)
+ 5.05 x 10-3(:
- 1. 56
)
(pH - 10) (T
- 45) (3. 1)
where mw/m s represents the ratio of the mass of water to the mass
of solvent in each batch extraction and { denotes the percentage of the
total Fenway acid which is found in the aqueous phase. According to
the statistics group, this equation is valid within 90 per cent confidence
limits for the ranges of the variables over which the data were taken.
Vapo r-Liquid Equilibrium Dat
Vapor-liquid equilibrium data were obtained for aqueous solutions
contaiing various amouts of aromatic solvent and Fenway acid.
These data showed that Fenway acid is vrtually nonvolatile at the
temperatures of interest in this problem (approximately 105C). Data
were obtained not only for mixtures containing various amounts of
Fenway acid but also for mixtures havng different values of pH in the
aqueous phase. As shown in Fig. 3-1, the relative volatility of the sol

vent mixture is quite unaffected by the variations in PH and acid con

centration, and all the data may be correlated by a single line.
Cost Correlations ad Desig Instructions
The following approximate cost-correlation equations may be use
ful in evaluating the economics of any of the process designs to be
considered:
1. D = 0. 47 (V)1/
2
2. Cost of distillation tray = $16(D)
3. Cost of reboiler = $470(V
)
O. 2
4
4. Cost of condenser = $15(V)
O. 6
58
C/apter 3
0.008
I
0.0
7_S)mbl
4
d
0.006
c-
W
0.00 5
Ys
0.004
0.0 3
0.00 2
I
'AW
0_0117
0.0197
0.0117
0.0197
/
0.0 1 /
/
0.0002
IpH
:/
8
/
8
12
12
/
V
/
V
.
+
0.0004 0.0006 0.0008 0.001
'SW
Figure 3 -1. Vapor-liquid equilibrium data for
solvent-water mixtures containing Fenway
acid salt.
5. Cost of miing tank = $150(C)
o.6
6. Cost of settling tank = $40(C
)O.
7
7. Cost of liquid pumps = $300(m)O
.07
where D = distillation column diameter (in.) ,
V = vapor rate (lb/hr),
C = capacity of item (cu ft),
m = mass flow rate (lb/hr) ,
8. Cost of controls for a mixer-settler combination = $3,000/
extraction stage.
It is to be noted that these equations yield estimates for the uninstal
led equipment costs. The installed equipment cost may be approxi
mated by multiplying the uninstalled cost by four.
Altough the residence time in the mixer-settler equipment will
eventually have to be ascertained by a detailed mass transfer analy
sis, the following approximate values may prove useful for prelimin
ary desig calculations: mixer residence time = 10 minutes, settler
residence time = 20 minutes. At these values of residence time, an
over-all stage efficiency of 100 per cent may be assumed. Reference
(7) may be consulted for a discussion of the factors that may affect
the efficiency of mixer-settler units. An average specifiC gravity of
1. 18 for the two-phase mixture may be used in computing the
required equipment capacity.
Low-pressure steam is available at $0
.
80/1000 lb, and cooling
water charges are approximately $0.02/1 000 gallons. The cooling
water is available for 80F, and it is desired to maintain the tempera
ture rise of the cooling water at values below 40F.
PYocess ImpYOl'eIllCII/ foy Liqllid-Liqllid Extractioll 59
PREPATION OF A PRELMIARY DESIGN
Exmiation ad Evalution of Exerimental Data
Before proceeding with an analysis of preliminary process de
signs, an examination and evaluation of the exerimental data is in
order. One important factor in determining the process economics
will be the extent to which water and the aromatic solvent are mutual
ly soluble. The mutual solubilities will be particularly critical, since
the final product specifications allow only a small amout of solvent
in the aqueous extract that is fed to the amidation reaction step. Any
excess solvent must be removed, and this should probably be accom
plished by a stripping or distillation process.
The original memorandum indicates that at 20C water is soluble
in the solvent to the extent of 0. 24 g/100 g and the solvent is soluble
in water to the extent of 0. 19 g/100 g. Ater converting these values
to a molar basis, it is useful to compare the mutual solubilities of
the pure solvents with those obtained from the liquid-liquid extraction
studies in which the acid salt was present. This comparison is imple
mented i Table 3-2.
Table 3-2. Mutul Slubilities of Water and the Aromatic Slvent
Exeriment Temp.oC pH nl
w
/m
s
x
sw
x
ws
Pure solvents
20
7 0. 00028 0. 0161
(lit. values)
I
20
8 2. 15 0.00089 0. 0272
II
20
8 1.0 0. 00054 0.0269
II 20
12 2. 15 0. 00058 0.0249
IV
20
12 1.0 0. 00035 0. 0233
V 70
8 2. 15 0. 00142 0. 0250
VI 70
8 1. 0 0. 00051 0.0306
VII 70
12 2. 15 0. 00101 0. 0222
VII
70
12 1. 0 0. 00077 0. 0323
The data at 20C reveal that the presence of the acid salt and the
variations in pH have only a modest effect on the mutual solubilities
of the two solvents. At constant temperature, the variations in pH and
salt concentration do not appear to produce any Signiicant or consis
tent changes in the mutual solubilities. The scatter in the data ob
tained at 70C may arise from difficulties in maintaining accurate
temperature control at the higher levels of temperature. For the pur
pose of a preliminary design, the effects of PH and salt concentration
will be neglected, and average values of the mutual solubilities will
60 CI/pter 3
be assumed to apply at each temperature level. These mean values
of x s
w
and x
w
s are as follows:
Temperature
xsw x
ws
0.00059 0.0256
0.00093 0.0275
After the design has been completed, it will be important to inves
tigate the effect of variations in the values of x
sw
and xws on the
estimated process economics. If variations of the magnitude shown in
Table 3-2 are found to affect the estimated financial return signifi
cantly, a more exact exerimental determination of the mutual solu
bilities would be required.
Consideration of the Liquid-Liquid Equilibrium Data
The application of factorial experiments is a very powerful tech
nique in obtaining and correlating experimental data for design work.
Using statistical methods, a great deal of information may be gathered
by expending only a minimum of exerimental time and effort. These
methods are particularly helpful when the theoretical basis of a pro
cess operation is poorly understood. References (J, 3, 4,5) serve as
an excellent introduction to statistical designs and their application
to practical chemical engineering problems.
By use of statistical methods, Eq. 3. 1 has been derived as the
proper correlating equation for the exerimental results show in
Table 3-1. This equation relates the fraction of the total Fenway acid
found in the aqueous phase to variations in the temperature, pH, and
amounts of solvent present. Its use in carrying out design calculations
presents two very Significant difficulties. First, the validity of the
equation is limited to the ranges of the variables for which the ori
ginal data were obtained. It is quite possible that the design specifi
cations will require the application of the correlating equation well
beyond the range of variables for which the data were taken. For ex
ample, the ratio of water to solvent was varied by only a factor of 2 in
the laboratory exeriments, and it is unliely that the optimum con
tinuous process would happen to correspond to solvent flow rates in
this range. Application of the equation outside these limits would be
quite risky and could lead to highly erroneous conclusions.
The second difficulty, somewhat related to the first, results from
the poor selection of variables that were used in setting up the ex
perimental design and in expressing the experimental results. The
selection of temperature and pH as variables appears to be reasonable;
however, the use of the ratio mw/ms has no reasonable basis, and the
quantity ( does not appear to be an appropriate form of the response
variable. For example, if very large values of 1I1W/1I1S (outSide the
range of the original data) were inserted in Eq. 3. 1, values of ( > 100
per cent could be obtained. This sort of physical impossibility might
Process Improvement for Liquid-Liquid ExlYaclion 61
have gone unnoticed i the equation had been combined with other alge
braic expressions. In view of these observations it is worthwhile to
search for a more satisfactory basis of expressing the experimental
results before proceeding with the design considerations.
In a thermodynamic system in which a solute is distributed between
two immiscible solvents, equilibrium will be obtained when the chemi
cal fugacity of the component in one phase becomes equal to the fugac
ity of the same component in the other phase. Wen this concept is
applied to the present case,
which in turn may be exanded as
fA'AlVxAlV fAos 'AS xAS
Therefore, at equilibrium one obtains
(3. 2)
Now fO, the fugacity of a component in its standard state, is independent
of composition, and at constant temperature Eq. 3. 2 may be written
as
(3.3)
In dilute solutions, the activity coefficients may be assumed to be in
dependent of concentration, and a further simplification is effected,
whereby
(3.4)
The quantity] is termed the distribution or partition coefficient, and
Eq. 3. 4 is conventionally know as Nernst's law. It proves quite use
ful in correlating the results of liquid-liquid extraction exeriments.
It is to be recognized that the distribution of Fenway acid between the
two phases is not determined solely by physical phenomena. Since a
neutralization reaction is also takig place, the simple form of Eq.
3.4 is not strictly applicable. However, as an initial approximation, it
may result in a data correlation that will be useful for preliminary
design calculations.
Casting the data of Table 3-1 in the form suggested by Eq. 3. 4
results in the data listed in Table 3-3. They indicate that the parti-
62 Clapter 3
Table 3-3. Liquid-Liquid Exerimenta Data Expressed
in the Form of a Partition Coefficient
Experiment Temperature,OC pH K = X
A
W/X
AS
I 20 8 0. 655
II 20 8 0. 656
II 20 12 0. 699
IV 20 12 0. 647
V 70 8 0. 661
V 70 8 0. 411
VI 70 12 0.782
VIII 70 12 0. 503
tion coefficient is independent of pH and the relative amounts of sol
vent. The average value of the partition coeficient at 70C is 0.590,
while that at 20C is 0.663; only a 10 per cent variation is found over
a 50C variation in temperatre. This is probably not satistically
sigiicant in view of the scatter in the data at the higher temperatre
level. Therefore, Eq. 3-1 may be discarde, and the liquid-liquid
extraction data may be exressed in terms of a single average value
for the partition coefficient, 0.63. Since the partition coefficient has
been found to be quite invariant over a wide range of exerimental
conditions, the designer may use this value conidently for vrtually
any desig condition. This broad application would not have been
possible i the design calculations were to be based on Eq. 3-1.
The results of this discussion emphasize the need for god judg
ment i the selection of the proper method for correlatig exeri
mental data. The use of statistically desiged exriments is an i
vluble aid, but it is not a substitute for a meanigul analysis of a
problem. I should b pited out tat the results of the ty show
i Table 3-1 are convntionally correlated in the form of a tringlar
diagram as illustrated by Hougen, Watson, and Ragatz (). However,
i the present situation the concentrations of te solute are s small
that a triaglar diagram would be neither meaningul nor useful.
Vapor- Liquid Equilibrium Data
The vapor-liquid equilibrium data appear to be well presented and
properly correlated. Figure 3-1 should be directly applicable to the
design calculation. In particular, the relative invariance of the rela
tive volatility with changes in pH and Fenway acid concentration is a
very convenient feature of the data and will grea tly reduce the re
quired computational effort.
Process Improvemellt for Liquid-Liquid Extractiol/ 63
PRELIAY PROCES DESGN CALCULATIONS
To produce a Fenway acid product having the specified purity, a
process must accomplish the following obj ectives:
1. Extract the acid salt from the aromatic solvent.
2. Remove the aromatic solvent from the aqueous acid salt
solution prior to the amidation step.
Since the original memorandum specified that a continuous mixer
settler design should be conSidered, a process of the type shown in
Fig. 3-2 should be appropriate. In this design scheme, the feed stream
containing the crude Fenway acid salt is mixed with recycle and auxil
iary water, and the combined stream is metered continuously into the
Feed stream
A' 680 Ib/hr
5' 3,00 Ib/hr
Organic
phose
RofflOQle streom
(Solvent recycle)
A' 141b Ihr
W'
S' 2,994.4lb/hr
Auxillory
water
Organic
phose
Aqueous
recycle
Aqueous
exlract
Figure 3-2. Block flow diagram of a one
stage continuous extraction process.
mixer. The effluent from the mixer is separated into organic and
aqueous phases in the settler. The small amount of dissolved organic
solvent in the aqueous phase is removed by a stripper from which the
final desired product is removed. It is important to note that the
aqueous stream fed to the stripper is saturated with organic solvent,
ad further that the mutual solubility of the water and solvent is quite
insensitive to temperature vriations. Therefore, i a rectiying sec
tion were added to the stripper, all the plates i such a section would
contai a two-phase liquid and no enrichment could be effected. For
this reason, the process will have to be designed so that the desired
product purity can be achieved by using only a stripping section. The
overhead product from the stripper is condensed and separated. The
64 Chapter 3
aqueous phase from the condensate settler serves as the recycle
water stream and is returned to the mixer
,
while the organic streams
leaving both the first and second settlers are combined, sent to a de
hydration still, and returned to the sulfonation reactor.
The present design will consider all aspects of the separation pro
cess except the dehydration unit for the aromatic recycle; the costs
associated with this latter operation are not expected to be signiicant.
As a first consideration, a design will be analyzed that incorporates
one mixer and two settlers, as shown in Fig. 3-2. Clearly the econo
mics of the process design could be altered Signiicantly by the use
of additional pairs of mixers and settlers. This concept will be ex
amined in a later stage of the analysis. Depending upon the number of
stages required by the mixer-settler operation, one might wish to
consider the use of an extraction column. Again this decision will be
deferred until the computations for the preliminary mixer-settler de
sign cases have been completed and analyzed. As a further initial ap
proximation, the process will be assumed to operate at 70C and a pH
of 8. However, j udging by the results shown in Table 3-3, variations
in these parameters would not be expected to affect the economics of
the process seriously.
The analysis of the design shown in Fig. 3-2 proceeds in a man
ner quite analogous to that for most separation unit operations-by the
alternate application of material balance and equilibrium relation
ships. First of all, the composition of the recycle stream may be
established partially by over-all material balances on components A
and S:
A lost in recycle of organic solvent = 680 - 666 = 14 lb
/
hr
S in recycle = 3, 000 - 5. 6 = 2, 994. 4 lb
/
hr
No over-all balance in water is yet pOSSible, since the flow rate of the
make-up water stream has not yet been determined.
Focusing attention on the mixer-settler unit, i equilibrium is at
tained, the relationship summarized by Eq. 3-4 may be assumed valid:
( . nAI )
xA
W nAI
+
nsl + nl
-= = 0. 63
xAS /lAS
(nAS
+
/lws
+
nSS )
(3.5)
The quantities designated by /
l
represent the molal flow rates of the
streams leaving the mixer-settler combination.
As a first approximation, the mutual solubilities of water and solvent
may be assumed to be negligible. The validity of this approximation
will be determined by a second iteration on the calculation. Equation
3. 5 may be rewritten as:
(3. 6)
Process Improvement for Liquid-Liquid Extraction 65
where
n
ss
= 3,000/123 = 24. 4 lb moles/hr
nAW = 666/246 2. 71
n
AS
= 14/246 = 0. 057
Therefore,
nww = 2. 71 (
429
- 1)
0. 63
= 1, 841 lb moles/hr
or mww = 33,200 lb/hr
With this initial approximation it is possible to account for the
mutual solubilities of water and solvent. In the analysis of the exeri
mental data taken at 70C, the following mutual solubility values were
established:
Xs
w = O. 00093
xws
= 0. 0275
Now the following values may be computed:
nsw = 0. 00093 (2. 71 + 1,841 + 1sw) 1. 72 lb moles/hr
111
sw = 212 lb/hr
nws
= 0.0275 (24. 4 + 0. 057 + nws
)
I
l
ws
= (24.46) (0. 0275) /0. 9725 = 0. 692 lb moles/hr
mw
s
= 12. 5 lb/hr
With these values, the final material balance on the process may be
established:
n
ss
= 24. 4 - 1. 72 = 22. 7 lb moles/hr
n
ww
= 1841 - 0. 692 = 1840 lb moles/hr
The auxiliary water stream may now be determined approximately
by an over-all material balance as:
Aux. water 1995 + 12. 5 = 2007 lb/hr
= 111.6 lb moles/hr
and the recycle water is established as
Recycle water = 1,841 - 111. 6 = 1729 lb moles/hr
= 31, 200 lb/hr
66 ClzapLer 3
The feed to the strippig column may now be tabulated:
Component lb moles/hr lb/hr
A
W
S
2. 71
1840
1. 72
666
33,100
212
and the vapor feed to the condenser may be approximated from a
material balance calculation:
Component lb moles/hr lb/hr
A
W
S
o
1729
1. 68
o
31,200
206. 4
As a further step in the calculation, an approximate material
balance on the condensate settler results in:
n
sw
= 0. 00093 (1729) 1. 61 lb moles/hr
ns
= 1. 68 - 1. 61 = O. 07 lb moles/hr
Now that the approximate flow rates of all streams have been estab
lished, a second iteration in the calculations should establish the flow
rates with sufficient validity for a preliminary economic evaluation.
First, the amount of solvent in the organic phase is recomputed, taing
into account the solvent in the recycle water stream:
nss
= 24. 4 - 1. 72 + 1. 61 = 24. 3 lb moles/hr ^ 2990 lb/hr
And the water flow rate in the aqueous phase is established as:
[1 + (24. 3/0. 057) ]
I
l
w
w =
- 1. 0
0. 63
2. 71 = 1, 832 lb moles/hr
33, 000 lb/hr
Pursuing the calculation procedure utilized previously:
I
l
sw
0. 00093 (2. 71 + 1,832 + IZ
s
w) 1. 706 lb moles/hr
In
sw
= 210 lb/hr
(0. 0275)
I
l
ws
= (24. 3 + 0. 057) - = O. 688 lb moles/hr
0. 9725
111
W
S = 12. 4 Ibs/hr
Process /1II/JrOL'Clllelll for Liq/lid-Liquid Exlraction
67
Aux. water = 1, 995 + 12. 4 = 2, 007 lb/hr = 111. 6 lb moles/hr
Recycle water ^ 1, 832 - 111. 6 = 1720 lb moles/hr = 31, 000 lb/hr
The second approximation of the feed to the strippig column is
delineated i Table 3-4.
Table 3-4. Feed to te Strippig Colum
Component lb moles/hr ;V lb/hr
A 2.71 .0014 7 666
W 1831 . 99 76 33,000
S 1. 706 . 00093 210
Total 1835.416 1. 00
The iterative procedure could be pursued, but the values tabulated
do not deviate Significantly from those obtained in the first iteration.
The material balance calculations have now been completed.
Stripper Desig
The feed to the stripper is specified by Table 3 -4, and the minimum
requirement for product purity is shown in Fig. 3-2. Since most of the
aromatic solvent will be taken overhead, the following molal flow rates
are readily established:
F = L = 1835. 4 lb moles/hr
V = 1721. 7 lb moles/hr
B = 113. 7 lb moles/hr
and a material balance on the solvent yields
1. 706 = 1721. ty
v + 113. 7X
B
(3. 7)
Now, there are limits to the values that can be assumed by xB The
maximum tolerable value of x
B
is specified as 0. 0004, and from
Eq. 3. 7 the corresponding value of)' v is 0. 000965. At the oter
extreme, if infinite stages were used, x
B
would be 0 and)' v would rise
to 0. 00099. Clearly this range of concentrations does not allow a sig
nificant variation in the usage of steam; the boil-up rate is essentially
fixed by an over-all stripper balance which in turn results from the
specification of a one
-
stage extraction process. This being the case,
the stripper in this instance is designed to minimize its capital cost;
i.e., only one ideal stage is utilized. Based upon one stage and an
L/V ratio of 1,835.4/1, 721. 7 1.067, the McCabe-Thiele diagram for
the stripper design is shown in Fig. 3 -3.
68
Glza/)ie)' 3
0.008
0.007
0.006
0.005
Ys
0.004
0.003
0.002
2) Twos109f case L/V. 1.85
3) Oneos1oQe cose L/V. 1.067
O.OI c
0 0008 0 01)10
Figure 3-3. McCabe-Thiele construction for the
closed steam strippers used in three extrac
tion cases. (Trial and error was used to ob
tain an integral number of stages for each
case with x B " O. 0004.)
The following compositions were derived by trial and error from this
figure and from Eq. 3. 7:
XB = 0. 000105
Yv
= 0. 000983
Qualitatively this result means that the large amount of water that
must be removed and the high relative volatility of the system cause
virtually all the aromatic solvent to be taken overhead in the stripper.
Actually, the material balance for the mier-settler units might be
modiied to include this change in flow rate, but such a computation is
probably not worthwhile.
SUMMAY OF TH MATEIAL BALANCE A STIPPER
CALCULATIONS
As mentioned before, the process shown in Fig. 3-2 may be modi
fied by using multiple mixer-settler units. Clearly the addition of
such units will decrease the required quantity of recycle water and
the corresponding steam usage. Figure 3-4 shows a three-stage
mier-settler process; an extension to other deSigns is obvious from
this figure.
Iterative material-balance calculations and stripper designs have
been completed for extraction processes having one, two, and three
PYocess Improveme1lt foy Liquid-Liquid Ext)-actioJl 69
Mixer
sellier
NO.t
Mlxer
sertler
No.2
Mixer
sertler
NO.3
Roffinote stream
(Solvent recycle)
Auxiliary
water
Aqueous
recycle
Aqueous extract
Condensole
sellier
Stripper
Product stream
Figre 3-4. }'low diagram for a three
stage extraction process.
stages. The computations for multiple extraction stages are quite
analogous to those already presented for the single-stage extraction.
The calculations for the stripper designs are summarized in Fig. 3-3.
In the stripper design for the two-stage extraction, a situation arises
which is analogous to that encountered in the stripper design for the
one-tage extraction; i.e. , the upper and low limits on Yv are nearly
equivalent. As before, a stripper having one ideal stage is indicated.
However, in the stripper design for a three-stage extraction process,
the limits on Yv vary from O. 0049 to O. 0079, and a suboptimization
is required for the rigorous determination of the desired number of
stripper stages. Even so, in this instance the capital and operating
costs for the stripper are so minimal that the number of stripper
stages may be arbitrarily selected within the limits on ,. v set by the
material balance. As such, the L/V ratio in this case has been varied
by trial and error until conditions for two ideal stages have been
established.
As indicated in the original memorandum, a stage efficiency of
100 per cent is assumed in the mixer-settlers. In the case of the
strippers, a more conservative and realistic value of 50 per cent is
assumed for the over-all stage efficiency. Usig these efficiencies,
one obtaino the results summarized in Table 3-5.
70 Chapter 3
Table 3-5. Results of Desig Calculations for Cases Havng One,
Two, ad Three Exraction Stages
Number of extraction stages, n 1 2 3
Water recycle rate, lb/hr 31,000 2,473 255
Auxiliary water, lb/hr 2,007 2,007 2,007
Total water rate, lb/hr 33,007 4,480 2,262
Vapor rate in stripper, V,lb/hr 31,000 2,480 265
Stripper overhead composition, Yv 0. 000983 0. 00160 0. 00505
Solvent content of product, xB 0. 000105 0. 00018 0. 000295
Number of trays in stripper 1 1 3
Number of pumps required 6 9 12
With this information, it is a simple matter to implement the de

sign criteria presented in the original memorandum and to estimate

the capacities required for the mixer and settler units and for the
stripping columns. The capital costs for process equipment then fol
low directly from the cost-correlation equations available in the
memorandum. After the caital estimate, the principal operating costs
for these design cases may be established. The major direct opera
ting costs are those for steam and cooling water; from the information
supplied in the memo, these costs may be estimated as
Steam cost 6. 4 V dollars/year
C IV cost 0. 48 V dollars/year
where V is the vapor rate in the stripping unit exressed i lb/hr.
The results of all the preliminary economic calculations are listed
in Table 3-6; it should be noted that operating costs such as labor,
overhead, and maintenance are not shown. However, the maj or vriable
operating costs have been taen into account, and calculations of the
type shown provide a reasonable basis for comparing various process
alternatives.
OPTMIZATION OF THE MIER-SETTLER PROES
Table 3-6 indicates that the economics of the proposed extraction
process are a very strong function of the water-recycle rate. It
appears that additional economies could be effected by further re
ductions in the water flow rate with corresponding increases in eqUip
ment capital costs. In the limit, i an infinite number of stages were
used, the following implementation of Eq. 3. 6 would apply to the first
Process illlprOl'emelli for Liquid- Liquid ExiYaclioll 71
Table 3-. Economic Evaluation of Processes Having One, Two,
and Three Extraction Stages
Numbr of exraction stges, I 1 2 3
Ca2ita cost estimate
No. Item
, Mier $2,130 $1,726 $2,136
, + 1 Settlers 2,880 1,500 1,600
1 Reboiler 5,640 3,060 1,800
1 Condenser 7,400 1,635 430
3(1 + 1) Pmps 3,760 5,080 6,620
Fi. 3-3 Trays 1,300 374 390
n Controls 3,000 6,000 9,000
Total purchased equipment 26, 110 19,375 22,106
Total installed equipment $104,440 $77,500 $88,424
Oerating cost estimate
Depreciation, 15 per cent $15,680 $ 11, 600 $13,300
Steam, 6. 4 V 198,500 15,900 1,700
Cooling water, O. 48V 14,900 1,190 130
Total operating cost, $229,080 $28,690 $15,130
dollars/year
mixer-settler stage (the stage to which the fresh Fenway acid-sol
vent mixture is fed):
1 + (3,000/123)/(680/246)
0.63
1 + n
w
w/(666/246)
Solving this equation shows that a minimum water flow rate of
39. 6 lb moles/hr or 713 lb/hr is required to complete the separation,
even if an infinite number of stages is utilized. Since this value is
less than that required in the fial product, it is clear that the amount
of auxiliary water will have to be reduced when the use of larger num
bers of stages is to be examined. In such instances, make-up water
could be added to the product stream after the stripping has been com
pleted. It may be show quite easily that, i more than three extraction
stages are used, a water flow rate of less than 2007 lb/hr will be re
quired and a corresponding decrease in the auxiliary flow must be
effected.
72 ClwjJ/er 3
In order to complete the selection of the optimum process, cal
culations have been carried out for four- and eight-stage extraction
processes. For these conditions, the following results have been ob
tained.
Number of extraction stages, n 4 stages
Total water flow rate, lb/hr 1565
8 stages
1000
In both systems, the water flow rate is reduced far below that re
quired in the final product stream. The present calculations assume
that the solubility of Fenway acid salt is not being exceeded. The
validity of this assumption will have to b checked i any processes
having a large number of stages appear economically significant.
Since direct operating expenses are negligible for the four- and
eight-stage operations, the fiancial evaluation for them is straight
forward. Combining these evaluations with those summarized in
Table 3-6, the process economics have been optimized according to
Fig. 3-5, which shows that either a three- or a four-stage unit could
be used with comparable economic results. In order to avoid any
problems that may result from exceeding the solubilities of the Fen
way acid salt, the three-stage unit is a good selection for this pre
liminary design.
!
i


!
M P O
+ +
4~~~ "
*
*

1-(0 f!! c.,,,

I
__Steam and coollng"wa,er
costs
Totol woler flow rote
t
Ib Ihr
Figure 3-5. Financial evaluation of te
extraction process.
SUMMAY OF THE DESIGN RESULTS
The results summarized in Fig. 3-5 indicate that the required
separation could be accomplished by a three-stage mixer-settler
process scheme. The required capital outlay would be approximately
Process Improvement for Liquid-Liquid Extraction 73
$88, 000, and the computed annual operating cost for the extraction
process is about $15, 000, which corresponds to $0. 003/lb of acid
salt. As mentioned previously, this operating cost does not include
several important items, such as labor, overhead charges, equipment
maintenance, and solvent losses. I particular, the cost of solvent
losses is often quite an important contributor to the operatig costs
of an extraction operation. I the additional cost items were included,
the resulting total conversion cost probably would approach $O. Ot/lb
-a value close to those usually incurred for the type of extraction
operation being considered in the present case. Since most of the
additional costs are quite insensitive to the number of extraction
stages, it is unlikely that the inclusion of these factors would cause a
significant alteration in the optimum process design tat has been
selected.
Since only three mixer-settler stages are required, it is doubtful
that a liquid-liquid extraction column would have any significant ad
vantages over the proposed design. However, Table 3-6 shows that
the cost for controlling the mixer-settler operation represents a
large fraction of the total capital investment. This investment might
well be reduced by use of an extraction column; this possibility should
be examined in subsequent design calculations.
Some assumptions made in the preliminary design analysis should
also be investigated in later, more detailed design efforts. First of
all, the residence-time values used in sizing the mixing equipment
should be recalculated. The values used are thought to be reasonable,
but they and the corresponding estimates for efficiency should be re
estimated using literature values of the mass-transfer coefficients
and settling times that are correlated for equipment of the type pro
posed for the present process. Such calculations would bp !artir1l1r
ly important because depreCiation charges constitute a major portion
of the total conversion cost. A second assumption, implicit in the siz
ing of the mixing and settling equipment, was that little or no difficulty
would be encountered in separating the phases; i.e., the two phases
are not easily emulsified. Evidently no dificulty of this type was
encountered in the laboratory. However, small amounts of surfactants
in the plant water system could cause substantial difficulties in the
commercial operation. This possibility needs to be studied thorough
ly before finalizing the design specifications.
A third assumption was that of constant mutual solubilities of the
two solvents. I solvent losses prove to be an important economic
factor, it may be worthwhile to establish these values more accurately.
However, since both the capital and operating costs for the stripper
are quite small in the optimum process, there does not appear to be
much j ustification for further exerimental work at this stage in the
design.
In summary, it is felt that the preliminary economic evaluation
calculations summarized in Fig. 3-5 have very likely established the
optimum design scheme. The clarification of some of the assumptions
and the inclusion of some of the costs not yet determined may alter
the size of the equipment and the operatig costs for the ultimate pro
cess. However, it is unliely that any of these factors will cause the
74 Chapter 3
optimum type of flowsheet to be greatly different from that of the
three-stage unit here established.
REFERENCES
1. Atkinson, A. C., Cher. Eng. 73, No. 10, 149 (1966).
2. Hougen, O. A., K. A. Watson, and R. A. Ragatz, Chemical Process
Principles, Part I, John Wiley & Sons Inc. , New York (1954).
3. Hunter, J. S. , Chem. Eng. 73, No. 7, 111, (1966) .
4. Hunter, W. G. , and A. C. Atkinson, Chem. Eng. 73, No. 12, 159 (1966).
5. Hunter, W. G. , and M. E. Hoff, Ind. Eng. Chem. 59, No 3, 43 (1967).
6. Sherwood, T. K., and R. L. Pigford, Absorption and Extraction,
McGraw-Hill Book Company, Inc. , New York (1952).
7. Treybal, R. E. , Mass Transfer Operations, McGraw-Hill Book
Company, Inc. , New York (1955).

4. Catal
y
tic Reactor Design for Benzene
H
y
drogenation
This case examines set'eral process allemalil:es for hydrogellalillg
ben wile /O j)rodllce cyclohexalle. The highly exolheYillic nature of Ihe
Yeaclioll necessi/ales cOllsideration of alleY/la/h'e lIlelhods of 1'e
JJ/o1'ing /he heat of 1"eaction. While liIlle cOllsideralioll is gil.'ell to
eualliatillg the ecollomics of the variolls processes considered, primar),
emphasis is placed on a deterlllillation of tile imjJOrtant deS(1/ l'ari
abIes alld on a prelimillary selection of all appropriate reactor COIl
figuration. ExtensiL'e lise is made of illterpolatil/g betll'eell alld extra
polatillg frolll felt' des(1Z condiliolls tllal are treated ill detail. lVi/h
ill the lim it s of tile a/>proxilllaliolls IIsed, prelim illary desigll scheme
is selected.
Cher-Fiber Cororation
Houston, Texs
To: Fred Schwartz, Process Desig
From: Ciyde Fox, Group Leader
Rc: Cyclohexane Production
The planned exansion of the company's nylon production will re
quire an additional 40 million lb/year of cyc10hexane as a raw mate

rial in the manufacture of adipic acid. Primary consideration is being

given to the possibility of buildig a new plant to product cyclo
hexne by the catalytic hydrogen3tion of benzene. Your Process
Desig group is asked to prepare a preliminary design for the hydro
genation process. At this early stage it would be most desirable to
develop the basic equipment requirements.
1b
/0 Clapler 4
For the time being it will not be necessary for you to complete
an economic evaluation of the process. The Market Research group
is currently negotiating with raw-material suppliers as well as with
possible outside sources of cyclohexane. When this group has assem
bled the appropriate market information, the process recommended
by your group will be evaluated. In this way a decision regarding the
i-house manufacture of cyclohexane can be made.
The Research Laboratory has come up with a new catalyst that
seems to be as good as or better than any we might buy or license.
This catalyst is basically nickel on alumina and would be provided in
the form of 3/16-in. pellets at a cost of $1. 40/lb (bulk density 88 lb/
cu ft). It is rugged and can be regenerated by burning off the carbon
in inert gas containing abut 1 to 2 per cent oxygen.
The laboratory has made a thorough study of the performance of
this catalyst, using bench-scale flow reactors. The results obtained
confirm the literature reports for platinum, palladium, and nickel
)4 S)the hydrogenation has been found to be first-order in
hydrogen pressure and zero-order in both benzene and cyclohexane.
The activation energy is found to be 12 Kcal
/mole, confirming the
literature data. Results from a representative laboratory test are
attached.
The reaction proceeds readily to clean and complete conversion of
the benzene to cyclohexane, with only small amounts of byproducts,
such as methylcyclopentane. The design problem hinges on the fact
that the reaction is highly exothermic and must not be allowed to
"run away." Designs must be developed to maintain stable operation
at an appropriate temperature; your group is asked to determine the
best approach to the deSign.
The Process Development group recommends that the tempera
ture of the reacting gases be held to a maximum of 6000K because
hydrocarbon decomposition resulting in excessive carbon deposition
occurs at higher temperatures. Carbon deposition also appears to
result from high pressures, although high pressure helps the con
version. For the present, plan on a maximum pressure of 150 psia.
DATA A ASSUMPTIONS
Summar
y
of a Representative Labrator
y
Run
The reactor was a tube of 1. O-in. i.d. contaiing a J_-in. o.d. axial
thermowell. The annulus was packed from a length of 36 in. with the
3
/
16
-in. nickel-on-alumina catalyst pellets. The reactor was wound
with heating tape and the resistances so adjusted that the reactor
operated adiabatically.
In one representative test the hydrogen and vaporized benzene
were preheated and fed to the reactor at 227C. The benzene feed
rate was 620 g/hr, and the molal ratio of hydrogen to benzene H the
feed was 10/1. The total pressure was 10 atm. The liquid product
collected in a dry-ice and acetone trap was found to contain 0.798
mole fraction benzene and 0.202 mole fraction cyclohexane.
L0!0!yC 00COI L0Sg
The Research Laboratory suggests that the work of Amane and Par
ravano [J),Kassel (5)and Smith [/} be consulted in order to provide
additional insight into the reaction kinetics for the vapor-phase hydro
genation of benzene. In addition, the work of Smith and Meriwether
(8) is helpful in that it presents results for the liquid-
phase hydro
genation. Finally, the text of Grifith and Marsh (4) provides an excel
lent discussion of the hydrogenation mechanism H terms of the geo
metrical characteristics of various catalysts.
Mass Transfer to and Difusion Withi Catalyst Pellets
It should be noted that the effects of mass transfer to the catalyst
pellet surface and diffusion within the catalyst pellet have been
analyzed. It has been found that neither of these afects the observed
results significantly.
Thermodynamic Data
(From American Petroleum Institute, "Selected Values of Physical
and Thermodynamic lroperties of Hydrocarbons, " Project 44,
Carnegie Press, Pittsburgh, Pa.1953.)
Cyclohexane
Alf
HH
T,oK
Kcalg mole gcal/g mole log
l
O
/ )
0 -20. 01 0
298 -29.43 4, 237 5.5605
400 -31. 70 7,352 -11. 2861
500 -34.08 11,425
-14.8932
600 -35. 57 16, 404 -17.4318
700 -36.59 22, 148 -19.3103
Benzene
0 24.00 0
298 19.82 3, 401 -22. 7143
400 18.554 5, 762 -19.1271
500 17.536 8, 750 -17.1521
600 16.711 12, 285 -15.9040
700 16.040 16,267 -15.0519
o Claptey 4
Hydrogen
0 0
298 2,024
400 2, 731
500 3,430
600 4, 130
700 4, 832
Heats of Vaprization
Benzene at 298K: 8030 cal/g mole
Cyclohexane at 298K: 7895 cal/g mole
Raw-Material Values
The raw-material costs are subject to the results of current
negotiations. However, the following approximate prices may be use
ful:
Benzene: $0.04/lb
Hydrogen: $0.40/MSCF
Oeration
The process should be operated 8000 hr/year.
Suggested Nomenclature

h
All!
k
Activation energ for the hydrogenation reaction, equal to
12 Kcal/ g mole
Equal to exp (12, OOO/RT) evaluated at absolute tempera
ture T
Temperature interval used in the evaluation of an integral
by Simpson's Rule, oK
Molal enthalpy of component l, g cal/g mole
Molal enthalpy of a compound or element in the gas phase at
any temperature when the molal enthalpy of the compound or
element is taken to be zero at a temperature of OOK, g cal/g
mole
Isothermal enthalpy change for the formation of a pure com
pound from its elements at any temperature, g cal/g mole
Reaction rate constant, lb mole/(hr) (atm) (cu ft of catalyst
bed)

P:
Q
?

1
V
B
W
7
Lulu]C Reactor Dcsi~t /9
Pre-exonential factor, lb mole/(hr) (atm) (cu ft of catalyst
bed)
Equilibrium constant for the formation of a compound from
its elements at any temperature, dimensionless
Total catalyst bed length, ft
Moles of hydrogen per mole of benzene in the feed stream
Molal flow rate of component in a catalytic reactor,lb
moles/hr or g moles/hr
Partial pressure of component l atm
Rate of heat exchange between a thermodynamic system and
the surroundings, g cal/hr
Rate of benzene hydrogenation, (lb mole benzene reacted)/
(hr) (cu ft of catalyst bed)
Universal gas constant, equal to 1. 985 g cal/(g mole) (OK)
Absolute temperature, oK
Volume of catalyst bed, cu ft
Rate of work interaction between a thermodynamic system
and the surroundings, g cal/hr
Distance from the inlet of a reactor, ft
Fraction of benzene converted to cyclohexane
Total pressure, atm
Subscripts

D
1
X
0
J,
1:
Refers to benzene
Refers to cyclohexane
Refers to a condition in the feed stream to a reactor
Refers to hydrogen
Refers to a condition i the product stream from a reactor
tube of length
Refers to a condition in a reacting stream at a distance 7
from the inlet of the reactor
Refers to conditions in the feed stream to reactor 1 in Fig.
4-4
Refer to the reactors identiied by these numbers in Fig.
4-4
Superscripts
, (prime) Refers to an intermediate feed stream as shown in Fig.
4-4.
80 Clwp/er 4
AALYSIS OF LABORTORY KETIC INFORMATION
Before proceeding to the reactor design, it is necessary to analyze
the kinetic data obtained in the laboratory. The kinetic constants ob
tained from the analysis may then be used in the subsequent design
calculations.
The chemical reaction that takes place in the reactor is as follows:
O
+
3H2O
4.
The use of the stoichiometry of Eq. 4. . helps D tabulating the flow
and enthalpy quantities in the manner shown in Table 4-1. This sort
of table is useful in accurately preparing the heat and material
balances necessary for the data analysis and design calculations.
Table 9-1. Molal Flow Rtes 0 Enthalpies
Basis: . mole feed
Compound Feed
Hydrogen ....
Benzene ....
Cyclohexane ..
Total ...
At conversion .
...... .... .
.... .
.....
..... .
Molal enthalpy
''
I
H
(H H+ Alj}B
|l/ 1l + Al)
J L
An approximate calculation of the compressibility factor for the
gas mixtures under consideration indicates that the ideal gas law may
be used with negligible error. Therefore, at any position 7 in the
reactor tube the partial pressure of hydrogen may be exressed as
..
For the particular conditions of the laboratory exeriment, wherein
a hydrogen/benzene molal feed ratio of .. was used, Eq. 4. . reduces
to
..... ..... .
)

. ......
(4.3)
where . denotes th degree of conversion of benzene to cyclohexane.
It is useful to evaluate the hydrogen partial pressure of the inlet and
outlet streams of the laboratory reactor. Since has a value of unity
Catalytic Reactor Desi 81
for the feed and a value of O. 202 for the product stream, the hydrogen
partial pressures are computed as
P
F
~ 0.9091 atm
P
_
_

O. 9091 - 0.2727 (0. 202)

0.903 atm
1. 0 - 0.2727 (0.202)
Since the hydrogen partial pressure vries only slightly through the
reactor, an average value of O. 9061 may be used for PJ and may be
assumed to be constant for purposes of calculation.
In order to proceed with an analysis of the laboratory data, it is
helpful to visualize the exerimental apparatus as shown in Fig. 4-1.
Writing a material balance on the differential element, one obtains
cNe = -dN
B
~ r(dV
where the reaction rate may be written as
Combining Eqs. 4. 4 and 4. 5, we get
-N
B
= kp (dV
O
_

kP
I
dV
Figure 4-1. Schematic diagram of hydrogenation reactor
tube.
(4.4)
(4.5)
(4.6)
As indicated in the original memorandum, the kinetic constant may be
exressed by an Arrhenius exression utilizing an activation energ
of 12 Kcal/g mole. Also, as noted previously, the partial pressure of
hydrogen may be equated to 0. 906 + Therefore, Eq. 4.6 reduces to
_0.?0?
q1
?,000/u

_
~
0.906
1ko
V
0
N
B
F
(4.7)
8
Chc[Icr 4
To proceed, the degree of conversion must be related to temperature.
This relation is accomplished by writing a first-law energy balance
on the portion of the reactor tube up to length X!
It is to be assumed that the system is adiabatic and that there are
no work interactions between system and surroundings. In addition,
at steady state the time-rate variation of the internal energy of the
system is zero. Therefore, the first law reduces to an eqution of
the feed-stream enthalpy with the enthalpy of the partially reacted
stream at any pOint H the tube. Hence,
(4.8)
where from Table 4-1
and
1
_ItF F
0.9091
|
- It)
IIF
0.0909 [H " I[
*
All,)
BF
N
TF
1
_Iixx = (0.9091 - 0.2727 O )
[
" It)
N
N
T
z
+ 0.0909 (1 " o ) (Ir - H
*
Al,)
B
+ O. 0909 O (Il" H AII
f
) (4.9)
In the laboratory experiment a feed temperature of 5000K was used;
the enthalpy of the feed stream is therefore computed as
1
N
T
F
_ Hi
F
l. = 0.9091 (3430)
+ 0.0909 (8750 + 24,000)
= 3118 + 2977
6095 cal/g mole of feed
Utilizing Eqs. 4. 8 and 4. 9, one obtains
5095 - 0.9091 [H - H)
O =

0.0909 (Il It
T I
f)
C
- 0.0909 [H - I[ AIlf)

- 0.0909 |' l[ AIj)

- 0.2727 [ -
I
)
(4. 10)
CaIu!y!ic Reactor Design 83
Over the temperature range liely to be encountered in the laboratory
or i a plant-scale reactor, the heat capacity of each component may
be assumed constant and each of the enthalpy terms may be assumed
to be linear in temperature. With these assumptions, O is also linear
in temperature. It now becomes convenient to evaluate the numerator
and denominator of Eq. 4.10 as functions of temperature; this evalua
tion is carried out in Table 4-2. For an analysis of the laboratory
reactor, only enthalpy values of 500 and 600" need be considered.
However, enthalpy values at other temperatures will be useful in later
stages of the design and have been included in the table. In line with
the assumption of constant heat capacities, the enthalpy at 550" has
been computed by averaging the values at 500 and 6000K.
Table 9Z, Computation of for Adiabatic Hydrogenation
Temperature, oK
298 500 550 600
[H 1I 2,024 3,430 3,780 4, 130
- H T
Ij)
B
27,401 32,750 34,516 36,285
(Ir - H
T
Ij) c -15,773 -8,585 -6,095 -3,606
0.9091 [Il - H
I
D
3,118 3, 755
0.2727 (H - H
D
935 1,126
O. 0909 (H It)
B
2,977 3,298
0.0909 (H - H - I
l
)
c
-780 -328
Numerator of Eq. 4.10 0 -958
Denomiator of Eq. 4.10 -4,692 -4,752
O, Eq. 4.10 0 0.2016
The fact that a value of zero is computed in Table 4-2 for at
5000K is consistent with the experimental situation; the reactant mix
ture was fed at 5000K, and at that temperature no reaction had taen
place. From a first-law balance, a conversion of about 0.202 was
computed at 6000K; therefore 6000K must have been the temperature of
the product stream leaving the reactor. Moreover, since the molal
enthalpies are assumed to be linear functions of Tthe results of
Table 4-2 may be summarized as
O =0.002016 [T - 500)
dO = 0.002016 d1 (4.11)
The results of the energy balance may now be combined with the
material balance by substituting Eq. 4. 11 into Eq. 4. 7:
a00
0.906 ko V

q
i?,
000/ _
T
=
mmmm
s00
2.016 Y 10-
3
NB
F (4.12)
oJ Chapter 4
The left side of Eq. 4. 12 may be integrated numerically by using
Simpson's rule; the calculations are shown in Table 4-3.
Implementing Eq. 4.12 to obtain the pre-exponential factor, one
obtais
H
('/
]

[
:
4
d.0[53 'q It
620
NBF
= = O. 0175 Ib moles/hr
(78) (454)
600
e12,000/R
T
500
h

T = [1
500
* 160
0 *
4(
fs10 *J530
*
.
`
1
590)
3
T
2
(
s20 *
fS40
T
...
IS
8
0)]
10
= - [201, 823
*
4(358,192)
*
2(266,839)]
3
10
= - [2,168,269]
3
= 7,227, 563K
(7,227,5 63) (2.016 ^ 10-
3
) (0.0175)

9.06 (0.0153)
= 1838 lb mole benzene/[hr) [cu ft of catalyst)
(atm)
It is interesting to determine the degree to which the rate constant
varied within the laboratory reactor during the exeriment just
analyzed. The following ratio is readily formulated:
R [ 1 1 ]

=
e12,0001R
1 T
k
L
P
For the particular case in question,

_ [ 600 - 500 ]
=
e12,00
0/1
.9
8
S
(600)(500
)
k
F
(4.13)
=
e202
7.5
U variations in the hydrogen partial pressure are neglected, it is
seen that the chemical reaction rate at the reactor outlet was more
than seven times faster than that established when the feed mixture
first encountered the catalyst. This sort of variation not only causes
Catalytic Reactor Desig 85
difficulties in analyzing the performance of a reactor for hydrogena
ting benzene but will also be particularly significant in the economic
optimization of a commercial scale process.
Table 9~, Evluation of the Itegral H
Q. 9.JZ by Simpson's Rule
T 12,000/RT

q
i?,ooo/u

500 12.090 178,081

510 11.853 140,507
520 11.625 111,831
530 11.406 89,848
540 11.194 72,673
550 10.991 59,338
560 10.795 48,743
570 10. 605 40,330
580 10.422 33,592
590 10.246 28,169
600 10.075 23,742
GENERL DESIGN CONSIERA.TIONS
Before considering the design of the complete process, it is useful
to establish the constraints that must be imposed upon the composi
tion of the gas mixture fed to the reactor. To this end, the thermo
dynamiC data given in the original memorandum have been used to
construct Fig. 4-2, which shows the effect of temperature and feed
composition on the equilibrium conversion of benzene to cyclohexane.
Examination of Fig. 4-2 shows that:
(i) U a temperature level of 6000K is to be used to activate the
reaction, equilibrium considerations suggest that at least five moles
of hydrogen per mole of benzene must be included in the feed stream.
(ii) From the viewoint of chemical equilibrium, the advantages of
increasing the hydrogen concentration diminish rapidly after a value
of ~ 5 has been reached.
(iii) With a fixed pressure ceiling of 10 atm, equilibrium dictates
that a few per cent of the benzene will remain unconverted if the maxi

mum temperature of 6000K is utilized.

(iv) For total pressures i excess of a few atmospheres, equili
brium limitations only start to become important at approximately
600oK.
80 Chapter 4

0F 6

8 0
O.I| {
j

Lne

8
C
0

f
6
1k n
500 7
500 5
500 5
600 W
600 7
600 5
600 5
_0.0| m
O
t
0.00| -- -- ---
0.l |0 |00 .O0 |0,O0
1_ lOlol pr088ut60!m
Figre 4-2. Equilibrium conversion of benzene for
various process conditions.
Moreover, this sort of high-temperature design would require a
large investment in heat-transfer area and control equipment. Thus,
an economic balance should be struck that accounts for the capital
costs for catalyst, reactor volume, heat transfer, and reactor control,
and for the important running costs-in particular the benzene and
hydrogen recycle costs.
As an illustration, the reactor volume can be reduced by increas
ing the hydrogen partial pressure in the feed. However, this gain
tends to be offset by the increased operating costs for hydrogen re
cycle. The present problem is typical of many design studies in that
a number of counteracting factors have to be balanced in order to
establish the over-all economic optimum.
The total catalyst volume requirement will be used as the initial
indication of the desirability of various types of design. Clearly
other costs, such as those for heat-transfer surface, hydrogen and
benzene recycle, and reactor controls, will signiicantly affect the
final design selection. However, for the preliminary analysis re
quested in the original memorandum, the catalyst volume should serve
as a sensitive indication of the relative values of various types of
design.
ANALYSIS OF LIMITING DESGN CASES
Destablishing reasonable estimates for the reactor volumes requir
ed by various process alternatives, it is desirable and exedient to
utilize some approximations. First it will be assumed that the tem
perature and feed compositions to be used will be such as to allow
complete conversion of benzene to cyclohexane. The greatest error
caused by this approximation will be that at 600oK, where a small per
-
Catalytic Reactor Design 8/
centage of the benzene will remain unconverted. This is a negligible
error for purposes of establishing reactor volume, although this small
percentage could greatly magnify the separation and recycle costs.
Second, it will be assumed that the zero reaction-order observed for
benzene H the laboratory run has application over the entire range of
benzene conversions, i.e., for 0 1. There is danger in extra
polating a kinetic exression outside the concentration range where
the data were taken; indeed, H the present case the reaction order un
doubtedly changes in the range of high conversions. However, in the
absence of more extensive experimental data, the zero-order exres
sion will be assumed to apply. Finally, all designs presented will as
sume the availability of pure hydrogen to be used in the feed. In many
commercial installations the available hydrogen contains substantial
amounts of ierts, and these have to be removed by a bleed stream on
the hydrogen recycle. The flow sheets presented in this study will pro
vide for such a bleed stream, but no consideration of hydrogen purity
will be made in the design calculations.
With these thoughts H mind, it is now useful to establish the re
quired reactor volume if the laboratory reactor were scaled directly
to commercial size. In this instance, the catalyst volume required for
the specified production rate follows directly as
( 40 Y 106 ) ( 0.0153 )
V 258 cu ft
8000 X 84 0.0175 X 0.202
Under the conditions of the laboratory test, the temperature of the
reaction mixture rises adiabatically from 500 to 6000K, and a total
operating pressure of 10 atm is used. The hydrogen partial pressure
falls from 9.09 atm H the feed gas to 9.03 atm at the exit of the reac
tor.
Clearly some advantages can be realized by operating at the maxi
mum allowable temperature and with large excesses of hydrogen. U
the limit, if a reactor were operated isothermally at 6000K, with a
hydrogen partial pressure deviating only slightly from 10 atm, the
following bed volume would result:
and
8 1838
q1
?,000/a00)

1838
23,742
0.0774 lb moles/(cu ft) (atm) (hr)
V
(40 X 106 ) ( 1
) 77 cu ft
8000 84 10 X 0.0774
This value constitutes the lowest attainable limit for the catalyst
volume under the restrictions of the specified design criteria.
85 Chu[Icr
There is no upper limit on the bed volume. However, an arbitrary
upper limit can be established by specifying an isothermal operation
at 5000K with a large excess of hydrogen having a partial pressure of
7.0 atm. Since the reaction rate is first order in hydrogen, 7.0 atm
would appear to be a reasonable lower limit for the values of ]
which should be considered in the design. Under these conditions, the
bed volume is computed as
and
1838
R ~ ~ 0.01033 lb moles/(hr) (cu ft) (atm)
178, 081
(10 0.0774 )
V ~ 77 - X
825 cu ft
7 0.01033
Thus by allowing the feed mixture to react adiabatically from 500 to
6000K, the bed volume is reduced from 825to 258 cuft, and a significant
step has been taken i reducing the catalyst volume to its theoretical
lower limit of 77 cu ft.
An adiabatic reactor would not require the substantial heat-trans
fer surface that would be necessitated by the isothermal operation.
Moreover, the temperature level of the isothermal operation would
require a high-temperature heat
-transfer fluid (e.g., Dowtherm or
Therminol) together with a specially designed boiler system. These
exenses can be avoided or at least substantially reduced by maing
proper use of adiabatic reactor operation.
It has been demonstrated that reactor volume can be minimized
by operating as closely as practicable to the maximum allowable tem
perature. Moreover, it obviously would be advantageous to devise a
design that would allow a stable, high-temperature operation without
the need for large quantities of heat-transfer surface. These objec
tives can be accomplished by the general type of design shown i
Fig. 4-3. In such a design, only a portion of the benzene is fed to the
inlet of the reactor. The remaining benzene enters as a liquid at
regular intervals along the length of the reactor. In this way the
latent and sensible enthalpy needed in bringing the benzene to reaction
temperature balances a portion of the heat of reaction and serves to
stabilize the reactor operation.
At this point it is well to emphasize that the over-all heat-tras
fer requirement (Btu/lb of product) is independent of whether an
adiabatic or isothermal design is used. The type of design show in
Fig. 4-3 has as one of its primary objectives that of reducig the
cost of removing the heat of reaction; the net amount of heat cannot
be changed.
In the reactor desig shown, a large excess of hydrogen may be
required to maintain a high hydrogen partial pressure and a cor
respondingly high reaction rate. This practice, of course, leads to an
increased requirement for hydrogen recycle and to correspondig
Catalytic Reactor Design 89
0m~
t
Figure 4-3. Adiabatic reactor design.
cooling, recompression, and heating expenses. The hydrogen-recycle
costs can be reduced by feeding liquid cyclohexane, along with hydro
gen and liquid benzene, at regular intervls along the reactor. The
advantages of reduced recycle costs are balanced at least partially
by the need for an increased reactor volume and by the increased
separation costs resulting from cyclohexne recycle.
Clearly an infinite variety of reactor designs may be conceived
that would incorporate one or more of the aforementioned modifications.
One of these, show schematically in Fig. 4-4, will be considered to
demonstrate the advantages of a split liquid feed.
The design shown in Fig. 4-4 maes use of five fixed-bed reactors
with equal amounts of benzene fed to each reactor. The gas entering
the first reactor is to be preheated to 550oK, and the temperature of
the gas leaving each reactor is specified as 600oK. The total pres
sure of the gas in each reactor is taken to be 10 atm, and the hydro
gen partial pressure will be speciied as 7 atm at each reactor out
let. All the benzene entering an individual reactor is assumed to be
completely hydrogenated i that reactor.
^se
a:.:+^: a,.^;,.^-a ;,.;,.s ;+.;+.+

H80y0|8 H,
l88d S\t80m
H80y0!8 D8n28n8
uSSum8d SmoII(
Ly0lOh8x0n8 prOdu0!
6OO
Figure 4-4. Five independent adiabatic reactors.
90 Chu[l
q
r d
Any number of reactors could of course be used H a design of the
type show in Fig. 4-4. A large number of reactors would lead to an
isothermal operation at 6000K ( and to the previously calculated mini
mum total bed volume i _ were maintained near 10 atm} . However,
as the number of reactors is increased, so are the piping and instru
mentation requirements. An optimum number of reactor stages
should be established for the final design, but the five-reactor system
show in Fig. 4-4 should effectively illustrate the inluence of divid
ing the fied bed into a number of individual sections. To determine
the degree of this influence, the design calculations for the selected
case will now be completed.
DESIGN CALCULATIONS FOR THE FIVE-REACTR SSTEM
As a basis for the design, it is convenient to choose
(4.14)
Since the benzene is assumed to react completely in each section of
the reactor, it is noted that
(4.15)
With the basis selected, the analysis of the reactor system follows
directly.
First Reactor
Since the benzene is assumed completely converted, a material
balance is quite simple:
An additional relation is obtained from the limitation imposed upon
the hydrogen partial pressure:
a

N
IO
-
3
N
B
O

0.7
Nlo -
3
N
B
O
+
Nco
+
N
B
O
and, from Eq. 4.14
,

I
O
(
1
6/3) + (7/3)
Nco
(4. 16)
Maing use of the enthalpy quantities given in Table 4-2
,
Eq. 4.8 may
be applied to the first reactor as follows:
34, 516
N
B
O
-
6095
Nco
+ 3780
I
O
= 4130 (I
O
-
3
NBo
)
-
3606
(
N cO
+
* BO
)
or
50
,
512
-
2489
Nco
- 3
50
i
I
O
0
Catalytic Reactor Design V1
(4.17)
Combining Eqs. 4. 16 and 4.17
,
values for
N
c
, '
I
O
'
NCl
' and
I
1 are
obtaied; these values are show in Table 4-4.
Table 99, Smmary of the Flow Qantities for the Adiabtic
Reactor System Sown H Fig. 9~
B
asis:
NBO

NBI
N
2

N
3

NI
4 1. 0
NBl

NB2

NB3

NB4

NBS

-
0
Reactor Stream
NB
N
c
N
n
1
Feed, N 1.0 14.67 39.54 7.15
Product,
Nl
0.0 15.67 36.54 7.0
2 Feed,
N
l
' 1.0 0.966 7.58 7.15
Product, 0.0 17.64 41.12 7.0
3
Feed,
N2
' 1.0 0.966 7.58 7.14
Product,
N
3
0.0 19.60 4
5
.70 7.0
4
Feed,
N
3
' 1.0 0.966 7.58 7.13
Product,
N
4 0.0 21. 57 50.28 7.0
5
Feed,
N
4' 1.0 0.966 7.58 7.11
Product,
Ns
0.0 23.53 54.86 7.0
Second Reactor
Applying material balances around the first mixing point and the
second reactor, one obtains
NC2
=
NCl
+ NB'l
+ N
el
= 16.67 + Nel
12

I1
+
!1 3Nn'l
33.54
+ ;1
(4. 18)
(4. 19)
Again applying the specification on the hydrogen partial pressure and
utilizing Eqs. 4.18 and 4.19
,
there results
33.
5
4
T
;l
. 7
~
50.21 +
!l
T N
e
l
or
(4.20)
An enthalpy balance on the second reactor yields the final required
equation:
92 Clu[Ic: 4
(4.21)
Note that the specific enthalpy quantities used in Eq. 4. 21 have been
obtained from Table 4-2. As mentioned previously, it is anticipated
that the use of a linear average to obtain the enthalpy values of 5500K
should be quite accurate. By combining Eqs.4.18 to 4.21, the values
for
N
dl' :l' N_

,and

have been obtained and are tabulated in

Table 4-4.
Other Reactors
With use of methods analogous to those just presented, the material
balances have been completed for all five reactors shown in Fig. 4-3;
the results are also summarized in Table 4-4.
In order to proceed with te chemical kinetic calculations, it is
necessary to establish the temperature at the entrance to each reac
tor. The temperature of the feed miture entering the first reactor
has been specified as 550"K, and the temperature of the product mix
ture leaving each reactor has been fixed at 600oK. However, the tem
peratures of the mixtures entering reactors 2 to 5 must be computed
by a series of enthalpy balances. The calculations summarized in
Table 4-5 serve to facilitate the estimation of the temperature at
the inlet of the second reactor. The molal quantities used in Table
4-5 were taken from Table 4-4, while the enthalpy values were deri
ved from Table 4-2.
Table 9~, Calculations for Estimation of the
Entrace Temperatre for Second
Reactor
Enthalpy Quantity 1 5500K T~ 6000K
(NB1
+
N;l
)
lB
34,516 36,285
(NC1
+
Nc)l
c
-101,400 -60,000
(.n
T
i/l)/
166,800 182,500
Total enthalpy 99,916 158,785
Catalytic Reactor Design 93
With use of the assumption of liear enthalpy values, the total
enthalpy of the stream entering the second reactor may be exressed
as

_ (
N
il T 2)li 158
,
785
-
1177 (600
-
T

___ l
(=3
Since the flow rate, composition, and temperature of the stream leav
ing the second reactor have already been established, the total enthalpy
in that stream is computed i a straightforward manner to be 106, 400
g cal. Because the reactor is operated adiabatically with no work
interactions with the surroundings, the enthalpies of the inlet and out
let streams may be equated, and the entrance temperature for the
second reactor is obtained:
T 600 - (
52
.
385
) 555. 5K
IT
1177
By analogous methods, the inlet temperatures for the remainig three
reactors are computed as 560.2,564.9, and 569. 6K for the third,
fourth, and fifth reactors, respectively.
An exression for the volume required in each of the five reactors
is easily obtained by returning to Eq. 4.7, whereby
(4.22)
The temperature in each reactor must be related to the degree of
conversion to complete the integral. This is easily accomplished
by reference to Fig. 4-5, since it has been shown that Tis linear in
& With the aid of this figure, the integral in Eq. 4. 22 may be evalua
ted either graphically or numerically. To compare the results for
the two extreme situations, the integration has been carried out
graphically in Fig. 4-6 for the first and fifth reactors.
The two curves shown in Fig. 4-6 were integrated graphically, and
the following results were obtaied:
1
Reactor 1:
_12000/!I_g ~
38
,
308
0
1
Reactor 5:

_12000/ _Q
3
1
,
168
0
Since a basis of one mole of benzene feed to each reactor has been
used in the present analysiS, the appropriate value of W__ for use in
Eq. 4. 22 is computed as
40 Y 106
11. 9 lb moles/hr
(8000) (84) (5)
V+ Chapter 4
Z .9 .b . .
w=,0
Figure 4-5. Temperature profiles
in the five reactors. (Each
curve is identified with a reactor
number.)
By use of this feed rate, the hydrogen partial pressure values from
Table 4-4, and the values from the graphical integration, the volumes
required for the first and fifth reactors may now be evaluated as
follows by means of Eq. 4. 22:
(11. 9) (38,308)
!
t
= ~
35. 1 cu ft
(1838) (
7
.07)
(11.9) (31,168)
!

=
(1838) (
7.05)
= 28.6 cu ft
These two values are not greatly different, and it is clear that the
volumes required for the second, third, and fourth reactors will be
distributed approximately linearly between these two extreme values.
Therefore, with reasonable accuracy, the total reactor volume for the
design shown in Fig.4-4 may be computed as
V

% (28.6+ 35.1) = 159 cu ft

.
|
0
~

60
^
6!0
5
\
1"0' I
'
"
Lul0]!C Reactor Dcsin 95
5
`
'
40
3J0
..
`
`

`

5

.:


5
2.2I0
0 0.J 02 0.3 0.4 0.6 0.6 0.7 0.6 0.9 |0
6gt66 0 C00v6r5t00
t
Q
Figre 4-6. Curves for the graphical integration
1
of ]
_12000m

_
_
0
This is a substantial reduction from the 258 cu ft that was obtained by scal

ing up the conditions from the laboratory experiments. Table 4 -4 shows

that the hydrogen partial pressure deviates only slightly from 7.0
atm.
DISCUSSION OF THE COMPUTED RESULTS
Figre 4-7 provides a convenient means for comparing the results
of the design calculations that have been completed. It shows reactor
volume plotted as a function of the hydrogen partial pressure; results
for various types of reactor operations are presented. The curves
for the isothermal operations are easily computed since, for these
cases, reactor volume is inversely proportional to p. The point con
ditions for the scale-up of the laboratory run and the five-reactor de
sign are also shown. In both of the adiabatic cases it is possible to
extrapolate the results to lower pressure by assuming that reactor
volume is inversely proportional to hydrogen partial pressure. This
corresponds to a reduction in total pressure while maintaining a
constant mole ratio in the feed. By this means of extrapolation, the
enthalpy balances used in determining the two point conditions re-
main valid at lower pressures. A simple extrapolation to higher hydro
gen partial pressures is somewhat tenuous, since the limitation of
10 atm on the total pressure would require a change in the enthalpy
Vb Chapter 9
balances. However, the effect of hydrogen on the enthalpy balance is
probably not particularly influential, and extrapolation over a modest
range of pressures, as show in Fig. 4-7, is probably warranted.
'0Or
900
sc

001
_-c~

borhetmoi tecc,,
6C0*8

O s 9 |
AverOge hydrogen pOrIi0I res8ureo!m
Figure 4-7. Reactor volume as a
function of hydrogen partial
pressure.
Up to this poit, much emphasis has been placed on the estimation
of the catalyst volume necessary to produce the required amout of
cyclohexane. Although catalyst and reactor capital costs may be
significant factors in the economic evaluation of a design, there are
other important factors that warrant detailed consideration. Oly in
Significant cost information was given in the problem memorandum,
and, in fact, an economic evluation of the reactor design was not re
quested. With this situation in mind, other factors that will eventually
have impact on the final economic evaluation will now be considered
i a qualitative sense.
With reference to Fig. 4-7, the advantages of a high-temperature
operation in reducing the reactor volume are obvious. However, as
mentioned prevously, a high-temperature isothermal operation would
require large quantities of heat-transfer area; this sort of design is
also vulnerable from the standpoint of possible stability problems. It
is seen that the design incorporating five individual, adiabatic reac
tors achieves nearly every possible economy in reactor volume while
Catalytic Reactor Design V
simultaneously providing the basis for a stable reactor operation. It
also should be noted that the adiabatic design requires only a minimal
quantity of heat-transfer area associated directly with the reactor
system.
There are approaches to the heat-removal problem other than the
one suggested in Figs. 4-4 and 4-5. For example, reactor temperature
could be controlled by the injection of liquid water at various points in
the process stream. Or intermediate waste-heat bilers could be em
ployed between reactor sections. These alternatives are attractive
in that they avoid the increased separation charges associated with
recycling cyclohexane. The practicality of utilizing these alternatives'
might well be considered in subsequent work.
A Significant subject for a detailed economic evaluation lies in the
selection of the exact type of reactor system to be used. For example,
Fig. 4-7 indicates that about 170 cu ft of catalyst can be saved by using
the five individual adiabatic reactors rather than the single adiabatic
unit scaled up from the labratory run. Obviously this advantage is
offset to some degree by the increased costs associated with building
and controlling five individual reactor units. By balanCing the con
flicting catalyst and equipment costs, an economic optimum can be
established, and the proper reactor system can be selected.
U addition to the designs already conSidered, Fig. 4-7 indicates
that the case of a single adiabatic reactor with a feed temperature
of 5500K should be investigated. This analysis will lead to a curve
falling between the 5500 and 6000K isothermal curves shown in Fig. 4-7.
This case would have the advantage of a single-unit construction;
however, the process control requirements for the single reactor
might be more difficult to satisfy than in the case of five individual
reactors.
The original temperature limit of 6000K resulted from the neces
sity of maintaining catalyst activity. Figure 4-2 shows that the ther
modynamiC limitations on the conversion per pass also become im
portant in the same range of temperature. It would be very interest
ing and challenging to discard the temperature limit imposed by the
catalyst and to balance the effects of temperature on equilibrium and
kinetics so as to achieve an optimum deSign. For example, a portion
of the reactor product could be cooled and recycled; the determination
of the optimum flow rate and temperature of the recycle would be a
very meaningul and enlightening project. Boonstra and Zwietering
[2) mention a second alternative wherein high temperatures are used
in the inlet sections of the reaction rates, while the downstream sec
tions are cooled to promote complete conversion in one pass. Dufau,
l aZ. (3) and Stobaugh [9) also provide useful discussions of the econ
omics of various cyclohexane production methods. These references
are helpful in giving perspective to the present design and in provid
ing additional background information for subsequent design projects.
With the proper reactor system selected, the next step in the de
sign procedure should be a consideration of the recycle costs. The
reduction in reactor volume achieved by increasing hydrogen partial
pressure must be interpreted in light of the limitation imposed on the
total system pressure. The advantages evdenced in Fig. 4-7 for
Vo Chapter J
higher hydrogen partial pressures must be weighed against the in
creased recycle costs caused by rising hydrogen concentrations in
the product mixture. Again an economic balance is indicated.
Several factors remain which must be given consideration as the
design proceeds to more advanced stages. For example, the pressure
drop i each reactor unit must be established. Too high a gas velocity
would lead to large pressure drops and to high rates of catalyst at
trition. Low flow rates require a large reactor cross section and re
sult in corresponding dificulties with bypassing and effective
catalyst usage.
It will be remembered that the effects of mass transfer and dif
fusion within the catalyst pellets were not thought to D influential
in the laboratory tests. The possible influence of these factors in a
commercial reactor should be ascertained. For example, the pres
sure drop in a reactor can be reduced by increasing the catalyst pel
let size. Since a large pellet diameter increases the resistance to
intraparticle diffUSion, the influence of this effect on reactor perfor
mance would have to be ascertained.
It is clear that a good deal of analytical work remains to be done
before a final design can be selected. Several factors, prinCipally re
cycle, heat transfer, and control costs, have not yet been given ade
quate consideration. Nevertheless, the design calculations summari
zed in Fig. 4-7 and the subsequent qualitative reasoning serve to
focus attention on those designs that are most promising. The type of
analysis set forth in the present design study illustrates the great
advantages which are to be gained by the use of simplified models as
limiting cases for a complex problem. A great deal was learned by
comparing the results of a single detailed analysis of a complex adi
abatic reactor system with the results of a straightforward analysis
of an isothermal system. In deSign work, such an approach is often
useful in minimizing the computational load while still obtaining
the required process information.
1IL1
1. Amane, A., and G. Parravano, "Advances in Catalysis" [Proc.
IntI. Congress on Catalysis), A. Farkas, Ed., Vol. 9, Academic
Press, New York (1957), p. 716.
2. Boonstra, H. J., and P. ZWietering, Cher. Ind. [London), 1966,
p.2039p
3. Dufau, F. A., F. Eschard, A. C. Haddad, and C. H. Thonon, Cher. Eng.
Progr. 43 (1964).
4. Grifith, R. H., and J. D. F. Marsh, "Contact Catalysis," Oxord
University Press, London (1957).
5. Kassel, L., U. S. Patent 2, 755, 317 (1956) [assigned to Universal
Oil Products Co.).
6. Reilly, J. W., and M. C. Sze, Cher. Eng. Progr. No. 6, 73 (1967).
Catalytic Reactoy Desil 99
7. Smith, H. A., in "Catalysis," P. H. Emmett, Ed., Vol. 5, Reinhold
Publishing Corp., New York (1957), Chapter 4.
8. Smith, H. A., and H. T. Meriwether, J. Am. Chem. Soc. 71, 413
(1949).
9. Stobaugh, R. . Hydrocarbon Process. Petrol. Refiner. 99 No. 10,
157 (1965).

5. Evalu
ation of New Methods for Sulfur
Transportation
As tle lcarld-Icide demalldfor fertilizer alld sulfuric acid has becollle
more lridesjJYead, tle need for more convellient alld economical sul
fur trans/Jortalion methods has grol('n more acute, Tllis case exalllines
sel'eral allerllatiL'e tralls/Jorlatiol! melhods ill a cursory fashioll alld
/Jrovides a more delailed allalysis of a /Jipeline syslem for lIle (011-
t'eymlce of liquid sulfur, The pipeline ecollolllics are o/Jtimized lci/h
respect to /Ji/Je diameler,line /J)'essure, number of /JIIJIl/Jilig staliolls,
and insulation Illicless, A short cOIlI/mler /Jl'og/'a/ll is /J)'ol'ided to
facilitate the design calculations, Tlie economics of the O/JtiIlWIll lille
are com/Jared Icilll those for olher /Jossible tralls/Jorlalion techniques,
LBCBS8u buUuT LCDQBB
CTl LBCBSC8u Cu1S1BB
To, Wilbur Wolf, Chief Design Engineer
From, Clyde Fox, Vice President
Rc: New Methods for Sulfur Product Transportation
The Exploratory Division of the company has recently discovered
8 new sulfur field near Pogo Lae. This find lies approximately 50
miles to the north of our shipping port, Port Calcaseau, and should in
crease our company sulfur reserves by at least 15 per cent.
Over the past several years, the Board of Directors has become
increasingly dissatisfied with our methods of transporting sulfur. As
you know, most of our inland sulfur production is now transported as a
liquid to Port Calcaseau by insulated trucks and,or railroad cars. The
I00
$/t/)ur Trats]orlu/iott Tcchniucs 101
costs of such transportation methods have been rising rapidly, and the
company is interested in developing new methods of delivering sulfur
from the mine to Port Calcaseau.
It is desired to use our new strike near Pogo Lake as a test area.
The Process Design group is requested to develop a list of possible
transportation methods complete with a brief description of the ad
vantages and disadvantages of each. The Board of Directors will meet
in three days, and I should like to have your list ready for presentation
at that time. Only a qualitative description of several possible methods,
along with your opinion regarding the probable merits of each method.
is desired. U the Board agrees to proceed, we shall probably request
a more quantitative study of one or more of the suggested techniques.
This investigation may well grow to a scale far beyond that of
the Pogo Lake project. U a good method is developed, it could well
be applied by many of our foreign affiliates to transport sulfur from
port cities to inland customers. For the present, however, concen
trate your efforts on the Pogo Lae project, and assume an annual
production of 375, 000 tons.
GENERL CONSlDER TlONS
Sulfur is conventionally produced by the Frasch process, whereby
high-pressure steam is injected into a sulur-bearing ore formation.
LiqUid sulfur resulting from the hot steam collects in a pool at the
bottom of the well; compressed air is injected beneath the surface of
the pool to lift the sulfur to ground level by a " bubble-pump" mech
anism. Once at the surface, the liquid sulfur is pumped to a heated
relay pan where noncondensibles are vented; the product is then fil
tered to remove any carbonaceous materials present. U the sulfur
well is reasonably near a shipping port, the purified product is piped
to a storage area where it is cast in large solid blocks. Wen a ship
ment from this storage is required, the solid sulfur is "mined" with
explosives to produce lumps of sulfur that can be easily handled by
standard power-shovel equipment.
On the other hand, if the sulfur well is located some distance from
a shilping port, the product is often shipped as a liquid in insulated
truckS, rail cars, or ships. Thus the problem put forth involves a
general consideration of methods of sulfur transportation when the
source and the shipping port for the sulfur are widely separated.
Upon initial conSideration, several possible transportation methods
come to mind that differ from the standard techniques used in the in
dustry. Some of these are here listed:
1. Pump liquid sulfur through 8 pipeline
2. Granulate solid sulfur and convey it pneumatically through a
pipeline
3. Store solid sulfur in capsules and convey the capsules pneu
matically through a pipeline
.

Slurry solid sulfur in water and pump the slurry through a
pipeline
5. Transport solid sulfur on a covered conveyor belt
1o Chapter b
These methods represent only a portion of the possible alternatives.
However, all have found application in various specific commercial
situations. Each has its advantages and disadvatages; some of these
will be discussed, and the more promising techniques selected for
further study.
Alternative No.1
The idea of pumping liquid sulur offers several possible advan
tages. Obviously, material-handling costs will be substantially re
duced by the elimination of solidification and granulation process
steps that are required at the wellhead by the other alternatives.
Clearly, however, a pipeline for liquid sulfur must be operated at a
high temperature. The problems attendant to this type of operation
must be given detailed consideration before such a pipeline can be
considered feasible.
It is interesting to note that one of the major sulfur companies has
recently installed a liquid-sulfur pipeline for connecting an offshore
drilling area with a major shipping port.
Alternative No.2
The technique of pneumatic conveyance of granular material is
quite well known, but typically l is used only for short distances, e. g.,
from one section of a plant to another. Starting up a long-distance
pneumatic process might present dificulties, and the prospect of a
plugged section of the line in a remote, swampy area is not particu
larly appealing. McCabe and Smith [6) have provided an excellent dis
cussion of the pneumatic method and have listed some of the opera
tional problems that may be encountered. Of particular interest here,
McCabe and Smith point out that erosion of the pipe walls can lead to
serious difficulties. This problem is especially pertinent in our case,
where many sections of the pipeline would be located in inaccessible
areas.
Alternative No.3
The technique of pneumatically transporting solid materials stored
in capsules has been given serious consideration in applications where
the solids are relatively fragile. For example, the possibility of ship
ping encapsulated wheat from interior agricultural regions of Canada
to Vancouver has been under study for quite some time. However, the
advantage of protecting the material being transported would be value
less in the sulfur transportation situation. As such, the large capital
and operatig exenses of this method would probably eliminate it
from consideration.
Alternative No.4
The prospect of pumping a sulfur-water slurry is attractive in
many respects. For example, the high temperatures ad start-up prob-
Sulfur Transportatioll Techlliques 1J
lems involved in a liquid-sulfur pipeline are avoided. The fact that
the method is feasible is attested by the near adoption of a plan to
transport coal in a water slurry from West Virginia to the great East
ern population centers. Even more recently, some industrial concerns
have given thought to the partial distillation of coal to produce a
"pumpable" slurry of partially coked coal in an aromatic liquid. How
ever, in the case of a sulfur pipeline the problems of separating and
drying the sulfur would entail some economic and technical difficul
ties. Frthermore, a water-sulur slurry would be highly corrosive
to mild steel and would necessitate a more costly construction
material.These problems tend to detract strongly from the advisablity
of this method.
Alternative No. 5
At first thought, the prospect of a fifty-mile conveyor belt appears
to be highly impractical. However, recent developments in conveyor
technology have shown that continuous belt systems of this length are
worthy of consideration. For example, the Societe des Potasses de Navar
re, a Spanish firm, is currently using a conveyor for transportig 620
tons per hour of potaSSium ore over a distance of four kilometers [3).
Cost figres for this method are available for comparison with the
costs for competitive techniques. The capital exenditures for the
conveyor unit were $ 400, 000, or $100 per linear meter ( 1964 costs) ;
this figre includes the entire metal frame, all mechanical compon
ents, roofing for the rubber belt, and the cost of transporting and erec
ting all equipment. The operating costs were found to be $0.01 per
ton-kilometer, including power, depreciation, maintenance, and super
vision.
CONCLUSIONS
The results of the foregoing considerations are far from definitive.
However, it appears that alternatives 2,3, and 4 would present more
operational and start-up difficulties than would alternatives 1 and 5.
Therefore, in the absence of detailed cost information about the first
four techniques, qualitative considerations suggest that alternatives
2,3, and 4 be discarded in favor of alternatives 1 and 0. By comparing
cost estimates resulting from preliminary designs for these latter
alternatives with standard rail and truck freight rates, a reasonable
assessment of the economic viability of the new transportation tech
niques should be realized.
1J Chulcr
Calcaseau Chemical Compay
Lae Calcaseau, Louisiana
To: Wilbur Wolf, Chief Design Engineer
Frolll: Clyde Fox, Vice President
Ater considering the vrious alteratives describd in your pre
limiary report, the Bard of Directors wishes you U prepare a
more detaied desig of a liquid-sufur transportation system. It is
aticipated tt a bried, jacketed pie will b used with 85 psig
steam provded in te anular spce to maitain the slur i a ml
ten state.
Naturally, it wll be desirable to compae te estimted capital
and runing costs for the liquid-sulr piplie wit the costs for 8
conveyor blt process ad te standrd shipping rates for rail and
trck transprtation.
Suggested Nomenclature
D Inner diameter of sulfur pipe, inches
. Oter diameter of steam pipe, inches
Di
Outer diameter of the insulation that surrounds the steam pipe,
inches
1 Distance between sulur pumping stations, miles
Maximum allowable pressure in sulur pipe, psia
l Thickness of the sulfur pipe wall, inches
Suggested Costs
1. Steel costs $250 per tOll plus $60 per ton for delivery to site,
unloading, placement, etc.
2. Engineering, surveying, etc., may be taken as $10, OOO
/
mile
3. Ditching, welding, installation of exansion bends, backilling,
cleanup may be taen as BOOD, $ /mile
4. First cost of a pumping station is $25, 000 plus $345 per bhp
5. Labor, supervision, and salaries are $2, 600 per month for a
400-bhp station; assume $600 of this independent of station size
and the rest proportional to station bhp
6. The annual maintenance cost of a pumping station may be taken
as 20 per cent of the first cost of the station
7. Insulation costs $14. 70/ cu ft (istalled cost)
. Steam costs $0. BO/
1000 lb [total cost including fixed plus
variable costs)
9. Electrical power is avilable at $0. 01
/
kwh
Sggested Assumptions
Sulfur Transportation Teclzniques 1b
1. Assume that the steam passes through an annulus 1 in. wide
2. Assume that the wall thickness of the steam lie is 0. 2 in.
regardless of the value of D. This should be large enough to
allow for any reasonable steam pressures and to provide the
necessary structural support for the pipelie
3. The annual fixed costs of capital equipment may be taken as
. . per cent of the initial instlled cost
. The sulfur pumps are to be powered by electric motors; the
motor-pump combination may be assumed to be . per cent
efficient
. The maximum allowable stress in the steel sulfur lie should be
taen as . .... psi
. The thickness of the sulfur pipe may be computed according to
the standard "hoop stress" formula
The Fanning friction factor may be taen as ) ^ .. . .

8. The sulfur pipe will be buried a depth of feet in earth whose
thermal conductivity may be taken as . .. Btu
/
(hr) (ft) (OF).
The pipe will be insulated with material whose conductivity is
. . . Btu/[hr) (ft) (OF)
DESIGN OF A PIPELNE FOR TilE TRNSPORTATION OF LIQUID
SULFUR
As mentioned previously, several serious dificulties are envision
ed in building a pipeline for the transportation of liquid sulfur. In
particular, the problems caused by the proposed operation of the line
at high temperature must be given serious consideration. Some of
these dificulties and considerations will be treated very briefly,
and then a preliminary process desig for the installation will be
completed. The prelimiary computations should bring to light the
critical parameters that should be examined more closely in subse
quent detailed design work.
Gneral Desig Coniguration
The typical pipeline system consists of long lengths of buried pipe
interconnected by a series of pumping stations. At each pumping
station the fluid being transported is raised to its maxmum pressure
and the line is designed such that the fluid pressure drops to a speci
fied minimum value at the entrance to the station. In the present case,
some means of heating the pipeline must be provided in order to main
tain the sulfur in a liquid state. Since the pipeline will be constructed
at ambient temperature, proper allowance must be made for exan
sion of the pipe as it is heated to operating conditions. Normally, the
10 Chapter
termal expansion problem is surmounted by the installation of ex
pansion jOints at predetermined intervals. However, these joints are
expensive, and recent development work has led to the technique of
preheating (and thereby pre-expandig) sections of pipe as they are
laid. The pre-exanded sections are then anchored in place and al
lowed to cool to ambient temperature. When during start-up the pre
stressed line is reheated to operating temperatures, the pipe assumes
a completely relaxed configration. Despite its obvous importance,
mechanical construction of this type will not be consi
d
ered in the
present computations. It will be assumed that the present preliminary
design will pOint up the critical parameters concerning the fluid
mechanics of the system. The mechanical aspects of the preliminary
design would then need to be investigated in a separate study.
As pointed out in the Freeport Sulfur Handbook (3) , liquid sulfur
is quite corrosive in the pre sence of small amounts of water. I our
desig it will be assumed that water concentrations can be kept low
and that carbon steel pipe may be used for construction. Carbon steel
has been shown to be satisfactory for handling most grades of liquid
sulfur product.
If air pockets are allowed to build up at high pOints in the lie,
an exlosion hazard could be created. However, it is anticipated that
this problem will be much less acute than that encountered in oil
pipelines. By proper grounding of the line (since it will be thermally
insulated) and by careful design and construction practices, exlosion
hazards may be effectively eliminated.
Heatig ad Isulatig Techiqes
Two fundamental heating methods may be considered for main
taining the desired operating temperature: steam ad electricity. In
a large-volume application such as the one under consideration,
steam would undoubtedly be the less expensive source of heat, par
ticularly in southern Louisiana where natural gas is plentiful. Thus,
steam boilers could be conveniently located at pumping stations, and
fuel could be made readily available.
Assuming that steam is to be used as a heating medium, the logis
tics of handling the boiler water is not immediately apparent. For
example, the possibility of running a boiler-water return line along
with the main pipeline should be considered. The savings generated
from decreased feed-water requirements would have to justiy the
additional capital exenditure. Without going into a detailed treatment
of this problem it seems unliely that the savigs in water treatment
would justify the increased capital.
Pertient Properties of Liqid Sulfur
Two available references [J,') fully delineate the properties of
liquid sulur. Te primary property to be conSidered in the present
design is the viscosity of the liquid. Figures 5-1 and 5-2, taen from
reference (9), summarize all kown data describing the liquid phase
viscosity.
0.04
0.03

.
:
0.02
:
0.0 I
o
\
\

`
^
`
`
Sulfur Tral1sportati on Teclili ques 107
`
'
|
.
r
140 160 180 200 220 240 260 280 300 320
Temperature gC
Figure 5-1. Viscosity of liquid sulfur (low
temperature range). ( From Tuller,
reference )
70

60
C
W
.

50
40

D
;; 30
x
:
20
10
1
cc
\
\
\
`
`
400 500 600
Temperature 1C
Figure 5-2. Viscosity of liquid sulfur
(high-temperature range). ( From
Tuller, reference ')
1o
C/w/)/eY b
Both figures show a sharp increase in liquid viscosity at a tempera
ture of approximately 320F. This icrease has been attributed to the
rupturing of S
8
rings to form polymeric-type chains of various lengths.
Above 370F the chains are thought to break ito smaller segments;
the continued chain breakage would lead to the observed decrease i
viscosity.
In order to avoid severe problems in the operation of the proposed
pipeline, temperatures in excess of 320 F will thus have to be avoid
ed. By using steam of a speciic pressure and degree of superheat,
this temperature control can be readily accomplished. However, elec
tric heating elements could probably not be adapted easily and inex
pensively.
References (J) and [) contain all the iformation regarding the
properties of liquid sulfur which will be necessary for the desig
calculations.
Assumptions to be Used U the Desig
In addition to the approximations and Simplifications prevously
discussed, the following assumptions will be employed to complete the
preliminary design:
1. Saturated steam at 85 psig |Ta t " 327F) will be used as
heating medium, and water at 15 psig (Tsat 250F) will be with
draw from the annular section of the line. By this means, the sulfur
will be maintained above its melting point of 246F but below the tem
perature range that produces the sharp viscosity increase.
2. The rate of heat loss through the isulation on the steam line
will be assumed to be constant along the length of te line.
3. The fluid mechanics withi the steam annulus will not be in
vestigated in the present study. A theoretical solution for the pres
sure profile in the annulus would mae a relevant and interesting pro
ject; this analysis will be required in the final design stage. Because
of the simpliication used, the size of the steam annulus must be
specified. Therefore, an annulus of I-in. width will be taken for all
sulfur-pipe diameters. This should allow for suficient steam flow in
all situations. The validity of this assumption will be checked at a
later point.
4. In a pipeline system the cost of steel pipe constitutes a major
capital investment. To minimize this ivestment, bth the diameter
and schedule number of the pipe are varied along the length of most
commercial oil and gas pipelies. For Simplicity, this practice will
not be followed in the present preliminary desig, and both . and /
will be assumed constant in any one design. After examination of the
preliminary design results, a judgment may be made as to the neces
sity of considering variations in . and l,
Sulfur Trallsportation Tcclmiqucs 1V
DESIGN COMPUTATIONS
The sulur pipeline is show schematically as follows:
T> m s0t
were te cross section of te pipe may b vsuized as:
. |nsu|cli

02
5lwm u|0s
D
,
D
s
5u|fur ppe
Preliminary computations will be carried out in order to treat the
fluid mechanics and heat-transfer aspects of this problem. After
these analyses, the annual operating and fied costs for the system
will be exressed as functions of . . and . Maing use of this
expression, the optimum set of desig parameters will be selected.
Fluid Mechanics Considerations
The sulfur flow rate has been fixed at . x .. lb
/
year or
. lb,hr. Thus the velocity and Reynolds HUDDC1 are given, res
pectively, by

. ......
/ .. . fps
. ..... . . ...
D
Up .
'0c
P
..
.. ...
x-
. .. .

. X .
.
If it is assumed that the gauge pressure in the sulfur line drops to
zero just as the sulfur reaches each new pumping station, the Fanning
equation may be written as

.

P
., .
11 ChoIcx o
where ] ..

Implementing this equation, we have

1 ... .

... . x ..

.. ...
.. ..
D
4 . ... .
. . X ..
psmile of pipe
.
The total power for pumping is given by
"
!
.
H ^ .. = .
0
...
. .. . ...
W= J
. ... ... . . .
^ .

.. hp (O. . .. hp per station)
The total annual cost for pumping S given
.
. ..
L_____

. . . . ... ... dollars

/year

.., . ... ...... . .....

Sulfur Pipe. From the standard hoop stress formula, the wall thick
ness of the sulfur pipe is given by
.. ..
l = -^ HI.
..
.....
and the .., of sulfur pipe is computed as

t (D + l)(7.7) (62. 3) (5280) (50)

( )
1 =
= 1380t

t U08
2000 (144)
....., the wall thickess in favor . and assuming that
l . we have
H_ ... .. . tons
Steam Pipe. Assuming that . wall thickness of the steam pipe is
. . we can calculate the weight of steel in the steam pipe as
(0.2) .( 2.4) (7.7) (62.3) (5380) (50)

=
200(144)
=
275 (D + 2.4) tns
Sulfur Transportation Teclmiques 111
Total Steel. The total amount of steel is thus written as
"steel =
004 D + 275 (D 2.4) tons
Heat-Transfer Considerations
We have the following heat-transfer problem:
`b0rth wrfe 0t
tepr0ture Tg
|
0
.
where steam is condensing on the iner wall of the steam pipe havng
a diameter _.
For steady-state conduction from a buried pipe, McAdams [4) gives
the relation

1 4Z
In-

t
Btu
/
hr
where Z` _ (see reference 1 for a derivation of this equation).
Note that P_ is the thermal conductivity of the earth, given as 0. 40
Btu
/
hr-ft OF and that 2 is the total length of pipe, equal to 50 miles.
Similarly, for heat transfer through the insulation, it is well known
that
Btu
/
hr
Since the steady-state heat-transfer case is being conSidered, the
two resistances may be summed and divided into the total driving
force to yield
Btu
/
hr
11Z
Cl[pter b
Taking average values of Ts as 285F, T
A
as 60F, I<
1
as 0. 40, Il
;
as
0. 05, Z as 6 ft, we have
Btu
/
hr (5.2)
Total An ual Cost for the Sulfur Pipeline
Making use of the foregoing analyses and the cost data given in the
problem statement, the costs for the pipeline project are computed
as
CFixed, Steel
0. 15(310) [0. 046D 2p 275(D 2. 4)]
2. 14D2p 12, 800(D 2. 4)
Cixed, Eng.
0. 15(50) (10, 000) 75, 000
CFixed,Di tehing
0. 15(800) (50) D ~ 6000 D
CFixed, Pump Sta.
0.15
_
[25, 000 345(0. 074 P)]
0 2000 50
CLabor Pump Sta

-
(600) (12)
! (0. 074 P)(12)
-
,
.
L 400 L
CMaint.
,
Pump St
a
.=
0. 20

[25
,
000 345(0. 074P)]
C team
CFixed,lnsul.
Cpump Power
1. 49 K 108 1 0. 80

8750
288 D; 1000 1000
In 8 n
D D
s
1. 04 X 106
288
D.
In
D
;
8 In

1 D.2 - D 2
= - '
s
(50) (5280) (14.70) (0.15)
4 144
3. 18 K 103(D
;
2 - Vs2)
241 P/L (see Eqs. 5. 3J.)
Sulfur Trallsportation Techniques 11
Summing all these costs, we obtain
7. 97 K 105 910 .
L_ ~ 105, 700 2. 14 D2p 18, 800

1
1. 04 K 106

288
In
D
i
8 In
3. 18 Y 103 (D, - Ds2) dollars
/
year
(5. 3)
Equation 0.` expresses the total annual cost of operating the pipe
line as a function of D, Ds' Di, . and . Some of these latter variables
are dependent upon each other; for example, Eq. 5. 1 relates . and .
In addition, a I-in. steam anulus has been specified for the design.
Thus, assuming a value for l of about 0. 1 in. , we get
(5.4)
By combining Eqs. 5. 1,5.3, and 5.4, an exression may be generated
for the total annual cost as a function of D, D
i
, and J.
Now that the appropriate cost relations have been developed, two
basic courses of action may be followed. First, by selecting various
insulation thicknesses a series of total cost equations may be devel
oped for each thickness. By partial differentiation of each equation,
the optimum combination of D, . and may be obtained for each in
sulation thickness. The total costs for each such combination can
then be calculated and compared in order to determine the over-all
optimum.
However, a knowledge of the shape of the total cost curve may be
even more important than the exact location of the optimum. For ex
ample, if the total operating cost is only moderately afected by one
of the variables, then the exact specification of that variable is ob
viously not important. L the other hand, an independent variable that
markedly affects the total cost curve may require more detailed ex
amination. Thus, a second approach to the use of the cost equations
would be to examine the sensitivity of the total and individual costs as
the independent design parameters are varied.
The two approaches just described for the cost optimization will
now be applied to the present design situation.
Otimization for a Speciic Isulation Thickess
For the pipeline sizes likely to be encountered i the present
analysis, an insulation thickness of 1. 5 in. appears reasonable. For
this thickness, it may be written that
D
i
D 5.6 (5. 5)
14 Chuplcr b
Substituting Eqs. 5. 1, 5. 4, and 5. 5 into Eq. 5.3, one obtains
p = 183,900 2. 14 D^p 37,900 D
2. 48 K 107 1.04 K 106
2. 18 X 1010

5. 67 7 I(D + 5. 6) 8 I(D 2. 6) (5. 6)
By use of the methods illustrated by Sherwood [},Eq. 5. 6 may be dif
ferentiated partially with respect to each independent variable and the
resulting partial derivatives set equal to zero. Following this pro
cedure to find the condition of minimum cost, one obtains
and
..
2. 18 X 1010
2.14 ^- 0
_
^
A.

1. 10 X 105

?-4
d
/
10. 5 10
1
0

4.28 DP + 37, 900 -

d J*
11. 9 X 107
:
1. 04 X 106
7
~
8

+ 5. 6 L 2.6
+ """"""" = 0
[5.67 7 In[ 5. 6) - 8 In[ 2. 5)]
2
Substituting Eq. 5. 7 into Eq. 5. 8 results in
6.08 X 105 1 1. 9 X 107
37, 900 - "
^:
1. 04 X 106
8

7

2. 6 5. 6
+ =0
[5.67 7 ln(D 5.6) - 8 ln(D 2 6) ]^
Solving this equation by trial and error, one finds
4. 1 in.
And, from Eqs. 5. 1 and 5. 7
J ~ 840 psi
= 27 miles
(5. 7)
(5. 8)
Sulfur Transportation Teclniques 11b
Taing as 25 miles (to give an integral number of pumping stations)
the following optimum conditions are found for a pipeline with a I-in.
steam annulus covered with 111 in. of insulating material:
L ~ 25 miles
J ~ 990 psia
D ~ 3. 9 in.
By using this set of optimum design conditions, the previously
derived cost equations may be applied to estimate capital and operat

ing expenses. These computations have been carried out and are sum
marized in Table 5-1, an examination of which reveals several inter
esting facts. First, it will be noted that the direct operating exenses
(labor, steam, and power) are much smaller than the indirect operat -
Table 5-1. Capitl and Oeratig Cost Estimates for
the Otimum Pipelie Havg 11/2-U Insulation
and ~U. Steam Anulus
Capital costs
Item
Steel pipe, 3. 9-in. diam Y 50 miles (sulfur)
Steel pipe, 6. 5-in. diam Y 50 miles (steam)
Engineering and survey
Ditching costs
Pumping-station costs, 2 @ $ 50,300
Insulation, I_-n thick Y 50 miles
Subtotal
Contingency, 20 per cent
Total capital estimate
Annual operating exenses
Item
Labor and supervision
Steam costs
Pumping power costs
Pumping-station maintenance cost
Depreciation, 15 per cent of capital
Total estimated operatig cost
Cost
$214, 500
537,500
500, 000
156, 000
100,600
1,018, 000
2, 526,000
505, 200
$3, 031,200
Cost
23,190
161, 400
9,550
20, 100
454, 680
$668,920
11O
Clwpler b
ing cost [depreciation.. Since the cost optimization was carried out
in order to minimize the total annual operating expense, it is clear
that the optimization essentially reduces to a minimization of the
capital requirement.
The capital cost estimate shown in Table 5-1 shows that the cost
for insulation is by far the largest contributor to the capital outlay.
Since an arbitrary insulation thickness was assumed, designs using
other insulation thicknesses should likewise be examied. In par
ticular, the use of a thinner insulation layer might lead to a more
economic over-all design.
Gneration of Cost Otimization Curves
Instead of summing all the individual cost equations to obtain a
total cost relation, as in Eq. 5. 3, it is instructive to exmine the im
portant individual costs. To accomplish this, these have been group
ed as shown in Table . all costs included in Eq. 5. 3 have been in
cluded in this grouping.
Table 5-2. Identification of Idivdual Costs
Iem
FORT
Symbl
Dpreciatin chrge CBNG
for engineering
80 srvey wrk
Depreciation charge CPIPE
for pipe costs and
ditching work
Power cost plus CPUMP
depreCiation
charge on pumpig
station
Depreciation charge CINSU
for insulation
Steam cost CSTM
Formula
T5,000
.

.. .. + .... + ...... + ...

7. 97 X .. ...
-

3. .8 X .. Y. . .. . ~ 6. 76)
.
.. X ..
... ._
In - +81n
'

D?.
By applying Eq. .. to the formulas given in Table .. the following
individual cost equations are derived:
CENG = 75, ...
CPIPE = ... Y ...

+ ..... . 30, ..
Sulfur Trans/Jorlatioll Techniques 117
CPUMP ^ Y 105 L + 2. 48 X 107 D
CISU ^ 3. 18 Y 103 [D D 5. 2 D . 7)
CSTM ^ 1. 04 Y l06/
ln

2
88
+
8 In

D

where
D
:
+ 2.
CTOT ^ CENG
+
CPIPE
+
CPUMP + CISU CSTM
In order to facilitate computations, a FORTRAN program has been
prepared using the foregoing equations; this program is listed in the
Appendi.
By using the computer program show in the Appendix, the individ
ual and total costs may be computed for any specified set of deSig
parameters. Retaining for the moment the assumption of a . 5-in. in
sulation thickness, the costs for operating a sulfur pipeline have been
computed as functions of pipe diameter and distance between pumping
stations, with the results summarized in Figres 5-3 and 5-4.
I
|

1

-
\
~
-
"
~
<
C

~
m
..
-

\
\
E . _"_
-.-

"
-
I
10 2 4 6 B 1 OO
Sulfur-pIp dIae. Ins
Figre 5-3. Individual and total costs as a
function of pipe diameter for the case:
(i) One-inch steam annulus
(ii) One-and-one-half-inch insulation
thickness
(iii) Two pumpig stations
118 Chapter b

1


T
S






T

s


|
5_
Sl .
J
Nti c
r
i.
"
'
t
_
.lIli

To ..
_I
-
.
.C
E-..
"
~
'

o 10 20 30 40 5
Dste bt_ lo stt ,mtl
6
Figure 5-4. Effect of number of pump
ing stations on costs for 4.0 in.
and an insulation thickness of 1. 5 in.
Figure 5-3 shows that the economics of constructing the pipeline
is a strong function of the pipe diameter. The capital cost for insu
lation and the steam costs are minimal, but the pumping costs and the
capital costs for pipe are exceedingly high at these small diameters.
As the diameter is increased, the pumping charges drop rapidly and
approach an asymptote equal to the fixed charges on the pumpig
station. The capital costs for pipe also decrease rapidly with increas
ing diameter, but these costs pass through a minimum and begin to
rise again. The extremely high pipe costs at small dimeters are the
result of the high pressure [and correspondingly thick pipe walls) re
quired to maintain the required sulfur flow rate. The steel required
to construct steam pipe with a O. .-in. wall thickess icreases with
. The charge for the steam pipe eventually exceeds that for the sul
fur pipe and causes the pipe cost curve to pass through a minimum.
On the other had, Figure 5-3 shows that the steam and insulation
costs increase continuously with increasing pipe diameter. The sum
of all these effects causes the total cost curve to be minimized at a
diameter of about 4 in. This diameter is i agreement with the result
computed analytically.
Further confirmation of the analytical computation is obtained
from Figure 5-4, where an optimum value for L of about . miles is
found. However, in this situation it is clear that the minimum in the
total cost curve is very broad. This breadth certainly indicates that
only a single pumping station at the inlet to the pipeline may be necessary.
The minimum in the curve occurs at an 1 value of . miles, but the
Sl/lfur Trw/spor/a/ion Teclmiql/es 119
minimum cost is only about $3000 per year less than the cost that
would be incurred if no intermediate pumping stations were used.
The preliminary cost estimate in Table 5-1 idicates that insula
tion costs constitute a major contribution to the total cost. Calcula
tions carried out to determine the effect of insulation thickness on
the economics of the design are plotted in Fig. 5-5. An optimum in
sulation thickness of about one third of an inch is found, with de
creasing insulation costs being offset by icreasing steam costs as
the pipe diameter is decreased. The figre shows that the balance be
tween these factors is nearly exact and that the effect of insulation
thickness having the specified conductivity is not particularly critical.
It would be interesting to investigate the effect of more expensive
types of insulation having lower thermal conductivities.
1
"




5
4
C
0

"

I
.

0
I

I
o
a
o
o
o

o
U




4
"
I
I
.
.
/
I

05
.
. .
m

>
>
.G
...
w
.
..
1.0 1.5
Inahon ICln , .
Figure 5-5. Effect of insulation thickness
on costs for D 4.0 in.; 1 .25 miles.
U consideration is limited to insulation having a conductivity of
0. 05 Btu
/(
hr
)(
tt
)(
q F
)
, then Fig. 5-5 indicates that the use of isulation
should probably be eliminated in order to simpliy construction. Cal
culations have been carried out in order to optimize the pipe diameter
for the case when only one pumping station is used [L 50 miles) and
no insulation is employed. The reslts are summarized in Fig. 5-6.
The elimination of insulation has only a minor effect on the over-all
120 Chc[ler b
economic evaluation of the pipeline, but a new optimum pipe diameter
of about 5. 0 in. has been arrived at. Therefore it may be concluded
that, within the accuracy of the preliminary design calculations and
the approximations that have been employed, the optimum pipeline
would be 5 in. in diameter, have an uninsulated steam jacket, and be
equipped with only one pumping station.
10
,

,
2
2
9

7


4
,
2
,
10
o
,
Totol
`
=

Slom
~
-

~

pe
~
(0$15
\
'
\
4
Enqlneennq an
survey coslS
p"m
""
CO55
8 12 16 20
SulfurpJpe dlomeler. inches
-

.
24 28
Figre 5-6. Indivdual and total costs as
fuctions of pipe diameter for the case:
(i) One-inch steam annulus
(ii) No insulation
(iii) One pumping station
FINAL CONSIDERTION OF SULFUR -TRANSPORTATION ECONOMCS
A liquid-sulfur pipeline has been desiged in order to take ad
vantage of the comparative product-handlig ease. Several major
design questions remain that must be resolved before the feasibility
of such a pipeline can be ascertained. U particular, it is necessary
to analyze and account for the mechanical construction that allows
for thermal expansion of the pipe. Also, the fluid mechanics in the
steam annulus must be studied, and the effect of variations in the an
nulus width should be investigated in some detail. As mentioned pre
viously, the economics of boiler-water recirculation should also be
evaluated.
Sulur Tmnsporlatioll Techniques 121
With the proviso that the mechanical construction problem and
steam-heating design will be treated in subsequent, more detailed,
investigations, the following optimum pipeline speciications result
from the preliminary design:
Sulfur pipe diameter 5. in.
Number of pumping stations 1
Insulation thickness in.
The cost estimate for these design conditions is given in Table 5-3.
Table 5-3. Preliminary Economic Evluation of the
Otimum Pipeline Havng No Isuation ..
Using Oe Pumpig Sttion
Capital costs
Item
Steel pipe, 5. O-in. diam Y 50 miles (sulfur)
Steel Pipe, 7. 6-in diam Y 50 miles (steam)
Engineering and survey
Ditching costs
Pumping-station costs, 1 @ $40,400
Insulation costs
Subtotal
Contingency, 20 per cent
Total capital estimate
Annual operatig exenses
Item
Labor and supervision
Steam costs
Pumping power costs
Pumping-station maintenance costs
Depreciation, 15 per cent of capital
Total estimated operating cost
Cost
$214,500
631,000
500,000
200,000
40,400
1,585,900
317,000
$1,902,900
Cost
9,880
286,000
2,920
8,080
285,500
$592,380
A comparison of Tables 5-3 and 5-1 shows that the estimated
total capital exenditure has been reduced by about one million dol
lars through the elimination of the insulation and the removal of the
intermediate pumping station. This reduction is partially offset by an
increased steam usage amounting to abut $125, OOO/year. The
/2Z
Chc[!c p
estimated annual operating expense for the design with no insulation
is about $ 75, 000 less than that summarized in Table 5-l.
Since the depreciation costs constitute such a large percentage of
the total operating cost, it is clear that the choice of the optimum de
sign will be a strong function of the rate at which the project is
charged for the capital it consumes. Figure 5-7 shows the influence
of the depreciation rate on the economic evaluation of the designs sum
marized in Tables 5-1 and 5-3. When capital must be charged at
8.5 per cent, the two designs are equally attractive from an economic
viewpoint. As depreciation rates vary from 8. 5 pel cent, one design
might be preferred over the other. Thus, the final design of the sul
fur pipeline could be altered radically, depending upon the required
rate of depreciating the capital invested in the project.
As a final step in the financial evaluation, the economics for the
sulfur pipeline should be compared with the standard rates for rail
and truck shipment as well as with the costs given in reference (2)
for conveyor-belt operation. Rail and truck transportation costs for
packaged chemicals are given by Tighe [8). Table 5-4 compares
these costs with those for the other transportation methods under con
sideration.

U
D

U
|200
|JJJ
Deslon Case I
3.9-ln pipe diameter
= 8
L

L
U
_6
8
On. umplnQ sialion
%
L
5
4

J
Deprecooloon rOle. per cenl
Figure 5-7. Effect of depreCiation rate
on economic evaluation.
Sulfur Tnills/)orlalioll TecJmiques 123
Table 5-4. Comparison of Trasprtation Costs
Method
Rail
Truck
Conveyor belt
Liquid pipeline
Cost per 100 lb/mile
$0.002 to $0. 003*
$0.003 to $0.004*
$0.00085
$0.00158
* Approximate 100-mile rates taken from reference (8).
Both the conveyor belt and the pipeline appear to show substantial
economic benefits over more conventional transportation methods.
The use of a conveyor system undoubtedly would result in higher
material-handling charges than would the pipeline system. These
charges would tend to equalize the economics of the two methods. For
convenience, the pipeline would probably be preferred. Further work
should be done to ascertain the feasibility of operating a long-diS
tance conveyor system in the humid climate found on the Gulf Coast.
The application cited in reference [3) is located in a comparatively
dry region. The mechanical difficulties of operating a conveyor de
vice in the geographical area contemplated in the present design
would probably eliminate this method from consideration. However,
in other parts of the world it may well be worthy of further investi
gation.
One additional fact should be pointed out: the possibility that the
rail or trucking firms would reduce their rates when faced with the
competition from a pipeline. I any event, a preliminary design such
as the one illustrated here is quite vluable as a preliminary means
of comparing a pipeline system with alternative transportation methods.
REFERENCES
1. Eckert, E. R. G., and R. M. Drake, Heat and Mass Transfer,
McGraw-Hill Book Company, Inc., New York (1959).
? Economic Section of the :rench Embassy in the U.S.A., French
Technical Bulletin No.6, 1, (1964).
--
3. Freeport Sulfur Company, Freeport Sulfur Handbook, Freeport
Sulfur Company, Technical Servce Department, New York (1959).
4. McAdams, W. H., Heat Trasmission, 3rd Edition, McGraw-Hill
Book Company, Inc., New York (1954).
5. McCabe, W. L., and J. C. Smith, Unit Oerations of Chemical
Engineering, McGraw-Hill Book Company, Ic., New York (1956).
121 Clwpler
6. Perry, J. H., Chemical Engineers' Handbook, McGraw-Hill Book
Company, Inc., New York (1950).
Sherwood, T. K. , A Course in Process Design, The M.I.T. Press,
Cambridge, Mass. (1963).
8. Tighe, F. C. , Chem. and Eng. News 31,3542 (1953).
9. Tuller, W. N., The Slfur Data Book, McGraw-Hill Book Company,
Inc., New York (1954).

6. Vacuum Fractionator Design for the
Purification of Styrene Monomer
This is tile first of tlCO cases dealing ICi/1 the /JroductiOIl of styrene
Iil 0/10111 er from ethylbellzene. This chapter describes tlte design of a
fractionating colulIlll for the separation of styrene and ethylbellzelle,
wllile C/w/ter 7 lcill deal wi th the dehydrogenatioll reactor for the
COil version of etllylbellzene to styrene.
In tie present chapter several allenzalive techniques are used ill
tlte COlUlIl1l design. Since the distillation must be carried Ollt llnder
mcuul/2, special cOllsideration is directed to the mechanical problems
engendered by this type of opera/iOIl. / cOIll/mter program l5 pre
sented to complete the design calculations by Sorel's method. This
/Jrogram accollliis for tie s/Jecial equi/)lJ1ent limitation illl/}osed by
mClllllll o/Jeratioll alld allons tile o/) til/2l1l/2 distillation s)stem to be
selected for any specified set of feed conditions.
Vandalia Chemical Company
South Chicopee, Mississippi
To: Wilbur Wol, Chief Design Engineer
From: Herman Fox, Vice President
Re: Styrene Production from Ethylbenzene
A recent market study by the company's Economic Planning group
indicates that the demand for styrene monomer will increase at a
substantial rate over the next several years. Since the economic fore
cast looks so promiSing, it is desired that your group prepare a pre
liminary design for a process to produce styrene via ethylbenzene.
Mter completing the design, you are to estimate the capital and oper
ating expenses for the process in order to assess the feasibility of
proceeding with this project.
125
126 Clapter 6
As you know, ethylbenzene is produced by alkylating benzene H the
liquid phase over an AICl3 catalyst at 95C. The ethylbenzene may
then be dehydrogented over an alumina catalyst at approximately
600C to produce styrene monomer. Many dehydrogenation processes
require large quantities of steam to be mixed in the feed. However,
the catalyst developed in the company laboratories functions most
effectively when pure ethylbenzene is used as a feed. It may be neces
sary to regenerate this catalyst periodically by means of steam.
While another design group will be studying the alylation reaction,
it is desired that your group concern itself with the dehydrogenation
reactor and with designing the fractionatig equipment necessary for
producing pure styrene monomer from the dehydrogenation reaction
product.
The Economic Planning group estimates that the proposed plant
should be designed to produce 20 million lb/yr. Some relevant data
and notes are herewith furnished.
Data 0 Notes
The heterogeneous reaction kinetics for the dehydrogenation of
ethylbenzene have been studied by Wenner and Dybdal (10), who pos
tulated that the following reactions are important:
MIO, cal! g mole
527C 627C
'uVe
29,715 29,824
24,472 24,282
29,770
(6.1)
24,377
(6.2)
C
S
H5C
2
H5 H
2
C
S
H5CH3 + CH -15,222 -15,537 -15, 380
(6.3)
The equilibrium constants for these reactions, shown in Fig. 6-1,
indicate that only in the case of Reaction 6. 1 is the reverse reaction
important. Some of the kinetic data obtained by Wenner and Dybdal
are shown in Table 6-1. Since only Reaction 6. 1 has a signiicant
reverse reaction, the authors analyzed their data by assuming the
following kinetic expressions for Reactions 6.1,6.2, and 6.3:
r
2
=
k2PE
r3 =llEP/
(6.4)
(6.5)
(6.6)
T
a
b
l
e

C
h
e
m
i
c
a
l

K
i
n
e
t
i
c

D
a
t

f
o
r

t
h
e

D
e
h
y
d
r
o
g
e
n
a
t
i
o
n

o
f

E
t
h
y
l
b
e
n
z
e
n
e

b
y

M
e
a
s

o
f

t
h
e

T
y
e

1

C
a
t
l
y
s
t

o
f

W
e
n
n
e
r

a
n
d

L
y
D
a
l

(
1
0
)

C
a
t
a
l
y
s
t

C
o
n
v
e
r
s
i
o
n

B
y
p
r
o
d
u
c
t
s

B
a
s
e
d

o
n

P
e
r
f
e
c
t

E
n
t
r
a
n
c
e

F
e
e
d

R
a
t
e

N
i
t
r
a
t
e

B
e
d

B
a
s
e
d

o
n

R
e
c
o
v
e
r
y
,

l
b

m
o
l
e
s

p
e
r

R
u
n

P
r
e
s
s
u
r
e

(
l
b
.

m
o
l
e
/

B
a
t
h

T
e
m
p
.

T
e
m
p
.

P
e
r
f
e
c
t

!
b

m
o
l
e

e
t
h
y
b
e
n
z
e
n
e

f
e
d

N
o
.

(
a
t
m
)

h
r
)

(
O
C
)

(
O
C
)

R
e
c
o
v
e
r
y

B
e
n
z
e
n
e

T
o
l
u
e
n
e

T
a
d

%

Y
i
e
l
d

R
-
2
2
8

1
.
0
2

0
.
0
2
0
7

5
5
6

5
5
2

0
.
2
2
5

0
.
0
0
5
4
0

0
.
0
0
6
3
6

0
.
0
0
4
9
2

8
2
.
4

R
-
2
2
9

1
.

0
0

0
.
0
1
4
1

5
5
7

5
5
5

0
.
2
2
7

0
.
0
0
9
4
8

0
.
0
0
8
2
5

0
.
0
0
4
7
3

8
0
.
7

R
-
2
2
7

1
.
0
2

0
.
0
2
1
9

5
5
7

5
5
5

0
.
2
2
6

0
.
0
0
6
3

0
.
0
0
6
8
9

0
.
0
0
4
5
5

8
2
.
5

R
-
2
3
0

1
.
0
0

0
.
0
1
3
1

5
5
7

5
5
6

0
.
2
2
7

0
.
0
0
6
0
4

0
.
0
0
9
3
6

0
.
0
0
4
7
3

8
1
.
9

R
-
2
3
7

1
.
0
4

0
.
0
2
4
3

6
0
4

5
9
8

0
.
2
4
6

0
.
0
0
5
8
7

0
.
0
0
8
1
7

0
.
0
0
3
6
2

8
2
.
2

*

R
-
2
3
9

0
0

0
.
0
1
2
8

6
0
4

6
0
0

0
.
2
6
9

0
.
0
0
5
3
4

0
.
0
1
8
9

0
.
0
0
4
5
2

8
1
.
0

i

R
-
2
5
2

1
.
2
0

0
.
0
2
3
8

6
8
1

6
7
6
*

0
.
3
6
6

0
.
0
3
7
7

0
.
0
4
4
7

0
.
0
1
2
1
1

7
3
.
1

R
-
2
5
4

1
.
2
0

0
.
0
2
3
5

6
8
1

6
7
6
*

0
.
3
6
0

0
.
0
3
9
6

0
.
0
4
1
2

0
.
0
1
3
2
8

7
3
.
0

"

R
-
2
4
2

1
.
0
0

0
.
0
0
6
4
0

6
0
4

5
9
8

0
.
2
6
0

0
.
0
1
0
6

0
.
0
1
3
5

0
.
0
0
3
9
6

8
0
.
6

:

R
-
2
4
0

1
.
2
7

0
.
0
3
2
1

6
5
4

6
5
0

0
.
3
1
0

0
.
0
1
6
7

0
.
0
2
0
4

0
.
0
0
6
0
6

7
9
.
4

R
-
2
5
1

1
.
2
0

0
.
0
2
2
7

6
7
7

6
7
1
*

0
.
3
5
5

0
.
0
4
3
0

0
.
0
4
5
9

0
.
0
1
1
4

7
2
.
2

0 "

R
-
2
5
5

1
.
0
0

0
.
0
1
3
1

6
7
4

6
6
8
*

0
.
3
7
2

0
.
0
5
6
1

0
.
0
6
9
6

0
.
0
1
6
8

6
8
.
2

C

R
-
2
3
2

1
.
0
0

0
.
0
0
6
8
9

5
5
7

5
5
5

0
.
1
9
5

0
.
0
0
2
9
8

0
.
0
1
7
7

0
.
0
0
3
6
5

7
9
.
1

7
o

*

E
s
t
i
m
a
t
e
d

t

A
s
s
u
m
e
d

m
o
l
e
c
u
l
a
r

w
e
i
g
h
t

1
8
0
.

-

128
Clapter 6
L

N
x

!
_

t F + !
_

|HccJion |1

2
c
o

L
z

x
[I
_
_
g /!
_
"
.,, .,
_. ..
IHccon .1

_
_ _ __ __ __
_

_

0m0t01ufL
Figure 6-1. Equilibrium constants for
dehydrogenation. (After Wenner and
Dybdal, reference 10.)
o

S
Maing use of the data of Table 6-1 and rate CQU!OuS 6.4,6.5, and
6.6 they obtained the following exressions for Ill, Il
2
' and Il
3
:
-11,370
log
IO
'
.
1
=
+ 0.883
4.575T
-50, 800
log
IO
k
2
=
+ 0.883
4.575T
-21,800
log
IO
'
.
3
=
1 9.13
4.575T
(6.7)
(6.8)
(6.9)
The data points used in determining Eq. 6. 7 are shown in Fig. 6-2.
All of the kinetic data gathered by Wenner and Dybdal were obtain
ed using an alumina catalyst whose particle size ranged from 4- to 8-
mesh standard Tyler screen size. The bulk catalyst density was
O1 lb/cu ft.
-l
M

8
..

1.0
-
'

Z
1.1 1.2
IO/T wtfhT M "K

Vacuum Fractiol1ator Desigll 129

..
1.3
Figure O-Z. Kinetic constants for the
dehydrogenation of ethylbenzene.
(After Wenner and Dybdal, ICCICnC6
10.)
Assumptions for the Distillation Colum Desig
1. The maximum allowable temperature in the column is 90C,
because the styrene will polymerize if this temperature is exceeded
(3). In fact, a nonvolatile reaction-retarding agent must be continu-
0usly fed to the top plate of the column in order to make possible an
upper limit of 90C.
2. A vacuum system operating at 40 torr is available.
3. The pressure drop across each tray will be assumed constant
at Z in. of water, and the over-all column efficiency will be taken as
70 per cent. Actually, some of the more modern sieve-tray designs
allow pressure drops signiicantly below the assumed value (1).
Also, special trays of the type licensed by Linde Division of Union
Carbide have tray efficiencies far higher than 70 per cent (). The
assumed pressure-drop and efficiency values may therefore be
thought of as conservative estimates. Once a procedure has been
established for the colunm design, it will be of interest to investigate
the effects of improving the pressure-drop characteristics and
efficiency of the trays to be used.
4. The composition of the distillate and bottoms must be
XD = 0. 98
2_ ~ 0. 01
where 2 refers to the mole fraction of ethylbenzene. Common in
dustrial practice might dictate a limit of 0. 003 for 7
D
(1). As in the
case of the tray hydraulics, the design computer program may be
1o Chaptel'
used to test the effects of bottoms composition on the process econo
mics.
5. U optimizing the distillation column, the following may be
assumed:
a. Capital cost for column ~ 12 d 1.2 dollars/tray, where 0 is
equal to the column diameter expressed in inches.
0. Assume that the column may be depreciated at 20 per cent
per year of 8000 operating hours.
L. Steam costs may be taken as $0. 80/1000 lb.
d. Cooling water costs $0.02/1000 gallons. The cooling water
is avilable at 70F; in order to minimize scaling problems,
this water should not be heated above 125F.
Sggested Nottion
D Flow rate of material in the bottoms stream, lb moles/hr
C Cost; CD refers to a depreciation cost, L_ the cost for steam, and
Ccw to the cost for cooling water, dollars/year
0 Column diameter, it
Flow rate of material in the distillate stream, lb moles/hr
Flow rate of material in the feed stream, lb moles/hr
k Kinetic constant for a heterogeneous reaction, lb moles/hr-atm
lb catalyst for reactions 6. 1 and 6. 2, and lb moles/hr-atm2-lb
catalyst for reaction 6. 3
'
Equilibrium constant for styrene formation, atm
Number of columns
W Number of plates
L Flow rate of liquid in the rectifying section of a distillation
column, lb mole s/hr
P Partial pressure, atm
o
Vapor pressure of a pure liquid, torr
Pressure, torr
Pressure drop through the column, equal to
B
- PD, torr
7 Rate of reaction, lb mOles/hr-lb catalyst
Universal gas constant, cal/g mole-OK
o Number of stages in a distillation column, equal to W 1
T Temperature, oK
O Vapor velocity in a distillation column, ft/ sec
V Boil- up rate in a distillation column, lb moles/hr
7 Mole fraction of ethylbenzene in the liquid phase
Y Mole fraction of ethylbenzene in the vpor phase
VacuulIl Fracliollator Design 131
Relative volatility, equal to P/ 15
Mr Standard state enthalpy change for a chemical. reaction, cal/g
mole
Heat of vaporization, cal/g mole
L
Density of liquid phase, lb/cu ft
7
Density of vapor phase, lb/ cu ft
Subscripts
rLJ
Refers to the actual number of trays
Refers to the bottoms stream; DJ refers to the bottoms stream
of column 1, etc
L
Refers to the critical vapor velocity i a distillation colum as
limited by entrainment consideration
/ Refers to a distillate stream; /] refers to the distillate stream
of column 1, etc
@ Refers to ethylbenzene
Refers to the feed stream
@ Refers to hydrogen
Refers to component l
m
Refers to the minimum reflu ratio
0IB0IB

Refers to an equilibrium property

DESIGN OF THE REACTOR A FRACTIONATOR FOR ETHYL
BENZENE PRODUCTION
A preliminary flow sheet for the production of monomer grade
styrene from ethylbenzene is sketched here. No provision has been
made in this sketch for the removal of tars from the reactor product
stream nor for that of the hydrogen produced in the reactor.
Feed
stream
Reactor
Reboiler
To
vacuum
Ethylbenzene
to recycle
Styrene
product
132
Clwpter
In addition, the removal and separate purification of benzene and
toluene should be undertaken. Generally, these steps are accom
plished prior to the ethylbenzene-styrene separation. Finally, the
tars are usually removed by use of a finishing tower operating on
the styrene product from the ethylbenzene-styrene separation (4).
The process shown in the flowsheet makes use of the indirect
addition of heat, in contrast to direct heat addition which can be
accomplished by injecting steam into the feed mixture. Ohlinger
and Stadelmann (5) have described some alternative schemes simi
lar to that suggested here. A study of their work is most helpful in
providing a broader understanding of this particular type of reactor
design. In addition, Carra and Forni (1) have provided kinetic data
which are useful in interpreting the processes described in refer
ence (5).
The memorandum asks that a design be completed for both the
reactor and fractionation units shown in the block flow diagram.
Obviously, the designs of these two units cannot be carried out inde
pendently, since feed rate and composition for the column depend
upon the operation of the dehydrogenation reactor. To circumvent
this problem and to illustrate the design method used, a feed rate
and composition typical of the reactor operation will be assumed,
and the distillation column will be designed. A generalized computer
program will then be written which will carry out the design calcu
lations for the fractionator for a wide variety of feed rates and feed
compositions.
Next, the kinetic data presented by Wenner and Dybdal (10) will
be examined in detail. With these data, as modified by the present
study, the design of the dehydrogenation reactor will then be com
pleted. Again, by means of a computer program the reactor design
calculations may be carried out for a wide variety of conditions.
By combining the computer program for the reactor design with that
for the column design, the design of the reactor-column system as
a unit may be optimized.
DISTILLATION COLUM DESIGN
Development of Vapor-Liquid Equilibrium U
The vapor pressure data for ethylbenzene and for styrene which
are necessary for the column design are tabulated in Perry's Hand
book (7. These data may be checked for consistency when it is re
membered that the data for each pure component must be well cor
related by the Clausius-Clapeyron relationship. Thus, Fig. 6-3 was
prepared by plotting the logarithm of the vapor pressure for each
component against the reciprocal of absolute temperature. The data
fall on a straight line, ad the heats of vaporization as computed
from the slopes of these lines agree well with experimental values
VuCuuI Fractiollator Design 133
4
5
^
-,

_
N
` `
. -

`.

4
"

-`

|
`

l0
2.70 2.B0 2.90 3.00 3.I0 3.20 3.30
I000'J wI1D |D "h
Figure 6-3. Vapor pressure data for styrene
and ethylbenzene. (Perry's Handbook,
reference 7
given in the literature (7). Therefore, the vapor pressure data shown
in the figure will be used in the design.
Since the molecular structures of ethylbenzene and styrene are
very similar and since both materials are reasonably nonpolar, one
might anticipate that liquid mixtures of these two components would
be ideal. Fortunately, the validity of this supposition can be verified
by referring to the paper of Chaiyavech and Van Winkle (2), who per
formed a very extensive study of styrene-ethylbenzene vapor-liquid
equilibria at reduced pressures. For variations in the total pressure
of 10 to 200 torr and at temperatures from 26 to 97C, these investi
gators have shown that the vapor-liquid system is ideal one. The
r\n:~11 deViR!\On from idealitv. as measured by the activity coeffi
cients, was less than 2 per cent, and this discrepancy was within the
limits of probable experimental error. From the assumption of
ideality and the vapor-pressure data of Fig. 6-3, the relative vola
tility of the styrene-ethylbenzene system may be computed as a
function of temperature. This calculation has been carried out for
temperatures ranging up to 90C, the maximum allowable tempera
ture for liquid styrene in the present process, and is summarized
in Table 6-2.
134 Chapter 6
Table Z, Relative Volatilit of Ethylbnzene-Syrene
Miures Computed by Assuig a Idea
Liquid Phase
T (OC) T (oK 1000iT (oK
p
o
b
p
o

=
PEIP'
. ,- , . _ _-- --
40 313 3.19 16.4 22.0 1.34
50 323 3. 10 25.1 34.3 1.365
60 333 3.00 40.8 55.9 1. 37
70 343 2.92 60.0 82.3 1.37
80 353 2.83 92.8 128.0 1.38
90 363 2.75 137.0 190.0 1. 385
Since the variations in relative volatility with temperature are
so slight, an average value of may be selected at a temperature
level typical of that to be encountered in the process. This value
may then be assumed constant for all calculations. In this manner,
an average value of 1. 37 is selected for the relative volatility, and
the design will be completed on this basis.
Cacuation of Minimum Number of Plates
Since is assumed constant, the minimum number of plates to
complete the desired separation can be calculated using the Fenske
equation (8), whereby
log
1
: j|
W + 1 .- -------
log
N + 1
=
log
0
.
9
8

0.
99

0.02 0.01
log 1. 37
N + 1 .27
Twenty-seven theoretical stages would therefore be required to carry
out the specified separation when operating the column at total reflux.
Calculaton of Minimum Reflu Ratio
The minimum liquid-vapor ratio in the rectifying section of the
column is easily computed from the following equation:
where
(X
F
)
Y
(e -l)X
F
+ 1
and in the present case,
(1. 37)(0. 70)
Y ~
(0.37)(0.70) + 1
Y = 0.762
Therefore,
(a/
V
) . =
0
.
9
8
- 0.762
un
0.9
8
- 0.70
~ 0.778
VuCuHD JIuClOlI0lOI Design 135
From this value, the minimum reflux ratio, (
O
/D)l
I
lin' can be cal
culated as follows:
(O/
V
)
min
0.778
(a/D)
i
= = -=3 50
m
n
1
_
(
O
/
V
)min
0.222 .
Cacuation of te Act Nubr of Trays
Once the minimum number of trays and the minimum reflux ratio
for the desired separation have been established, there are several
available methods for computing the number of trays required as a
function of reflux ratio. Since these methods vary in their complexity
and in the accuracy of the resulting answers, two of them will be
illustrated here. In addition, the results of a third one will be pre
sented.
1. Gilliland Method. Using normalized, dimensionless variables,
Gilliland has plotted the number of fractionation stages versus reflux
ratio. Using his correlation, we may find the number of stages re
quired at 1. 3 times the minimum reflux ratio by the following com
putations:
(a/D) = (1
.
3) (3. 50) = 4.55
From Fig. 12-2 of Robinson and Gilliland (8),
(a/D) - (O/D)
m
in
4.55 - 3. 50
=
-- = 0.18
9
(
a/D) + .M 4.55 1.0
136
Chapter 6
whereby from the chart
~ S
m
i
n

27. 44
T 0. 44
and = T 9 theoretical stages
0. 56
Erbar and Maddox have presented a correlation which for some sys
tems appears to be more accurate than that of Gilliland. Oliver (6)
provides an excellent discussion C this new correlation, which may
be used to provide additional insight into the present problem.
2. McCabe-Thiele Method. For a binary system for which the
equilibrium data are well known, the most accurate technique for
computing the number of trays in a column is the McCabe-Thiele
method, which is merely a graphical representation of the Lewis
Matheson computational technique.
With equilibrium data computed using a constant value for of
. 37, Fig. 6-4 was prepared to illustrate the McCabe-Thiele method
for the case of (a/D) = . 3 (O/D)mi
n
. As can be seen, this method
indicates that 44 theoretical stages are required for the separation.
|.
O.
O.
>-
o
g .7
.
O!
.D
w
: .A
9

O.
a
T
:;

Lowet otioo ot
sppoq scoo
C.C5 C1C
O. .Z O. O O.D O.! O.7 O. O. l.O
Mo|e froc!ion o! e!hyIbenzene in |quid X
Figure 6-4. McCabe-Thiele design of the
distillation column.
\CCuuD r0ClOn0lOT DesiKIl 137
3. Other Techniques. Robinson and Gilliland (8) present several
other techiques
f
or te computation of the number of required stages.
For the most part, the calculations required are quite tedious, and it
is usually best to rely on the McCabe-Thiele method unless an excep
tionally large number of trays is needed.
One of the methods described in reference (8), that of Lewis, was
used in the present study for illustrative purposes. It indicates that
46 stages are required for DID = . 3 (OID)m
in; this number com
pares well with previous computations.
4. Summary of Results. Table 6-3 has been prepared to allow
comparison of the various calculational procedures. The number of
actual plates required for the separation has been calculated for
each case by assuming an over-all tray efficiency of 70 per cent.
Tale , Stes Required for the Fractionation of Srene ad
Etylbenzene when DID ~ . (OID
>in
Number of
t\DOU Theoretical Stages
Gilliland 49
McCabe-Thiele 44
Lewis 46
Nuber
Actual Stages
70
63
66
The three methods used are in reasonably close agreement. Gilliland's
method is the most convenient to use but also probably the least
accurate. If the equilibrium 7-Y diagram is carefully prepared, the
McCabe-Thiele method is probably the most accurate; however, when
large numbers of plates are involved, as in our case, significant
errors may result from using the McCabe-Thiele technique. The
Lewis method is a bit cumbersome, but for systems requiring large
numbers of trays it is quite accurate. For the present case, it will
be assumed that 66 actual stages are required when the reflux ratio
is 30 per cent greater than the minimum.
Pressure and Temperature Considerations in te Desig
Since the design instructions require a 90C maximum tempera
ture in the column, it is necessary to determine temperature and
pressure profiles for each column design. Because the proposed
column is to operate under a vacuum, the pressure drop in the column
will be quite important in determining the temperature level on each
tray. For example in the 66-stage column just designed, the pressure
drop may be computed as follows:
N o- = 65
138
Chapter
where it is assumed that
MIN T 2 in. of water
760
T 2 T 3. 7 torr
34 X 12
and M T 3. 7(65) T 240 torr
Since the pressure at the top of the column is 40 torr, that in the
reboiler is computed as
J
0
T 240 40 = 280 torr
However, if the temperature in the reboiler were equal to 90C, the
pressure in the reboiler would be
J
0
T

x _ p T (0. 01) (205) + 0. 99(138) T 139 torr

Thus, at the bottom of a tower with 65 trays the temperature
would be greater than 90C, and clearly an alternative design scheme
must be proposed if the restriction on the reboiler temperature is
to be met. An alternate scheme that might satisfy the temperature
limitation would use two towers (Fig. 6-5) .
|eed
sIreom
T0 v0cuum
system oI ~~~~ ~~
4 I0rr
0 vocuum
Boltoms
streom Io
0!IercooIer
ond stor0ge
Figre 6-5. Two-column system for the fractiona
tion of ethylbenzene-styrene feed.
To check the feasibility of such a process scheme, the tempera
ture and pressure in the bottom of the first column must be com
puted as a function of the number of trays in that column. The
\uCuuD JIuClOnu|Or LcSllI 139
results of this computation are show in Table 6-4. In these calcu
lations,X
E
is read directly from the McCabe-Thiele diagram (Fig.
6-4) , and total pressure at the bottom of the first column is computed
on the assumption that T
B
= 90C.
Table 9,
Assumed
N
l
40
30
35
36
Temperatue 80 CompOSition on Btom Tray
of First Colun
/_ =
PE
l
=
3. 7
'1
40
=1
148 torr 188 torr
111 torr 151 torr
130 torr 170 torr
133 torr 173 torr
0. 7 N
l
28
21
24. 5
25
PB
2
7_ at 0.7 N
l
= 205X
E
(from Fig.6
-
4) 1O2

0. 4 165
0. 67 182
0. 56 176
. 54 174
Hence, a two-tower system with 36 trays in column 1 would satisfy
the restriction on the reboiler temperature. Note that the tempera
ture at the top of the second tower must be low enough so that J__
will be 40 torr. This temperature is calculated by trial and error
from vapor pressure data (Fig. 6-3) and the composition of the liquid
stream at the 37th tray (Fig. 6-4) .
Assumed
Temp. oC
_o

]
o

.54 p 0.46 p
P
D
2
=

Pi
55
44 33 23. 8 14. 5 38. 3
56 46 34 24. 8 15. 6 40. 4
Therefore, the liquid leaving the first column at point! in Fig.
6-5 must be cooled to 56C before being fed to the top of the second
column.
Checking pressure drop through the second column, we find that
P
B
2
=
(30)(3. 7) 40 = 154 torr
Since the X
E
at the bottom of the second column is 0. 01, P
B
2
may
be compared with 1_
o
which is 1D+ torr at 93. 5C. Therefore, at a
reflux ratio of 1. (0/ D) three towers would be required to com

plete the separation without exceeding the maximum temperature

specification of 90C.
14 Chapter
RESULTS FOR OTER REFLU RATIOS
The results for other reflux ratios are shown in Table 6-5,
Table , Resuts for te Nubr of Trays ad the
Nuber of Colus
(O/D)/(O/D)m
i
n (O/D)
*ACT
Number of Towers
1. 2 4. 20 77 3
1. 3 4. 55 66
3
1. 45 5. 07 63
2 or 3
1. 70 5. 95 60
2
which maes clear that for all (O/D)/(O/D)min > 1. 45 two columns
will be required, while for all such ratios < 1. 45 three columns are
needed. For the purpose of this design, the number of columns will
be arbitrarily limited to a maximum of three, although more than
three columns obviously would be necessary at very low OlD ratios.
At the other end of the spectrum the minimum number of theoretical
trays in the system is 26; this corresponds to 37 actual trays. Thus,
it is clear that at least two columns will be required to carry out
the desired fractionation, regardless of the reflux ratio.
OPTMIZATION OF THE SYSTEM
It will be remembered that a feed composition of 70 mole per
cent of ethylbenzene has been assumed. The distillation column
system will now be optimized while maintaining this assumed feed
composition. A computer program will then be developed that will
be useful for the entire range of feed compositions as well as for
variations in other important design parameters.
Computtion of the Colum Diameter
In order to ascertain the diameter of the column system, the
hydraulics of the liquid and vapor flow within the column must be
considered. The limitation on the size of the column is imposed
by determining the maximum superficial vapor velocity that can
be allowed. In turn, the vapor velocity is limited by the liquid
handling capacity permitted by the tray design or by the increase
in liquid entrainment as the vapor velocity is increased. As men
tioned previously, a complete consideration of the tray hydraulics
is beyond the scope of the present design; however, the computation
of the entrainment limitation should serve as a good criterion for
this preliminary tye of design.
VuCuuD Tu0lOnu|OT Desi 141
In the design of a vacuum column where the absolute pressure
varies significantly through the column, it is common to specify a
larger column diameter in the rectifying section than in the stripping
section of the column. This practice is necessary to maintain the
vapor velocity within reasonable limits. U the present design were
being carried out in detail, the column diameter would undoubtedly
be varied through the column. However, to simplify the calculation,
a single diameter will be computed at the average column pres-
sure of 90 torr. U a tray spacing of 12 in. and liquid seal of 0.5 in.
are specified, the correlation based on entrainment limitations indi
cates that (8)
O
0

"

= 0. 13
LV
Taing the specific gravity of the liquid mixture to be 0. 8 and com
puting the vapor density at 90 torr, we have

49.
9

O
0
= 0. 13 = 5. 5 ft/sec
0. 0279
Using this critical velocity, the column diameter may be computed
once the boil-up rate has been specified.
More recent correlations have been established for the critical
velocity. The most important of these are discussed by Fair and
Bolles in Smith's text (9). In pursuing subsequent, more detailed
analyses of the tray hydraulics for this design, the reader would be
well advised to consult this excellent reference.
Heatig ad Cooling Duties
To visualize the heat exchange problem for this column, it is
useful to construct a "pseudo" pressure
-
enthalpy diagram for the
system, as shown in Fig. 6-6. The numbers on this diagram cor
respond to those on Fig. 6-5.
|74
3

=
;

-
[n!h0lpy (tu/1b ~
Figre 6-6. Pseudo pressure-enthalpy
diagram. Numbers correspond to
numbers in Fig. 6-5.
142 Clmpler
Since the (O/V)
m
i
n
ratio in the column may be taen very roughly
as unity, the heat exchanger shown in Fig. 6-5 may be used to cool
the hot liquid from the first column by heating the condensed vapor
from the top of the second column. Clearly, however, this heat ex
changer can account for only the sensible heat changes of the two
streams, and a vaporizer must be supplied to boil the feed to the
bottom of the first column. Hence, when two columns are required
for the separation, the reboiler and condenser duties are effectively
doubled. This is a good assumption, since for liquid ethylbenzene
styrene mixtures the enthalpy of vaporization is an extremely weak
function of temperature and composition.
Thus, for (O/D)/(O/D)mi
n
< . 45, the reboiler and condenser
duties will be taen as twice those for a single column. For
(O/D)( O/D)
m
i
n
< 1. 45these duties will be assumed to be three
times those for a single column.
Optimization Computtions
Restrictions 7g = . 70
X
D =
0. 98
X
B
=
0. 01
(1 X
B
)B ~ 20 ^ 106 lb/year
For 8000 hr/year operation
(20 Y 106)(0. 99)
= = 24. lb moles/ hr
(8 Y 10
3
)(104. 02)
Now to determine V, the boil-up rate, as a function of (O/D) and D.
V = L(O/D 1)
where L = D and X
F
T D X
D
X
B

Hence, in general,
and V

X
B
-
X
F
(O/D 1)
X
F D
Vacuum FracLionalor Design 143
For the present case:
0. 01 - 0. 70
V = (24. 3) (O/D 1)
0. 70 - 0. 98
~ 60. 0 (O/D 1)
NOw, the velocity to be used in the specification of a column dia
meter is to be computed at an average pressure of 90 torr. Thus,
760326
(60. O) (O/D 1) (359)
90 273
76. 9 (O/D + 1)
O =
3600 (1 / 4) ^
=

^
Setting O equal to the maimum allowable velocity of 5. 5 fps, one
obtains
0 = 3. 74 (O/D 1)1/ I = 44. 9 (O/D + 1)1/
2
in.
and the yearly depreciation of the column is
C
D
= N
{
12
[
44. 9 (O/D 1)1/
2
]
1
. 2
}
(0. 2) = 232N(O/D 1)
0
.
6
Note that the capital cost for the column system is taen to be
a function of the number of trays only and not of the number of
columns. This assumption is obviously a great simplification of
the true situation.
For styrene, from Fig. 6-3,
bH C In
p
O In 200

In 20
-
=

= 47900K
C (l/T) (2. 67 - 3. 15) K 10-3
bH = 9500 cal/g mole or 17,100 Btu/lb mole
The steam costs for boil-up are
(60. O) (O/D 1)(17,100) (0. 80) (8000)
/
C
s
= - 6780 (0 D 1)
(970) (1000)
dollars/year
where the heat of vaporization for steam is taken as 970 Btu/lb. It
must be remembered that this steam cost is that required for a single
column. Costs for condenser cooling water may be estimated as
C
_ (60.0)( OlD + 1) (17, 100) (0. 02) (8000)
LW
(55) (8. 35) (1000)
= 358 (O/D 1) dollars/year
Thus, for a single column,
C
s +
C
c = 7140 (O/D 1) dollars/year
144 Chapter
Now all costs for the distillation system are known in terms of
the reflu ratio and the number of trays required for the separation.
The computed costs for typical reflux ratios are shown in Table 6-6.
These costs may be plotted versus reflux ratio in order to obtain the
optimum reflux ratio for this system (Fig. 6-7) .
Because of the increased heating and cooling duties required in
the three-column system, a discontinuity in the total cost curve fixes
the optimal point at (OlD) = 5. 07. In the actual desi n a reflu rS1^
somewhat higher than this should be used, as slight variations in pro
cess conditions could lead to temperatures above 90C if the system
were operated exactly at the optimum point.
18
16
140

120

"
".10

. 80
i
C

u
%
Z

1.0
'

,
,
./"
,
,
,
b/
a
V
,
,
,
>
,
I
< ,

.
\
`

I-
e
-
0=
1.2 1.4 1.6 1.8
(0/0)/(0/0) ..
2.0
Figre 6-7. Economic optimization
of the distillation column sys

tem.
DeSign conditions: Legend
7
|
~ 0. 70 Total
annual
cost
D
~ M.d A Steam
costs
7
0
~ M.M1 Deprec-
iation
costs
~ 24. 3 lb Cooling-
moles/hr water
costs
, g
T
a
b
l
e

C
o
s
t
s

f
o
r

t
h
e

D
i
s
t
i
l
l
a
t
i
o
n

S
y
s
t
e
m

O
l
D

=
N
o
.

O
/
D
m
i
n

o
f

T
o
w
e
r
s

(
O
l
D
)

(
O
l
D

1
)
0
.
6

1
.

2
0

3

4
.
2
0

7
7

2
.
6
9

1
.

4
5

3

5
.
0
7

6
3

2
.
9
5

1
.

4
5

2

5
.
0
7

6
3

2
.
9
5

1
.

7
0

2

5
.
9
5

6
0

3
.
2
0

L
D

L
3

C
C
I
V

$
4
8
,
2
0
0

$
1
0
5
,
8
0
0

$
5
5
8
0

4
3
,
1
0
0

1
2
3
,
7
0
0

6
5
7
0

4
3
,
1
0
0

8
2
,
4
0
0

4
3
5
0

4
4
,
6
0
0

9
4
,
3
0
0

4
9
8
0

L
T

$
1
5
9
,
6
0
0

1
7
3
,
4
0
0

1
2
9
,
9
0
0

1
4
3
,
9
0
0

" o

1 Chapter
REVIEW OF THE DESIGN TECHNIQUE
The two-tower distillation system as discussed in this chapter
is a classical design that has been applied to styrene processes for
decades. The optimization calculations summarized in Fig. 6-7 are
the result of several approximations that were used to simplify the
presentation and to emphasize the more important aspects of the
design technique. From the results obtained, it is clear that the
assumed values for over-all column efficiency and for tray pres
sure drop are critical in determining the ultimate design economics.
For example, the use of a tray design that would provide a higher
efficiency at a lower pressure drop would allow a distillation system
requiring only one column. Indeed, recent advances in tray design
have made this feasible, and the two-tower distillation system has
been replaced by single-tower fractionators in virtually every
modern styrene installation (4).
The column design has been carried through to the development
of a rather elaborate computer program, which is here presented
so that it may be used to investigate further the economics of the
styrene purification process. In particular, the effect of variations
in the tray hydraulic characteristics should be studied because of
their potential impact upon the process economics.
It has been pointed out that the feed concentration to a commercial
ethylbenzene-styrene tower is usually 45 to 60 mole per cent ethylben
zene (4). Many of the newer plants apparently find it advantageous
to operate in this range despite the yield losses that must result
from the use of higher conversions. Since the present design has
assumed 70 mole per cent of ethylbenzene in the feed, the computer
program should prove valuable in investigating the economic effects
of variations in the feed composition.
DEVLOPMENT OF A COMPUTER PROGRM FOR TI
DESGN OF THE DISTILLATION COLUM SYSTEM
It will be remembered that the distillation system was designed
after fixing the feed composition. At this point, a computer program
will be written so that computations may be completed for a wide
range of feed compositions and reflu ratios. This development will
assume that the reader is familiar with the fundamentals of machine
programming. In addition, some familiarity with the FORTRAN auto
matic coding system will be necessary. The reader who is not
familiar with the languages and strategy of programming is referred
to the comprehensive series of reference manuals published by the
International Business Machine Corporation.
Figure 6-8 shows the flow diagram for a FORTRAN program for
the computation of the number of trays and number of columns re
quired to produce a styrene monomer product of the desired purity.
The program requires as input the flow rate and composition of the
feed stream; these quantities will be established only after the reac
tor design has been completed. For the present, however, the nature
Input
Material
balance
Establish
operating lines
InitialIZe
variables
compute number
of trays by
LPwI5NDThP5OR mPThO0
compute actual
number of trays
and costs
Pr int output
\u0|thII FraclioJlator Design 147
RRMI, f, Xf, X0
RmN1NI, ANT1NI
fT1N1CT1NI,
CM1N1,TC1NI,
W1NI
T
Figure OO. Flow diagram for the FORTRAN computer
program. (Circled numbers refer to statement num
bers in the FORTRAN program.)
of the feed stream will be assumed to have been established. Also
required are the bottoms (styrene) flow rate and composition. The
styrene stream has of course been fixed in the original design
specifications at 20 Y 106 lb/year of 99 mole per cent pure product.
After reading the input variables, the program then completes a
material balance and specifies the distillate flow rate and composi
tion. The minimum reflux ratio for the required separation is then
computed.
The method used is conceptually very straightforward; if desired,
the program can be shortened substantially if the tray-to-tray calcu
lations are replaced by the Smoker equations or the Underwood
148 Chapter 6
equations in their binary form (8). Also, since the general shape of
the optimization curves (Fig. 6-7) is already known, the computing
effort could be reduced to a narrow rage of (O/D) /(O/D) min and
the slopes of the cost curves could be tested to determine the opti

mum. Either of these approaches would constitute desirable design

projects.
Next the program enters a "DO-loop" in which the required num
ber of trays and number of columns are computed for each specified
value of the reflux ratio. Actually, the calculation is carried out for
a wide range of values for the variable RRCN(I) , which represents
the ratio of the operating reflu ratio to the minimum reflux ratio.
For each value of RRCN, the equations of the upper and lower operat
ing lines are first developed. Then, starting from the top of the
column, plate-to-plate calculations are made using the method of
Lewis and Matheson (8). Since constant molal overflow has been
assumed, this method merely involves alternate applications of the
equilibrium relationship and the material balance (operating line)
expressions. At the conclusion of this calculation, the number of
theoretical trays and the feed plate location have been specified.
As a final step in the "DO-loop," the column diameter is computed,
the required number of columns and the actual number of trays are
calculated, and the capital and operating costs are specified. In
these calculations the maximum number of trays per column has
been set arbitrarily at 33. This of course is an approximation, but
it seems suficiently accurate for the purposes of the present work.
As the last step in the machine program, a table is printed which
shows the number of trays, the feed tray, and the operating and capi
tal costs as functions of RRCN. The Appendix contains the actual
FORTRAN program used as well as the output from this program,
which corresponds to a feed stream flow rate of 71 Y 106 lb/year
having an ethylbenzene composition of 70 mole per cent. The rela
tive volatility has been taen as 1. 37 for this computation. The re
sults shown in the output table compare well with the manually com
puted results shown in Fig. 6-7.
LP Notation
A Slope of upper operating lie when the column is operating
at miimum reflux
ALPHA Relative volatility
AT Actual number of trays required for the desired separation
CEPTL Intercept of lower operating line on the Y-ais of a McCabe-
COST
Thiele diagram
Total annual operating cost for a distillation column, equal to
the sum of the costs for cooling water, steam, and deprecia
tion,dollars/year
CPCOS Annual depreciation charge for the capital invested in the
column, taken as 20 per cent of capital, dollars/year
V0CuuD 10COn0O1 Desi 14
CWCOS Annual cooling water costs, dollars/year
D Distillate rate, lb moles/hr
F Feed rate, lb moles/hr
FT Feed tray location
I Subscript corresponding to plate number
N Subscript corresponding to the particular reflu ratio that
is being considered
NC Number of columns
NFT Theoretical number of trays
NT Actual number of trays
QCEPT Value of Y which corresponds to the itersection of the upper
and lower operating lines
QN Liquid flow rate in the rectifying section of the column
RR Oerating reflux ratio
RRCN Ratio of the operating reflux ratio to the minimum reflux
ratio
RRM Minimum reflux ratio
SLOPL Slope of lower operating line
STCOS Steam cost, dollars/year
TNT Total numbe: of theoretical trays
TIDX Internal diameter of column, ft
USLOP Slope of upper operating line
UCEPT Intercept of upper operating line with the Y axis
V Boil-up rate, lb moles/hr
XB Mole fraction of ethylbenzene in the bottoms
7 Mole fraction of ethylbenzene in the distillate
XF Mole fraction of ethylbenzene in the feed
Y(I) Mole fraction of ethylbenzene in the vapor phase
YCEPT Equilibrium value of Y corresponding to XF, equal to
REFERENCES
1. Carra, S. and L. Forni, Ind. Eng. Cher. , Process DeSign and
Development 4, 3, 281 (1965).
2. Chaiyavech, P. , and M. Van Winkle, J. Cher. Eng. Data 4, 1, 53
(1959)
lO Clza/ier 0
. Faith, W. L.,D. B. Keyes, and R. L. Clark, Industrial Chemicals,
John Wiley & Sons, Inc., New York (1950) .
4. Kehde, H., Styrene Technology Center, Midland Division Research
and Development, Dow Chemical Company, Midland, Mich. per
sonal communication, October 1967.
5. Ohlinger, H., and S. Stadelmann, Cher. Ing.-Tech. 361 (1965) ,
6. Oliver, E. D., Diffusional Separation Processes: Theory, Design,
and Evaluation, JohWiley & Sons, IncNork
(T
966).
7. Perry, J. H., Chemical Engineers' Handbook, McGraw-Hill Book
Company, Inc., New York (1966) .
8. Robinson, C. S., and E. R. Gilliland, Elements of Fractional Dis
tillation, McGraw-Hill Book Company, Inc., New York (1950).
9. Smith, B.D., Design of Equilibrium Stage Processes, McGraw
Hill Book Company,lnc.,New York
(
1963) .
10. Wenner, R. R., and E. C. Dybdal, Cher. Eng. Progr. 44, 275 (1948)

. lOCBSS LBSgD BDU CVBUB1lOD Ol 1DB
lOUUC1OD O 1lBDB ODODBl
This case foclises the design of a yeact oy for tile catalytic dehydro
genatioll of ethylbenzelle to /)Yoduce styrene. Initial elllphasis is placed
tlle analysis of laboratory Idlletic data and on ascertailling Ille iu
fluence of mass t ransfer effects Ihese data. Afler an acce/Iable
correlalion of the dala is achieved, a cOlllmercial-sclle dehydrogena
tion react or is modeled bJ' a set of nonlinear diferential equalions. .
com/mler program is IcriltCIl for tie solution of Ilese equations. The
effects of mass transfey in the bulll gas and diffusion within calalyst
pellets are found lO be particularly significant faciors in tie reactor
operalion.
The computer program from Chaptey is combined with the reac
tor design /)Yogram prepared in Chapter 7, so thai a filla/lcial evalua
tion of Ihe reactoY-fraclionatoy combination call be completed. The
optililum process is selected the basis of this el'aluation.
Vadalia Chemical Compay
Suth Chicopee, Mississippi
To: Wilbur Wolf, Chief Design Engineer
Fyom: Herma Fox, Vice President
C Styrene reactor design
The preliminary design work for the ethylbenzene-styrene frac
tionating unit which was prepared by your group has been received
and is now being reviewed by my staff. Your design appeared to ac
count for all the important operating variables in the system with
the exception of the feed rate and feed composition. Naturally, a de
tailed consideration of these parameters was not warranted because
the reactor design had not yet been completed.
151
1O<
Chapter 7
It is now desired that your design group proceed with the reactor
design. After completing consideration of the reactor it should be
possible to optimize the reactor-fractionator process unit. Most of
the information necessary for the reactor design is available in my
original memorandum, entitled "Styrene Production from Ethylbenzene. "
For the phase of the design involving the reactor, you may find some
useful information in the references listed. D addition, some data and
notes are given that pertain to the experimental work done in studying
the chemical kinetics for the catalytic dehydrogenation of ethyl
benzene.
Data ad Notes
The following information describes the bench-scale reactor used
by Wenner and Dybdal (12) in obtaining kinetic data for the conver
sion of ethylbenzene to styrene.
Reactor type: Catalytic, fixed bed
Reactor diameter: . 75 in. Ld.
1. 00 in. o.d.
Over-all lengh: 78 in.
Preheat section: 1. 5 ft packed with 6-mm porcelain Berl
saddles
Catalyst section: 4 ft
Catalyst volume: 348 cc
Temperature was measured in the catalyst bed 1 2 in. from the
exit.
Because the terms "conversion" and "yield" have been defined in
the literature in a number of diferent ways, the following definitions
are set forth as those to be used consistently throughout the present
discussion:
Conversion: Moles of ethylbenzene reacted per mole of ethyl
benzene which entered the reactor.
Yield: Moles of styrene monomer formed per mole of ethyl
benzene which reacted.
NOTATION
Specific surface area of catalyst, sq ft/lb catalyst
A Cross-sectional area of reactor tube, sq ft, or in Eq. 7. 33 the
total surface area in the shell-and-tube reactor, sq ft
L Flow rate in bottoms stream of fractionator, lb moles/year
CeC
l
Capital cost for two catalytic reactors, dollars
C
p
Heat capacity, Btu/(lb) (OF)
'
n
Annual cost for recycling ethylbenzene, dollars/year
Cy Annual cost for yield loss, dollars/year
Process EmIlio/ion /01' Styrel/c Prodllctio/l 1OJ
D Binary diffusivity, sq it/hr
d___ Average diameter of reactor tube, ft
d_ Pellet diameter, it
J
1. 35/z1 P [( 1 * 7 - y - ) - (P/ll) (x (x ~
z
)]
1 Total ethylbenzcne feed rate to a reactor tube, equal to ( / * 7)/
B000 number of tubes, lb moles/(hr) (tube)
/ Ethylbenzene fresh feed rate, D moles/year
Mass velocity, lb/(hr) (sq ft)
J Enthalpy change of Reaction 6. 1, Btu/lb mole
/o
Diffusion j-factor, defined in Eq. 7. 2, dimensionless
: Boltzmann's constant
:
P
Kinetic constant for Reaction 6. 1, g mole/(sec) (g mOle/cm3)
(cm3 of pellet volume)
k
y
Gas phase mass-transfer coefficient, lb moles/sq ft (atm) (sec)
:
1
Kinetic constant for Reaction 6. 1, (lb moles)/(hr) (atm) (lb cata
lyst)
IZ
2
Kinetic constant for Reaction 6. 2, (lb moles)/(hr) (atm) (Ib cata
lyst)
11
3
Kinetic constant for Reaction 6.3, (lb moles)/(hr) (atm
2
) (lb cata
-
lyst)
'
j
Equilibrium constant for Reaction 6.1, atm
l c//ax, see Eq. 7. 2 5
1 Lengh increment n numerical computation, it
L Lengh along catalyst tube, ft
L Mean free path of a gas molecule, cm
Molecular weight
P Mass flux, lb moles/(sq ft) (sec)

Knudsen number, 2y
e
/ L dimensionless
,__ Reynolds number, d_ G/,dimensionless
[__ Schmidt number, l/pD
P
Partial pressure, atm
P Total pressure, atm
J_ Critical pressure, atm

g
Reduced pressure, atm
r Reaction rate, D moles/(sec) (lb catalyst)
`8
Mean pore radius, cm
1.54 ChuIcr 7
Universal gas constnt, cal/(g mole) [K);in computation of
effectiveness factor, equal to catalyst pellet radius, cm
R Ethylbenzene recycle rate, lb moles/year
b
y
Catalyst surface area, m
2
/g
T Temperature, oK in kinetic discussion, 1 in reactor
-design cal -
culations
Over-all heat-transfer coefficient, Btu/(hr} (sq ft) (OF)
Superficial velocity of reacting gas in tube, ft/sec
\
a
Specific volume of catalyst bed, cm3/g catalyst
!
Critical volume, cm3/g mole
'
Specific pore volume, cm3 of pore volume/g catalyst
r
Specific volume of solid material which comprises the catalyst,
cm3/g
z
Bulk void volume of catalyst bed, taen as . 3BV
a
, cm3/g catalyst
Degree of conversion of Reaction 6. 1
2 Mole fraction of ethylbenzene in the liquid phase
)
Degree of conversion of Reaction 6. 2
I Mole fraction in vapor phase
Degree of conversion of Reaction 6.3
Z Compressibility factor
a(x) m for flrSt dmerence,egml l0 (X
I
,_)/Q
L

/
/8
/
/T

Lennard-Jones force constant, erg

Effectiveness factor; ratio of the observed reaction rate to the
rate that would be observed in the absence of internal difusion
effects, dimensionless
Porosity of a catalyst pellet
Dynamic viscosity, cp
Growth factor in numerical solution
3.1416
Bul gas density
Bulk density of catalyst bed, lb m/cu ft
Density of catalyst pellet, g/cm3
Intrinsic density of solid material that comprises a catalyst
pellet, g/cm3
Lennard-Jones force constant,
Tortuosity of pores in a catalyst pellet, minimum value equal to
2, dimensionlesss
lOccoS It|ulO|I (5r $lrcnc JOt/t|cI1O|t
JOO
T
Thiele modulus, VPo__ dimensionless
1 l+x +.

c
Diffusional collision integral, dimensionless
Subscripts
ave Refers to average value
Refers to composition of bottoms stream in fractionation unit
0 Refers to critical property
Refers to ethylbenzene
E0 Refers to equilibrium property
L2 Refers to experimental property
O1 Refers to effective diffusivity
J
Refers to log mean property within the film

Refers to condition in bulk gas stream

Refers to property of flue gas

Refers to property of hydrogen

Refers to property at outer surface of catlyst pellet
Integer subscript referrig to position within reactor tube
n1 Refers to an intrinsic chemical kinetic constant
D Refers to a Knudsen diffusivity
Refers to a property of styrene
FORTRA Nomenclature
A Total surface area for heat transfer in shell-and-tube
reactor, sq ft
ADENt Density of reacting gas at entrance to a tube, lb m/cu ft
ADEN2 Density of reacting gas at exit of a tube, lb m/cu ft
ADEN (ADENt * ADEN2)/2. 0
AMW Average molecular weight
APRESS Absolute pressure at tube entrance, psia
AREA Internal cross-sectional area for flow withi a tube,
sq ft/tube
ATEMP Temperature of ethylbenzene feed, 1
ATEMPG Exit temperature of flue gas, `1
AVEL Average superficial velocity within a tube, ft/sec
] Clza/>le /
BZ Mole fraction of benzene in gas phase
CATCST
CATDEN
CATDIA
CATPRC
CPEB
CPFG
DEN
EB
EBPRCE
EK1
EPS
ETA
FRATE
H
HYD
I
J
K
M
MDEL
Total capital investment required for catalyst in two
reactors
Bulk density of catalyst bed, lb m/ cu ft
Diameter of catalyst pellet, ft
Price of catalyst, dollars/lb
Heat capacity of ethylbenzene, Btu/(lb) (OF)
Heat capacity of flue gas, Btu/(lb) (OF)
Density of reactig gas mixture at any point in the reac
tor, lb/ cu ft
Mole fraction of ethylbenzene in vapor phase
Price of ethylbenzene, dollars/lb
Equilibrium constant for Reaction 6.1, atm
Bulk void fraction of packed bed, dimensionless
Effectiveness factor, dimensionless
Feed rate of ethylbenzene, lb/(hr) (tube)
Over-all heat-transfer coefficient, Btu/(hr) (sq ft) (OF)
Mole fraction of hydrogen in vapor
Integer subscript controlling the numerical calculations;
I advances by 1 for each hal step in distance for which
the calculations are completed.
Integer variable that controls the value of the temperature,
pressure, and composition variables to be used in taing
the next distance step. When the value of I is an odd num

ber' J is set equal to unity, and the values of temperature,

pressure, and composition at the beginning of a distance
increment are used to compute the new temperature,pres
sure, and composition at the midpoint of the distance incre
ment. Similarly for even I, J equals 2 and the values of
the midpoint are used to complete the values at the end
of the distance increment (K C 3).
Integer variable used in tandem with J. A K value of 2
corresponds to the midpoint of a distance increment where
the average values of temperature, pressure, and compOSi
tion are established. Similarly, a K value of 3 corresponds
to the end of a distance increment .
Integer equal to 1/2; when M exceeds MM, the temperature,
pressure and composition at that point in the calculation
are printed.
Input integer variable which controls the frequency of
printing composition, temperature, and pressure values
MM
NSLICE
PRESS
PRESSG
PREXG
PREXT
RATEFG
RCCOST
RCPRICE
RCTCST
RK1
RK2
RK3
SLICE
STY
TDEL
TEMP
TEMPG
TEMPK
TID
TLEN
TOD
TOL
TOTCST
Process El'aluation for Styrene Production 757
as the calculation proceeds. MDEL was set equal to 10
for the example shown in the appendix.
Integer used to control printing of output, see definition
of M
Number of distance increments used for the numerical
calculation. A value of 100 was used in the example
shown in the appendix.
Gauge pressure of reacting gas at any point in the reac
tor tube, psig
Absolute pressure of reacting gas at any point i the
reactor tube, pSia
Gauge pressure of reacting gas at exit of reactor tube,
psig
Absolute pressure of reacting gas at exit of reactor tube,
psia
Flow rate of flue gas, lb/(hr) (tube)
Cost of recycling ethylbenzene, dollars/year. Used ID com
putations for first economic basis
Price of recycling ethylbenzene, dollars/lb
Capital cost for two shell-and-tube reactors, dollars
Kinetic constant for Reaction 6. 1, lb moles/(hr) (atm)
(lb catalyst)
Kinetic constant for Reaction 6.2, lb moles/(hr) (atm)
(lb catalyst)
Kinetic constant for Reaction 6. 3, lb moles/(hr) (atm
2
)
(lb catalyst)
Equal to NSLICE
Mole fraction styrene in vapor
Length increment in tube, equal to TLEN/SLICE, ft
Temperature of reacting gas mixture at any point in the
reactor tube, of
Temperature of flue gas at any point in the reactor, `1
Absolute temperature of reacting gas at any point in the
reactor tube, oK
Internal diameter of reactor tube, in
Total tube length, ft
Outer diameter of reactor tube, in
Mole fraction of toluene in vapor
Total operating cost for first economic basis, equal to
RCCOST + YLCOST, dollars/year
J58 Chapter 7
TUBES
VEL
VOID
X
X
XMOLE
Y
Number of tubes required for the reactor
Velocity of reacting gas at any position in reactor tube,
ft/sec
(1. 0 EPS)/EPS**3
Total convei'sion of ethylbenzene, equal to STY + BZ +
TOL
The reciprocal of the fraction of TDEL to be used in the
numerical calculation. For odd I, XI is set equal to 2. 0,
and a "half-step" in distance is taken; for even I, XI is set
equal to unity and a complete step in distance is taken.
1. 0 + BZ ' STY
Number of distance increments for which calculations
have been completed at any stage in the numerical solu
tion
YIELD Per cent styrene yield
YLCOST Annual cost for ethylbenzene that is degraded to the
undesirable by-products benzene and toluene, computed on
the basis of $ . 1 2/lb ethylbenzene lost, dollars/year
Z Distance along the reactor tube, ft
DESGN OF TIlE REACTOR FOR CONVERSON OF ETIIYLBENZENE
TO STYENE MONOMER
With the fractionator design now complete, it becomes appropriate
to consider the design of a chemical reactor for the conversion of
ethylbenzene to styrene. Before proceeding with the design calcula
tions, it is necessary to review the possible influences that mass
transfer and diffusion within the catalyst may have upon the per

formance of the reactor. With these thoughts in mind, the data of

Wenner and Dybdal (12) will be examined to determine their appli
cability to the present design. Following these necessary preliminary
calculations, the actual reactor design will then be completed. As an
integral part of the reactor design, it will be necessary to consider
various techniques in the numerical solution of differential equations.
Proper use of these techniques will allow a computer program to be
prepared to carry out design calculations for wide ranges of design
parameters.
Properties D te Catalyst Used DI the Dehydrogenation D Ethyl
benzene
The data of Wenner and Dybdal are to be used to design a catalytic
reactor for the dehydrogenation of ethylbenzene. Before the design
can be carried out, it is necessary to determine whether or not mass
transfer had any effect on their experimentally determied reaction
rate data. Both the effects of mass transfer to the catalyst pellet sur-
Process Elyliualioll foy Styrene Production 1OV
face and diffusion within the catalyst pellet must be considered. U
mass transfer is found to have limited the rate of reaction, then the
data must be corrected for these effects before the design can be
i mplemented. To ascertain the influence of mass transfer, it is im
portant that the properties of the catalyst pellets used in their tests
be established. The following information is available from the orig
inal article:
Bulk density: 61 lb m/cu ft
Pellet size: 4-8 mesh, standard Tyler screen size.
From Perry's Handbook (],we find that
4 mesh' 4. 70 mm ' . 185 in. ' 3/1
6
in.
8 mesh' 2. 36 mm ' 0. 093 in. ' 3/
3
2
in.
As an average pellet diameter we may then take
1 _
3
3
_ C -+ _ in.
. 1 40 in. C 0. 0117 ft
2 16
32
This information constitutes the only speciications given by Wenner
and Dybdal for the catalyst used by them. Thus, some "engineering
approximations" must be made in order to proceed with the analysis.
To carry out a dehydrogenation reaction, one might well turn to a
platinum catalyst deposited on a silica-alumina carrier. From Sat
terfield and Sherwood [9),Table 3 - 1, p. 72, the following data are found
for a typical silica-alumina catalyst used in the dehydrogenation of
cyclohexane to benzene:
Sg 7 240 m
2
/g
C 0.59
n,

1. 33 g/cm3
r
7 3 . 2 5 g/cm3
where S is the surface area of the catalyst, the porosity of a
catalyst pellet, _ the density of a catalyst pellet, and
n
r
the intrinsic
density of the solid material that comprises the pellet. For this
catalyst, the pore volume per mass of catalyst, _,is given by
1 1
_
C
/,
C
- . 59/1. 33 C 0. 443 cm3/g catalyst

, n
r
Now consider a container of volume
\
a
,packed full with one gram of
catalyst pellets. U the pellets are assumed to be spheres, it is reason-
J ChaI /
able to tae the vid vlume btween pllets, Fas O. 3of the ttl
vlume of the contier.
To find the total vlume 0 the contaier:
_ F O. 443 cm3/g catlyst
F 4 .3F_
F
t
F I/g 4 .3
F_ 4 .
1
5I .3Fg cm3/g catlyst
Hence,
F_ 4 1. 21 cm3/g catalyst
or the bl density of the catlyst bd is
Q F _(6.4) F 5I. 5 l/cu
Tis numbr is smewht less m the value of 01 l/cu gven in
te origi article. However, te property vlues

"
.5
P "
I
-
51
P
t
" -
I5
are taen, the container volume can be computed in a similar maner
to be 1 . 034 cm; /g catalyst . A value of 61. 0 lb/cu ft then follows for

B
. Since only minor modifications were necessary in the listed pro
perties of a "typical" dehydrogenation catalyst, the last set of proper
ties in this computation will be taken as those of the catalyst used in
reference (1:).
General Examination of te Kinetic Dat
Before proceeding to analysis relating to the mass transfer and
difusion reSistances, it is interesting to examine the kietic data
show in Table 6-1 for the possibility that chemical equilibrium may
have limited some of the observed conversions. The equilibrium ex
pression for styrene formation is written as
Process Evaluation for Styrene Production 1O1
Values of _ are shown in Fig. 6-1. U the side reactions are neglec
ted for the moment, the conversion of ethylbenzene may be calculated
as
_2
b

1 - _
The equilibrium conversion of ethylbenzene, as computed with this
equation, is shown in Table 7
=
1e From the comparison of [Y)

and
[Y)

_
1
it is clear that the measured conversions in runs 227, 228,
229, and 230 correspond almost exactly to those computed using the
assumption of chemical equilibrium. Thus, even though the analysis
of the data set forth in reference [J3) does take the reverse reaction
hydrogenation) into account, there would be no way of analyzing the
data from these runs to obtain accurate values of u_+ For example,
by doubling the length of the experimental reactor, the same equilib
rium conversion would be obtained; however, the computed value of
u from this hypothetical experiment would be quite different from that
obtained from runs 22 7 to 230.
Table 1~1. Comparison of Observed Conversions with
Those Computed Assumig Chemical
Equilibrium
Run Number T (OC) _ (atm)
\s!sp \ssxr
227 555 0. 082 0.246 0.226
228 5 52 0. 082 0. 246 0. 22 5
2 29 555 0.082 0.2 50 0. 227
230 556 0. 082 0.2 50 0. 2 27
237 598 0. 2 2 0.3 6 7 0.246
239 600 0. 2 2 0. 3 72 0. 269
2 52 676 0. 95 0. 5 80 0. 366
2 54 676 0. 95 0.580 0. 3 60
242 598 0. 22 0. 3 72 0.260
246 650 0.64 0.500 0. 3 10
2 5 1 671 0. 93 0. 575 0. 3 5 5
255 668 0.93 0. 601 0. 3 72
232 555 0. 082 0. 2 50 0. 1 95
Followig up on the observation that equilibrium was attained in
four of the experimental runs, it is interesting to examine the Arrhe

nius plot for the kinetic constants obtained by Wenner and Dybdal. The
authors did not relate each of the plotted values of /shown in Fig. 6 -2
1OZ Cha/)ter 7
to a specific experimental run summarized in Table O1. However,
an attempt is made in Fig. 7-1 to associate the P_ values with speciic
exerimental runs. This association was accomplished by comparing
the temperatures from the data of Table 7-1 with those of the pOints
shown in Fig. 7 - 1 . The anticipated discrepancy for runs 227 to 230
i s seen in Fig. 7-1. There is great scatter in the data at a 1 000/T
value of approximately 1. 2 1 . It seems clear that the constant conver
sion obtained when the feed rate was increased from . 1 4 to 0. 21 lb
moles/hr led to the computation of P
j
values that were too high. There
fore the data from runs 227, 228, 229, and 230 will be discarded i
future computations.
3Oj
}
. ...
. tt t t.}
K IT wtIh T lf "K
Figure 7-1. Kinetic constats for the
dehydrogenation of ethylbenzene.
[Werer and Dybdal, reference 12.)
Identification of experimental runs
has been inferred by examination of
data tabulation in reference 1Z
Table 1.
Before proceeding, it should be noted that a rather serious dis
crepancy remains. A comparison of runs 227 and 232 indicates that
when the reactant feed rate was decreased, with all other variables
held constant, the conversion was reduced by almost 1 5 per cent.
Naturally, by decreaSing the flow rate and increasing the residence
time, one would expect an icreased conversion of ethylbenzene to
styrene. Other than exerimental error, no reasonable exlanation
for this behavior is apparent.
Gneral Considerations U the Aalysis of Catalytic Reactions
D analyzing a catalytic reaction, seven possible resistances to the
progress of the reaction must be considered:
1 . Transfer of reactants from the bul medium to the catalyst
pellet surface
2. Difusion of reactants from the pellet surface to the interior
Process Em/l/ati ol1 jm' Styrene Productioll
163
of the pellet where most of the catalytically active area is
found
3. Adsorption of reactants on the catalyst surface
4 . Reaction upon the catalyst surface
5. Desorption of the products
6. Diffusion of products from within the pellet to the pellet sur
face
7. Transfer of the products from the pellet surface to the bulk
medium
A detailed analysis describing the kinetics of adsorption and a
consideration of several possible mechanisms for the surface reac
tion are beyond the scope of the present work. Smith's text (J0) pro
vides an excellent discussion for the reader interested in this sub
ject. The kinetics of adsorption and surface reaction will be analyzed
by the models originally postulated in reference [J2). However, com
putations will be made to analyze for possible mass transfer or di
fusional resistances that may have inluenced the exerimental data
of reference (12).
Mass-Transfer Rates at the Surace D Catalyst Pellets
The system to be analyzed is a packed bed of catalyst pellets
through which a reacting gas stream is passed. The rate of mass
transfer from the bulk gas to the surface of a pellet in the bed may
be expressed as the product of a mass transfer coefficient and a
partial pressure driving force:
( 7. 1 )
Mass transfer coefficients have been conveniently correlated i n terms
of the Chilton -Colburn ) factor. Experimental data have shown that
the best exression for ) is that given by (9)
k A1
]
C
. NO
.66
7 7 0 989
N-O
.
41 for ,y_ ` 3 50

cC O
( 7. 2a)
7 1. B3 for t\ < 3 50 ( 7. 2b)
As the first step in the computation of ]
c
it is necessary to estimate
the Reynolds number for the flow over the catalyst pellets. In this
calculation, the lowest mass flow rate obtained will be used (run 242);
this is the condition at which mass transfer could have had the great
est influence on their experiments. The calculation of each of the
variables necessary for the estimation of is
(a) Mass velocity,
For run 242,
(0. 0064) ( 106)
C 7 0. 061 3 lb m/sec-sq ft
(0. 00307) (3 600)
1OJ
C|u[Icr 7
(b) Viscosity,
The viscosity of ethylbenzene vapor is not stated in the
literature; therefore, it will be estimated by two different
techniques and these estimates will be compared.
(i) The viscosity of ethylbenene is probably quite similar
to that of toluene, which is listed in Perry's Handbook
(7 ) as 0.0202 cp at 598C. Note that no pressure
correction is necessary for this value, since
P/PC=PR 1.
(ii) The method of Bromley and Wilke, as described by
Reid and Sherwood (8) , is a convenient method of
estimating viscosity; by this method
(7.3)
where from reference (7)
7_ = 346C and 1_= 38.1 atm.
For a Z_ value of 0.27 (equal to that for toluene) ,
Z_R7,
0.27 (82.05)(619) cc

, =

= =360
P
,
38. 1 g mole
Therefore,
0.00333 (106 X 619)
1
/ 2 (1.137)
/
=
/
= 0.0191 cp
(360) 2
which agrees quite well with the value for toluene.
(c) Reynolds number and jD
4_G (0.0117)(0.0613)

_ =
~ 56

-
(0.02)(6.72 X 10-4)
Hence, from Eq. 7. 2b,
jD = 1. 82(56)-0.51 = 0.234
To use this value in calculating the mass transfer coeffi
cient' it is first necessary to estimate the diffusivity of
ethylbenzene in styrene. This will now be carried out using
Hirschfelder's equation, as given in reference 8):
0.001858 T3
/
2
(
M

1
/
2
D
1
2
=

Pa
2
iD
(7.4)
Process Evaluation for Styrene Production 1OO
The Lennard-Jones force constants are estimated by the
conventional empirical method, whereby
a
1
2
0.833 V
/
3
c/u C . 77T
c
Since no data on the critical constants of styrene are readily
available, the critical values for ethylbenzene will be
assumed to apply for the mixture. Therefore, the force
constants are computed in a straightforward manner:
a
1
2
~ 0.833 (360)
1/
3 5.92
c/
u 7 0.77 (619) C 476K and
/:I/c C 871/476 1. 83
Thus, from Reid and Sherwood (8) the collision integral is
found to be
QD C 1. 116 - 0.6 (0.011) 7 1. 109
The diffusion coefficient can now be computed as
0.001858
(
871)
3
/
2
104
+ 106
l
1
/2
D
ES
C
104 K 106
:"-0.168 cm
2
/sec
(1. 0)(5. 92)2(1. 109)
And the Schmidt number is found as
0.0191

8
C
J
/
p
D
ES
C 7 0.772
C
102 (1. 47 K 10-3)(0.168)
The mass transfer coefficient can now be calculated:
(0.234)(0.0613) lb moles
u
7 ~ 1. 73 K 10-4
y
(105)(0.9)(772)2/3 sq ft-atm-sec
To compute the possible diffusional resistance in the kinetic
experiments, the superficial area of the catalyst pellets must
be determined. U it assumed that the pellets are uniform
spheres of diameter then the porosity of the bed is 0.38
and the total superficlal area may be computed:
(0.62)(7)(0.0117)2
C
/
5.12 sq ft/lb catalyst
1 67(0.0117)361
1
Clu[Icr 7
The observed reaction rate in run 242 is
0.00640 (0.26) lb moles EB reacted
7 C C 6.19 K 10-7
3600(0.749) lb catalyst-sec
Since at steady state the mass transfer rate must be equal
to the observed reaction rate, the partial pressure driving
force for mass transfer is found from Eq. 7. 1 to be
6.19 K 10-7
v -
p
.)
= = 7.00 K 10-4 atm

(1.73 K 10-4){5. 12)
Hence, with an insignificant partial pressure drop due to
mass transfer, it is concluded that no important resistance
exists for diffusion through the boundary layer at the pellet
surface.
It may be validly pointed out that a detailed calculation of the dif
fusion coefficient is unwarranted, since only a rough estimation of
the mass transfer rate is required. The details arc presented for
illustrative purposes.
Diffusion Witin the Catayst Pellet
The catalyst used in this work has most of its effective area within
the interior of the porous pellets. Thus, limitations due to diffusion
within the pores must be considered. The effect of pore diffusion is
evaluated in terms of the dimensionless parameter q the effective
ness factor. The parameter q is equal to the ratio of the actual reac
tion rate to the reaction rate that would have been observed i all the
interior catalyst surface were exposed to a reactant of the same con
centration and temperature as found at the outer surface of the cata
lyst pellet. For a first-order irreversible reaction, 7 is found to
be (9)
3
1
=-

--
c tanh c (7.5)
where
(7.6)
The variable c is conventionally termed the Thiele modulus; it may
be thought of as the ratio of the chemical reaction rate to the diffu
sion rate within the pellet. To facilitate calculations, Eq. 7. 5 has
been presented graphically in Fig. 7-2.
Process Et:tIuolioit/or Styrene 1Od/IcOH 1&
|.0
`
\
09
0.0
\
0.4
0.5
0.2
0 2
\
A
`
`
"

4 6 0
Th8|8 m00u!u5g
`
`
|0 |2
Figre 7-2. Effectiveness factor as a
fuction of Thiele modulus.
The effective diffusivity is most frequently computed by assuming
that the resistances due to ordinary molecular diffusion and to Knud
sen diffusion act in series; therefore it may be written that
1 1 1
~4
-yy l

/-yy '8ne1t
(7.7)
The effective molecular diffusivity is defined as
D
12B/T, where g the
tortuosity, is the factor that corrects for nonlinearity and nonunifor
mity of the pores within the catalyst pellet. The effective Knudsen
diffusion coefficient is computed from
(7.8)
where the factor 20/S
0
may be thought of as the mean pore radius
in the pellet, T. Equaion 7. 8 has been derived from the kinetic
theory of gases for cases when 7 is less than the mean free path of
the gas under consideration.
From Eqs. 7. 5 through 7. 8 the value of q may be computed for
each experimental point of Wenner and Dybdal. This trial-and-error
computation is illustrated in the following section.
!63 Chu[lcr 7
Computaton of Efectiveness Factor for the Exeriment Dat
From the definition of (D1
2
)
e f f
there follows
(0.16(0. 5 0.084
(D
1
2
)
eff
=
= -cm
2
/sec

And from Eq. 7. 8 we have
(
0.59)
2
1871 0.00610
D ) = 19 400
.
= cm
2
/sec
hT L11
7(1. 33)(240 K 104) 105
Since D
k
)

__ is much smaller than

(D
12)eff' the Knudsen mechanism
constitutes the controlling resistance to diffusion within the catalyst
pellets. As a check on this conclusion it is interesting to compute the
dimensionless Knudsen number. First the mean free path for the
reacting gas is found as
1. 881
L= --=
1. 88 (1. 91 X 10-4)
.`
0
-
%
[871
]
1. 47 K 10-
3
-(8.29 K 107)
105
= , 933 K 10-5 cm = 933
From the definition of the mean pore radius,
29 2 (0.59)
r
e
=
-
=
., --
S_0
(240 K 104)(1. 33)
C 3.70 K 10-7 cm = 37
And the Knudsen number is found to be
2
r
e 2 (37)
* -= ~= 0.0793
L 933
With __ 0.1, Knudsen diffusion should be the controlling mecha
nism of diffusion. This calculation checks the previous result.
USing 8 as a typical value for the tortuosity, the effective diffusi
vity becomes
0.0061
Deff

-
8
-= 0.00076 cm
2
/sec
With this information, the following general relation for the Thiele
modulus may now be developed:
-
= 0.178
= 6.45.k
F
0.00076
(
7.9)
Process Evaluation for styYene Pyoduction 169
By combining Eq. 7. 9 with Eq. 7. 5 or Fig. 7-2, a trial-and-error
method is made available for the calculation of q. In this calculation
the experimental value
k
l is read from Fig. 7-1; the dimensions of
this rate constant must be converted to those of
k
F
g the variable
necessary for the computation of the Thiele modulus. The basis for
this conversion is
g moles
(7. 10)
The values of l?
F
computed by this expression represent the experi
mentally observed rate constants. By first assuming a value for 7,
the intrinsic rate constant {l?
F
)
in _ may be calculated by taking the
ratio
(
k
F
)
e
x
p
/7. Then, by using Eq. 7. 9 and Fig. 7 -2, the accuracy of
the assumed 7 value is checked. A sample calculation for the deter
mination of 7 is shown in Table 7 -2.
Table 7-2. Caculaton of Q for u Z1
I 871K
kl
7 1. 10 K 102 ( Fig. 7-1)
(kF)e
x
p
7 0
.
343 sec
l
(Eq.7.10)
Tassumed
(k
Ant
Eq. 7. 9
0.7 0.490 4.52
0. 5 0.687 5. 35
0.44 0. 782 5. 70
7, Fig. 7-2
0. 5
0.45
0. 44
The computed value of q indicates that internal diffusion within the
catalyst pellet does indeed provide a major resistance to the rate of
the dehydrogenation reaction. In fact , for run 237 the observed reac
tion rate was only about 44 per cent of the rate that would have been
observed if no internal diffusion resistance had existed. It must be
remembered that a great many assumptions have been made in com
puting 7. It would be interesting to determine the effect of variations
in these original assumptions
(
e.g. , _ or r} on the calculated value
of ].
Ater calculating q for each experimental run by these methods,
Eg. 7.10 is used to compute corrected values of !hC intrinsic rate
constant . These calculations are shown in Table 7 -3 and the cor
rected intrinsic kinetic constants are plotted in Fig. 7-3.
170 Chapter
`

F
-?
z&
4
`
= E
H'
r

E
`

|.0 i.t t. t.}
7T wt!h T |n "K
Figre 7-3. Chemical kinetic constants
for Reaction 7.1 after correction for
internal difusion resistance.
10 1, 0C1V0DCSS CDIS ad DIDS1C 0
LDDSD\S DI CIDCD\1 Dat
Run I,k
(
ki
nt (
lzl
)
i
n t
K 102
237 871 0.44 0.780 2.5
239 873 0. 44 0. 782 2.5
252 949 0. 28 2. 19 6.45
254 949 0. 29 2.02 5.94
242 871 0.44 0. 78 2.5
246 923 0.32 1. 70 5.15
251 944 0.33 1. 64 4.86
255 941 0.37 1.11 3. 30
232 828 0. 53 0.402 1. 36
These data are found to be best fitted by the equation
D kl 7
19
,
100 _
7.30
RI
[7. 11)
The activation energy of 19.1 Kcal is a much more realistic value
for a dehydrogenation reaction than the activation energy of 11.37
Kcal computed without correcting the data for diffusion resistance.
In fact, the relatively low value of the activation energy derived from
the original data would indicate in itself that diffusion might be exert
ing a controlling influence on the observed reaction rates. The fact
Process Elall/ation for Styrene Prodl/ctiol! 171
that Eq. 7.1 1 differs significantly from the rate exression obtained
in reference (12) should not be taen as a criticism of this work.
It must be remembered that the properties of the catalyst used in
reference [!2) are not known and that Eq. 7.11 would be strongly
influenced by variations in the assumed properties which have been
used in the present calculation.
Desig of a Fied-Bed Dehydrogenation Reactor
Because the dehydrogenation reaction is endothermic, heat must
be supplied to the reacting mixture; therefore, a shell-and-tube reac
tor of the type shown in Fig. 7 -4 has been selected for the present
design in order to provide a large heat-transfer area per unit volume
of catalyst. The catalyst pellets will be packed in the vertical tubes,
and the heat -transfer fluid will be passed through the shell side
countercurrent to the reacting gases. It is anticipated that a stream
of hot flue gas will be available to supply heat to the reactor.
EIhybezeoe-
!eed
.& . _
1'ue QU5
. . . . . _
ReocIo
producI
Figre 7 -4. Dehydrogenation
Reactor.
By referring to Eqs. 6. 1,6.2, and 6.3, the stoichiometry of the
reacting gas mixture may be summarized as in Table 7 -4. Here
the variables X)and represent the degree of conversion of
Reactions 6. 1, 6.2, and 6.3, respectively. With the stoichiometry thus
established, differential equations will now be derived to determine
variations in the composition and temperature of the reacting mixture
as it passes through one of the tubes packed with catalyst .
1. Material balances. At any arbitrary position along the cata
lyst tube, consider a section of tube dL in length. A steady-state
1
'
!c[lcx
Table 1~9. Stoichiometr of the Reacting Mie
Basis: 1. 0 lb mole ethylbenzene feed
Lb Moles at
Lb Moles Position L
Compound in Feed Catalyst Tube
Ethylbenzene 1.0
J_
)
:
Styrene 0 A
Hydrogen 0 X
Benzene 0
y
Ethylene 0
)
Toluene 0
Methane 0
1.0 1 X j
material balance written on styrene for this differential element
yields the following:
or
Input - Output
7 Accumulation

d
(Fx) JL -

|dL)0
a
] 0
dx
F
-
=
^a

1
dL
Substituting from Eq. 6. 4, there results
_
0
^

PP
dL
P

(7. 12)
(7. 13)
By similar consideration of Reactions O.` and 6.3, the following
equations are derived:
(7.14)
(7. 15)
Iroccss LIIIulO|I)r S/]rcnc Iruttcloti 173
2. Energy balaces. Considering a differential element of a
catalyst tube, the following energy balance may be written:
Input - Output

Accumulation
FIMC _(T 7
I ' '"
d
ave}
d
L(T

T) r

(AdL) _
-I
M
C peT - To}
'
[F1HC _ ( T - To)] O 7 0
If we assume that the quantity F111C
p
is constant and introduce
Eq. 7. 12, it follows that
dT
7
G___
1 T
ABb/(dX)
0 F1
M
C_

C_ d (7. 1 6)
It should be noted that the enthalpy changes due to Reactions 6. 2 and
6. 3 have been neglected.
Analogously, the energy balance on the countercurrent flue gas
stream is written
(7. 17)
Equations 7.13 through 7. 17 may now be used to determine variations
in temperature and composition within any one of the catalyst tubes.
The only information still needed for a complete solution to the prob
lem is a relation that predicts the bulk pressure drop through the
catalyst bed. A standard expression for pressure drop through a
packed bed will be introduced in a later portion of the analysis (1).
Equations 7. 13 through 7. 1 7 constitute a set of first-order, non
linear differential equations. A curs'JrY inspection reveals that they
are not amenable to an analytical solution and that thpy must be
approximated by a set of finite difference equations that can be
solved by numerical techniques. A trial solution will now be pre
sented to illustrate the general principles involved.
Tria Nuerica Soluton
The major disadvantage of solving differential equations numeri
cally is the fact that specific values of the equations' coefficients
must be selected and the numerical solution must be accomplished
for each desired set of coefficients. This is a difficulty that must
be tolerated when an analytical solution is not possible. The solu
tion to be presented here is based on the set of design parameters
used by Wenner and Dybdal in their original work [12). This set of
constants is:
Tube dimensions:
Feed rate:
Feed heat capacity:
d_

4.03 in; c

4. 50 in;

1 5 ft
F ~ 4. 0 lb mole/(hr)(tube)
0. 635 Btu/(lb) (OF)
J7 Chapt er 7
Feed temperature:
Heat of reaction:
Feed pressure:
Exit pressure:
Flue gas flow:
Flue gas exit temperature:
Flue gas heat capacity:
Over-all heat-transfer coefficient:
Catalyst pellet diameter:
Bulk catalyst density:
550C
53,600 Btu/lb moles
41 psig
5 psig
6520 lb/(hr)(tube)
1600F
0.285 Btu/(lbWF)
9.7 Btu/(hr)(sq ftWF)

__ in.
61 lb/cu ft
For this sample calculation the total pressure will be assumed to
vary linearly through the reactor. This assumption will be abandoned
in later calculations; it is used here only to simplify the illustration.
With the speCified design parameters and the stoichiometric relations
developed in Table 7 -4, Eqs. 7.13 through 7. 17 may be transformed
into the following difference equations:
l (x)
7 1. 351P[(1 - x ) - z) (x
)(x -z)]
l (y) 7 1.351z2P(1-x ) z)
l (z) 7 1. 35k3P (1 - x ) z)(x - z)
l (17 0. 0422 (T
G 7; 7921 (x
)
l (TG
) 7 00615 (Ta
- T)
(7.18)
(7.19)
(7.20)
(7.21)
(7.22)
The rate constants used in these equations must be the effective
values of these constants. Thus to determine k

an effectiveness
factor must be determined for the speCified design conditions. Since
the design temperature level is about the same as that used for the
laboratory exeriments, it is assumed that the only parameter that
Significantly affects q is the pellet radius H. The values of q deter
mined for the experimental data were about 0.4; the value of
corresponding to this value of q is about 6.2. Hence it may be
written that
ad
l/
o
,
ega
# 0.44
The vaue m

to b used on the desig is therefore ten to b

l

/,yy
F 0.44 (1.48 X 103 q~1.100/k1
(7.23)
Ircccss EtIttoIicn/cr Styrene IrcJttcIcn 175
As an approximation, Eqs. 6. 5 and 6. 6 will be assumed to describe
the rate constants for Reactions 6. 2 and 6. 3 accurately.
To continue the analysis, the reactor is now divided into a number
of sections, as indicated in Figre 7-4 . With the inlet conditions 1=0,
the temperature and composition at point = 1 are computed. The
new values of temperature and composition are then used to calcu
late the values for = 2; computations for subsequent points proceed
analogously. This comparatively simple technique, wherein the con
ditions at the end of the length increment are predicted by using
conditions at the beginning of the increment, is conventionally called
the "marching" method. Calculations using the marching technique
for a reactor subdivided into three length increments are shown in
Table 7 -5. Note that the procedure used here is quite simple, since
the dependent variables r. re known as = 0; in other words, this is
an initial value problem. The table shows that a negative value of \
is computed at position
2 in the reactor. Clearly this physical
impossibility is the result of some deficiency in the ability of the
infinite difference equations to model the differential equations.
Genera Considerations i te Solution of Diferentia
Equations by Finite Dference Techiques
Three basic criteria are conventionally used in evaluating the
effectiveness of finite difference techniques for solving differential
equations. The first, accuracy, is self-explanatory; the accuracy
with which the numerical technique models the differential equation
must be evaluated. In most instances, the accuracy of the solution
will improve as the increment in the dependent variable is decreased.
Obviously, however, there is a lower limit placed on !: L since the
number of calculations and the corresponding requirement for com
puter time increase as the number of increments in the independent
variable are increased. The amount of required computer time is
the second criterion.
The third criterion requires that the stability of the numerical
solution be examined. When a differential equation is solved by
numerical methods, small round-off errors are generated in the
last decimal place of each number used in the solution. In certain
instances, these errors can be magnified by the numerical technique
being used, and in extreme cases these errors increase until they
completely dominate the solution.
The accuracy of a finite difference technique is typically checked
by examining the solution to a problem similar in mathematical form
to the desired problem but analytically solvable. By comparing the
analytical and the finite difference solutions, the accuracy of the
finite difference solution is easily ascertained. The stability of a
numerical solution may be studied approximately by linearizing the
nonlinear portions of the original equation and analyzing the resulting
linear equation for its stability. Because the results of Table 7-5
indicate gross inaccuracies in the numerical technique, the system
of finite difference equations which yielded these results will be
examined for its stability.
I&
Chu[Icr
Table 1. Finite Diference Calculations for the Dehydro
genation Reactor Using Three Increments in the
Lengh Variable
Position, 0 1
X
I
0 0.237
Y
I 0 0.0011
ZI 0 0
T
I
, of 1022 962
(TG)I' of 1600 1618
PI (linear), atm 3.79 2.97
1 -(X+Y+Z)I 1.0 0.752
(x -z) I 0 0.237
i1,Eq.7.23 0.00925 0.00794
k
2
,Eq.6.5 4.3 K 10-5 1. 5 K 10-5
k
3
, Eq. 6. 6 10-
3
0.6 Y 10-
3
J{l' Fg. 6-1 8.4 K 10-2 3.8 X 10-2
1. 35 k1P(1 -X-Y -z) I 0.0473 0.0168
1. 35 k1P2xix-Z)I
/K1
0 0.097
1. 35 k
2
P (l-x-y
) I 2.2 Y 10-4
1. 35i
3
P2 (l-x-y -z) I
(X-Z)I 0
0.0422 (T
G
-T) I 25.5
792 ^](x) I 37.5
. 00615(TG -T) I 3.55
xI+1, Eq. 7.18 0.237 -0.163
Y
I+1, Eq.7.19 0.0011
Z I+ _ Eq. 7.20 0
TI+1, Eq.7.21, oF 962
(TG) 1+ __ Eq. 7. 22, of 161 8
A method of analyzing for possible instability in the finite differ
ence method used in Table 7 -5 may be illustrated by considering
Eq. 7. 13 and writing it in the more general form
dx
- =f(T,x,y, z)
JL
Process EI:aluatioll 101' Styrene Product ion 1
Expressing this equation in its finite difference form yields
XI+l
+
XI
-- = II
=
l
x
I
6L
where the preceding equation is written by assuming that the function
/ ca be expressed as a single-valued, linear function of x. The
equation then can be rewritten
Using a conventional technique, this equation is solved by assuming
that the solution for
^
I
7 A(. Substituting this assumed relation into
the difference equation, one obtains

=
( 1 + 6L) [7 . 34)
It is desirable to maintain the absolute value of the growth factor at
a level less than unity, for otherwise a slight perturbation or error
in the value of A would cause an unstable magnification of this error
with increasing values of . To satisfy this stability criterion, values
of the product 6L must be restricted as follows:
0 > 6L > -2
Since the value of is found to be negative, the first portion of the
criterion is automatically satisfied. In order to satisfy the second
portion, however, it is required that
~3
6L <
[7
:
35)
Equation 7 . 3 5 then provides the basic criterion for the stability of
the numerical solution. To implement it, 1 is first calculated as
]

7
-1. 3 5 1l
1
P + 3x

1
1
The largest value of K would b found when
Kmax -1. 3 5 (9 X 10-3) (3.79) 1
:
7
(
1
,1
8 K 10-2 J
And the maximum value of 6L to be used in the numerical calculations
is
3
(
6L) ma
x
7 1. 7 5 ft
1. 14
1 78 Cholcr7
Since a ( value of

ft was used in the calculations summarized in

Table 7-5, the approximate analysi s j ust carried out would indicate
that this value is too high. Smaller values of L will therefore have
to be used to mae a stable and accurate solution to the problem
possibl e. It was probably intuitively obvious that ( was too large
in the cal culations shown in Table 7

however, the foregoing ana

lysis provides at least a semiquantitative basi s for selecting an
appropriate value of the distance increment.
It should be pointed out that there are many other techniques that
mi ght have been applied in the numerical solution to this problem;
for example the method of lI ilne or that of Runge
-
Kutta might have been
used. Frequently these techniques are available in computation
centers in the form of " canned" programs. H applied in the present
problem such programs mi ght well reduce the requirements for
computation time.
Preparaton of a FORTR Program for te Reactor Design
A computer program for the reactor design is prepared by solving
the finite difference equations, Eqs. 7. 18 to 7. 22. One additional
equation, a relation for the pressure drop in the reactor, is requisite
for an accurate solution to the problem. The equation to be used in
the final design calculations is that given in reference [1) .
0

pu
z

O u_ 1 -
( 7. 26)
In solving the system of finite difference equations, the " marching"
technique used in Table 7

will be modified so as to insure a numeri

cal solution that i s not only stable but also highly accurate. This
improved method of solution, sometimes described as the improved
Euler technique , consists of the following:
1 . As before, divide the interval of solution into equal incre
ments of length.
2. Since this i s an initial value problem, values of all the inde
pendent variabl es are known at the inlet. Substituting these
values into Eqs. 7. 18 through 7. 22 and Eq. 7 . 26, the varia
tions in the dependent variables over half the first increment
are computed. Using the values of the dependent variabl es
computed at the mi dpoint of the first distance increment,
Eqs. 7 . 1 8 through 7 . 22 and Eq. 7. 26 are used to compute
the changes in the dependent variabl es over the entire first
increment.

. Steps ( 1) and ( 2) are repeated for all the distance increment s,

i . e. , until l

W.
This technique may be expressed mathematically as
^r +: ^j
)

7
\ j+i / z
^
j+: /z +: / z +: /z
( ( 7. 27)
Process EI' (lllalioJ l for Styrene Producli (/
1v
Analogous equations may be written for the other dependent variabl es.
By use of methods similar t o those previously illustrated, t he growth
factor for the numerical solution to Eq. 7 . 27 may be shown to be
.
'

=
1 + J l
@
1 + -
2
-
( 7. 28)
With a value of -1 . 1 4 for h, the maximum value of (J that will en
sure stability i s found to be 1 . 75 ft . Thus, the stability criterion has
remained unchanged from that for the simpler marching technique;
however, use of the improved Euler method should result in a marked
improvement in the accuracy of the solution.
A FORTRAN program may now be prepared so that design cal cu
lations can be carried out for wide range s of the parameters th< t
characterize the system. The improved Euler method is used for
the solution of the finite difference equations.
The computer program for carrying out the design calculations
is quite straightforward. To facilitate the interpretation of the pro
gram, a flow diagram for the programming logic i s shown on Fig. 7 - 5 .
The program first makes provision for reading the necessary
input data, including the flow rate , feed temperature, and exit pres
sure of t he ethylbenzene feed stream, t he rate and exit temperature
of the flue gas stream and the lengh and diameter of the tubes to be
used in the reactor. In addition, the program requi res values of the
catalyst bulk density and porosity, catalyst pellet diameter, and the
effectiveness factor for the dehydrogenation reaction as carried out
using catalyst pellets of the specified diameter. Finally, the number
of distance increments to be used in the finite difference calculations
must be supplied in the input data.
As a fi rst step in the calculation, the program estimates the pres
sure drop through the bed. This value i s computed from Eq. 7 . 26
with use of average values of the gas density and velocity. By adding
thi s approximate value of the pressure drop to the specified exit
pressure, the ent rance pressure is esti mated. Since the technique of
estimating is not exact, an iterativ, procedure is incorporated into the
program. By this procedure the initial estimate of the ent rance pres
sure is used and the enti re problem is solved to yield values of tem
perature, pressure, and composition at the downstream end of a reac

tor tube . If the value of the exit pressure is sufficiently close to the
specified exit pressure ( 5 2 . 5 psig) , then the finite difference calcu
lations are assumed complete and the program moves on to the next
portion of the calculation. U the indicated agreement on the down
stream pressure is not aChieved, the entrance pressure is re- estima
ted and the problem is solved repeatedly until convergence is
obtained.
The actual finite difference calculations follow directly from the
methods previously discussed. In all the computer calculations, 1 00
distance increments were used; even with thi s relatively large num
ber the program requires only a modest amount of machine time .
180
Chu[Icr7
lnIfohz vorIobIe8
IZI KI * IZ I I I + TDEL/Xl o F2 1 1
TEMPI KI TEMP I I I + TDEL /XI . F4I1
PESSI KI. PESSI I I - TL/XI . 'I II
WI gW fm0fWMm
wf8 of WK
fflox tdfs
Figre 7 - 5 . Flow diagram for the programming logic of
design calculati on. | Circled numbers ref er to statement
numbers i n the FORTRN program. )
Multiplying , the ethylbenzene feed rate per tube, with the value
of found at the reactor exit yields a value for the styrene produc
tion rate per reactor tube. Then by dividing thi s value into the total
required production rate, values for the number of tubes needed in
the reactor and the required catalyst volume are e stablished. These
quantities are of course necessary in sizing the reactor and in e sti
mating the capital requirement s for the process. In order to allow
for l osses in the fractionator, a styrene production rate 5 per cent
Process Eualualion /0r S/rcnc Production 1 81
greater than the required 20 K 106 lb/year has been used to estab
lish the reactor si ze. This i s an important assumption, and its effect
on t he over-all design should be subsequently examined.
In order to consider the estimation of costs for the dehydrogena
tion of ethylbenzene, it is assumed that / lb moles/year of fresh feed
must be mixed with lb moles/year of a recycle stream and then
fed to the reactor. For purposes of illustration, the costs of carrying
out the operation were computed on two different bases:
First Basis-Optimization of the Reactor Only. On the first baSiS ,
the costs of operating the reactor were evaluated by taing a fixed
price for completing the separation of the reaction mixture; i . e . , the
economi cs of the distillation operation was not considered. On this
basis the operating cost was assum ed to be composed of the sum of
two factors :
1.
_
g
, the cost of separating ad recycling the unconverted
ethylbenzene, taen as $0.01 per lb of unconverted ethyl
benzene.
2. ,the cost of ethylbenzene converted to by-products that
were assumed to be valuel ess, taen as $0.12 per lb of
ethylbenzene lost.
All ethylbenzene was assumed to have been recovered in the separa
tion step. With the definitions given, it i s easily demonstrated that
/

'
,
(20 X 10')/(104)

(20 K 10')/(104)
_
R
= ' (106) (0.01)
_
_ =q
'
:) _ (/ ')(106)(0. 12)
(7. 29)
(7.30)
(7. 31)
(7. 32)
In addition to the running costs, the capital cost for the reactor and
catalyst also had to be estimated. Here it was assumed that two
reactor units would be required and were to be operated in tandem
with one uni t on stream while the catalyst in the other is being re
generated. The capital cost for these units was estimated by fitting
an analytical expreSSion to the cost data for stainl ess steel, shell
and- tube reactors given by Chilton J) . A correction for the ENR
price index was applied to these data to update the costs to ENR
index of 980; the result of these operations was
L_ __ 7 2340Ao
.
57 1 (7.33)
where
_
_ _
P
i s the installed cost of two identical reactors with A sq ft
of surface area. The capital required for purchasing the catalyst
was also estimated by assuming that silica alumina catalyst of the
desired type could be purchased for $3. OO/lb.
1 83 Clapter 7
Second Basis-Optimization of the Reactor - Fractionator System.
I n the second cost -estimating procedure the same method of estimat
ing Cy was used. However, instead of charging for recycle on a fixed
price basi s, the distillation program previously presented was applied
to each reactor design case. In each case a di stillation column was
designed that would separate the ethylbenzene - styrene reaction mix
ture and produce a product stream containing 99 mole per cent sty
rene. The reactor design provided a feed stream of known flow rate
and composition. Ey specifying the bottoms stream flow rate and compo

sition as in the distillation program, the material balance for the

column is fixed. Then, as illustrated in the fractionator design,
column design calculations were carried out for a wide range of
reflux ratios. Thus, for each of 36 selected reactor cases, distilla
tion columns were designed for each of 60 different reflux ratios,
maing a total of 2 1 60 indi vidual reactor-
fractionator designs. For
each reactor case, the optimum fractionation reflux ratio was selec
ted. Thus, by adding the cost s f or steam, cooling water, and column
depreciation to the sum of Cy and the depreciation charge for the
reactor, the total running cost for each reactor -fractionator syst em
was computed, as was t he capital requirement for each process sys
tem.
The appendix contains a listing of the FORTRAN program that
corresponds to the block flowsheet shown in Fig. 7 - 5. Included in
this appendix i s a printout of the results obtained for a typical reac
tor design calculation. In addition, the design computations are dis
played for the fractionator required for the reactor being considered.
The fractionation program was written as a subroutine of the reactor
program, whi ch was called into effect each time the reactor program
had iterated to provide an outlet pressure within the prescribed
limi t s.
From results of t he type shown in t he appendix t he capital and
operating costs for each desired reactor design were computed.
These costs were estimated both on the first basis where the reactor
was considered independently, and on the second basis in which the
optimum fractionation uit was selected for each reactor.
Selection of the Reactor-Fractionator Desig Cases
Using the program in the appendix, it is possible to provide a
complete design for the dehydrogenation reactor. From the descrip
tion of the program, however, i t i s obvious that the large number of
design variabl es must be specified before completing the design and
that a complete consideration of all these variables would be imprac
tical. For example, if 15 different parameters were varied indepen
dently, the interpretation of the results would become quite complex,
even in a situation such as the present where a computer program is
available to carry out all the calculations. For this reason, three
design parameters whi ch were believed to be t he most critical in
determining the economi cs of the reactor design were selected for
a detailed investigation. These variabl es, together with the specified
values of these variabl es, are shown in Table 7 -6.
JIOCCSS L//lt0lOl//cr CljICn( JrOdtt Cl lOu 1oJ
Table 7-6. Desig Variables Ivestigted U the Present
Std
Variable
Tube l ength
Values Studied
1 0, 1 5, 20 ft
1 , 2, 4 in. Tube diam eter
Inlet temperature 842, 932, 1 022, 1 1 22F
By taing all combinations of the values specifi ed in thi s tabl e, 36
different cases were made available for the characterization of the
dehydrogenation reactor. In order to proceed with the solution,
values of the other design parameters had to be selected. Values for these
parameters which appeared to be reasonable were determined ad
are shown i n Table 7-7.
Table 11. Paraeters Fied for All Desig Calcuations
Catalyst bul density
Catalyst pellet diameter
Mass velocity of feed stream
to reactor
Outlet pressure of reaction
mixture
Flue-gas exit temperature
Over- all heat-transfer coefficient
Effectiveness factor
Flue- gas feed rate
61 lb/cu ft
1
/_ in.
4 800 lb/( hr) (sq ft)
5. 0 4 2. 5 psig
1 600F
8 Btu/(hr) ( sq ft) ("F)
0. 44
6520 lb/( hr) (tube) for the
4-in. tube; other flow rates
taen as proportional to
the square of the outer tube
diameter so as to maintain
the ratio ! _ approxi
mately constant.
With the specified values for the parameters shown in Table 7 -7,
the reactor design was completed [or all 36 cases summarized in
Table 7 - 6. After the calculations were completed for each case, the
costs of carrying out the reaction in each reactor was computed on
each of the two cost bases previously described.
Results and Discussion of te Economic Evluation
The results of the design calculations for each of the 3 6 reactor
cases are summarized in Table 7-8. Sufficient results are provded
to allow an economic evaluation to be completed on either the first
or the second basi s.
|

T
a
b
l
e

T
-

d
.

R
e
s
u
l
t
s

o
f

U
e

D
e
s
i
g
n

C
a
l
c
u
l
a
t
i
o
n
s

f
o
r

R
e
a
c
t
o
r
s

W
i
t
h

1
-
m
.

T
b

D
i
a
e
t
e
r

<

O
p
e
r
a
t
i
n
g

C
o
s
t

o
n

~

F
i
r
s
t

B
a
s
i
s

-

F
l
u
e
-
G
a
s

F
r
a
c
t
i
o
n

R
e
a
c
t
o
r

I
n
l
e
t

E
x
i
t

I
n
l
e
t

I
n
l
e
t

E
x
i
t

U
n
c
o
n
v
e
r
t
e
d

F
r
a
c
t
i
o
n

N
u
m
b
e
r

o
f

R
e
c
y
c
l
e

Y
i
e
l
d

C
o
s
t
,

`

T
e
m
p
.

T
e
m
p
.

T
e
m
p
.

P
r
e
s
s
.

P
r
e
s
s
.

E
t
h
y
l
-
S
t
y
r
e
n
e

T
u
b
e
s

i
n

C
o
s
t
,

E
q
.

C
o
s
t
,

E
q
.

E
q
.

7
.
3
3
,

f
t

Q

p
s
i
g

p
s
i
g

b
e
n
z
e
n
e

Y
i
e
l
d

R
e
a
c
t
o
r

7
.
3
1
,

$
/
y
r

7
.
3
2

$
/
y
r

D
o
l
l
a
r
s

1
0

8
4
2

1
3
2
5

1
6
8
2

4
0
.
7
9
6

3
.
9
1
3

0
.
6
0
7

0
.
8
6
3

2
7
4

3
8
3
1
9
7

4
0
8
4
9
4

1
1
6
0
8
9

1
0

9
3
2

1
3
3
5

1
6
7
6

4
2
.
2
8
7

4
.
9
3
6

0
.
5
5
9

0
.
8
4
6

2
4
9

3
2
0
9
9
4

4
6
7
7
0
4

1
0
9
9
6
3

1
0

0
2
2

1
3
4
6

1
6
7
1

4
3
.
7
3
8

5
.
5
9
0

0
.
5
0
8

0
.
8
2
7

2
2
8

2
6
7
0
3
5

5
3
8
3
9
3

1
0
4
6
1
6

1
0

1
1
1
2

1
3
6
0

1
6
6
5

4
5
.
1
5
3

5
.
8
5
5

0
.
4
5
2

0
.
8
0
4

2
1
0

2
1
9
6
5
3

6
2
4
2
4
3

9
9
9
7
7

1
5

8
4
2

1
4
3
0

1
7
1
1

5
7
.
9
7
3

6
.
7
0
1

0
.
2
6
2

0
.
7
3
2

1
7
2

1
0
4
0
5
0

9
3
9
6
4
2

1
1
2
1
9
3

1
5

9
3
2

1
4
4
0

1
7
0
3

5
8
.
8
4
5

5
.
4
0
5

0
.
2
3
0

0
.
7
1
9

1
6
8

8
9
3
5
9

1
0
0
3
6
2
1

1
1
0
6
7
6

1
5

1
0
2
2

1
4
5
2

1
6
9
6

6
0
.
6
6
4

7
.
0
2
5

0
.
1
8
9

0
.
6
9
9

1
6
4

7
1
4
9
2

1
1
0
8
1
8
7

1
0
9
1
7
6

1
5

1
1
1
2

1
4
6
6

1
6
8
8

6
1
.
4
3
6

4
.
5
3
3

0
.
1
5
7

0
.
6
8
3

1
6
1

5
8
1
8
4

1
1
9
1
1
0
8

1
0
8
1
1
9

2
0

8
4
2

1
5
6
0

1
7
3
0

7
3
.
6
9
4

6
.
5
2
5

0
.
0
3
2

0
.
6
1
3

1
5
6

1
1
6
9
9

1
6
2
3
8
0
3

1
2
5
3
3
4

2
0

9
3
2

1
5
6
7

1
7
2
0

7
4
.
8
8
4

6
.
5
6
5

0
.
0
2
4

0
.
6
0
2

1
5
8

8
6
7
2

1
6
9
7
0
2
5

1
2
5
9
4
1

2
0

1
0
2
2

1
5
7
6

1
7
1
1

7
6
.
0
1
2

5
.
9
8
9

0
.
0
1
7

0
.
5
9
1

1
6
0

6
1
8
4

1
7
7
4
9
4
5

1
2
6
7
2
9

2
0

1
1
1
2

1
5
8
5

1
7
0
1

7
7
.
0
8
1

4
.
7
0
3

0
.
0
1
1

0
.
5
8
0

1
6
2

4
2
4
9

1
8
5
9
7
6
6

1
2
7
7
3
3

T
a
b
l
e

7
-
8
a

c
o
n
t
i
u
e
d

O
t
i
m
u
m

C
o
l
u
m
n

C
o
r
r
e
s
p
o
n
d
i
n
g

t
o

E
a
c
h

R
e
a
c
t
o
r

C
a
t
a
l
y
s
t

C
o
l
u
m
n

^
z
a

O
/
/
)

D
e
p
r
e
c
i
a
t
i
o
n

H
e
a
t
i
n
g

C
o
o
l
i
n
g

T
o
t
a
l

O
e
r
a
t
i
n
g

L

C
o
s
t
,

F
e
e
d

R
a
t
e
,

S
s

i
n

M
i
n
i
m
u
m

C
o
s
t
,

C
o
s
t
,

C
o
s
t
,

C
o
s
t

f
o
r

C
o
l
u
m
n
,

i
t

D
o
l
l
a
r
s

l
b

m
o
i
e
s
/
y
r

i
n

F
e
e
d

D
i
s
t
i
I
I
a
t
e

R
e
f
l
u
x

'
'
/
/
/
-
,
-

T
r
a
y
s

$
/
y
r

$
/
y
I

$
/
y
r

/
x
r

1
0

3
0
0
9

5
6
3
4
2
9

0
.
6
4
2

0
.
9
7
4

3
.
8
3
4

.
3
0

6
6

3
8
0
6
5

6
2
8
4
6

3
3
3
5

1
0
4
2
4
7

1
0

2
7
3
6

5
0
4
7
4
8

0
.
6
0
0

0
.
9
6
9

4
.
0
7
9

1
.

3
0

3
5
3
9
5

9
4
0
2
3

6
6

5
5
6
7
2

2
9
5
4

1
0

2
5
0
7

4
5
3
8
4
:

0
.
5
5
5

0
.
9
6
3

4
.
3
8
0

1
.

3
0

6
5

3
2
4
6
3

4
9
4
4
1

2
6
2
4

8
4
5
2
9

v

7
6
5
4
6

P

1
0

2
3
1
5

4
0
9
1
4
3

0
.
5
0
6

0
.
9
5
5

4
.

7
6
2

1
.

3
0

6
5

3
0
2
5
3

4
3
9
5
9

2
3
3
3

3
0
0
0
8
3

0
.
3
2
7

0
.
9
0
9

7
.
0
1
2

1
.

2
5

2
3
7
6
5

2
9
4
0
1

1
5
6
0

5
4
7
2
7

r

1
5

2
8
3
4

6
5

1
5

2
7
6
7

2
8
6
2
2
4

0
.
2
9
5

0
.
8
9
5

7
.
6
6
8

1
.

2
5

6
3

2
2
2
2
4

2
7
6
9
8

1
4
7
0

5
1
3
9
3

~
.

1
5

2
7
0
2

2
6
9
3
6
8

0
.
2
5
0

0
.
8
7
2

8
.

7
8
0

1
.

2
0

6
6

2
1
7
1
2

2
4
6
5
4

1
3
0
8

4
7
6
7
5

6
1
5

2
6
5
6

2
5
6
8
1
4

0
.
2
1
4

0
.
8
4
6

9
.
9
7
9

1
.

2
0

6
5

2
0
5
5
9

2
3
0
9
1

1
2
2
5

4
4
8
7
6

2
0

3
4
4
3

2
1
2
9
6
0

0
.
0
5
2

0
.
4
8
6

2
3
.
3
4
2

1
.

5
0

3
3

9
3
3
4

9
5
8
4

5
0
8

1
9
4
2
7

~

2
0

3
4
7
2

2
1
0
1
0
4

0
.
0
3
9

0
.
3
9
3

2
4
.
9
7
7

1
.

3
5

3
3

8
2
6
7

7
8
2
9

4
1
5

1
6
5
1
3

2
0

3
5
1
0

2
0
7
7
5
7

0
.
0
2
8

0
.
2
8
8

2
5
.
0
1
8

1
.

2
0

3
3

7
0
1
9

5
9
6
0

3
1
6

1
3
2
9
6

C

2
0

3
5
5
9

2
0
5
9
3
2

0
.
0
1
9

0
.
1
7
7

2
1
.

4
4
9

1
.

2
0

4
6
3
7

9
3
1
6

2
6

4
4
4
3

2
3
6

<

~
.

r

T
a
b
l
e

7
-
8
b
.

R
e
s
u
l
t
s

o
f

t
e

D
e
s
i
g

C
a
l
c
u
l
a
t
i
o
n
s

f
o
r

R
e
a
c
t
o
r
s

W
i
t
h

2
-
i
.

T
b

D
i
a
m
e
t
e
r

~

O
e
r
a
t
i
n
g

C
o
s
t

o
n

F
i
r
s
t

B

s
i
s

`

F
l
u
e
-
G
a
s

I
r
a
c
t

O
n

H
e
a
c
I
O
I

I
n
!
e
I

E
X

I
n
l
e
t

I
n
l
e
t

E
X
i
t

U
n
c
O
n
v
e
r
I
e
d

F
r
a
c
t
i
o
n

N
u
m
b
e
r

O
!

R
e
c
y
c
l
e

Y
i
e
l
d

C
o
s
t
,

L

T
e
m
g
.

T
e
m
p
.

T
e
m
g
.

P
r
e
s
s
.

P
r
e
s
s
.

E
t
h
y
I
-
S
I
y
r
e
n
e

T
u
b
e
s

C
o
s
t
,

E
q
.

C
o
s
t
,

E
g
,

E
q
.

7
.
3
3
.

f
t

O
F

O
F

-
I

p
s
i
g

g
s
I
g

b
c
n
z
c
n
c

e
!
d

R
e
a
c
t
o
r

7
.
3
1
,

$
/
y
r

7
.
3
2
,

$
/
y
r

D
o
l
l
a
r
s

1
0

8
4
2

1
1
4
8

1
6
5
6

3
6
.
9
7
6

4
.
2
7
8

0
.
8
4
0

9
5
.
6
1
8

1
5
6

1
1
7
5
5
4
3

1
1
7
7
0
2

1
1
8
1
3
7

1
0

9
3
2

1
1
7
1

1
6
5
1

3
8
.
2
1
3

4
.
0
7
0

0
.

7
9
4

9
3
.
9
7
6

1
2
3

8
7
6
0
9
8

1
6
4
6
5
0

1
0
3
1
9
6

1
0

I
D
2
2

1
1
9
4

1
6
4
7

3
9
.
5
4
4

3
.
8
4
9

0
.
7
4
0

9
1
.

8
4
1

1
0
0

6
6
4
6
3
0

2
2
8
1
7
9

9
1
7
2
3

1
0

1
1
1
2

1
2
1
6

1
6
4
3

4
0
.
9
9
3

3
.
6
3
7

0
.
6
7
9
5

8
9
.
0
9
3

8
4

5
0
9
3
3
8

3
1
4
4
4
9

8
2
7
6
5

1
5

8
4
2

1
2
2
0

1
6
7
9

4
9
.
9
7
3

4
.
6
4
4

0
.
6
8
0

8
9
.
3
1
6

8
4

5
1
0
0
5
7

3
0
7
5
2
0

1
0
4
2
9
4

1
5

9
3
2

1
2
3
6

1
6
7
3

5
1
.

8
4
5

5
.
5
9
4

0
.
6
2
7

8
6
.
6
7
7

7
4

4
1
4
9
4
8

3
9
4
8
0
7

9
7
1
4
5

1
5

1
0
2
2

1
2
5
4

1
6
6
8

5
3
.
6
6
4

6
.
0
2
9

0
.
5
6
7

8
3
.
6
5
6

6
6

3
3
5
6
0
1

5
0
1
7
9
1

9
1
1
1
0

1
5

1
1
1
2

1
2
7
4

1
6
6
3

5
5
.
4
3
6

5
.
8
8
2

0
.

5
0
1

8
0
.
2
9
3

6
0

2
6
7
9
4
6

6
3
0
4
1
3

8
6
0
0
3

2
0

8
4
2

1
2
8
2

1
6
9
9

6
2
.
6
9
4

6
.
4
1
1

0
.
4
9
4

8
0
.
2
4
6

5
9

2
6
0
1
8
9

6
3
2
2
6
8

1
0
0
5
0
4

2
0

9
3
2

1
2
9
8

1
6
9
3

6
3
.
8
8
4

4
.
2
6
2

0
.
4
4
4

7
7
.

8
8
6

5
5

2
1
9
0
8
7

7
2
9
2
5
2

9
6
8
6
4

2
0

1
0
2
2

1
3
1
5

1
6
8
6

6
6
.
0
1
2

5
.
2
3
1

0
.
3
8
2

7
4
.
6
1
9

5
2

1
7
6
6
9
7

8
7
3
6
3
8

9
3
4
2
4

2
0

1
1
1
2

1
3
3
4

1
6
8
0

6
8
.
0
8
1

5
.
3
3
9

0
.
3
1
5

7
1
.
2
4
4

4
9

1
3
8
2
0
9

1
0
3
6
7
1
4

9
0
5
3
0

T
a
b
I
O

7
-
8
b

c
O
n
t
i
n
u
O
d

C
a
t
a
l
y
s
t

I

C
o
s
t
,

U

D
l
l
a
s

1
0

6
6
7
2

1
0

5
2
6
3

1
0

4
2
8
0

1
0

3
5
7
4

1
5

5
3
6
1

1
5

4
7
3
3

1
5

4
2
3
0

1
5

3
8
2
3

2
0

5
0
2
4

2
0

4
7
1
0

2
0

4
4
2
0

2
0

4
1
8
3

C
o
l
u
m

A

F
e
e
d

R
t

A

i

l
b

m
o
l
e
a
/
y
r

i

F
e
e
d

D
i
s
i
l
l
t
e

1
3
1
0
9
2
5

0
.
8
4
6

0
.
9
9
1

1
0
2
8
4
3
0

0
.
8
0

0
.
9
8
8

8
2
8
9
3
3

0
.
7
5
6

0
.
9
8
5

6
8
2
4
3
0

0
.
7
0

0
.
9
8
0

6
8
3
1
0
9

0
.
7
0

0
.
9
8
0

5
9
3
3
8
3

0
.
6
6
0

0
.
9
7
6

5
1
8
5
2
7

0
.
6
1
1

0
.
9
7
0

4
5
4
7
0

0
.
5
5
6

0
.
9
6

4
4
7
3
8
5

0
.
5
4
9

0
.
9
6
2

4
0
8
6
0
9

0
.
5
0
6

0
.
9
5
5

3
6
8
6
1
9

0
.
4
5
2

0
.
9
4
5

3
3
2
3
0
9

0
.
3
9
2

0
.
9
3
0

M
i
i
m
u
m

(
P
/

R
e
f
l
u

(
P
/

I
a
;
a

2
.
9
6
0

1
.

4
5

3
.
1
0
7

1
.
4
0

3
.
2
8
9

1
.
4
0

3
.
5
1
6

1
.
3
5

3
.
5
1
5

1
.
3
5

3
.
7
3
6

1
.
3
5

4
.
0
1
3

1
.
3
0

4
.
3
7
4

1
.
3
0

4
.
4
2
7

1
.
3
0

4
.
7
6
7

1
.
3
0

5
.
2
7
4

1
.
2
5

5
.
9
8
4

1
.
2
5

O
i
m
u
m

C
o
l
u
m

C
o
r
r
e
s
p
n
i

m

E
a
c
h

R
e
a
c
t

r

D
p
r
e
c
i
t
i
n

H
e
a
t
i
g

C
o
l
i
g

T
o
t

O
e
r
a
t
i
g

C
o
s
t
,

C
o
s
t
,

C
o
s
t
,

C
o

f
o
r

C
o
l
u
m
,

^

T
a
y
s

$
/
y
r

$
/
y
r

$
/
y
r

$
/
)

6
6

6
8
7
0
1

1
6
8
1
3
7

8
9
2
4

2
4
5
7
6
2

v

6
6

5
6
0
4
1
2
6
9
5
8

6
7
8

1
9
1
7
4
2

6
5

4
9
8
9
1

1
0
1
1
9
2

5
3
7
0

1
5
8
5
4

<

6
6

4
3
9
3
7

7
9
8
2
0

4
2
6

1
2
7
9
9
3

-

6
6

4
3
9
6
5

7
9
9
0
5

4
2
4
1

1
2
8
1
1
1

6
5

3
9
5
2
1

6
8
6
2
6

3
8
4
2

1
1
1
7
9
0

6

6
6

3
6
0
3
4

5
7
3
5
8

3
0
4
4

9
8
4
3
7

6
5

3
2
5
0

4
9
5
4
7

2
6
3
0

8
4
6
8
1

6
5

3
2
1
5
0

4
8
6
5
0

2
5
8
2

8
3
3
8
3

6
5

3
0
2
2
6

4
3
8
9
4

2
3
3
0

7
8
5
0

6
6

2
7
9
9
8

3
7
6
6
6

1
9
9
9

6
7
6
6
4

e

6
6

2
6
0
7

3
3
3
0
8

1
7
6
8

6
1
0
8
3

T
a
o
l
e

&
.

R
e
a
u
l
t
a

o
f

m
e

u
a
Q

C
a
l
c
u
u
t
i
o
o
a

f
o
r

R
e
a
c
m
r
a

w
i
t
b

1
-
m
.

T
b

o
a
m
e
t
e
r

<

O
e
r
a
t
i
n
g

C
o
s
t

o
n

"

F
i
r
s
t

B
a
s
i
s

`

F
l
u
e
-
G
a
s

F
r
a
c
t
i
o
n

R
e
a
c
t
o
r

I
n
l
e
t

E
x
i
t

I
n
l
e
t

I
n
e
t

E
x
i
t

U
n
c
o
n
v
e
r
t
e
d

F
r
a
c
t
i
o
n

N
u
m
b
e
r

o
f

R
e
c
y
c
l
e

Y
i
e
l
d

C
o
s
t

L

T
e
m
p
.

T
e
m
p
.

T
e
m
p
.

P
r
e
s
s
.

P
r
e
s
s
.

E
t
h
y
l
-
S
t
y
r
e
n
e

T
b
s

m

C
o
s
t
,

E
q
.

C
o
s
t
,

E
q
.

E
q
.

T
.

J
J

f
t
.

O

p
s
i
g

p
s
i
g

b
n
z
e
n
e

Y
i
e
l
d

R
e
a
c
t
o
r

T
.
J

$
/
y
r

T
.
J
Z
,

$
/
y
r

D
l
l
a
r
s

5
1
Z

J
J

J
b
.

5
J

1
.
b
J
T

.
9
Z

9
5
.
Z

Z
b
Z
J
T
1
1

1
b
1
Z

9
J
Z

J
b
.
1
9
Z

1
.
Z
5
5

.
5
T
9

.
9
J
J

b
1

J
9
b
T

Z
b

9
Z
5

Z
Z

9
Z

Z
5

J
T
.
5
T
5

J
.
9
T
J

.
5
J

9
1
.
9
1
Z

J
9

9
9
b

5
1

T
T

5
9

Z
b

Z
b

J
9
.
J

J
.
b

.
T
T

9
Z
.

Z
Z

T
5

J
b

9
b

5
1
Z

T
Z

1
T

b
.
1

.
5
1
9

9
b
.
b
9
1

1
1

Z
9
9

5
J
T
b

J
T
J

9
J
Z

1
1

1
5
.
Z
1
b

b
.

.
5

9
J
.
Z
J

J
1

9
Z
5
1
J
Z

5
b
J
Z
J

5
9
9
b
1

Z
Z

1
9
.
9
9
J

1
.

T
1

Z
5

b
9

Z
5
J

9
Z

T
9
b
5
T

J
T

1
.
b
Z

T
9

5
b
.

Z
J

b
Z
b

b
J

b
J
J

5
1
Z

\
b

T
b
Z

b
.
J
T

.
T
T

.
J

T
5
9
T
T
T

Z
1
1

5
5

9
5
5
Z

9
J
Z

1
b

b
7

b
5
.
b
9

1
.

T
9
1

.
T

5
T
.
9
T
T

Z
b

T
T
Z
b

J
b

5
9
b

Z
Z

b
Z

Z
1

1
.
1
T
Z

b
J

5
J
.
9

Z
Z

1
5

5
1

1
9
Z
b
9

5
b

9
b

1
5

Z
.
9
J

1
.
b
T

.
b
T
T

T
9
.
Z

Z
9

b
T
1

J
Z

T
b
b
J

T
~
&

c
o
t
i
u
e
d

C
a
t
a
l
y
s
t

I

C
o
s
t
,

f
t

D
l
l
a
r
s

1
0

1
3
0
'
1

1
0

8
7
4
7

1
0

6
3
1
7

1
0

4
8
3
5

1
5

1
0
6
3
4

1
5

8
2
8
1

1
5

6
6
8
0

1
5

5
5
4
0

2
0

9
7
6
6

2
0

8
2
1
0

2
0

7
0
1
7

2
0

6
0
9
5

C
o
l
u
m

F
e
e
d

R
t
e
,

A

I
b

m
o
l
e
s
/
y
r

i

F
e
e
d

D
i
s
t
i
l
l
a
t
e

2
5
8
2
8
1
3

0
.
9
2
2

0
.
9
9
6

1
7
2
4
5
2
4

0
.
8
8
3

0
.
9
9
4

1
2
3
9
2
0
3

0
.
8
3
7

0
.
9
9
1

9
3
9
5
2
6

0
.
7
8
5

0
.
9
8
7

1
3
9
3
4
2
4

0
.
8
5
5

0
.
9
9
2

1
0
7
7
8
0
2

0
.
8
1
3

0
.
9
8
9

8
5
8
9
9
9

0
.
7
6
5

0
.
9
8
6

6
9
8
1
7
8

0
.
7
1
1

0
.
9
8
1

9
4
6
9
9
6

0
.
7
8
7

0
.
9
8
7

7
8
4
6
8
3

0
.
7
4
3

0
.
9
8
4

6
5
5
5
5
1

0
.
6
9
2

0
.
9
7
9

5
5
0
6
3

0
.
6
3
3

0
.
9
7
3

M
i
i
m
u
m

(
P
/

R
e
f
l
u

(
P
/

l
a
t
a

2
.
'
2
9

1
.
5
5
"

2
.
8
4
3

1
.
5
0

2
.
9
9
0

1
.

4
5

3
.
1
7
6

1
.
4
0

2
.
9
3
1

1
.
4
5

3
.
0
7
5

1
.
4
0

3
.
2
5
4

1
.
4
0

3
.
4
8
6

1
.
3
5

3
.

1
6
9

1
.
4
0

3
.
3
4
6

1
.
3
5

3
.
5
7
4

1
.
3
5

3
.
8
8
2

1
.

3
0

O
t
i
m
u
m

C
o
l
u
m
n

C
o
r
r
e
s
p
n
d
i
n
g

t
o

E
a
c
h

R
e
a
c
t
o
r

D
p
r
e
c
l
t
l
n

H
e
a
t
i
g

C
o
l
i
g

T
o
t
l

O
r
a
t
i
n
g

C
o
s
t
,

C
o
s
t
,

C
o
s
t
,

C
o
s
t

f
o
r

C
o
l
u
m
,

:

T
r
a
y
s

$
/
y
r

$
/
y
r

$
/
y
r

$
/
y
r

C

6
6

1
0
'
6
5
4

3
5
5
4
4
'

1
8
8
6
5

4
8
1
9
6
'

v

6
6

8
2
7
3
8

2
2
9
2
0
9

1
2
1
6
5

3
2
4
1
1
2

6
6

6
6
3
4
0

1
5
8
6
1
9

8
4
1
9

2
3
3
3
7
9

6
6

5
4
8
3
6

1
1
5
4
7
7

6
1
2
9

1
7
6
4
2

6
6

7
1
3
5
1

1
7
9
0
8
5

9
5
0
5

2
5
9
9
4
2

6
6

5
9
7
7
7

1
3
3
3
3
4

7
0
'
7

2
0
0
1
8
8

S

6
5

5
1
0
3
1

1
0
5
0
7
6

5
5
7
7

1
6
1
6
8
4

6
6

4
4
5
8
7

8
1
7
9
9

4
3
4
2

1
3
0
7
2
8

`
. -

6
6

5
5
1
1
0

1
1
6
4
4
2

6
1
8
0

1
7
7
7
3
2

6
6

4
8
0
5
3

9
2
6
6
6

4
9
1
8

1
4
5
6
3
7

3

6
6

4
2
8
1
2

7
6
4
4
2

4
0
5
7

1
2
3
3
1
1

C

6
6

3
7
4
6
9

6
1
2
1
3

3
2
4
9

1
0
1
9
3
1

~

1 90 Clapter 7
To illustrate the effect of the three critical variables upon the
maj or costs associated with the reactor, one typical set of cases was
selected for presentation. In this example, a base case of a 1 5-ft ^
2 in. reactor tube with a feed-gas temperature of l 022F was selec
ted; subsequent cases were e stablished by varying the design para
meters both above and below the values used in the base case . Cost s
were computed on the fir st basis where independent consideration i s
given t o the reactor; the depreciation charges were found t o be negli
gible compared to the yield and recycle costs and were not consider
ed i n the optimization. The results of these calculations are shown in
Fig. 7-6, 7-7, and 7-8. U addition, a plot of ethylbenzene conversion
ve rsus styrene yield for all 36 design cases i s presented in Fig. 7-9.
)O
|pOO

000

600
8
400
6

C
200
O
O

h
\

\
%

% 1eto| epetettog :e._

\
X
^
^

^ N
- w *
\

\
e|d cest y
/

\
X
e

'/

^
c)cM cest p

Z 4
NOm| nC| tuD di umtt
t
i nCh5
Figure 7 - 6. Reactor operating costs as a
function of tube diameter for a 1 5 -ft tube
with a feed temperature of l 022F.
In Fig. 7-6 the effect of tube diameter upon operating cost is illus
trated. Wile the cost of yield loss decreases with increasing tube
diameter, the effect of diameter upon recycle cost i s just the opposite.
The explanation of these effects lie s in the fact that for a given tube
length and inlet temperature the heat supplied per unit volume of
reacting gas decreases with increasing tube diameter. The reaction
t emperature i s therefore lower at larger tube diameters. Since the
conversion increases with increasing temperature, it can be conclud
ed that as the tube diameter increases, the conversion is lower and the
Prcccss E//i|alictt}cr CIIun0 PrcJitc/ici| 1 !
r ecycle cost is higher. The effect upon the cost of yield loss is the
opposite because, as illustrated in Fig. 7- 9, conversion and yield are
inversely related.
The effect of t he reactor tube length on the operating cost i s plot
ted in Fig. 7- 7, indicating that for a given tube diameter and inlet
temperature, ethylbenzene conversion increases with tube length. It
therefore follows directly that recycle costs decrease with increas
ing tube length . Si milarly, it i s obvious that yield loss due to side
reactions must inc rease as tube lengh increases .
C
a

|zuu
00
uu
6

_ 6uu
C
C

T
8
o
= 4uu

o
C
zuu
0

TOlol optrofln9 cost .
'
e
#
e
e
e

-
a
a
7
=

" ~ ~a =q=
/
/

Y' tl d COSt , C y /
` 1 /
\
/
Z

Z
co" , c.

o | l 4 6 lo zu zz
uD0 l 0ngI h , I00I
Figure 7- 7. Reactor operating costs
as a function of tube length for a
2 -in. -diam tube and feed tempera
ture of l 022F.
In Fig. 7 8
)
the costs are plotted as functions of inlet temperature.
Sice t he conversion increases with temperature, t he effect of in
creasing temperature is generally similar to that of increasing tube
length.
Figures 7 -6 to 7 -8 show that each of the selected design param

eters causes independent variations in the recycle and yield costs,

so that the total operating cost passes through a minimum. However,
in selecting an over -all optimum design, all combinations of the de
sign parameter s must be taen into account, not j ust independent
variations in the parameters concerning a selected base case. These
conSiderations have been made using t he data from Table 7-8; the
J
2 Chapter 7
.p
_0uu

Oa
.
I
1cro| oe onog ::._ _
O
=

~ ^"
.. __

uu

c
C
C
.4uu
:

-

8
C
C

zuu
"

8ec,c| e covtC
q
~
=
'
.

0
uu

OO | uO
j
| | uu
|ee0 | empr0I ur8 "|
Figre 7- 8. Reactor operating cost s as a
function of feed temperature for a
1 5 -ft ^ 2 - in. tube.

|Ou~
0~
u

8
zc~

==

u
=

u
=

O
j
5| yr8n8 y| e' 0 ,per cent
Figre 7-9. Ethylbenzene conver
sion as a function of styrene
yield.
Process Evaluation for Styrene Production J VJ
result s are plotted in Fig. 7 - 1 0, which shows the total operating costs
for each diameter-length combination versus inlet temperature . The
total operating costs were computed by adding the yield-loss cost
and the recycle cost, the latter computed on the fir st baSiS, an assum
ed cost of $ 0. 01/lb of ethylbenzene recycled. On the first baSis for
e stimation, the optimum reactor is one having 249 tubes, each ten feet
by one inch, and is operated with an inlet feed temperature of approxi
mately 900F. The effect of inlet temperature on operating cost is
quite small; this observation should have important ramiications in
the control of the reactor.
C

s
<

,oo
2, zuu
_| , 0uu
c
C
S
c
' , 4uu
C
C

C
u
|00d I0mp0rOI ur0 , " |
Figure 7 - 1 0. Total reactor operating costs
for 3 6-tube geometries.
To illustrate the effects of the fractionator design on the over- all
proce ss de sign, Fig. 7- 1 1 was prepared. I is analogous to Fig. 7 - 1 0
in all respects except that here the second economic basis was used;
i . e. , the optimum reflu ratio was selected for the reactor-fractiona
tor combination, and the total running cost for the combination was
computed in each case . Figure 7 - 1 1 shows that the combination of
1 0-ft by 2 - in. tubes with a feed temperature of approximately 9aaoF
provdes the opt imum process desi gn. For this particular reactor,
73 Chapter 7
0` r I
*
| ,800P-

U
2

::
L

c
L

. :::
~o

' ::: -
~
:=

:
--- ----

:
-

|000 ! 0J0| 0 I ur 0 " |

Figure 7 - 1 1 . Total operat ing cost s computed
by using the optimum fractionation unit
for each dehydrogenation reactor.
two 33 -tray distillation columns are required which ar e to be opera
ted at a reflux ratio of 4. 3 5, or 1 . 4 times the minimum ratio . Table
7-8 reveals that neither that particular reactor nor that particular
distillation unit is an individual optimum, whereas the combination of
both gives the optimum over-all process unit. Figure 7- 1 1 also shows
that the effect of inlet feed temperature on the total conversion cost
is quite minimal and in fact can be neglected over the temperature
range of 8420 to 1 022F. This characteristic will allow extremely
flexible control and operation. By maintaiing the appropriate reflux
ratio in the distillation unit so as to ensure proper product quality,
the over- all cost of the conversion of ethylbenzene to styrene will
not be strongly influenced by the operating conditions in the reactor.
Some interesting conclusions about the economics of the separation
process can be drawn by comparing Fig. 7 - 1 1 with Fig. 7- 1 0. Not only
does the consideration of the distillation costs lead to the selection of
a different reactor, but also the total conversion cost i s significantly
lower than that computed by assuming a flat rate of $ O. Ol/lb for
separating the styrene. This point s up the danger of using approxi
mate rules of thumb in estimating conversion costs.
Process E|/ttalion/or Styrene 1roJuc//on
J95
Economic Evluation of the Proj ect
Table 7-9 provides a rough e stimate of the capital requirement s
for this proj ect as based upon the cost e stimates previously present
ed. I n preparing this tabulation, the preli minary estimation method
recommended by Happel (1) was used.
Table 7-9. Equipment Specifications and Cost
Total
Item Number Unit Cost
Istalled
Cost
Reactor, shell ad tb 2 $ 51, 598 $1 03, 196
160 tbes; 10 ft K 2 i.
stinless steel
Catalyst 1 754 lb
5, 262
Fractiontion colum 2 145, 1 1 2 290, 224
33-tray vcuum colum
Colum diameter-S. 75 D
Rebiler and condenser 100, 000
Pmps 20, 000
Istruments (10 per cent of 51, 870
abve)
$ 570, 552
Item Material Lbor
Isulation $47, 000 $ 70, 500
Piping 23, 000 23, 000
Foudations 23, 000 34, 500
Structres 19, 000 3, SOO
Fireprofing 4, 700 28, 200
Electrical 28, 000 42, 000
Paitig and cleaup 4, 700 23, 500
Subtotals 149, 400 225, 500 $374, 900
Estimated costs of equipment ad material, istlled 945, 452
(A)
Oerhead, 30 per cent of A 283, 636
Total erected cost
, 229, 08S
Engineerig Fee, 13 per cent of A 122, 909
Contigencies, 13 per cent of A 1 22, 909
Total capital for process uit $ 1, 474, 906
Mnuacturig bildigs $, 000
1 V0
C/apt er 7
The optimum process, as selected from Fig. 7 - 1 1 , would produce
the required amount of styrene at an annual operating cost of about
$ 3 60, 000, which corre sponds to $ O. 01 8/lb of product. This figure
does not include depreciation of the reactor and catalyst; if these items
are computed on a basis of 7 per cent of the capital estimate in
Table 7-9, an additional $ 1 00, 000 per year must be included. Thus, a
total cost of $0. 023/lb is computed for the conversion of ethylbenzene
to styrene . Because of the proprietary nature of the information, no
direct comparison with actual commercial costs i s possible; however,
this figure does appear reasonable for a continuous proce ss of this
type.
The capital e stimate summarized i n Table 7-9 is presented as an
example of the detailed itemization that must be considered in evaluat
ing critically the economics of the proposed proj ect . As subsequent
steps in the evaluation, the potential market volume s and estimated
selling prices must be forecast for styrene as well as for the import
ant raw materials and by-products. With such information, a sal es
proj ection can be prepared and a return on investment estimated. A
great deal of additional work would be required to complete this task,
but these considerations are beyond the scope of the present evalua
tion. The reader is referred to the works of Happel [4) and of Vil
brandt and Dryden [JJ) for excellent discussions of the estimating
techniques that must be used. One must always be alert to the fact
that the accuracy of the available cost and market inormation i s usual
l y greatly inferior t o t he reliability of t he process data. Despite the
detailed process consideration pre sented here, the viability of the pro
j ect would hinge upon the availability of reliable market and cost in

formation. Even with the development of modern technology, including

electronic computers, the most important considerations in evaluating
many proj ects must be derived from the semiquantitative estimations
of exerienced engineers and busiessmen.
General Comments on the Desig
It is felt that the design calculations j ust completed provide a good
example of the power of digital computation in the completion of com
plex de sign problems that involve many design var iables. In the ab
sence of computer facilities the amount of computati on work evi
denced in the present design could have been accomplished only by the
investment of a great many technical man-hour s . The use of the ma
chine to ease the computation burden allows the engineer to apply his
talent s in the most meaningul way, i. e . , in the interpretation of the
result s.
The emphasis i n t he pre sent discussion has been upon the technical
merits of the process and on the invention and analysiS of a proce ss
scheme. As mentioned previously, the market s for the styrene. end
product must be given at least the same degree of attention before
capital funds are requested. Actually, an economic evaluation of the
process as presented here would not in itself have a great deal of
meaning, since the economics would undoubtedly dictate that the equip
ment considered should be a part of an over -all petrochemical com-
Process Evaluation for Styrene Production 1V
plex, which would utilize a raw petroleum feed stock and would pro
duce, among other things, a final polystyrene product .
It is helpful to compare t he result s of t he present study with the
current industrial practices for styrene manufacture (2, &). The com
petitive economics of the petrochemical industry dictates that styrene
be made in very large equipment, and usually in a single train 5) .
The capital costs for reactor construction can be greatly reduced by
using a self-regenerative catalyst that requires steam to be added to
the ethylbenzene feed. This type of operation eliminates the need for
two parallel reactors; this can be very important in modern, large
volume installations. Thus it is in the selection of a catalyst that the
present design depart s most significantly from current industrial
practice. Also, in the modern, large - scale operations the values of the
by-product s from cracking become very significant economic factors.
U the present study i t was assumed that benzene , toluene, and hydro
gen were to be discarded because they were produced only in small
quantities. The value of these materials generated from a large - scale
proce ss cannot, however, be neglected. Furthermore, it has been point
ed out that the assumed value for ethylbenzene of $ O. 12/lb is high
[5) and that the effect of a reduction in this cost should be investiga
ted in subsequent studi es. The suggested change s in these factors
would tend to yield an optimum process that operates at higher ethyl
benzene conversions but lower styrene yields . This trend i s in agree
ment with the point made in Chapter 6 that the maj ority of industrial
ethylbenzene - styrene fractionator s operate with feed concentrations
in the range of 45 to 60 mole per cent ethylbenzene. These values are
to be contrasted with the 82 mole per cent feed interpolated from
Table 7-8 for the optimum system selected in the present study.
As a final i ssue, it has been pointed out that standard commercial
practice dictates the use of reactor pressures as low as pOSSible, with
values seldom exceeding 20 psig (5) . The late st designs appear to
have reduced this even further. This reduction in reactor pressure
is probably one of the advatages realized from the development of
more active catalysts for the conversion of ethylbenzene to styrene.
The optimum process selected in the present study requires pres
sures as high as 3 8 pSig.
Returning to the optimization calculations summarized in Fig.
7- 1 1 , it is clear that a number of critical assumptions have entered
into the selection of the " optimum" process unit. Fir st, the assumed
temperature and flow rate of the flue gas and the assumed over -all
heat -transfer coefficient were critical in obtaining the result s pre
sented here. By using the program given in the appendix, variations
in these parameters would be investigated readily. The catalyst pel
let diameter and effectivess factor used were important i n deter
mining the results previously pre sented. The effect of varying these
parameters should al so be investigated by further use of the pro
gram provided. For example, the economics of trying t o develop a
better catalyst could be studied; this type of investigation would
i llust rate the ki nd of work that must be completed by a di rector of
research in attempting to place priorities on the different re search
proj ects he must evaluate and manage.
1VO Process Evaluation for Styrene Production
In addition, it will be remembered that the number of reactor tubes
was established by requiring a reactor production 5 per cent greater
than the necessary 20 million lbs/year of styrene. This assumption
in essence fixes the composit in of the distillate stream in the frac
t ionation unit. Since it probably has a substantial effect upon the pro
cess economics, i t should be the subj ect of further investigation. No
consideration has been given to means for removing the by-products
from the reaction mixture. The separation of hydrogen, benzene, and
toluene should not provide any signiicant problems; however, some
thought must be directed toward this matter before the economics of
the process can be properly evaluated.
Investigation of some of the process variables described may well
lead to a result that is superior to the " optimum" case recommended
in the pre sent study. Nevertheless it is felt that within the limits of
the assumptions made, the result of the present study provide a reali s
tic, feasible, and economic process for the production of styr ene mono
mer.
REFERENCES
1 . Bi rd, R. B. , W. E . Stewart, and E . N. Li ghtfoot, Transport Pheno -
mena, John Wi ley Sons, Inc . , New York ( 1 960) .
-
2 . Carra, S. , and L. Forni, I . and E. C. , Process Design and Develop
ment 4, 3, 281 ( 1 965) .
3 . Chilton, C. H. , Cost Engineering in the Process Industries, McGraw
Hill Book Company, Inc . , New York 9T).
4 . Happel, J. , Chemical Process Economics, John Wiley & Sons, Inc. ,
New York ( 1 958) .
5. Kehde, H. , Styrene Technology Center, Midland Division Re search
and Development, Dow Chemical Company, Midland, Mich . , per
sonal communication, October, 1 967.
6. Ohlinger, H. , and S. Stadelmann, Cher. Ing. - Tech. 1, 361 ( 1 965) .
1. Perry, J. H. , Chemical Engineers' Handbook, McGraw-Hill Book
Company, Inc. , New York ( 1 950) .
8. Reid, R. C. , and T. K. Sherwood, Properties of Liquids and Gases,
1 st Edition, McGraw- Hill Book Company, Inc. , New York ( 1 958) .
9. Satterfield, C. N. , and T. K. Sherwood, The Role of Diffusion in
Catalysis, Addison-Wesley Publi shing Co. , Inc. , Reading, Mass.
(1 963) .
10. Smi th, J. M. , Chemical Engineeri ng Kinet i cs, McGraw-Hi ll Book
Company, Inc . , New York ( 1956) .
1 1 . Vilbrandt, F. C. , and C. E. Dryden, Chemical Engineerig Plant
Desi gn, McGraw-Hill Book Company, Inc. , New York ( 1 959) .
1 2. Wenner, R. R. , and E. C. Dybdal, Cher. Eng. Progr. 44, 27 5 ( 1 948) .

Appendix
L N Program O Sample Printout for Chapter D

C
C
C
C
C
0C1
?
IDC3
PP
PPP
DCf6
DDT
^P8
DCP9
01^
(nll
PP1?
D1
P"\*
001
P^J^
rOI7
0C1P
onl9
p^]n
P71
P"Z
0}
F"?*
^75
PD6
C077
f^
/^7
P"^
DSiG W MW 'IPELlNE
0 - SULfUR 'IPF OIAfTEI IN INHS
OS. QTEI OIAETEI OF STEAN 'I'E I . IN"ES
0# QTEI OIANETER OF INSULATION IN IN"ES
T"ICKNES OF INSUATIO IN IN"ES
, - NNIFI OF PUNPING STATIONS
$ PNZPeZ3
FONAT'I"I.'"SULFU 'l'fLINE DESIGN CALCULATIONS/I" II" f
3 Foa T"X.2HNIEI PN'IN STATIONS - ,F,.I.I" II
q FQI NAT.5X,""INSUATIO T"ICKNSS IN INHf5 - ,F'.l.lH II" II
5O PDNT'X.6'HANUAL COSTS AlE SHW IELOI AS FUNTIONS OF SULFU ,
II,E OIANETEI,IH fIH J
T0 FOMATI, I'H'I'E OIA,INH,TX.llHNGlNEIING,9X.9'I'E COST,'X,.H'
IUN' CnST X.IOHINSUATIO ..... IOHSTfAN COST IOHTOTAL COST'
80 FONATI7FI Z'
FN-7CIn.0
90 D 11" ,T
P fY loo.lno
I0 P1M Z
'1INT 50sP
'liNT "."T
"IUNT
PR1Nf T0
0 OO I-I. ZO
P1 fJ
05-0+2.6
DI.OZ.O.T
DIST-(.""
'l'E-I'.200.0_015T"10 __ Z '+1.10.0_0+30700.0
PUN'-7.700.0/0IST+Z".OOC.O/O"".8
GN31B+01 __ Z-0S __ ZJ
T.-IO"00O.(/IALOGI2.e.OIDI,+e.0.ALOGI0I/OSI'
TOTA-fNG+PI'E+'UN'+CINSU+STN
'RI . T eO.n.fNG."I'P,'UN',CINSU.STN.TOTAL
70 CDTlI
TO Y
P
END
Z
oo Appendix

1
T
I
1
P

P
I
P
H
I
'
F

D
E

T
G
N

C
A
L
C
U
L
A
T
I
O
'
S

'
U
M

F
R

n
F

D
U
M
P
I
N
G

S
T
A
T
l
O
'
S

I
N
5
U
I
A
T
I
O
.

T
H
1
I
K
N
E
5
5

I
'
C
H
f
S

A
N
N
U
A
L

c
n
S
T
S

S
H
O
W

B
E
l
O
W

A
S

P
I
P
F

D
I
A
(
\
N
C
H

f
'
r
I
N
E
E
R
I
N
G

1
.
"
'

7

"
C
I
.
r
Q

+
D
^
7
5
^
^
P

J
&
P
^

7
'
O
"
(
.
,
r

=
"
"

7
'
"
(
"
,
n
l

"

7
5
"

"
,
^
\

.
P
"

7
-
n
/
, .
I
'

7
,
D
0

7
"
1
(
'
,
'
'

A
.
n
n

7
"
"
"
'
+

C
N
'
.
N
'

I

=
"
'

^
^
P

^
^

,
"
"

7
1
C
[
C
.
(
n

1
'
.
"
"

7
1
"
'
''
.
'

I
3
+
"
l

7
5
D
C
,
C

J
A

P
'

J
5
^
n
^

^
^

,
P
^

n
"
c
c
.
c
"

l
^

'
'

7
5
P
^
"

1
7
.
"
"

7
,
n
(
n
.
N

!
"

7
^
D
0
+
P

l
R
,
^
^

7
5
P

D
,
^
D

?
^
_
o
o

7
6
^
O
^

^
P

"
+

F
U
N
C
T
I
O
N
S

O
F

S
U
L
f
U
R

P
I
D
F

C
O
5

1
'
"
5
(
!
.
0
l

4
8
6
1
9

7
2
1
3
6
.
4
4

1
B

R
9
.
.+
4
4

1
5
6
1
2
0
.
1
2

I

Z
T
T
R
+
3
1

I
T
4
B
P
&
4
4

1
1
9
7
1
4
a
1
5

?
D
6
O
9
4

5
6

U
H
I
2
.
"
C

'
4
1
(
)
1
.

2
'
"
1
6
1
.
1
2

"
"
7
1
1
2
.
1
1

2
9
5
6
.
"
.
'

J
l
4
1
8
1
.
'
4

"
"
"
'
6
.
'
4

]
5
\
J
4
'
.
1
I

"
'
'
8
'
.
4
4

J
"
I
6
6
4
.
5
D

4
P
7
1
1
7

P
I
P
E

O
I
A
M
E
T
E
R

P
U
P
P

L
O

l
N
5
U
L

1
J
O
N

S
T
U
"

L

T
A
L

L
O

2
4
1
1
5
"
'
6
,
0
"

U
+
C

2
3
7
5
J
.
0
6

2
1
1
"
6
0
.
0

"
(
6
\
11
.
1
2

.
r

2
5
1
1
"

.
6
Y

1
1
1
1
7
1
!
O
O

1
4
J
l
1
6
.
7
5

~

9
0
+

7
0
n
I
5
.
5
6

4
7
8
'
6
.
1
'

.
n

2
7
5
4
J
\
.
6
'

5
6
4
2
2
5
.
0
0

2
6
1
8
9
.
5
5

-

.
o

Z
B
0
1
Z
&

5
4
4
8
3
1
.
6
2

2
1
5
0

~
P

2
'
6
1
'
5
.
1
'

5
5
4
4
1
1
.
2
5

1
8
1
1
7
.
6
2

-

.
C

1
0
5
1
1
4
.
6
2

T
3

J
Z
s

1
7
0
1
7
.
1
5

+
P

)
1
4
'
5
0
.
5
0

5
'
6
7
5
2
.
)
7

J

1
s
7
0

3
2
'
'
'
O
.
J
I

6
2
1
4
7
6

6
2

1
6
1
3
1
.
(
'

)
)
2
1
6
.
6
2

6
4
6
'
9
1
.
6
2

1
6
1
8
'
.
1
5

J
4
0
6
6
0

6
'

6
7
Z
1
U
.
I
'

I
I
I
(
.

0
Z

$
4
1
7
6
6

5
7

6
9
"
'
1
8
.
1
1

1
6
0
5
1
.
5
6

+
C

,
'
6
6
9
1
.
0
0

7
2
5
0
5
4
.
1
1

1
6
0
1
1
.
1
7

*

+
D

J
6
4
4
5
1
.
)
I

7
5
1
1

e
1
'

1
5
9
'
6
.
1
J

)
7
2
0
8
4
.
1
7

7
7
1
2
6
2
.
"

1
'
'
8
1
.

I
I

-
"
.
(

3
7
'
5
0
'
.
1
1

8
0
n
0
T
.
1
1

1
5
9
1
0
.
T

+
D

$
.
.
.
.
.
.
6
2

1
2
.
'
0
1
.
1
'

1
5
9
6
"
'
9

-
1
.
0

3
.
4
2
1
7
.
4
4

8
5
5
2
4
0
.
2
5

1
5
'
5
1
.
'

'
&
'

4
0
1

0
2
.
'
4

.
1
1
1
2
5
.
4
4

1
'
.
5
4
.
1
1

=
D
,

4
0
8
6
4
2
.
6
.

'
0
6
.
'
1
.
"

FORTRN IV Program and Sample Printout for Chapter 5 203

U
I
F
U
R

P
I
P
E
L
I
N
E

D
E
S
I
G
N

C
A
L
C
U
L
A
T
l
n
N
S

A
N
N
U
A
L

c
n
S
T
S

A
R
F

S
H
O
W
N

B
E
L
O
W

A
S

F
U
N
C
T
I
O
N
S

U

S
U
L
F
U
R

P
I
P
E

D
I
A
M
E
T
E
R

T
P
F

0
1
A

1
N
C

t
'
N
E
f
q
I
N
G

P
I
P

L
O
5

P
U
M
P

C
O
S
T

1
,
"
(

1
5
t
O
(
.
"

2
'
5
'
5
(
r
.
(
l

5
3
b
+
C

=
"
"

n
,
`

4
8
6
B
'
.

'
7

^
1
*
T

1
Z

n
r
_

2
2
1
3
6
2
.
4
4

1
4
3

7
6
.

7

=
"
"

7
"

"

"

1
b

.
4
4

T
8
9

+
S
9

7
<

>
6
0
Z
"
+

A
8
q
.
,
S

6
.
"
'

"
"
P
+

l
f
2
7
7
R
.
l
1

7
.
f

7
5
(
"
(
.
'
'
'
1
7
4

2
(
.
4
4

!
Y
1

7
+

_
c
o

7
K
^
P
P
,

'

B
9
T

T
"

T
+

"

7
1
"

r
.
(
'

1

?
4
.
5
6

I

S
9
l
.
7
6

!
^

,
^

!
"
"
=

3
I

l
A
3

J
.
"
S

1
1

"

7
5
0
^

2
4
1
1

1
.
7
'

1

.
7
S

l
7

"
"

T
5
P
^
n
_

1
"
'
1
6
8
.
1
2

l
b
1
0
J
+

l
J
.
f
^

T
P
D
C
.
C
P

2
7
7
3
1
2
.
)
1

!

"
5
1
+

1
4
.
1
"

7
1
(
f
"
.
"
'

'
Q
'
6
q
"
.
7

I
h
I
1
8
.
1
7

t
5
.

'

7

"
"
0

.
C
C

3
1
4
1
8
1
.
9
4

S
S
b
+
1

=
r

7
1
"
n
t
.
"
"

H
2
7
3
6
.
'
4

l

S
8
1
+
1
8

!
?
,

P
C
P
,
^
0

1
5
1
3
H
.

1

I
S
9
7
"
.
7

^
^

7
-
(
(
(
.
"
'

3
6
'
'
f
'
.
4
4

1
5
9
6
3
.
3
9

l
S
,
^
^

C
"
C
=

3
8
R
6
6
4
.
5
C

I
,
Q
5
8
.
C
S

7
"

n
o

7
'
1
'
1
.
'
1

4
1
7
1
6
2
.
1
'

1
5
q
S
4
.
1
1

I
N
S
U
L
A
T
I
O
N

2
6
1
1
5
.
9
8

H
4
3
5
.
9
q

3
5

V
'
.
9

+
9
V

1
5
1
'
.
9
q

T
8
7
.
0
k

Z
3
+
S
Y

"

+
9
A

1
6
9
5
,
.
9
4

R
B
l
S
.
9
4

8
q
6
7
S
.
9
4

Q
6
!
3
5
.
9
4

1
0
2
3
9
4
.
6
9

I
I
R
7
5
S
.
R
I

l
1
S
1
1
5
.
8
1

1
2
1
4
7
5
.
8
1

l
Z
1
8

.
S
6

1
3
4
1
9
S
.
8
1

1
4
0
5
S
S
.
8
1

1
4
6
'
H
S
.
8
1

S
T
E
A
M

C
O
S
T

!
T
0
5
5
1
.
C
O

1
8
8
6
1
5
.
0
6

2
(
4
3
1
'
.
7
5

2
1
8
3
2
'
.
6
3

2
1
1
1
0
5
.
6
1

2
4
2
'
2
2
.
4
4

2
5
3
'
7
8
.
'
4

2
6
4
4
1
8
.
3
1

2
7
4
3
4
'
.
'
4

2
8
3
.
5
2
.
6
'

2
9
2
'
'
4
.
8
7

3
0
1
8
2
7
.
0
6

3
1
0
3
9
2
.
0
0

3
1
8
7
2
2
.
3
1

3
2
6
8
.
8
.
8
1

1
3
4
7
'
5
.
3
7

3
4
2
5
'
3
.
5
6

3
5
0
2
2
'
.
6
9

3
5
7
7
"
'
.
5
0

3
6
5
1
5
7
.
4
4

T
O
U
L

C
O
S
T

2
8
0
"
7
0
. 0
.
0

I
6
B
B
3
T

0
B
Z
5
9

8
7

5
5
2
2
7
8
.
8
7

5
"
1
3
3
1
.
1
9

5
5
9
0
8
0
.
5
0

5
8
6
1
5
Z
.
'

6
1
6
.
1
6
.
1
2

6
U
'
9
1
.
1
'

6
8
1
.
1
3
.
6
2

T
1
9

1
&
3
I

1
4
8
0
3
"
.
'
"

7
8
1
1
5
0
.
5
6

8
1
4

8
4
7
1
4
.
6
9

8
7
'
'
8
'
.
2
5

9
1
2
7
3
4
.
6
9

9
4
5
3
7
8
.
3
1

'
7
7
'
2
7
.
1
1

1
0
1
0
3
8
'
.
5
0

o
7 Appendix

I
1
J

F
l
i
P

P
I
P
f
1
.
I
N
F

D
F

T
1
N

C
A
I
.
C
U
L
A
T
l
n
N

N
U

n
F

D
I
I
M
O
J
N

S
T
'
T
r
n

&

I
N

U
l
'
T
'
O
N

'
W
I
C
K
N
F
S

!
N
L
H
F

I
.

N
N
A
L

5
T
S

4
R
E

S
H
O
W
N

B
e
l
O
W

S

F
U
N
C
T
I
O
N
S

O
F

S
U
L
F
U
R

P
I
P
E

O
J
A
M
F
H
R

P
I
P
E

f
U
.
I
N
C
W

F
'
(
'
'
H
R
I
N
r

P
I
P
E

C
I

T

P
U
M
P

C
O
S
T

I
N
S
U
L
A
T
I
O
N

S
T
E
A
M

C
O
S
T

T

T
m

5
T

&

7
C
"

"

"
'

&
^

9
E
1

&

&

1
2
0
'
1
0
.
0
6

2
1
0
6
1
3
0
4
.
0

?
&

'

}
q
(
r
n
_

9
B
6
\

&

^
B

&

1
b

&

U
5
9
2
0
.
2
5

1
,
,
9
6
2
7
.
0
0

&

\
&

9
9

&

&

1
4
9
7
1
7
.
6
2

7
2
4
6
.
. .
7
5

4
.

'

<
^
r

&
*
r

.
4
4

9

b
&

A
B
&

1
6
2
5
1
1
.
9
4

6
0
5
l
n
.
2
5

5
+
"

7
5
P

+
P
P

^
B
Z
\
&
1

E
H
&

b
Z

&

1
7
4
1
8
7
.
0
0

6
0
6
7
2
4
.
6
2

&

&

&

&
B

&

1
1
5
5
1
1
.
9
4

6
3
6
5
0
9
.
0
0

t
.
e
n

7

+
P
C

9
*
Z

&

&

&

1
9
6
0
5
1
.
3
7

6
7
5
7
7
7
.
3
1

+
^
^

7
5
^
D
^

1
9
&

&

2
3
"

6
7
.
9
4

2
0
6
0
7
1
.
1
1

7
1
1
7
4
1
.
6
2

V

"

&

&

b
^

&

&

2
1
5
6
1
5
.
1
3

T
6
n
0
9
.
'
1

!
^

^
r

!
"
"
&
P
P

3
1
+
^
"

3
&

&

2
2
4
9
0
0
.
0
6

l
o
n
n
.
o
o

!
1

"
l

T
"
"
"

"

&

B
^
&

b
&
O

U
l
I
O
I
.
1
7

1
5
4
1
2
9
.
0
6

Z
=

"

7
-
"
0
"
.
"
'

'

6
R
.
1
2

&

1
F
&

2
4
2
4
2
0
.
0
0

1
9
9
7
7
1
.
5
0

&
^
^

y
r
^

,
P

&

1
&

Z
b

b
b

2
5
0
"
1
.
7
5

9
4
5
4
1
9
.
0
0

!

&

"

&
D
P

&

1
^
&

&

2
'
'
9
)
5
.
6
2

9
9
0
9
9
9
.
0
0

l
'
.
"
1

7
5
^
"
C
=

9
1

&

b
&

3
6
4
4
2
8
.
2
5

2
6
6
8
1
5
.
5
0

I
n
6
4
9
l
.
1
1

!
^
=

"

'
"
"

!
"

b
&

&

B
&

2
7
4
6
5
1
.
'
"

1
0
1
1
1
.
4
.
0
0

7
,
C
"

"
'

&

&

&

&

2
I
Z
Z
7
)
.
_
4

1
1
2
7
1
7
6
.
0
0

1
A
_

+
F
O

^
9
9
8
9
,
4
A

1
^

&

b
b

&

Z
9
7
4
'
.
M

1
1
7
2
)
6
5
.
0
0

9
,

C
=

B
h

^
9
&
^
1

*
&

9
9

&

2
9
7
0
.
.
.
9
4

l
Z
I
7
4
"
.
0
0

"
,
^
r

T
5
P
"
,
P
0

9
I

&

"

9
&

7
&

3
1
4

J
l
'
.
6
1

1
2
6
2
4
5
9
.
0
0

U
L
f
U
R

P
I
P
f
L
I
N
f

D
E
S
I
G
N

C
A
C
U
L
A
T
I
O
N
S

N
U
M
B
E
R

U
F

P
B
P
T
N
G

S
T
A
T
I
O
N
S
.

Z
+
O

I
N
S
U
L
A
T
I
O
N

T
H
I
C
K
N
E
S
S

I
N

I
N
C
H
E
S

0
.
0

A
N
N
U
A
L

C
O
S
T
S

A
R
E

S
H
O
W
N

B
E
L
O
W

A
S

F
U
N
C
T
I
O
N
S

O
F

S
U
L
F
U
R

P
I
P
E

D
I
A
M
E
T
E
R

P
I
P
E

D
I
A
.
l
N
r
H

E
N
G
I
N
f
E
R
I
N
G

P
I
P
E

C
n
S
T

P
U
M
P

C
O
S
T

I
&

7
5

"
0

1
5
(

5
0
0
.
(
O

2
1
8
3
1
8
7
2
.
"
0

2
.
(
'

7

2
7
7
2
1
9
.
6
9

Y
Z
Z
1
Z
1

1
Z

D
0

7
5
0
0
0
.
0
0

1
5
4
2
3
1
.
1
9

1
5
9
0
1
6
.
7
5

.
(
n

7
5
n
f
(
.
"
(

+
1
H

6
3
8
3
6
.
8
9

5
&
D
D

7
5
0
0
0
.
0
(

1
4
0
7
6
0
.
0
6

4
2
8
2
'
.

5

6
.
n
"

7
5
0
(
0
.
(
0

1

3
1
3
'
.
1
2

3
6
"

8
0

3

0
O

7
5
0
0
0
.
0
0

1
6
8
5
6
1

I
9

3
4
0
5
7
.
6
2

=
D
O

(
(
(
.
(
(

U
5
4

7
.
3
1

3
0
2
7
.
1
5

9
.
0
n

7
5
0
0
0
.
0
0

2
0
2
H
9
7
.
2
5

3
2
5
3
1
.
7
6

I
n
.
n
t

7
5
0
"
O
.
/
(

2
2
1
0
0
6
.
(

3
2
2
1
3
.
0
5

I
I
.
n
n

7

0
M
.
o
n

2
3
9
2
6

.
8
7

3
2
1
2
8
.
7
5

1
2
.
"
"

7

D
O
+
P
O

2
5
7
M
"
.
"
6

3
?
0

3
.
8
2

I

"
"

U
P

2
7
6
2
M
.
1
2

3
H
H
9
1
.
5
6

I

.
n
"

1
'
(
(
0
.
C
'

2
9
4
7
9
8
.
8
7

3
1
9
5

.
1
7

j

D
D

7
5
0
1
0
.
0
1

1
3
4
4
0
.
9
"

1
1
9
3
6
.
1
3

^
+
"

7
5
"
n
(
.
M
I

3
Z
1
1

+
^

3
1
9
2
1
.
1
8

(
3

0
D

n
0
0
.
n
n

5
0
S
2

.
S
3

3
1
9
1
0
.
7
8

l
f
.
"
n

T

)
6
9
5
"
4
.
6
'

3
1
9
C
'
.
3
9

1
9
.
0
n

7
5
1
0
0
.
0
(

3
8
8
2
8
2
.
'
5

3
1
8
9
8
.
0
5

?
5
P
"
U
=
D
C

4
(
7
"
3
I
.
n
6

3
1
8
H

.
1
1

I
N
S
U
L
A
T
I
O
N

S
T
E
A
M

C
O
S
T

-
0
.
0

2
3
1
3
3
3
.
0
6

2
5
1
3
9
5
.
6
9

-
0
.
0

2
6
3
9
1
6
.
3
8

-
0
.
0

2
7
5
H
I
.
6
9

-
0
.
0

2
8
6
1
2
2
.
0
0

-
0
.
0

2
9
6
1
9
5
.
1
9

-
(
.
0

3
0
5
7
H
.
6
2

-
0
.
0

3
1
"
'
5
0
.
5
0

-
0
.
0

3
2
3
1
9
0
.
3
1

-
0
.
0

3
3
Z
3
h
6
.
6
2

~

&

H
0
6
6
0
.
6
9

-
0
.
0

H
8
1
6
6
.
3
1

-
0
.
0

3
5
6
6
9
1
.
0
0

-
D
.
0

3
6
4
"
5
7
.
3
1

-
0
.
0

3
7
2
0
8
4
.
8
7

-
0
.
(

3
7
H
5
8
9
.
8
1

-
0
.
0

3
8
6
9
8
6
.
6
2

-
0
.
(

3
9
"
2
8
7
.

"

0
1
5
0
2
.
9
1

-
0
.
0

4
0
8
6

2
.
6
9

2
6
6
.
.
.
.
.
0

1
'
2
5
"
"
.
0
0

0

"
0
1
6
6
.
1
5

,
,
"
1
1
1
.
5
6

5
6
0
"
1
.
0
6

0
1

2
.
.
.
0
0
1
3
0
.
0

.
,
.
,
1
9
.
3
1

6
6
0
6
2
'
.
6
2

6
1
7
0
5
5
.
1
1

7
1
,
.
,
,
.
2
5

T
.
.
.
.
.
0

7
6
6
2
1
"
.
1
1

7
9
2
"
6
1
.
M

1
1
1
6
2
'
.
)
7

1
9
9
U

.
2

0
T

B
Y
0
0
B
S
s
1
Y

'
1
2
5
6
7
.
1
.

g

;

4

.
.

g

1

5
U
L
F
U
P

F
J
F
f
L
1

D
E
S
I
G
N

C
A
L
C
U
L
A
T
I
O
N
S

N
U

1
P

D
F

F
J
P
P
T

S
T
A
T
I
O
N
S

=

z
.
r

I
N
S
l
I
l
A
T
l
O
"

T
H
l
f
K
"
F
S
S

l
N
C
H
F
S

1
.
5

A
N
N
U
A
L

C
O
S
T
S

A
R
E

S
H
O
W
N

B
E
L
O
W

A
S

F
U
N
C
T
I
O
N
S

O
F

S
U
L
F
U
R

P
I
P
E

D
I
A
M
E
T
E
R

N
C
'

F
N
0
T
N
b

T
N
G

F
1
F
1

C
O
S
T

F
U

C
O
S
T

1
.

'

7
5
"

"
.

r

1
5
0
4

.
n
r

2
4
8
3
1
6
H
.
O
O

"

7
s
n
n
,
.,

o

2
7
7
Z
I
9
.
6
9

'
7
2
1
Z
I
.
1
2

3
,
P
P

7
S
(
"
(
.
O
(

1
5
4
2
3
1
.
1
9

1
5
9
C
1
6
.
7
5

4
.
(
'
7
1
"
(
(
.
(
"

n

a
q
"
.
l
q

,
"
3
6
.
8
9

"

7
5

"
"
=

1
4
0
1
6
0
.
1
6

4
2
8
2
9
.
5
5

6
.
"
"

1

^
n

1
5
1
0
'
.
1
2

3
6
4
4
3
.
8
"

7
.
"
(

7
5
0
(
(
.
'
'
'
1
6
"
5
6
"
.
1
9

3
4
C
5
7
.
6
2

n
c

)
6
^
^
^
,
P
"

)
B

"
T
+
3
?

3
1
(
2
1
.
1
5

q

.
"
!

7
5
0
0
D
.
O
(

2
0
2
9
9
7
.
2
5

3
2
5
3
1
.
7
6

'
"

1
6
^

1
7
1
"
n
6
.
"

'
2
2
7
1
.
n

1
1
.
(
0

7
5
(
'
C
.
(
0

2
3
q
z
6

.
8
7

3
2
1
2
8
.
7
5

1
,
.
"
,

7
"
"
'
'
',
(
1

2
5
7
1
8
4
.
0
6

3
Z
t
4
3
.
a
2

l
3
=
"
"

7
5
0
0
(
.
1

2
7
6
2
0
6
.
1
2

3
1
9
9
1
.
5
6

1
4
.
1
t

7
s
n
n
n
.
(
n

7
Q
4
7
'

.
R
7

3
1
9
5
8
.
1

7

7
5

C
C
.
l
n

3
1
3
4
4
0
.
9
4

3
1
9
3
6
.
1
3

l
l
.
"
I

1
5
n
n
c
.

3
1
I
l
R
.
4
4

3
1
'
2
1
.
1
8

1
7
.
0
(

7
5
D

t
.
P

3
5
0
8
1
1
.
8
7

3
1
H
O
.
7
8

l
R
.
"
f

1
1
'

P
"

"

1
6
'
5
4
4
.
6
'

3
1
9
(
3
.
1
'

l
C
.
I
'
'
7
5
0
D
(
.
o
r

3
R
R
2
8
2
.
7
5

3
1
8
'
8
.
1
5

?
"

"
"

7
$
^
^
n
_
o

4
r
7
0
J
I
.
"
6

3
1
"
9
4
.
1
1

I
N
S
U
L
A
T
I
O
N

S
f
E
A
M

C
O
S
T

9
7
3
0
7
.
B
8

1
2
0
5
8
0
.
0
6

1
1
6
3
8
1
.
9
4

1
3
5
'
Z
O
.
2
5

1
3
5
4
6
7
.
9
4

H
9
7
I
T
.
6
2

1
5
4
5
4
8
.
0
0

1
6
2
5
3
1
.
9
.

1
7
3
6
2
8
.
0
0

1

7
.
,
8
7
.
0
0

1
9
2
7
0

.
(
(

1
8
5
5
U
.
9

2
1
1
7
8
7
.
9
4

1
'
6
0
5
1
.
3
1

2
3
n
8
6
7
.
9
4

2
0
6
0
7
8
.
8
1

2
4
9
9
4
7
.
9
4

2
1
5
6
1
5
.
1
3

2
6
9
0
2
7
.
9
4

2
2
.
9
0
.
1
6

2
8
8
1
0
6
.
1
9

2
1
3
8
0
1
.
8
7

3
C
7
1
8
6
.
5
6

Z
U
.
Z
O
.
O
O

3
2
6
2
6
6
.
3
7

2
5
0
7
8
8
.
7
5

3
4
5
3
4
7
.
5
0

2
5
8
9
3
5
.
6
Z

3
6
4
4
2
8
.
2
5

2
6
6
8
8
5
.
5
0

3
8
3
5
n
R
.
2
5

2
7
.
6
5
8
.
"

4
C
2
5
8
8
.
2
5

2
8
2
2
U

4
Z
1
6
6
7
.
5
0

2
8
9
7
.
5
.
5
0

4

O
H
7
.
5
0

2
9
7
0
"
.
9

4
5
'
8
2
7
.
5
0

3
0
4
3
1
5
.
6
2

T
L

2
6
6
2
'
2
3
2
.
0
0

1
5
2
6
6
.
1
.
0
0

6
U
5
0
3
.
5
0

5
'
1
8
1
5
.
0

6
0
6
6
0

5
6

6
U
8
'
.
1
I

6
"
.
5
1
.
1
2

U
0
3
1
1
.
2
5

1
1
6
1
5
2
.
0
6

8
2
2
2
0
1
.
0

8
6
8
3
0
3
.
1
.

.
"
'
3
3

9
6
0
Z
5
2
.
8
1

1
0
0
6
0
.
0
.
1
2

1
0
5
1
6
9
0
.
0
0

1
0
'
7
2
0
6
.
0

1
1
4
2
5
'

1
1
8
7
8
6
1
.
0

1
2
3
3
0
1
6
.
0
0

1
2
7
8
0
6
1
.
0
0

.

FORTRAN IV Program and Sample Printout for Chapter o

U
L
F
L
J
R

P
I
P
F
L
I
N
F

n
F
S
l
r

C
A
L
C
U
L
A
T
I
O
N
S

N
L
J

A
F
R

P

P
T
N
6

S
T
A
T
[
"
N
S

1
N
S
U

A
T
J
O
N

1
L
X
N
F

I
N

I
N
C
H
F
S

.
r

A
N
N
U
A
L

C
O
S
T
S

A
R
E

O
W
N

B
E
L
O
W

A
S

F
U
N
C
T
I
O
N
S

U
f

S
U
L

F
U

P
I
P
E

E
T
E
R

P
I
P
F

n
I
A
.
r
N
C

F
N
C
N
F

P
T
P

L
O

P
U
"
P

L
U
b
1

I
N
S
U
A
T
I
O
N

S
T
E
A
I

C
O
S
T

T
O
T
A
L

C
O
S
T

I
.

n

7
5
D
"
"
.
^
P

1
5
(
.
5
0

.
O
(

J
?

2
5
U
5
5
.
9
.

9
2
.
6
2
.
2
5

2
6
7
5
5
6
4
8
.
0

,
.
n
n

7

^
^

^
(

H
1
?
1
9
.
(
9

"
n
I
2
1
.
1
?

2
9
0
(
1
5
.
9
.

I
O
U
0
1
.
3
1

1
6
6
8
5
5
8
.
0
0

3
&

"

"
P
"

1
5
4
Z
l
I
.
1
9

S
U
!
^
&

3
2
8
1
7
'
.
9
4

1
1
'
0
7
1
.
4
4

U
l
4
n
.
i
l

k
,
P
^

7
5
^
O
t
,
[
(

1
1
5
8
9
8
.
1
9

6
1
R

.
A
9

3
6
6
3
1
5
.
9
4

1
2
5
2
6
.
8
l

1
6
6
3
n
.
8
l

?
&
*

7
5
"

"

1
4
"
1
(
1
.
1
6

>

4
0
.
4
9
5
.
9
4

1
3
.
9
1
0
.
1
2

1
9
8
0
0
5
.
6
2

^
n

^
^
{

^
r

1
'
1
U
9
.
1
1

3

k
k
3
,
F
D

4
.
2
6
'

.
9
4

1

1
2
3
.
9
.

8
5
1
3
6
2
.
1
5

7

7
5
P
"

1
6
B
5
6
P
.
1
9

-
^
Z

U
/
8
l
'
.
9
.

1
5
2
9
4
6
.
1
5

9
1
1
3
8
0
.
5
0

P
,
^
o

'
"
"
.
P
"

I
M
.
"
1
.
H

7
+

"

5
1
8
9
1
4
.
6
2

1
6
1
4
4
2
.
0
6

9
1
)
8
5
1
.
1
9

9
&
0
0

"

2
C
Z
'
9
1
.
2
5

>
3
1
e
7
^

5
5
1
U

1
6
9
6
5
1
.
6
.

1
0
3
7
3
1
5
.
1
9

I
P

"
"

)
6
^
r
^
_
^
r

7
2
1
"
"
6
.
(
"

J
?
T
+

>

5
9
5
2
<
4
.
3
1

1
1
7
6
1
0
.
"

1
1
0
1
1
8
3
.
0
0

1
1
&
"
"

7
5
^
^
P
.
C
0

2
3
9
2
6
'
.
8
1

J
Z
!
2

6
3
3
.
5
4
.
1
9

1
8
n
.
6
.
?
5

1
1
6
5
1
.
5
.
0
0

Z
+
"
"

7
6
^
O
^

^
r

7
M
4
.
1
6

Z
P
4

,
*
Z

6
7
1
6
1

8
1

1
9
2
8
8

6
9

1
2
2
9
2
2
7
.
0
0

1
?
.
D
P

!
O
,

C
^

1
1
6
2
1
6
.
1
2

3
1
S
9
1
+

7
(
9
1
7
4
.
6
9

2
0
0
2
4
4

1
2
.
3
2
1
6
.
0
0

I

+
P
"

^
^
O
_
+
r

7
'
4
7
9

.
8
1

I
"

.
1
7

1
4
7
9
H
.
8
1

2
0
7
.
4
4
.
6
2

1
3
5
7
1
]
6
.
0
0

I
*

^
^

P
"

+
P
P

3
1
1
.
4
1
.
'
4

J

9
3

+
1

7
8
6
0
9
5
.
8
l

2
1
H
9
'
.
5
6

1
4
2
0
.
7
1
.
0
0

I
A
_

)
<
^
r
n
_
^
r

1
3
1
I
1
A
.
4
4

1
S
1

!
B

!
2
4
1
5
5
.
8
l

2
2
1
U
3
.
0
6

I
U
.
7
1
8
.
o

7
,
C
^

"
"
,
P
P

3
5
0
1
1
1
.
8
7

1
9
1

8
U
4
1
5
.
8
1

2
2
8
2
2
7
.
0
0

1
5
4
u
n
.
O
O

F
+
P

7
S
^
^
C
.
"
"

1
6
Q
'
4
4
.
6
9

5
!
R
U
3
+

9
0
0
5
7
5
.
0
6

2
3
4
9
2
1
.
6
.

1
6
1
1
9
4
4
.
0
0

R
=
"
P

7
5

^
^

^
^

8
2
1
2
.
2
5

3
1
8
9
6
.
0
5

9
3
8
7
]
5
.
1
6

2
U
5
1
6
.
5
0

1
6
1
5
.
1
1
.
0

7
n
.
"
"

7
5
^
^
^
_
^
r

4
n
7
1
3
1
.
!
6

.
1
1

'
7
6
8
9
5
.
8
1

2
4
8
0
2
0
.
1
.

1
7
3
8
8
4
0
.
0
0

U
1
T
U
P

P
1
P
E
t

n
r
S
I
G
N

C
A
L
C
U
L
A
T
I
O
N
S

U
N
R
T
P

D
F

J
H
P
1
N
C

S
T
A
T

O
N

.
"

1
"

U
l
A
T
l
n
'

T
H
1
L
K
N
F
S
5

1
N

1
N
C
H
F
S

I
,
^

A
N

U
A
)

C
O
5
T
5

A
R
E

5
H
D
V
N

B
E
l
O
W

A
5

F
U
N
C
T

U
N
5

U
F

S
U
L
T
U
K

P
J
P
E

D
1
A
H
E
T
E
R

1
t

0
1

T
N
H

F
N

T
N
?

R
T
N
6

P
U
*
P

L
O
5
1

l
.
f
r

7

1
0
(
.
o
r

1

l
q
4
9

,
H

2
4
8
4
7
9
C
9
.

0
"

Z
&

P
P
&

2
n
7
5
7
9
.
6
9

9
l
A
n
6
1
.
t
2

3
&
P
"

'
5
(
4
t
.

1
3
1
8
5
4
.
1
2

1
7
4
9
5
6
.
7
5

4
.
"
"

7

I
"
O
.
(
"

l
2

b
9

,
R
)

7
9
7
7
6
.
8
7

o
"
"

"
O
"
O
.
(

"

+
t
S

5
e
7
6
9
.

.
"

0
M
.
0
P

I
A
9
9
?
f
.
^
6

2
1
8
3
.

1
.

t

1
5
,
(
0
.
(
'

l
b

T
=

4
9
9
9
1
.

T
5
P
0
P

R
P

I
?

7
1
.
^

k
b

7
=
1

9
&
D
^

1
0
o
c
.
r
f

Z
C
1

9
6
4
+

4
8
4
7
1
.
7
9

"

^
"
+

"

?
2
"
?
3
+

B
?
1

+
^

I
I
.
"
"

7

f
l
"
.
"
t

7
H

7
7
.
]

4
A
O
A
A
.
1
8

"

^
P

J
?
7

4
7
9

l
e
9

&
P

7
'
1
'
(
.
l
n

2
7
,
A
U
.
4
4

4
7
9
1
1
.
5
9

1
4
.
-
"

T

P
^
P

4
4

,
?
*

4
7
R
9
8
.
2
.
.
f
*

P
P

U
D
D
&
O
0

3
1

3
1
H
3

7
8
7
b
o
1
6

B
+
^

T
Y
^

3
3
1
9
1
2
.
1
1

J
6
6
1
+
2
0

1
7
.
n
t

l

I
'
O
.
(
1

l

4
7
.
9
4

4
7
8
5
0
.
8
1

8
,
0

T
5
O
O
O

P
P

3
6
9
3
9
6
.
4
4

J
6

3
+

1
9
.
"
1

n
l
l
O
.
l
n

8
F
1
^
4
+
B
\

4
1
8
3
A
.
r
7

"

"

"
(
'
.
"
(

!
^
S
?
V

b
S

3
^

1
N
S
U
L
A
T
1
O
N

S
T
U
N

C
O
S
T

5
8
5
1
1
.
9
3

1
3
9
1
3
3
.
1
2

7
l
2
3
1
.
9

1
5
6
0
6
2
.
5
6

6
)
9
5
1
.
9
4

1
7
1
1
1
2
.
1
5

9
b
6
7
1
.
9
4

1
8

7
6
6
.
1
2

1
1
9
3
9
2
.
(
0

1
9
1
)
3
9
.
1
2

1
2
l
1
1
2
.
0
0

2
0
9
0
5
2
.
1
9

1
3
4
8
3
1
.
9
4

2
2
0
0
6

1
4
7

5
1
.
9
4

2
3
0
4
9
6
.
3
8

1
6
0
2
7
1
.
9
4

2
4
0
4
4
0
.
0
6

1
7
2
9
9
1
.
9
4

2

9
9
6
8
.
0
6

1
8
5
7
1
1
.
9
4

2
5
9
1
3
8
.
6
9

I
9
8
4
3

,
6
9

2
6
8
0
0
0
.
0
6

2
1
1
1
5
C
.
5
6

2
1
6
5
8
9
.
8
1

2
2
3
8
7
1
.
6
9

2
8
4
9
4
1
.
8
1

Z
3
b

9
1
+
t
9

2
9
3
0
8
3
.
7
5

2

H
I
2
.

4

1
0
1
0
3
8
.
1
5

2
6
2
0
3
2
.
4
4

3
0
8
8
2
7
.
1
5

2
7
4
1
5
1
+
1
9

3
1
6
4
6
1
.
5
0

Z
b
J

+
b
9

3
2
3
9
1
3
.
6
2

3
1
0
1
9
1
.
6
9

3
3

L
l
5
9
.
W

T
O
T
A
L

C
O
S
T

2
6
1
3
9
9
2
0
.
0
0

1

7
9
n
,
o
o

B
T

5
6
2
1
1
3
.
1
5

5
1
5
9
0
1
.
3
1

6
0
M
l

,
0
6

6

U
6
1
.
"

6
8
5
9
1
1
.
0
0

7
2
6
1
4
8
.

6

7
6
M
I
O
.
3
1

8
0
6
5
9
6
.
6
2

8
4
6
6
'
7
.
2
5

8
8
6
5
0
9
.
"

9
2
6
2
L
L
.
0
0

9
6
5
1
U
.
5
0

1
0
0
5
1
2
4
.
6
9

1
0
4
4
3
5
8
.
1
1

1
0
8
3
4
5
9
.
0
0

l
L
2
2
4
3
8
,
O
O

1
1
6
1
3
0
6
.
0
0

|

;
:
P

.

L Y Program 8O Sample Printout for Chapter
nA
PV07
nrT3
0P04
P^P
0ODq
mDY
P9PB
^P
0O]P
"1
"J
^J
0^14
""1
"1&
PP3
PP
1B
P^#R
0#P
"^7
0
77
P"73
0^74
"7
P7
"^?T
CP7B
P^ZP
nO
0P
003#
P 33
0P4
P^3
P4&
PT
P0
O
39
0D4D
^41
0D4
P^,,
0P44
0P45
0P40
0P43
0P4
P^^
on<
\
7
""
n(-4
("5 5
PRnrRl FOR STYQFNF-ETHYL8FNlENF VACUUM OISTILlATION
C ROTH FfE kND BOTTOS RATES. F AND B. ARE H POUNDS PEK YEAR
nlFNION RRCNI!l. FTfll. X!Cl. Y!00l.ANT!l. COSTf6!l
nlHFNSION CI(OSf"! l+ SCOS111I. CPFO1t14

^ J&O "l &lPH&,F,X,R,R

IF fALPHA-I.e1 qqq.qq'.stl
6^1 F"ntF=F/ ('F*tC6 . (+tl . (-Xtt*1"4.nl
RHDlE=B/fXS.IC6.0+fl.C-XB'!04.(
I=FnlF-Rnl F
Xn=fFnlF*XF-RHnlEXR"D
YCFPT=AlDHA.,Fflt."+(AlPH4-J."t*XF,
A=fXO-VrFPTl/lxn-Xcl
.RIN=AIf! .-Al
RQrNf I I=!.S
PO DD N=I,6"
P.=RRMI.RRCMIN'
IISlOP=RA/IRR+!.'
l/rfPT=XDI I +1.'
DlCPT=lISl OP.XF+IJCEPT
npl,orPT-XR"IXF-XB'
f11j sY"* t I _ n~<r not t
TNT=C.f
v1 I I =vr
DO 599 l=7,FD^
XI ~1 I=Vf ~1IIf lPHl-(AlPHA-l."'.Vf I-I"
lT lY1111-X?1 5"7+P23
5^? lF lY1 f-tJ-XFt 5^. 'r4,tf'
5D Y111=|1SITP+Y1A~11+l"PT
CO Tn 99
54 lT r" I-? -XF' 5"6,DE+7
7 NT=frNT.,.nl/a7+1_"
FTf'I=NFT
oh Vfl'=SlOPl*Xfl-ll+CEPTl
qq T'T:T'T+t.(
n, Of=R".O
v:rON+I"(OI,
Nr=TT I. 7+ 1.1
"'TINI ='T
TfIX=5.SVf.5
NC=ANTINI/" .O+I.(
lNC=Nr
rPCOSINI=IZ.4.ITIOX'.*1.21*ANTfN'
TrnfN1=V.Ar.'1,_q
cwrnSIN'V.ANC.h.04
rnSTIN'=CPCOSINI+CWCOSIN'+STCnSINI
6(( .A(NIN.II=RPrNINI+.(
PRINT !P?s+TD1N
PRINT J^$
11 FnQATIFI7."
I' FOR_AT II'H tf RATf: .F9.0.72H ER FRACTION IN FEED= .F5.3.27H EB
! FD.c.,nN IN OISTIll'TF,FS.3.!2H IIMUN q. , Fb&3771
1"3 FnR.ATIXIH RR FACTnR,13XHTRAYS,9XqHFffO TRAy,qX9HOfPR COST.qX9H
\tA cnST,qXqHCll :OST,qXqHvFA COSTIII
PU 1"4 N=I,r
14 DqlNT l+ RPCNII,ANTINl.FTIN"CPCOSfN'.STcnSIN'.CWCOSIN'.COSTfN'
!C FORATI7fI8.211'
CO Tn "
9qq 5TOP
fN!
ZV
F
E

9
&
T

T
J

F
B

T
T
O
N

F
R

C
T

D
I
S
T
I
L
L
A
T
E
-
O
.
9
1
6

"
I
N
I

3
.

6
6

|

C
*
%

T
'
A
V
S

F
P

T
R
A
Y

P
R

C
O
5

U
U
L

L
O
5
T

Y
E
A
R

L
O
5
T

I
&
"

9
8

4
R

5
B
B
B
9

9
5
2
9
5
.
'
6

5
0
5
7
.
8
7

1
5
7
2
H
.
3
1

,

.
1
"

8

4
1
.
"
"

"

1
.
"

9
1
"

1
9

l
4
6
.
8
0

1
5
4
5
4
3
.
0
0

<

~
.

1
.
'
'

7

"

V
V

9
Z

l
O
2
4
1

.
8
1

5
4

5
.
7
3

1
5
5
1
3
0
.
6

I
.
'
"

T

.
"

?
&

5
4
1
.
'

l
(
9
1

.
'
0

5
6
2
4
.
6
6

1
5
6
1
4
6
.
1
2

I
e
7

9
&

3
1
.
"
"

"

.
6
'
1
1
9
5
'

.
1
9

5
8
1
3
.
5
9

1
5
8
8
9
3
.
3
7

1
.

"

A
_

5
9
.
1
0

1
1
1
(
9
1
.
1
1

6
0
0
2
.
5
2

1
6
1
5
5
5
.
6
2

I
s
3
5

3
(
^
^

Z
B
n

"

U

.
"
Z

1
1
1
6
9
.
0

1
2
7
.
6
3

1
2
3
1
8
6
.
6
2

t
+
4
"

<
+
^
^

1
1
1

.
"
"

I
O
l
U
.
(
6

4
'
5
3
.
5
9

1
2
5
7
6
9
.
6
9

I
+

1
+

"

Z
^
+
P
"

1
4
2

'
6

1
2
5
1
"
.
1
9

4
3
7
9
.
5
4

1
2
8
3
2
0
.
5
6

f

5
^

f
P

Z
^

0
P

1
1
5
"
.
"
1

1
4
1
1
1
.
2
5

4
5
0
5
.
4
9

1
3
0
1
4
8
.
8
\

1
.
5
.

9

^
?

o
n
2
.
"

8
1
2
6
1
.
3
1

4
6
1
1
.
4
5

1
3
2
6
2
5
.
3
1

I
&

0
&
P
P

.
"

'
'
9
6
6
.
1
1

1
9
6
1

.
5
0

4
7
5
7
.
4
0

1
3

3
5
8
.
1
2

I
+

0
&

]
3
u

9
1
.
l
l

n
O
O
T
.
5
6

8
8
3
.
3
6

1
3
H
8
8
.
6
9

&

]
/
s

"

'
6
.
'

9

"
1
"
.
1
'

5
0
n
9
.

1

1
3
9
8
7
6
.
8
7

1
.
1
"

5
3
e
T
P

2
'
.
1
0

'
9
0
0
.

I
T

9
6
1
5
'
.
1
1

5
1
3
5
.
2
7

1

1
4
8
9
.
3
1

,
A
n

3
=

Z
?
+
^
C

4
(
1
8

.
1
7

9
9
1
2
h

s
S
4

5
2
6
1
.
2
2

1
4
4
5
6
8
.
2
5

"

I
.
A

3
&

"

Z
=
D
C

4
1
7
5
4
.

?

I
C
1
5
0
0
.
0
1

5
3
8
7
.
1
7

1
4
7
6
H
.
6
9

g

'

C
?
t
I

?
1

4
1

4
3
.
9
1

1
1
3
8
7
3
.
1
2

5
5
1
3
.
1
2

1
4
9
9
J
O
.
1
2

s

!

1

9
5

"

?
\
,
^

4
1

9
7
.
1

1
0
6
2
4
6
.
1
9

5
6
3
9
.
0
8

1
5
2
9
8
2
.
3
7

"
:

+
"
"

\
_

^
n

4
"
8
4
4
.
'
C

I
I
R
6
1
9
.
3
1

5
7
6

.
0
4

1
5
5
2
2
8
.
8
7

"

"
*

\
&
"
^

2
1
.
n
r

4
1
1
7
7
.
7
(

1
1
0
9
9
2
.
4
4

5
8
9
0
.
9
9

1
5
8
2
6
1
.
1
2

1
.
"
"

'
l
.
n
l

4
1
o
n
"
.

1
3
3
6
5
.
5
6

6

1
6
.
9
4

1
6
1
2
8
8
.
7
5

I
.
?
1

9
,
^
^

!
R
+
P
!

4
C
T
6
6
,

1
1
5
7
3
8
.
6
9

6
1
4
2
.
8
9

1
6
2
6
4
8
.
0
6

~

+
P
"

!
R
=
^
P

4
1
2
f
5
.
9
9

I
I
R
I
I
I
.
7

6
2
6
8
.
8
5

1
6
5
6
4
6
.
5
6

2
.
?

4
c
.
(
f

1
9
.
"
1

4
1
7
6
1
.
4

1
2
0
4
8
4
.
8
7

6
3
9
4
.
8
0

1
6
8
6
4
1
.
1
2

"

~
.

=
"
"

^
P

!
P
,
Q
P

4
1
2
5
3
.
1
3

1
1
2
8
5
8
.
C
r

6

?
1
.
7
6

1
7
1
6
3
1
.
8
7

~

k
R

R
+
^
P

"
1
8
6
8
.
6
9

1
2
5
2
3
1
.
C
6

6
6
4
6
.
7
1

1
7
3
H
6
.
4
4

,
^
n

I
q
.
t
I
9
?
3

&
S
6

1
2
7
h
0
4
.
1
9

6
7
7
2
.
6
7

1
7
6
7
1
9
.
8
1

(

+
4
5

R
.
P
^

R
,

4
?

I
+
t
T

1
2
9
9
7
7
.
2
5

6
8
9
8
.
6
2

1
7
9
6
8
9
.
5
0

=
"
"

k
R

^
r

'
'
I
'.
Q
1

1

Z
3

3
7

7
0
2
4
.
5
7

1
8
2
6
5
5
.
8
7

Z
+

,
^

!
P
,
D
^

4
1
7
4
4
.
'
1

1
3
4
7
2
3
.
5
0

7
1
5
0
.
5
3

1
8
5
6
1
8
.
9
4

t

| |
t

~

t

Z
,

4
A
.
"
"

I
q
.
"
r

4
4
7

l 3
7
r
9
6
.
6
2

7
2
7
6
.
4
8

1
8
8
5
7
8
.
7
5

:
.

9
6
+

"

,
P

4
2

.
1
9

1
3
9
4
6
V
.
6
9

7
4
n
.
4
3

1
B
Y

T
h
&
3
1

$
.

.

Z
+
T
"

4
1
.
"
'

!
P

^
n

4
1
7
3
7
.
1
'

1

Z
+

2
8

3
9

1
9
2
6
0
8
.
8
7

?
7

1
,
6
.
(
1

1
"
.
"
"

4
'

7
n
.
l

1
4
4
Z
1
J
a

7
6
5
4
.
3
4

1
9
5

.
5
6

+
*
"

4
,
. .
"
"

|

,
^
P

4
4
1
1
(
.
"
7

14
6

8
9
.
(
O

7
7
8
(
.
3
0

1
9
8
"
6
Y
.
3
1

4
6
.

1

!

_
r
n

4
4
5
2
7
.

V
b
Z
+
1
Z

7
9
0
6
.
Z
b

1
3
Y
b

3
T

^
,
!
"

,
M

"
4
9
5
1
.
1
1

15
1
3
1

.
2
5

8
0
3
2
.
2
1

2
0
4
3
1
8
.
6
9

^
,
^
r

1
"
.
"
"

4
5
H
2
.
9

3
T
D

+
3
T

I
b
8
.
1

2
0
7
2
3
9
.
3
7

4
,
.
"
"

!

^
^

4
4
7
'
'
.
4
1

1

6
r
8
1
.
4
4

1
1

2
0
9
1
6
2
.
0
0

1
.
5
.
(
(

l
R
,

3
.
9
1

1
5
8
4
5
4
.
5
0

4
I

.
0
6

2
1
2
0

8
.
4
h

+
"
!

I
A
.
"
"

4
5

.
A
9

1
6
0
8
2
7
.
6
2

5
3
6
.
0
2

2
1
4
9
7
Z
.
b
0

4
1
.
t
'
'
I
q
.

o

4
M
I
I.

4

1
6
3
Z

"
.
7
5

1
.
9
7

h
.
2

+
"
"

+
"
"

4
M
l
I
.
Q
r

1
b

T
+

7
.
9
J

2
2
0
7
7
3
.
5
0

_
D
n

o

4
,

(
q
.

7
3

16
7
9
4
6
.
9
4

8
9
1
3
.
8
8

2
2
J
6
7

.
5
0

^
=
^
'

4
-
.
"
1

)
A
,
n
o

4
3
Z
^

+
A
5

1
7
0
J
0
.
C
6

'
0
3
'
.

2
2
6

6
5
.
3
7

&

"

4
t
.

R
+

4
7
'
q
q
.

1

1
7
2
6
9
3
.
1
'

9
1
6
5
.
1
9

2
2
'
H
1
.
9

g

'
.
'
.
"

4
1
.
'
1

P
,
^
P

4
7

.
4

1
7

C
6
6
.
3
1

n
9
1
.
7
5

2
1
2
3
4
8
.
5
0

;

4

3
.
4

"

l
P
,

"

4
.
3
7
9
.

1
6

1
7
7
4
3
'
.
3
7

9
'
1
+
7
0

2
3
5
2
3
6
.
6
9

.
..

4
'
.
"
"

F
+

"

4
6

9
q
.

"

1
7
9
8
1
2
.
5
0

9

4
3
.
6
5

2
3
5
9
5
5
.
6
2

A
<

'
"

R
,
"

4
6
9
6
1
.

n

1
8
2
1
8
5
.
5
6

9
6
6
'
.
6
1

2
H
8
2
2
.
6
2

9
*
&
^
"

1
,
.
1
'

4
7
3
n
.
6
b

1
8
4
5
5
R
.
6
9

9
1
'
5
.
5
6

2
4
1
6
8
7
.
8
7

^
+

4
3
,
P
P

!
A
+
^
"

4
7
6
'
7
.

8

1
8
6
9
3
1
.
8
1

9
9
2
1
.
5
1

2
H
5
5
1
.
1
9

"

^
+
P
"

1
,

4
8
1
6

.
3
n

I
R
9
3

4
.
9
4

1
0
1
4
7
.
4
7

H
1
U
2
.
6
9

- W

4
,
.
r
l
l
1
1
.
1
1

4
A
4
2
0
.

9

1
9
1
6
7
8
.
C
O

1
0
I
n
.

2

2
5
0
2
1
2
.
3
1

.
..

^
+

"

4
^
+
^
P

^
+
"
"

4
.
1
7

1
'
4
1
5
1
.
1
2

1
0
2
9
9
+
3
7

>
1
>
U

1
Z

.
.

.
.
3

4
'
.
"
1

l
^
=

"

4
9
1
3
6
.
7
1

1
9
6
4
2
4
.
2
5

1
0
4
2
5
.
3
3

2
5
5
9
B
6
.
2
5

'

!
.
?
^

4
"
.
"
'

1
"
.
l
f

4
'
4
'
2
.
0
7

I

9
B
7

?

3
\

1
0
5
5
1
.
2
8

2
5
8
8
4
0
.
6
2

;
3
.
q

l
A
+
C
^

4
9
R
4
5
.
7
I

2
0
1
l
7
t
.
4
4

1
0
6
7
1
+
2
4

2
6
1
6
9
3
.
3
7

.
.
W

4
,
"
'
'
l
^
+
^
"

5
P
1
9
7
.
6
4

?
(
3

4
1
.
5
n

1
1
8
1
3
+
1
9

2
6
4
5

.
3
1

Z1 Appendix
L IDgI Q Spe IUDu DI LDQ0I
000 1
ooz
0013
uOO
OOO
OOOb
0007
OOOH
0;09
00 1
(011
OOIZ
CU13
00 1
0,1 '
COlb
001 7
OOh
0019
OJlO
OOZI
OOZZ
OOZJ
( ..2
OOZ,
COlb
C027
002A
OOl9
t030
0 031
OOJl
033
C03
003
03b
0037
003H
C039
COO
C041
UCZ
003
OOH
(0'
10b
0047
OOH
(049
005
!C1
UOZ
0053
Oi'
u05
OO,b
00$1
CO,d
li9
OOb'
OObl
OOol
00b3
00b4
Oub
"Obb
00b7
L P<UCESS OESIC FU STVfNE PROOUCTIUN
UIMfNSIU E813I,TVI3I,BlI3I,TOLI3I,HVDI)I,TEMPI3I,TEMPGI)I,PRESSI
l I
LUU TlE,rIO,rOu,FRATE,rEMP,TEPG,CArOIA,CATOEN,ETA,fPS,H,ePEB,
ILP.G,<ATEFG,PR(XC,SLICe,SrV,TOL,dl,ES,PRESS,HVD,XOLe,PRESSG,V,M,
LI,,XI,K,J,VleLO,rUeL,RKI,RKZ,RK3,EKI,eONSTI,CONSr2,CONST3,eONsr4,
3VJli,ARA,AT(MP,ArePG
RtAD 13,CATO,p"Exlr,FGRATE,CPEB.CPFG,H,EPS
"fAD 13,CATDIA,TA
"tAi 1.SLICE,OeL
"tAu 13,CPRCf,tdP"CE,eArp"C
L fkMAT STATFMtNTS
Il .kMATIIH,ZX,fI0 . 3,5FI. ,fI 0 .Z,FIO. Z,Zf10. 31
] FJKMArI8FIC. 51
14 fJRAT181101
PRExG=PREXIT-14 . 7
Z READ IJ,TIO,TOO,REI
L CONTkOL tARO FU. STUPPING PkOGRAM
IF Irl01999,99 9,ZI
ZI RAU 13.ATEMP,ATfMPG
A.EA=3. 1419.Tlu**Z/57.
V10=11. -EPSI/EPS*3
SLIt E=NSLI CE
lu<L=TLE/SLltE
F{ATE-Z' . *lrID/4 . 031**Z.
RATEF=FCRATEITOO/.O'*l.
JO tALL PRTLAT
tOsr l=tArON*AA*IOb. I.RATf
CuNSTl-H*TUO/1 2.*3. lI;9/ItPES*FRATEI
CuNSTJ=;3bOu./ICPEB*IOb. 1
CJSr4=H.rOO/IZ. *3. 141;9/ICPFC*RATEFG'
LALtULATE ENrRANCE PRESSURE
AOlNZPREXlr / l. 7*8;. *.CbZ4/IC.99*0. 08Z0b*1 ATEMP +bOO.'*0.;;561
A.ESS3 l . *L. Olb7/CATOIA*F<ATf**Z/4Z;. **Z*0 . 0 8 8;**Z/AREA**Z*TLENI
1 1.'19 . 7
APkESS=I. lOAPPESS
U 100 l-l,ZC
AOtNI=APkESS/14 . 7*10o.*0 . ObZ4/10 . 9 9*O. 0 820b*1 ArEp +4bG. I*O . ;;bl
AvENIAOtNI+AD(ZIIZ.
AVtL=FkATE/IAD*3bOO. *AfAI
ID AkfSS-PRExlr'TLEN*0 . OOOZI84.1.7'/CATDIA.ADEN.AVEL**2*VOID
APRESS- l . I*APRESS
U TEMPI31=ATEMP
rtMPGI3'-ATEMPG
PrtSSI31 =APRESS
SrV( 3' =0.
ruLI3I=O.
bLI3I-0.
MM-O
N=Z*tSLltE+ l
UU so. I=I,
19 (FII-Z*(I/ZII 100,ZOO,jOO
OO tIZI=1. -(STVIZI+LI21+rLLIZI'
hVOIZI-STVIZI-TJLIZI
XI=l.
JZ
K=3
XMOLE- l B1IZI'STVIZI
b TO 4CC
3CO CALL RE INIT
IF IM-MMI j02,30 l,30 1
301 M= M"+MOLL
P<I NT 1 Z ,M, X, BI 11 ,S TV III ,811 I" hILI II ,HVDII' ,TEMP III ,Tf OPG I 11 ,
PRESSG,VIELI
JbZ XI=Z.
J=I
K-Z
CO rEMPK=ITEMPIJI+400 . I*C.;5;o
K(I=XP l- l904. /ll . 987*rEPKI+ 7 .Z9I'
kKl=ETAORKl
kKZ=XP 1-'0800. /11. 9 81.TfPKI+ZO. 9991
<K3=EXP I-ZI800 . /11. 98 7*TEMPKI+b. 3941
fl=XP 1-3.7377+(3. H1 0Z4/3bO.'*lrEMPIJI-q3Z. 11
40 1 tALL tHE M
I+ IP<tSSIKI-p"cXlr+2.510'.;00,lO
CU tJrl"UE
;40 lFIPRESIKJ-7 . ;-I". 71l,50 l,0
CObb
OObQ
COlO
0011
0012
0013
OJ14
COH
001b
0011
001.
0019
COSO
OiOI
0002
000)
0004
OOO
ODOb
0001
0008
C009
0010
0011
0012
0013
0014
Ol
uOlb
C017
Ool8
0019
C020
0021
022
"OZJ
u024
C02
002b
COZ1
OOlb
OJ29
03C
OOH
COll
0033
'034
003
013b
C037
C03S
0039
(040
(041
C042
,043
0001
0002
0003
C004
CJO
CUub
0001
Uo
(oeq
OOIC
COli
0012
FORTRAN IV Program and Sample Printout for Clzapter Z1J
L I_ITIAL PtURE TOO HIGH
U APE=APkt ~0. IPKESIK| -14. 5 . 0 l
Gj TO 40
;01 CAl REACT<IKCP" C .EdPRCE.CATPHC.TUSESI
70U CALL IITILIFRATE.TUBES.EB.STYI
GJ TO 20
>b l I TlE NXTL L /IO. 1 Jl.Ol.;Ob
>b AP Rt SS=APRESH.
GJ TQ bC8
;(1 "Pkt=APRt+I.
d PUT ll,M,X.fBtll,STYClJ,tH(1).rOlCl),rYOtl),TE"P(1),TE"PG(1),
I PKESG.YIELD
GU T O 40
VH luP
END
UoRUUIINE PRIDAT
o I ME NS IllN LbI 3 I STY l jI Sll 3 1 TOLl \I. H YD I 3 I 1EP 13 I TE PG l3 l PR ES 5l
131
tuMMUN TL N, 1 1U,TJu,f RAT E ,TEMP,TtMPG,CATDIA.CATDEN,ETA,EPS,H,CPEB.
lCPG,ATlfG.PExu,NSLICE,Sry,TOl,dl,Ed,PESStHYO,X"OLE,PESSG,y,,, ,
1,x,Xl,K,J,YlfLu,TO[L.1,Kl,K3,EKl,CONSTltCONSTZtCOST,CONST4,
3VJI,ARtA.ATMP,ATEPG
P"INT 4.TLEN
P"I'T 2.TUD
PHINT ),TID
P<tNT I.FRATE
PtINT 7,ATtMP
P41 NT /JA1VPG
P11 17.CATDIA
""INT Ib.CATUEN
PRINT 14.EU
PRINT 13.EPS
P{ I 1 19.H
PKIT 18.CHO
PINT 22.CPG
PRINT 20.ATEfG
P"INT . PREXG
P{INT 8."SLICE
"" INT 9
PRINT 10
PKI NT II
KE 'UR
fJKMATlbX.IIHf(EO RATE =.FI0.3.3x.25HLB ETHYL 8E NIE NE/HR/rU8E
fJKMATlbX.23hOUTSIDE TUBE DIAMETER =;11D+;3X+O1N l
FJMATlbX.22HINSIJE 1UE OI AETER =.FID.5.3x.3HIN I
UP*AT11H1+X+13HTU8 lEGTH =,F10.5,)X,)HFT )
> fkMATlbX.23HASSU"E EXIT PRESSUKE -+FD3X.5HPSIG I
1 f 3 kATlbX.2bI NI TIAL FEED TEPERATURE =.fIC.2.3X.IOHUEGREES f I
d rJKMATlbX.22HUBEK Of I NCEMENTS =.15.IH IIH IIH IIH IIH I
9 F,HAT I 3X,9H l NC ENT,/X,blENuTH,4X.5HLTHYL,4X,HSTYRENE.3XTHBEN
IltE.3X.7HTULUENE.2X.bHHYDUGEN.2X.SHKEACTION 2x.8HfLUE GAS.2X.SHP
2RESSUE.X.bHYIElD I
l )KMAT(5X.bHNUdER.5X,lHFT.5X.7HBENlENE.l'3x,8HETHYLENE,2X.7HTH4N
lE,14X,4HTEMP,2x,4X.4HTEMP.6X.4HS[ G,4X.14HlB MOLES STY IIH ,
II FJKMATI23X.H lB MOlE fOR"E PER L8 MOLE Of ETHYlBENIENE FED DE
I f. X.5HDE G F.IJx.IHlB MJLE EB RtACTED IIH IIH I
lb fJRMATlb 23HCATAlYST UlK DE NSITY= fIO.5.3X.9HlB/CU fT I
17 r JKMAT l b X.2bHCATALYS T PELLET DIAMETER =.fIOe 5.3X.3 rfT l
B JKMAT l bX.J2HHEAT CAPACITY Of ETHYL BENIENE =.fIO.5.3X.RHBTU/LB/fl
1 V JKMATlb'.3bHOVER-ALL EAI TASFER COEffiCIENT =.fIO.. )X.1 5HBTUI
lH7SU FT/f I
20 fJKMATlbX2CHfLUt GAS FLO" RATE =.fI0.3.)X.20HLR fLUE GAS/HR/TUBE I
21 f""MATlbXdIHASUIEJ EXIT TEMP iJF F LUE GAS =.FIO.3.3X.9HDEGREES FI
l2 JRMATlbX.21HHEAT CAPACITY uF f lUE GAS " .FIO . 5 s 3X.QHBTU/lB/f I
23 fUKMATlbX.29HVOIO fRACTIO Uf PACKED dED = .FID.5 .IH l
24 fjKMAT(bX.2JHEffECTIVENES ACTOR = fIO.2.IH I
ci+U
SuROUTI"E KtlNIT
U I ME NS I O LBI3 I 5TV I Jl.8111 1TUL I JJ H YO I 3 I 1L"PI 31 T EM PG 131 PR E S5 I
1;
CJMMUN TlE,T[U,TOO.fRAT[,TEMP,rEMPG,CATDJA,CATOEN,ETA.EPS.H,ePE6,
lCPfG,ATtFv,PRtXu,NSlICE,TV,TUL.Bl.E.PRESS.HYO.XOLE.PRESSG.Y..
21.x.xt ,K.J.YIEL0.TOEL,RK 1 .K2.kK3,Kl.COSTl.CONST2.CONST3.CONST4,
3VJIO.AREA
STYlll=TYI3I
lJLllI=W1I31
Bllll=BlIJl
T<MPIII=TEMPI))
TtMPGl1 I=TEMPGI 11
to111?FKLS11
Ed I I I = I . -I STY I II +I I I + T OLlII I
HY(III=IYIII-TJLllI
XMLltI.+SlIII+STYIII
Z1 Appendix
Ol)
0014
001
Olb
0017
0018
0019
OOZO
OOZI
0022
vOZ J
00Z4
OOZ
002b
0001
OOOl
0003
OJ04
OvO
OOOb
0007
COO
0009
0110
Gull
OOIZ
0013
COl4
0001
OOOZ
OOu)
0\04
COOS
COlb
0007
0008
0009
DOLO
DOLI
(OIZ
DOL)
OOH
OOI
OOlb
(017
0018
COl9
0021
COZI
0022
C023
(OZ4
VOZS
OOZb
0017
PtSSG=PRESSIII-14.7
X= S TYI 11 +BII l) +TOUl)
IfIX)j3.31.)52
31 YILO=O.O
Gu Tl 3S4
"l PINT 101
CALL bXIT
101 fJkMATIHNEG. SUMI
32 YIELD=STY(1)OIOO./X
34 Y=IIZ
M=I/Z
x=YOTOH
KdURN
U
SUoROUT I /E CHEM
01nN5 ION FI J1. SIYI I .B?I J1 lUll 31 H YO, 31 TEP I 31 r EMPG (31 PR ES S l
131
COMMUN llEN.TID,TUO.fRATE,TEP.TPG.CArOIA.CATOEN,fTA,EPS,H,ePEB,
lPG,KATEFG,PXG.NSlICE,STY.rO,dl.EB,PQESSfHYD,XMULE,PRESSG,y,M,
Ll,x,XJ,K.J,YIELU,TDEL,KK1,KK2,RK3,EK1,CONST1,cnNST2,CONST3.CONST4,
3VuIO.ARH
SIYIKI=SIYIII.rO(L/xl*cnNSTI'"KIOPIESS(J I /I14.70XOLE I oIEBIJI-STYI
J1 +HYO, J 10PI S S' J 1/1 E K 1* 14.70 'MO LE I I
blIKI=dllll+rOtL/IIOCUSTI'RKZ"PESSIJI/I14.70XMOLEloEBIJ I
1O11K 10 lull II +1tL7X l*CLJNST1"HK+1 PRE SS I J 1/ l 14. 70 XOLE I l""ZOEBI J 10
IHY! l J 1
TEMP I81 = TEMP l 11+ reEL! x I "COS TZ* I TEMPGI J I-TEM PI JI I -CONSH"I STY, K I -S
I TY I I I I
T.MPGIKI=TEHPGII I +TOEL/XI"CONST40ITEMPGIJI-TtMPIJ I I
A.=IEBIJ I '106.+STYIJ I 'IO.+HYOIJloZ.+TOLIJ I 0108.+BlIJ10106. I /XHOL
IE
OtoPRtSSIJI/14.1'AMWO.ObZ4/10.99"0.OB20b'ITEMPIJI+460.100.55561
VcLoFAIE/IOEN03000.0AEAI
PESSIK I =PESSII I -TOEl/XI'0.000215S401.7S/CATOIAoDENOVELOOZ_VOID
KEIU"N
cNL
SOUTINE REACTIRCPRCE.tdPR CE.CATPR C.TUBESI
UIMENSION E8131.STYI31.8LI31.TOLI31.HYO I31.TEPI31.TEMPGC)I.PRESSI
131
CJHO TlE.TIO.TOO.FR.TE.TM.TEPG.CATOIA.CATOEN.ETA.EPS.H.CPEa.
lCPfG.RATtfG.PREXG.Sl ICE.STY.TU.dl.E8.PRESS.HYO.Xl E.PRESSG.Y..
ll.X.XI.K.J.YIEl.TOl.KI.RK2.K).EKI.COS'I.CONS'2.COST3.CONS'.
3VI O. AREA
TUDES-l.0S*20000000.*IJb./lfRATE*STYlll*IO.*dOOO.1
PIT IS.TUBES
Kccosr-RCPKCE*20UOOOUO.*I.OS*I06./II 04.*YIElDI*EBIII/CI.-ERIIII
RCtOST-R CCOST*IO.
PINT Z5.RCPR tE.RCCOST
YLCOSTERPRC*200000UO.*1.05*106./IO.*ll00.-YIElOI/YIElO
RI' Z6.EBPRCE.YlCOST
TUTCST-YlCOS'.RCCOST
PKIT 27.TUTCST
A-3.1.19*'O/12.*TLE*TU8ES
PIT 28.A
RC TCST-2]u.*A**O.51
PK1J 29.RCTtST
tATCST3.14159*TID**2/516.*TLE*TUBES*CATOE*CATPRC
PINT lo.CArPRt.CATCST
REJUR
15 fJKMATIIH 18Xl1HNUMiR Of TU8 S -.FIO.3.IH I
Z5 f]KATI8X4ZHANNUAL COST Of RECYCLING ETHYL 8ElEE AT .F5.3.6HS/LB
I =. fl0.C.3x.5HS/YR I
= URMATldX.HHANNUAl CUST OF ETHYL BENlEfI E CONVERTED TO BY-PROOUTS
1 AJ ,FS.],CH$/l8 =+?1U&U+X+^H57YR l
T FJRMAT 18X29HTOTAl ANUAL OPEHATIG COST =.FI0.0.3x.H/YR I
fJKA' 18X25HSUFAC AEA UF tACIOR =.FIO.l.)X. 6 HSQ FT I
ZS fURAT 18X36HUST OF TO "EACTURS OF ABOVE SllE =.FIO.Z.3X.2H$ I
JU fJRMAT IbXI9HCOST Of L ATALYST AT.F.Z.HHLB =.FIO.3.3X.ZHI. IIH II
lH I
O
COOl
(002
l003
0004
(iO
000 .
0007
COOS
(009
0010
0011
,012
GOD
0014
001;
0016
0017
OOU
1019
C020
0021
0022
C023
C024
002;
0020
C027
1028
002<
(03C
lu31
0032
COB
(034
103;
003h
(on
0038
"u3
Ci4C
(041
(0'2
(043
CO'4
C04
1040
0047
0048
0049
OOO
00;1
00;2
OO!
(05 4
00;0
00;7
(0;8
UOS9
COoO
0001
0002
L Y Program 8O Sample Printout for Chapter
SUoUUTINe OISTILCFRATe.TUaS.eS.STV I
PMGKAH FOR STVNe-ETHVL8eNlENE VACUUM OISTILLATION
DIMENSION RRCNCbl,. FTCbl,. XC;OOI. VCSOOI.ANTCbOI. COSTCbl1
DIMENSCON CWCOSCbl'. STCUSCbll. CPCOSlbl1
IMENSION STVC31.E8131
F0FRATETUAeS.8000.J
fFED/IOo.*IE8ClloST VCIII
AI PHA-l.37
Xb=O.OI
X=F8Cll/CE8111+STYCIII
&=2COOOOOO./IO'. /CI.-XSI
IF I ALPHA-I.CI 9'9.999.;01
SOl X=IF*XF-B*XSI/CF-81
O-f-B
YCET-ALPHA*XF/CI.O+ I ALPHA-I.OI*XFI
A=CXDYCEPTI/CXO-XFI
KiMIN=AICI.O-AI
KCNC 11= I.C5
DJ bOO N-I,OO
kK=RRH[N*RRCNI1
USlO P=RR/CRR+I.OI
uCEPT-XD/CRRol.OI
UCtPT=USlOP*XF+UCEPT
SL"Pl=I CEPT-XbI/I XF-X81
CtPTL=XS*CI.G-Sl OPll
TNT-O.O
VC lI-XO
U 5 99 1-2.5CO
X I I-II=YII-ll/IALPHA-I ALPHA-l.OI*V I I-111
IF exc I-il-XBI 502.502.503
C3 IF CXCI-il-XFI 504.50'.505
Ub YCII'USLuP*XCI-II+UCePT
UJ TO 9'
5(4 IF CXCI-21-XfI 5C6.5(b.07
>UJ NFT-CYNT+l.0IIO.7+1.0
HCNINFT
5Co YII,=SLOPL*XCI-II+CEPTl
59 nr-TNT+ 1.0
02 =k*D
V= 10"+01/800C.0
NT=TNT/0.7+1.(
A.H l NI=NT
TIIX.8V"0.5
Nr=&N111J??.C+1.
ANt-NC
PLnINI=12.'.CTIOXI*1. 21*ANTI NI
STCOSCNI-V*ANC*113.8
C"tOCNI=V*ACO.O'
IJTCNICPCUSCNI+CWCUSCNI+STCOSCNI
000 KACNCN+II=RRCNCNloO.C5
P"INT 101
I^1 HlkHATI6X.;'HO ISTILLATION CALCULATIO NS FO R REACTOR oeSCPIBEO ABnVE
IIH IlH I
PIT l02.F.Xf.XD.RRM'N
PINT le3
le2 F)RMAT 112H FELD ATE= .F9.C.22H EA FRACTION IN FfEO= .FS.3.27H E B
1 FkACTIUN IN 0IsrILLATE-.FS.3.12H MINIMU R= Fb.3111
103 FJkATI 8X10H KR FACTnR.13x5HTRAYS.9X9HFEEO TdAY.9X9HOEPR COST.9X9H
IHFAT LUST+SA9HLUUL COS T.9X9HVEA CUST/II
UU 104 N=l,bC
10' PklNT 105. RRCNCNI.ANT1NI.FTI NI.CPCOS1N).STCUSCNI.CwCOSCNI.COSTCNI
IO FJHATC7FI8.2/11
G" 1U 5ec
999 STuP
SUO "E TURN
ENO
Z1
T
U
B
E

L
E
N
G
T
H

=

1
0
.
0
0
0
U
O

F
T

O
U
T
S
I
D
E

T
U
S
f

1
1
A

H
c

n
R

=

1
.
3
1
5
0
0

I

S
I
I
E

T
U
d
E

u
I
A
H
L
T

=

1
.
0
4
9
0
0

I
N

f

E
U

R
A
T

2
8
.
7
9
b

L
B

E
T
H
Y
L

B
E
N
l
E

E
/
H

/
T
U
S
E

I
N
I
T
I
A
L

f
E
t

n
M
P
L
k
A
T
U
R
E
=

9
3
2
.
0
0

O

F
E
S

F

A
S
S
U
M
E
D

b
4
l

T
E
M
P

O
F

f
L
U
L

G
A
S

=

1
6
0
0
.
0
0
0

D
E
G
R
E
E
S

F

C
A
T
A
L
Y
S
T

P
t
L
L
E
T

D
I
A
M
e
T
e
R

=

0
.
0
1
0
4
0

F
T

C
A
T
A
L
Y
S
I

d
U
L
K

D
E
N
S
I
T
Y
'

6
1
.
0
0
0

0

L
B
/
C
U

F
T

E
f
F
E
C
T
I
V
E
N
E
S
)

F
A
C
T
O
R

=

0
.
4
4

V
U
I
D

K
A
C
T
I
U
N

U
F

P
A
C
K
e
D

B
E
D

=

0
.
3
8
0
0
0

O
V
t

-
A
L
l

H
E
A
T

I
R
A

S
F
E
R

L
U
E
I
F
I
C
l
E

T
=

8
.
0
0
0
0
0

B
I
U
/
H
K
/
S
Q

F
T
/
F

H
E
A
T

C
A
P
A
C
I
T
Y

U
I

E
T
H
Y
L

B
t
N

E

=

O
.
6
3
5
0
C

B
T
U
/
L
8
/
F

H
E
A
l

C
A
P
A
C
I
T
Y

U

F
L
U
E

G
A
S

=

0
.
3
0
0
0
0

B
T
U
/
L

/
F

F
L
U
E

G
A
S

F
L

R
A
T
E
'

5
5
6
.
7
b
7

L
8

F
L
U
E

G
A
S
/
H
R
/
T
U
8
E

A
S
S
U
M
t
U

t
X
1
1

P

t
S
S
J
K
E

=

5
.
C
O
O
O
O

P
S
I
G

N
U
H

c
K

U
f

l
N
C
K
E
M
L
N
f
S
-

1
0
0

I
N
C
R
c
M
E
N
T

L
E
N
u
T
H

L
T
H
Y
L

N
U
M
d
E
o

F
T

b
L
N

L
N
L

S
T
Y
R
c
N
E

b
c
N

E
T
H
Y
l
E
N
E

T
O
L
U
E
N
E

M
L
T
H
A
l
E

H
Y
U

O
G
E
N

L
8

M
U
L
E
S

r
U

P
L
J

P
E
R

L
B

"
U
L
E

O
f

E
T
H
Y
L
8
E
N
l
E

E

F
E
D

0

0
.
0

I
.
O
O
O
O
C

C
.
O

0
.
0

0
.
0

0
.
0

1
0

1
.
J
v
l

u
.
9
8
0
7
C

C
.
O
I
'
l
l

0
.
0
0
0
0
6

0
.
0
0
0
1
l

0
.
0
1

0
0

2
0

l
.

O
O

0
.
9

O
0
7

0
.
0
4
d

7
1

0
.
0
0
0
4
1

0
.
0
0
0
1
5

0
.
0
4
8
0
?

3
0

3
.
0
0

0
.
9
0

d
3

0
.
0
8
&
5
2

0
.
0
0
1
2
5

0
.
0
0
2
4
0

0
.
0

4
1
3

4
0

4
.
1
0
0

u
.
b
o

0
.
1
2
8
"
4

0
.
0
0
2
8
8

0
.
0
0
5
1
1

0
.
1
2
1
6
3

5
0

5
.
0
(
(

0
.
8
1
1
9
0

0
.
1
7
3
2
0

0
.
0
0
5

1

0
.
0
0
9
4
0

0
.
1
6
3
8
0

6
0

&
.
(
0
0

O
.

7

9
0
5

0
.
2
1
7
4
1

0
.
0
0
9
2
5

0
.
0
1
4
2
9

0
.
2
0
H
2

7
0

7
.
U
O
O

0
.
7
0
&
1
5

0
.
2
&
.
2
5

0
.
0
1
4
1
3

0
.
0
1
9
4
6

0
.
2
4
0
7
9

8
0

8
.

v
O

;
.
6
5

0
0

0
.
3
0
0

9

0
.
0
2
0
0
6

0
.
J
2
4
3
5

0
.
2
7
6
2
4

9
0

9
e
0
0

.
6
U
7
b
2

0
.
3
3
7
0
6

O
.
C
2
6
1

o
.
n
a

8

0
.
J
O
d
6
8

9

.
9
0
0

J
.

1
8
1

0
.
3
6
4
,
2

0
.
0
3
2
8
7

0
.
0
3
0
d
O

0
.
3
H
l
l

0

0
.
"

1
.
0
0
.
0
0

0
.
0

0
.
0

0
.
0

0
.
0

1
0

1
.
0
0
C

w
.

9
d
U
3
1

0
.
(
1
9
4
3

0
.
0
0
0
0
8

0
.
0
0
0
1
2

0
.
0
1
9
3
1

2
0

2
.
0
0
1

0
.

4
9
3
1

0
.
0
4
9
4
'

0
.
0
0
0
4
2

0
.
0
0
0
7
9

0
.
0
4
H
O

3
(

)
.
U
I
O

0
.
9
0
8
5
9

0
.
0
8
7
b
3

0
.
0
0
1
2
6

0
.
0
0
2
5
2

0
.
0
8
5
1
1

4
0

4
.
u
J
O

J
.
6
b
1
1
b

0
.
1
3
0
3
6

o

.
O
C
2

O

0
.
0
0
5
5
1

0
.
1
2
4
7
9

5
0

e
0
0
u

J
.
8
0
9
6
8

0
.
1
7

9
0

0
.
0
0
5
5
5

0
.
0
0
9
8
1

0
.
1
6

0
3

1
0

o
.
O
v
O

0
.
7
5
o
2
2

0
.
2
l

1

0
.
0
0
'
3
3

0
.
0
1
5
0
4

0
.
2
0
4
3
7

7
0

1
.
0
u
O

v
.
7
0
2
5
2

0
.
2
6
2
b
4

0
.
0
1
4
2
8

0
.
0
2
0
5
6

0
.
2
4
2
0
8

8
0

8
.
t
O
(

u
.
6

0
2
8

0
.
3
0
3
5
5

0
.
0
2
0
3
2

0
.
0
2
5
8
5

0
.
2
7
1
1
0

9
0

0
u

0
.
6
0
1
4
9

0
.
3
4
0
9
5

0
.
0
2
7
2
1

0
.
0
3
0
3
4

0
.
3
1
0
6
1

1
0
0

1
0
.
O
L
O

0
.
5
5
'
2
4

0
.
H
l
d
1

0
.
0
3
4
4
0

0
.
0
3
3
5
0

0
.
\
3
9
3
7

R
E
A
C
T
I
O
N

f
l
U
E

G
A
S

T
E

P

T
E
M
P

D
E
G

f

D
E
G

F

9
3
2
.
0
0

1
6
0
0
.
0
0

1
0
1
1
.
9
3

1
6
1
0
.
4
2

1
0
1
4
.
2
8

1
6
1
9
.
8
2

1
1
2
1
.
2
1

1
6
2
8
.
4
7

1
1
6
3
.
0
5

1
6
3
6
.
5
3

1
1
9
1
.
0
8

1
6
4
4
.
1
1

1
2
2
7
.
H

1
6
5
1
.
7
9

1
2
5
5
.
5
3

1
6
5
8
.
1
0

1
2
8
2
.
4
9

1
6
6
4
.
5
7

1
3
0
9
.
4
4

1
6
7
0
.
1
9

1
3
3
5
0
1
8

1
6
7
5
.
9
1

9
3
2
.
0
0

1
1
0
0
.
0
0

1
0
1
1
.
1
9

1
6
1
0
.
4
2

1
0
7
3
.
1
8

1
6
1
9
.
8
3

1
1
2
2
.
5
0

1
6
2
8
.
4
9

1
1
6
2
.
2
1

1
1
3
6
.
5
1

1
1
9
6
.
1
5

1
1
4
4
.
1
6

1
2
2
b
.
4
2

1
6
5
1
.
3
5

1
2
5
4
.
3
9

1
6
5
8
.
1
8

1
2
8
1
.
1
0

1
6
1
4
.
6
7

1
3
0
7
.
5
1

1
6
7
0
.
8
3

1
3
3
5
.
2
5

1
6
7
1
.
6
4

P
R
E
S
S
J
R
E

P
S
[
G

4
1
.
2
8
7

3
9
.
3
6
5

3
7
.
2
2
0

3
4
.
8
1
9

3
2
.
1
1
7

2
9
.
0
5
8

2
5
.
5
5
6

2
1
.
4
8
5

1
6
.
6
2
5

1
0
.
5
2
9

2
.
9
8
3

4
2
.
2
8
1

4
0
.
4
0
0

3
8
.
2
9
6

3
5
.
9
4
5

3
3
.
3
0
5

3
0
.
3
2
3

2
1
.
9
2
5

2
2
.
9
9
7

1
8
.
3
5
2

1
2
.
6
3
3

4
.
9
3
5

L
0

L
t

T
Y

L
I

R

E
I

R
E
A
T
E
D

0
.
0

9
8
.
9
8
9

9
7
.
6
1
4

9
5
.
9
5
3

9
4
.
0
2
7

9
2
.
0
7
7

9
0
.
2
3
1

8
8
.
5
6
7

8
7
.
1
2
8

8
5
.
9
H

8
5
.
1
3
0

0
.
0

9
8
.
9
7
0

9
7
.
1
1
3

9
5
.
8
6
4

9
3
.
8
9
5

9
1
.
8
9
8

9
0
.
0
0
3

8
8
.
2
8
8

8
6
.
7
9
1

8
5
.
5
5
1

8
4
.
5
9
6

.

N
U
K
b

U
I

T
U
b
L

2
4
9
.
1
8
1

A

N
U
A
L

L
U

U
r

L
L
Y
L
L
l

L
1
H

b
L
N

b
N
c

A
T

0
.
0
l
0
f
f
L

3
2
1
0
1
8
.

$

A
N
N
U
A
L

C
U
>

U
F

c
H
Y
L

L
N
L

C
O
N
V
E
R
T
E
D

T
e

b

F
R
U
0
U
C
T
$

A
T

0
.
l

0
$
L

=

4
6
7
6
7
3
.

H
Y
R

I
U

A
L

A
N
N
U
A
L

u
P

R
A
T
I
N
G

C
U
$

=

7
8
d
6
Q
2
.

$

S
U

f
A
C
E

A
R

A

U
F

c
A
C
T
U
R

=

8
5
1
.
0
4
4

S
O

F
T

C
U
>

U
F

1
W
U

K
L
4

U
K
S

U
F

A
b
U
V
c

S
I
l
E

=

1
0
9
9
6
0
.
0
0

C
U
S

U
F

L
A
T
A
L
V
,
'

A
T

3
.
0
0
$
/
L
B

?

2
7
3
0
.
7
9
8

$

D
I
S
J
I
L
L
A
J
I
O
N

t
A
L
t
U
L
A
J
I
O
N
S

f
O
R

R
E
A
t
J
D
R

D
E
S
C
R
I
B
E
D

A
8
0
V
E

;

f
E
E
D

P
5
4
7
1
.

E
_

F
R
A
t
J
I
O
N

I
N

F
E
E
D
0

0
0

E
B

F
R
A
t
J
I
D
N

I
N

0
I
S
J
I
L
L
A
f
E
-
0
.
9
.
9

I
N
I
"
U
"

R
-
"
.
0
7
9

R
R

f
A
C
l
L
i

T
R
A
Y
S

F
E
E
D

f
R
A
Y

D
E
P
R

C
O
S
T

H
U
T

L
U

L
U
J
L

L
U
5
1

T
t

5
I

1
.
0
5

9

0
0

0
0

1
1
2

7
0
0
0
1
.
1
9

3
7
1
5
.
3
5

1
2
0
0
2
9
.
1
9

1
.
1
0

8
5
.
0
0

3
9
.
0
0

8
.
6
2

7
2
7
0
3
.
4


3
8
5
8
.
7
8

1
1
8
5
1
0
.
8
1

<

7
8
.
0
0

3
5
.
0
0

3
9
3
.
6
.
2
.

7
S
.
0
S
.
U

4
0
0
2
.
2
1

1
t
8
7
5
4
.
2
5

1
.
2
.

2
.
0
0

3
Z
.
0
0

3
7
0
5
.
0
5

7
8
1
0
8
.
1
2

4
1
4
5
.
6
3

1
1
9
3
4
8
.
7
5

1
.
2
5

.
9
.
C
O

$
1
.
0
0

3
U
8
Z
.
3
2

8
0
8
1
0
.
5
0

4
2
8
9
.
0
6

1
2
1
3
8
1
.
M

~
.

1
.
3
0

2
9
.
0
0

3
5
1
9
6
.
5
7

8
3
5
1
2
.
1
1

4
"
3
2
.
4
9

1
2
3
3
H
.
8
7

I
.
J
5

u
.
O
O

2
B
.
0
0

3
4
U
9
.
U

5
7
.
7
6
.
7
7

3
0
5
0
.
6
1

9
4
9
6
6
.
8
1

'

1
.
4
0

.
2
.
1
0

2
6
.
0
0

3
U
U
.
2
6

5
9
2
7
8
.
2
9

3
1
4
b
.
2
J

9
6
9
5
0
.
7
5

6
1
.
0
0

2
6
.
0
0

3
.

.
1
1

.
1
0
7
9
.
1

3
2
4
1
.
8
5

9
8
9
0
6
.
8
1

{

5
9
.
0
C

2
5
.
0
0

3
4
0
3
9
.

6
2
8
8
1
.
3
8

3
3
3
7
.
4
1

1

0
0
Z
5
8
.
5
6

1
.
5
5

S
8
.
C
O

2
5
.
0
0

3
4
0
'

7
.

.
.
6
B
2
.
9
6

3
4
3
3
.
0
9

1
0
2
1
5
0
.
7
5

l
.
b
J

5
<
.
C
O

2
3
.
0
0

l
.
b
5

5
5
.
C
O

2
3
.
0
0

1
.
1

5
5
.
C
O

2
3
.
0
0

1
.
1
>

5
3
.
V
O

2
2
.
0
0

1
.
8
0

5
3
.
0
0

2
2
.
0
0

1
.
8
5

5
2
.
C
O

2
2
.
0
0

1
.
9
J

5
2
.
J
0

2
2
.
0
0

)
1
.
C
O

2
2
.
0
0

-
v
J

.
C
O

2
1
.
0
0

2
.
0
)

5
1
.
0
0

2
1
.
0
0

T
U
B
I

L
E
N

T
H

=

b
.
0
0
0

F
T

O
U
T
,
I
U
E

T
U

E

U
I
A
M
I
T
E
R
.

4
.
5
0
0
0
0

I
N
,
I
U
E

T
U
d
e

U
I
A
M
E
T
E
R
'

4
.
0
2
6
0
0

I
N

f
L
0

A
I
E
'

4
2
4
.
1
5
.

L
a

t
T
H
Y
L

B
E
N
l
E
"
E
I
H
o
I
T
U
B
E

l
N
l
1
I
A
L

F
E
b
)

T
O
W
l
iU
T
U
R
E
.

1
0
2
2
.
0
0

D
E
G
R
E
E
S

F

A
>
S
U
M
I
D

E
X
I
T

c
H
P

U
F

F
L
U
E

G
A
S
'

L
m
0
o
0
0
0

D
E
G
R
E
E
S

F

C
A
T
A
L
Y
S
T

P
I
L
l
t
r

U
l
l
M
E
T
E
R
.

0
.
0
1
0
4
0

f
f

4
L
Y
S

U
L

U
I
N
,
I
T
Y
.

.
1
.
0
0
0
G
O

L
d
/
C
U

f
T

E
F
F
E
C
T
I

E
N
I
5
.

F
A
C
T
O
K

0
.
4
4

V
O
l
o

F
R
A
C
T
l
O
"

O
F

P
l
C
K
E
D

B
E
D
.

0
.
)
8
0
0
0

3
3
4
0
1
.
2
4

3
H
4
1
.
2
1

3
3
8
6
6
.
3
3

H
I
3
5
.
5
8

3
3
6
3
1
.
3
1

3
3
'

8
.
4
4

3
3
9
5
5
.
5
5

3
3
7
6
6
.
1
3

3
4
2
2
5
.
4
6

3
4
6
8
0
.
1
8

O
V
E

-
A
L
L

t
l
A
T

T
k
A

S
f
E
R

C
O
E
F
F
I
C
I
E
N
T
'

8
.
0
0
0
0
0

B
T
U
l
l
/
S
O

f
T
l
f

H
l
A
T

C
A
P
A
C
I
T
Y

O
F

"
T
H
Y
l

B
E
N
Z
E
N
E
'

0

3
5
0
0

B
T
U
I
L
B
l
f

H
I
A
T

C
A
P
A
C
I
T
Y

U
l

F
L
U
k

G
A
S
.

0
.
3
0
0
0
0

B
T
U
I
L
B
I
F

F
L
U
E

G
A
S

F
L
U
W

t
A
l
t
.

1
5
2
0
.
0
0
0

L
I

f
l
U
E

G
U
/
H
R
I
T
U
I
E

A
>
S
U
M
I
U

e
X
I

T

P
R
t
S

U
R
E
'

5
.
0
0
0
0

P
S
I
G

N
U
M
B
E
k

l
N
L

c
H
c
N

1
0
0

6
6
4
8
4
.
5
0

3
5
2
8
.
7
0

1
0
J
4
2
0
.
4
4

6
8
2
8
6
.
0
6

3
6
2
4
.
3
2

1
0
5
2
5
1
.
6
3

7
0
0
8
7
.
5
6

3
7
1
9
.
9
4

1
0
7
6
B
.
8
1

1
1
8
8
9
.
1
9

3
8
1
5
.
5
6

1
0
8
8
h
0
.
3
l

7
3
6
9
0
.
6
9

3
9
1
1
.
1
8

1
1
1
2
3
3
.
1
9

)
.

7
5
4
9
2
.
2
5

4
0
0
6
.
B
O

1
1
2
9
7
7
.
4
4

7
7
2
9
3
.
8
1

U
0
2
.
4
1

1
1
5
3
5
1
.
7
5

7
9
0
9
5
.
3
1

4
1
9
B
.
0
3

1
1
1
0
5
9
.
5
0

8
0
8
9
6
.
9
4

4
2
9
3
.
6
5

1
1
9
4
1
6
.
0
0

8
2
6
9
8
.
5
0

4
3
8
9
.
2
1

1
2
1
7
6
8
.
5
)

I

N
C
R
E
I
E
N
T

l
E
N
v
T
H

t
T
H
Y
L

'
T
Y
R
E
N
E

8
E
N
Z
E
N
E

T
O
L
U
N
E

H
Y
D
R
O
E
N

R
U
C
T
I
O

F
L
U
E

G
A

P
R
E
S
S
U
R
E

Y
l
E
L
'

N
M
8
E
R

F
T

8
b
N

E
N

E
T
H
Y
L
E
N
E

O
U
H
A
N
E

T
E
I

T
E
M
P

P
S
I
G

L
B

K
0
L
L
5

5
T
Y

L
8

M
L
c

F
O
R
M
E
D

P
b
K

L
0

M
O
E

D

E
T
H
Y
L
B
N
I
E
N
E

F
E
D

D
E
G

?

D
E
G

F

L
B

P
D
L
E

R
E
A
C
T
E
D

1
.
0
0
0
0
0

0
.
0

0
.
0

0
.
0

0
.
0

1
0
2
2
.
0
0

1
6
0
0
.
0
C

5
3
.
6
6
4

J
-
0

@

1
0

1
.
5
0
C

:
.
U
U
O

0
.
0
6
3
0

O
.
O
O
O
)
!

0
.
0
0
0
8
5

0
.
0
.
s
4
5

1
0
1
5
.
7
9

1
6
0
4
.
2
2

5
1
.
1
9
0

9
7
.
4
6
9

2
0

0
.
9
1
7
4
)

0
.
0
7
9
2
0

0
.
0
0
0
6
5

0
.
0
0
2
7
1

0
.
0
7
6
4
9

1
0
2
0
.
5
1

1
6
0
8
.
4
1

4
8
.
5
1
9

V
b
.
9
2
1

3
0

4
.
:
O
U

0
.
8
9
0
4
6

0

9
9

0
.
0
0
0
9
8

0
.
0
0
5
1
2

0
.
0
9
8
3
3

1
0
3
1
.
8
9

1
6
1
2
.
7
0

4
5
.
b
3
6

9
4
.
4
3
2

4
0

6
.
,
r

:
.
8
6
7
)
2

0
.
1
2
3
4
1

O
.
O
O
l
U

0
.
C
0
7
9
0

0
.
1
1
5
5
1

1
0
4
6
.
2
2

1
6
1
6
.
1
6

4
2
.
5
1
8

V
J
e
0

5
0

7
.
5
0
0

,
.
8
4
n
4

0
.
1
4
1
6
5

0
.
0
0
1
1
5

0
.
0
1
0
9
6

0
.
1
)
0
6
9

1
0
6
1
.
3

1
6
2
0
.
9
5

1
9
.
1
2
6

V
l
e

6
0

9
.
0
U
U

0
.
8
2
3
9
9

0
.
1
5
9
4
0

0
.
0
0
2
4
3

0
.
0
1
4
1
8

0
.
1
4
5
2
2

1
0
7
6
.
3
0

1
6
2
6
.
9
5

J
b
e
9
0

9
0
.
5
6
3

7
0

1
,
.
5
0
(

1
.
8
0
2
2
7

0
.
1
7
7
1
9

0
.
0
0
3
1
1

0
.
1
5
9
7
6

1
0
9
0
.
6
5

1
6
2
8
.
8
8

3
1
.
2
8
1

8
9
.
b
1
1

8
0

1
2
.
0
0

0
a
f
0

0
.
1
9
5
1
9

0
.
0
0
3
8
9

0
.
0
2
0
5
4

0
.
1
7
4
6
5

1
1
0
6
.
3

1
6
3
2
.
7
3

2
6
.
6
2
5

8
8
.
8
1
6

9
0

U
.
:
U

0
.
n
U
8

.
2
L
J
6

0
.
0
0
7
4

0
.
0
2
3
)
)

0
.
1
9
0
0
1

1
1
1
7
.
3
4

l
0
6
.
5
1

2
1
.
2
3
b

8
8
.
3
1
2

1
0
0

L

0
.
7
)
7
4
4

O
.
2
J

"

0
.
0
0
5
6
1

0
.
0
2
5
6
0

0
.
2
0
5
7
4

l
U
O
.
0
2

1
6
6
0
.
Z

1
4
.
7
0
9

8
8
.
1
1
2

0

"
.
0

1
.
0
0
0
0

0
.
0

0
.
0

0
.
0

0
.
0

1
0
2
2
.
0
0

1
6
0
0
.
0
0

4
8
.
9
9
1

0
e
0

1
0

1
.
5
0
0

0
.
9
:
:
0
8

0
.
O
U
8
8

0
.
0
0
0
7
1

0
.
0
9
J

8

1
0
1

.
6
6

1
6
0
4
.
2
1

4
6
-
J
3
2

V
1
e

J
o

0
.
9
2
0
.
1

C
.
0
7
6
4
9

0
.
0
0
0
6
3

0
.
0
0
2
2
7

0
.
0
7
4
2
2

1
0
2
2
.
5
0

1
6
0
8
.
4
5

4
3
.
4
4
1

9
6
.
3
4
9

K

4
.
5
0
0

0
.
8
9
U
9

0
.
1
0
U
'

0
.
0
0
0
9
4

0
.
C
0
4
2
9

0
.
0
9
7
0
9

1
0
)
)
.
2
8

1
6
1
2
.
6
7

4
0
.
2
9

V
>
e
0
V
9

;

6
0

6
.
0
0
0

0
.
8
6
9
9
7

0
.
1
2
2
U

O
.
O
O
l
n

0
.
0
0
6
6
0

0
.
1
1
5
!
3

1
0
6
6
.
9
3

1
6
1
6
.
1
l

3
6
.
8
6
0

9
3
.
9
2
1

5
0

7
.
5
0
0

0
.
8
4
8
2
1

0
.
1
4
0
9
8

0
.
0
0
1
7
4

0
.
0
0
9
0
8

0
.
1
)
1
9
0

1
0
6
1
.
5
5

1
6
2
0
.
9
0

J
-
J

9
2
.
8
1
4

K

6
0

9
.
L
O
O

0
.
8
2
1
0
3

0
.
1
5
9
1
2

0
.
0
0
2
2
5

0
.
0
1
1
6
1

0
.
1
4
7
5
1

1
0
7
6
.
1
8

1
6
2
4
.
9
0

2
8
.
8
3
9

9
1
.
9
8
8

7
0

1
0
.

S
O
l

0
.
8
0
0
7

0
.
1
7
1
0
6

0
.
0
0
2
1
4

0
.
0
1
6
0
4

0
.
1
6
3
0
1

1
0
9
0
.
6
2

1
6
2
'
.
1
3

2
4
.
0
2
1

9
1
.
2
9
7

R

8
0

1
2
.
0
0
0

0
.
7
1
5
4
4

0
.
1
9
4
.
8

0
.
0
0
3
4
8

0
.
0
1
6
2
1

0
.
1
7
1
"

1
1
0
6
.
2
3

1
6
3
2
.
6
8

1
8
.
3
6
2

9
0
.
8
2
6

9
0

L
e

1
.
7
0
5
7
3

0
.
2
1
2
2
2

0
.
0
1
7
9
1

0
.
1
9
4
)
1

1
1
1
7
.
8
9

1
6
3
6
.
6
7

1
1
.
]
2
0

9
0
e
b
V

9
8

1
4
.
7
0
0

"
.
7
:
1
7
0

0
.
2
2
4
9
0

0
.
0
0
4
6
1

0
.
0
1
8
7
9

0
.
l
0
6
1
1

1
1
2
9
.
6

1
6
3
9
.
6
6

]
.

6
6

9
l
.
5
1
3

;

0

0
.
0

1
.
0
0
0
0
0

0
.
0

0
.
0

0
.
0

0
.
0

1
0
2
2
.
0
0

1
6
0
0
.
0
0

4
'
.
9
9
]

J
e
0

1
0

1
.
s
t
l

0
.
9
l
U
l

0
.
0
9
4
9
1

0
.
0
0
0
3
6

0
.
0
0
0
7
.

0
.
0
.
3
6
8

1
0
1
7
.
2
5

1
6
0
6
.
2
1

4
1
.
3
7
4

9
7
.
6
4
2

2
0

3
.
0
1
0

1
.
9
1
9
9
0

0
.
0
7
7
1
1

0
.
0
0
0
1
3

0
.
0
0
2
1
6

0
.
0
7

.

7
5

1
0
2
2
.
0
.

1
6
0
8
.
6
6

4
4
.
5
1
4

9
6
.
2
b
O

1
0

4
.
5
1
0

0
.
8
9
2
7
1

0
.
1
0
1
8
8

0
.
0
0
0
9
5

0
.
0
0
.
.
6

0
.
0
9
7
6
2

1
0
3
2
.
9
"

1
6
1
2
.
6
7

l
e
9
b
0

9
4
.
9
5
6

6
0

1
0
0

b
Y
J
9

0
.
1
2
2
"
7

0
.
0
0
U
2

0
.
0
0
8
7

0
.
1
1
5
6

1
0

7
2

1
6
1
6
.
8
3

3
8
.
0
8
7

V
3
.
7
3
3

~
.

5
0

7
.
:
0
0

0
.
'
.
"
7

0
.
1
"
1
2
0

0
.

0
l
I
6

0
.
0
0
9
6
7

0
.
1
)
1
7
2

1
0
6
1
.
6

1
6
2
0
.
9
1

3
4
.
3
9
2

9
2
.
b
3
0

6
0

9
.
1
1
0

I
.
U
U
O

0
.
1
5
9
2
6

0
.
0
0
2
2
9

0
.
0
1
2
1
5

0
.
1

.
7
1
1

1
0
7
6
.
1
5

1
6
2
6
.
9
1

3
0
.
2
8
4

9
1
.
6
8
9

~

7
0

1
,
.

0
.
8
0
!
1
6

0
.
1
7
7
1
9

0
.
0
0
2
8
9

0
.
0
1
6
7
5

0
.
1
6
2
.
.
1
0
9
0
.
6
0

1
6
2
8
.
8
6

2
5
.

3
9

9
0
.
9
4
2

8
0

1
2
.
.
0
0

0
.
7
1
4
2
2

0
.
1
9
l
1
0

0
.
0
0
J
S
7

0
.
0
1
7
1
2

0
.
1
7
7
9
.

1
1
0

.
.

1
.

1
6
3
2
.
7
0

2
0
.
2
4
2

'
0
.
4
1
4

~

9
0

U
.
:
O

0
.
7
.
3
'
!

1
.
2
1
2
7
J

6
.
0
0
.
2
.

0
.
0
1
9
0
S

0
.
1
9
3
6
8

1
1
1
7
.
5
9

1
6
3
6
.
.
.

1
3
.
6
6
8

V
0
e

1
0
0

U
.
O
O
l

U
e

9
l
.
2

0
.
1
2
9
1
2

0
.
0
0
.
9
2

0
.
0
2
0
]
6

0
.
2
0
8
7
8

1
1
3
1
.
5
1

1
6
6
0
.
1
9

4
.
1
6
5

V
0
e
0
1
0

N
U
N
8
E
M

T
U
t
t
i

2
7
.
5
]
1

A
N
N
U
A
L

c
u
S
T

O
F

K
L
L
Y
C
L

I
N
G

E
T
H
Y
L

.
E
N
l
t
.
1
E

A
T

0
.
0
1
0
"
L
8

.
9
.
5
5
)
.

f
T
P

A
N
N
A
L

C
O
S
T

U
f

c
f
H
T
L

'

N
Z
t
N
t

C
O
V
E
M
T
E
D

T
O

B
Y
-
P
R
O
D
U
C
T

A
T

O
.
I
Z
O
.
/
L
'

2
8
]
1

9
.

$
f
Y
8

.

f
U
I
A
L

A
h
N
U
A
l

I
t
M
A
T
I
N
G

C
O

9
7
9
7
3
2
.

i
l
Y
R

~

S
U
R
F
A
t
t

A
I
E
A

a

K
E
A
t
T
O
R

4
8
6
.
5
0
9

S
Q

F
r

C
U

T

W

4
t
A
L
f

U
f

A
8
V

7
9
5
8
9
.
6
9

C
O
S
T

U
f

C
A
T
A
L
Y
S
T

A
T

]
.
O
O
.
/
L
I

6
6
8
0
.
8
9
8

D
I
S
T
1
L
L
A
T
I

N

C
A
l
C

l
A
T
I
U
N
S

F
O
R

R
E
A
C
T
O
R

D
E
S
C
R
I
B
E
D

A
8
0
V
E

F
E
E
D

R
A
T
E
-
d
,
9
J
4
8
.

L
b

F
R
A
C
T
I
O
N

I
N

F
E
E
O
-
0
.
7
6
b

f
M
A
C
I
l
U
N

l
N

D
I
S
T
I
L
l
A
T
E
=
O
.
9
8
6

P
l
N
l
R
U
R

K
*
3
.
2
5
4

K

f
k
L
I
U
M

I
8
k

0
L
P
M

L
U
S
f

H
t
k
f

L
U
5
1

L
U
J
L

L
U
5
r

Y
b

L
U
5

l
.

1
0
5
.
e
o

1
1
8
3
6
.
7
5

1
6
7
0
8
0
.
3
7

8
8
6
7
.
8
9

2
4
7
7
8
5
.
0
0

1
.
l

9
1
.
0
0

6
3
6
2
4
.
1
1

l
Z
9
9
Z
6
.
4
4

6
8
9
5
.
q
z

2
0
0
4
4
7
.
0
6

4

1
.
1
5

8
Z
.
0
0

5
8
5
4
5
.
1
1

1
3
4
5
4
Z
.
5
0

7
1
4
0
.
9
2

2
0
0
2
2
9
.
1
2

~

1
.
2
1

7
8
.
0
0

5
6
d
2
8
.
5
2

1
3
9
1
5
8
.
6
2

7
3
8
5
.
9
3

2
0
3
3
7
3
.
0
6

.

1
.
2

7
3
.
0
0

5
4
2
H
.
3
0

1
4
3
7
7
4
.
7
5

7
6
3
0
.
9
3

Z
0
5
6

2
.
9
4

1
.
3
0

1
1
.
0
0

5
3
1
6
1
.
1
4

1
4
8
3
9
0
.
8
7

7
8
7
5
.
9
3

2
1
0
0
2
7
.
9
4

1
.
3
,

6
8
.
0
0

5
2
4
H
.
6
8

1
5
3
0
0
7
.
0
0

8
1
2
0
.
9
3

Z
1
3
5
7
2
.
5
6

1
.
4
0

6
5
.
e
o

5
1
0
3
3
.
0
1

1
0
5
0
8
Z
.
0
6

5
5
7
7
.
2
9

1
6
1
6
9
Z
.
3
1

1
.
1
5

6
3
.
e
o

5
0
3
2
6
.
8
5

1
0
8
1
5
9
.
5
0

5
7
4
0
.
6
3

1
6
4
2
2
6
.
9
4

1
.
!
I

6
2
.
0
0

5
0
3
6
8
.
7
8

1
1
1
2
3
6
.
8
7

5
9
0
3
.
9
6

1
6
7
5
0
'
.
5
6

1
.
5
;

6
1
.
e
o

5
0
3
7
4
.
5
1

1
1
4
3
1
4
.
3
1

6
0
6
7
.
3
0

1
7
0
7
5
6
.
1
2

5
9
.
C
C

4
9
5
0
5
.
6
4

1
1
7
3
9
1
.
7
5

6
2
3
0
.
6
4

1
7
3
1
2
8
.
0
0

1
.
6
!

5
9
.
0
0

5
0
2
8
0
.
3
6

1
2
0
4
6
'
.
1
9

6
B
3
.
9
7

1
7
7
1

3
.
5
a

1
.
7
0

5
8
.
0
0

5
0
1
8
1
.
9
2

1
2
3
5
4
6
.
6
2

6
5
5
7
.
3
1

1
8
0
2
8
5
.
8
1

1
.
7
)

5
8
.
C
O

5
0
q
z
8
.
2
2

1
2
6
6
2
4
.
0
0

6
7
2
0
.
6
4

1
8
4
2
7
2
.
8
1

l
.
d
J

5
6
.
0
0

4
9
8
8
5
.
6
8

1
2
9
7
0
1
.
4
4

6
8
8
3
.
9
8

1
8
6
4
7
1
.
0
6

1
.
8
5

5
6
.
0
0

5
0
5
9
2
.
5
7

1
3
2
7
7
8
.
8
7

7
0
4
7
.
3
2

1
9
0
4
1
8
.
7
5

1
.
9
0

5
5
.
C
O

2
3
.
0
0

5
0
3
7
6
.
8
7

1
3
5
8
5
6
.
2
5

7
2
1
0
.
6
5

Y
5

t
.
v

5
5
.
0
0

2
)
.
0
0

5
1
0
5
8
.
4
6

1
3
8
9
3
3
.
6
'

7
3
7
3
.
9
8

1
9
7
3
6
6
.
1
2

2
.
G
O

5
3
.
0
0

2
2
.
0
0

1
9
8
5
2
.
8
1

1
4
2
0
1
1
.
1
2

7
5
3
7
.
3
2

1
9
9
1
0
1
.
2
5

g

2
.
J
5

5
3
.
C
C

2
2
.
0
0

5
0
4
1
8
.
2
3

1
4
5
0
8
8
.
5
6

7
7
0
0
.
6
6

Z
0
J

d
I
o
9
9

<

.
I
J

5
3
.
0
0

2
2
.
0
0

5
1
1
3
8
.
2
0

1
4
8
1
6
5
.
'
4

7
8
6
3
.
9
'

2
0
7
1
6
8
.
1
2

>

5
2
.
0
0

2
2
.
0
0

5
0
7
'
5
.

'
'

1
5
1
2
4
3
.
3
8

8
0
2
7
.
3
3

0
0

2
.
2
0

5
2
.
C
C

2
2
.
0
0

5
1
4
1
3
.
6
3

1
5
4
3
2
0
.
8
1

8
1
9
0
.
6
6

2
1
3
9
2
5
.
0
6

;

2
.
2
5

5
2
.
V
O

2
1
.
0
C

5
2
0
2
6
.
3
8

1
5
7
3
9
8
.
1
9

8
3
5
4
.
0
0

2
1
7
7
7
8
.
5
6

2
.
3
J

5
2
.
0
0

2
1
.
0
0

5
2
6
3
4
.
3
8

1
6
0
4
7
5
.
6
2

8
5
1
7
.
3
4

2
2
1
6
2
7
.
3
1

;

2
.
3
5

5
1
.
0
0

2
1
.
0
0

5
2
2
1
3
.
8
6

1
6
3
5
5
3
.
0
6

8
6
8
0
.
6
7

2
H
4
I
7
.
5
6

2
o

5
1
.

0

2
1
.
0
0

5
2
8
0
1
.
1
8

1
6
6
6
3
0
.
5
0

8
8
4
4
.
0
1

2
2
8
2
7
5
.
6
'

~
.

b
l
.
0
C

2
1
.
0
0

5
3
3
8
1
.
1
0

1
6
'
7
0
7
.
'
4

V
0

7
o
3
9

Z
J
2
0
'
9
.
3
I

0 ~

t
.
c
0

2
1
.
0
0

5
3
9
6
2
.
8
2

1
7
2
7
8
5
.
3
1

9
1
7
0
.
6
8

2
3
5
9
1
8
.
7
5

'

2
.

5
1
.
0
0

2
1
.
0
0

5
1
5
3
7
.
5
3

1
7
5
8
6
2
.
7
5

9
3
3
1
.
0
2

2
3
9
7
3
4
.
2
5

2
.
b
J

4
9
.
C
O

2
1
.
0
0

5
2
9
1
6
.
9
7

1
7
8
9
4
0
.
1
9

9
4
9
1
.
3
5

2

1
3

.
5
0

Z
o
b

4
9
.
V
O

2
1
.
0
0

5
3
4
9
1
.
4
4

1
8
2
0
1
1
.
5
6

9
6
6
J
.
6
9

2
4
5
1
6
9
.
6
9

2
.
7
0

4
9
.
C
O

1
'
.
0
0

5
4
J
3
2
.
3
1

1
8
5
0
9
5
.
0
0

9
8
2
4
.
J
2

2
4
8
9
5
1
.
3
1

|

2
.
7
5

4
9
.
0
0

l
V
.
0
0

5
4
5
6
9
.
5
4

1
8
8
1
7
2
.
4
4

9
9
8
7
.
3
6

2
5
7
7
2
9
.
3
1

~

2
.
8
0

4
9
.
C
O

2
.
8
5

4
9
.
0
0

2
.
9
0

4
8
.
0
0

2
.
9
5

4
8
.
0
0

3
.
C
O

4
8
.
0
0

T
U
8
E

L
E

6
T
H

2
0
.
0
0
0
0
0

F
T

O
U
T
S
I
D
E

T
U
B

D
I
A
M
T
E

2
.
3
7
5
0
0

I
N

I
N
S
I
D
E

T
U
'
.

D
I
A
M
E
T
E
R

2
.
0
6
7
0
0

I

1
9
.
0
0

1
9
.
0
0

1
9
.
0
0

1
9
.
0
0

1
9
.
0
0

F
E
e
D

R
A
r
E
.

L
L
L

L
B

E
T
H
Y
L

8
E
N
Z
E

E
/
H
R
/
T
U
B
E

I
N
I
T
I
A
L

F
E
E
D

T
E
M
P
E
R
A
T
U
R
E
.

8
4
2
.
0
0

D
E
G
R
E
E
S

F

A
S
S
U
M
E
D

X
I
T

T
E
M
P

U
F

F
L
U
E

G
A
S
.

1
6
o
.
0
0
D

Q
E
G
R
E
E
S

F

C
A
T
A
L
Y
S
T

P
E
L
L
E
T

D
I
A
M
E
T
E
R
.

0
.
0
1
0
4
0

F
T

C
A
T
A
L
Y
S
T

B
U
L
K

D

N
$
I
T
Y
.

6
1
.
0
0
0
0
0

L
B
/
C
u

F
T

E
F
F
E
C
T
I
Y
E

E
S
S

F
A
C
T
D
R

0
.
4
4

Y
O
I
D

F
R
A
C
T
I
O
N

U
t

P
A
C
K
E
D

B
E
D

0
.
3
8
0
0
0

5
5
1
0
)
.
1
9

5
5
6
3
3
.
5
2

5
5
0
1
4
.
3
7

5
5
5
2
7
.
4
6

5
b
0
3
7
.
l
q

O
Y
E
R
-
A
L
L

H
E
A
T

T
M
A

$
F
E
R

C
O
E
F
F
I
C
I
E
N
T
.

8
.
0
0
0
0
0

B
T
U
/
H
R
I
S
Q

F
T
/
F

H
E
A
T

C
A
P
A
C
I
T
Y

O
F

E
T
H
Y
L

8
E
N
Z
E
N
E

0
.
6
3
5
0
0

8
T
U
/
L
B
/
F

H
E
A
T

C
A
P
A
C
I
T
Y

U

F
L
U
E

G
A
S

U
.
3
D

0
0

B
T
U
/
L
8
/
F

F
L
U
E

G
A
S

F
L
O
.

R
A
T
E
.

1
8
1
6
.
1
4
3

L
8

F
L
U

G
4
S
/
H
K
/
T
U
8
E

A
S
S
U
M
E
D

E
X
l
r

P
R
E
S
S
U
R
E

5
.
0
0
0

O

P
S
I
G

U
M
8
E
R

U
F

I

C
R
E
M
E

T
S

1
0
0

I
N
C
R
E
M
E
N
T

L
t
N

I
H

E
T
H
Y
L

S
T
Y
R
E
N
E

B
e
N
l
E
N
E

T
U
L
U
E
N
E

H
Y
O
R
O
G
E

R
E
A
C

T
1
O
N

N
U
M

E
R

F
r

8
E
N
l
E
N
E

E
T
H
Y
l
E
N
E

M
H
H
A

E

T
E
M
P

L
a

M
O
L
E
S

f
O
R
H

v

P
E
R

L
B

M
O
L
E

O
F

E
T
H
Y
L
B
E
N
l
E
N
E

F
E
D

D
E
G

F

0

J
e

1
.
0
0
0
0
0

U
-
U

0
.
0

0
.
0

0
.
0

8
4
2
.
0
0

1
0

2
.
0
0
U

U
-
9
!
!
>
9

0
.
0
2
2
2
0

0
.
0
U
0
0
3

O
.
O
O
O
l
d

0
.
J
Z
2
0
2

9
>
o
3
8

2
0

4
.
0
U
U

J
.
9
4

H
2

0
.
0

5
7
1

0
.
0
0
0
1
7

O
.
O
O
I
H

0
.
0
5
4
4
0

9
9
1
.
1
5

3
0

b
.
O
O
O

;
.
8
q

S
8

U
=
U
9
>

0
.
0
0
0
6
0

0
.
0
0
4
)
1

0
.
0
9
0
8
9

1
0
4
5
.
4
3

4
0

d

u
+
d
>
U
I
>

0
.
1
3
7
8
7

0
.
O
C
I

2

0
.
0
U
9
6

0
.
1
2
8
1
1

1
0
Y

.
6
8

5
0

1
0
.
0
0
0

0
.
I
9
6
1
5

U
-
l
b

>

0
.
0
0
3
5
6

0
.
0
1
7
5
4

0
.
1
6
5
2

l
l
3
l
.
5
6

6
0

e
0

0
.
7
3
1
4
5

0
.
2
2
'
0
5

0
.
C
0
6
6
6

0
.
0
2
6
8
4

0
.
2
0

1
1
6
6
.
4
4

1
4
.
J
C
O

0
.
6
1
6
9
6

0
.
2
1
>
6
6

0
.
0
1
0
9
6

0
.
0
3
6
4
2

0
.
2
3
9
2
>

1
1
9
1
.
7
3

8
0

1
6
.
0
0
0

0
.
6
1
1
8
2

0
.
3
2
l
l
5

0
.
0
1
6
2
4

0
.
0
4
4
8
0

0
e
2
7
6
3
5

1
2
2
6
.
9
1

9
0

1
8
.
0
0
(

0
.
>
6
4
9
8

0
.
3
6
2
6
7

0
.
0

l
8
3

0
.
0
5
0
5
1

0
.
3
1
Z
1
5

1
2
5
6
.
)

9
)

1
8
.
b
O

0
.
5
,
1
9
0

0
.
3
7
)
2
5

0
.
O
Z
3
3
6

0
.
0
5
1
4
9

0
.
)
2
1
7
6

1
2
6
6
.
0
8

1
9
1
2
4
9
.
8
7

1
0
1
5
0
.
7
0

2
5
6
5
0
3
.
7
5

f

1
9
4
3
2
7
.
2
5

I
O
H
4
.

3

2
6
0
2
7
4
.
1
5

1
9
1
4
0
4
.
6
9

1
0
4
7
7
.
3
7

2
6
2
8
9
6
.
3
8

2
0
0
4
8
2
.
1

1
0
6
4
0
.
7
0

2
6
6
6
5
0
.
2
5

G

~
.

2
0
3
5
5
9
.
5
0

1
0
8
0
4
.
0
4

2
7
0
4
0
0
.
6
9

f
l
U
E

G
A
S

P
R

E
S
S
U
R
E

Y
I
E
L
D

T
E
M
P

P
S
I
G

L
B

U
L
E
5

5
T
Y

O
E
G

F

L
B

K
U
L
L

L
b

8
F
k
L
1
t

1
6
0
0
.
0
0

5
8
.
6
'
4

0
.
0

1
6
1
3
.
1
7

5
5
.
'
3
5

Y
Y
.
0
0

I
b
2
5
.
2
1

5
2
.
8
1
4

9
7
.
4

0

1
6
3
6
.
3
8

4
9
.
2
1
5

9

.
0
8

1
6
4
6
.
8
4

4
5
.
2
4
2

Y

e
J
1

1
6
5
6
.
6
9

4
0
.
6
0
3

8
9
.
6
5
1

6
6
6
.
0
4

3
5
.
1
8
4

8
7
.
2
3

1
6
1
4
.
9
5

2

.
6
9
0

8
5
.
3
3
3

1
6
8
3
.
4
7

2
0
.
5
2
6

8
4
.
0
3
0

1
6
9
1
.
6
2

9
.
0
4
1

8
3
.
3
6
9

1
6
9
3
.
9
8

4
.
1
1
3

J
.

9
0

J
+
0

1
.
0
0
0
0
0

0
.
0

0
.
0

0
.
0

0
.
0

8
4
2
.
0
0

1
6
0
0
.
0
0

5
9
.
6
9
4

0

e
u

1
0

2
.
0
0
0

0
.
9
1
7
3
3

0
.
0
2
2
4
b

0
.
0
0
0
0
3

0
.
0
0
0
1
9

0
.
0
2
2
2
7

9
2
5
.
1
9

1
6
l
l
.
1
7

5
6
.
9
7
3

Q
q
.
0
4
0

2
0

4
.
1
0
0

0
.
9
4
2
3
5

0
.
0
5
b
1
3

0
.
0
0
0
1
7

0
.
0
0
l
l
6

0
.
0
5
4
7
7

9
9
0
.
8
7

1
6
2
5
.
2
1

5
3
.
8
9
9

9
7
.
J
5
5

3
0

b
.
O
O
O

0
.
8
9
9
3
1

0
.
0
9
5
b
l

O
.
O
O
O
b
l

0
.
0
0
4
4
7

0
.
0
9
1
1
4

1
0
4
5
.
1
9

1
6
3
6
.
4
0

5
0
.
U
8

9
4
.
9
5
8

'
0

d
.
O
u
O

0
.
8
5
0
0
5

0
.
1
3
S
I
9

0
.
0
0
1
b
5

0
.
0
1
0
1
1

0
.
1
2
8
0
8

1
0
9
1
.
5
3

1
6
4
6
.
8
5

4
6
.
4
6
0

9
2
.
1
5
8

5
0

1
0
.
0
V
O

0
.
7
9
!
1
5

(
.
1
8
3
0
3

0
.
0
0
3
b
2

0
.
0
1
8
2
0

0
.
1
6
4
8
2

1
1
3
1
.
4
7

1
6
5
6
.
7
1

4
1
.
9
1
9

8

.

3
4
7

b
O

l

.
J
u
O

7
3
5
9

0
.
2
2
9
3
2

0
.
0
0
b
7
9

0
.
0
2
7
9
5

0
.
2
0
1
3
7

1
1
6
6
.

J
1

1
6
6
6
.
0
6

3
6
.
6
3
9

8
6
.
8
4
5

7
0

L
o
U
0
0

:
.
6
7
4
6
9

0
.
2
7
b
0
2

0
.
0
t
l
2
2

0
.
0
3
8
0
8

0
.
2
)
7
9
4

1
1
9
7
.
6

1
6
7
4
.
9
7

3
0
.
3
5
2

8
4
.
8
4
7

8
0

I
b
.
O
O
(

0
.
b
1
"
3
8

0
.
3
2
1
8
0

0
.
0
1
b
6
9

0
.
0
4
7
1
3

0
.
2
7
4
6
/

1
2
2
6
.
5
9

1
6
8
1
.
5
0

2
2
.
5
5
0

8
3
.
4
4
9

'
0

I
d
.
L
O
O

0
.
5
5
9
3
8

0
.
3
b
4
3
3

0
.
0
2
2
b
3

0
.
0
5
3
6
6

0
.
3
1
0
6
7

1
2
5
5
.
3
2

1
6
9
1
.
6
5

1
1
.
9
3
9

8
2
.

b
8
6

9
5

B
.
O
O
O

a
.
B
6
S
7

0
.
3
8
l
5
3

0
.
C
2
5
H

0
.
0
5
5
5
3

0
.
3
2
7
0
0

1
2
7
1
.
0
0

1
6
9
5
.
5
9

4
.
2
1
7

8
2
.
5

0

0
.
0

1
.
0
0
0
0
0

0
.
0

0
.
0

0
.
0

0
.
0

8
4
2
.
0
0

1
6
0
0
.
0
0

6
0
.
6
9
4

0
.
0

1
0

2
.
0
0
0

0
.
9
7
7
0
6

0
.
0
2
2
7
1

0
.
0
0
0
0
3

0
.
0
0
0
2
0

0
.
0
2
2
5
1

9
2
5
.
0
0

1
6
1
3
.
1
7

5
8
.
0
1
0

9
9
.
0
1

~

2
0

'
0
.
0
0
0

0
.
9
4
1
8
8

0
.
0
5
b
5
4

0
.
0
0
0
1
7

0
.
0
0
1
4
1

0
.
0
5
5
1
3

9
9
0
.
5
9

1
6
2
5
.
2
3

5
4
.
9
8
1

9
7
.
2
8
9

3
0

b
.
o
o
e

0
.
8
9
8
7
7

0
.
0
9
5
9
9

O
.
O
O
O
b
l

0
.
0
0
4
6
2

0
.
0
9
1
3
7

1
0
4
4
.
9
7

1
6
3
6
.
4
1

5
1
.
5
5
7

9
9
.
8
2
6

'
0

8
.
0
1
0

0
.
8
4
9
3
7

0
.
I
H
4
9

0
.
0
0
1
b
7

0
.
0
1
0
4
7

0
.
1
2
8
0
2

1
0
9
1
.
4
1

1
6
4
6
.
8
7

4
7
.
6
7
0

9
1
.
9
4
2

5
0

1
0
.

J
O
G

0
.
7
9
"
1
9

O
.

I

8
3
2
b

0
.
0
0
3
b
8

0
.
0
1
8
8
8

0
.
1
6
4
3
8

1
1
3
1
.
4
1

1
6
5
6
.
7
2

4
3
.
2
2
4

8
9
.
0
4
1

b
O

1
2
.
0
0
0

0
.
1
3
4
4
8

0
.
2
N
5
4

0
.
0
0
6
9
2

0
.
0
2
9
0
6

0
.
2
0
0
4
8

1
1
6
6
.
3
3

1
6
6
6
.
0
8

3
8
.
0
7
2

8
b
.
4
4
9

7
0

1
"
.
0
0
0

0
.
6
7
2
4
9

0
.
2
7
b
3
0

0
.
O
l
l
4
7

0
.
0
3
9
7
5

0
.
H
6
5
5

1
1
9
7
.
5
3

1
6
7
4
.
9
9

3
1
.
9
7
7

8
4
.
3
b
2

8
0

l
b
.
I
O
O

0
.
6
1
1
0
9

.
3
2
2
2
9

0
.
0
1
7
1
4

0
.
0
4
9
'
"

0
.
2
1
2
8
2

1
2
2
6
.
3
6

1
6
8
3
.
5
2

2
4
.
4
9
6

8
2
.
8
7
2

9
0

1
8
.
0
0
0

:
.
5
5
"
2
2

0
.
3
b

5
8

0
.
0
2
3
4
1

0
.
0
5
/
7
9

0
.
3
0
8
8
0

1
2
5
4
.
5
6

1
6
9
1
.
6
9

1
4
.
5
8
9

8
2
.

0
t
t

9
7

1
9
,
"
0
0

0
.
5
2
1
0
8

0
.
3
9
1
b
7

0
.
e
2
7
5
1

0
.
0
5
9
7
4

0
.
3
3
1
9
3

1
2
7
5
.
9
1

1
6
'
7
.
1
7

4
.
3
3
3

8
1
.
7
8
?

0

0
.
0

1
.
0
0
0
C
O

0
.
0

0
.
0

0
.
0

0
.
0

8
4
2
.
0
0

1
6
0
0
.
0
0

6
1
.
6
'

0
.
0

1
0

2
.
0
0
0

0
.
9
7
6
8
1

.
0
2
2
9
b

0
.
0
0
0
0
3

0
.
0
0
0
2
0

0
.
0
2
2
7
6

9
2
4
.
8
2

1
6
1
3
.
1
7

5
9
.
0
4
6

9
9
.
9
9
5

2
0

4
.
0
0
0

0
.
9
"
1
"
3

0
.
0
5
b
Q
"

0
.
0
0
0
1
7

0
.
0
0
1
4
6

0
.
0
5
5
4
8

9
9
0
.
3
2

1
6
2
5
.
2
3

5
6
.
0
6
1

9
7
.
Z
2
2

)
0

b
.
J
O
O

0
.
8
9
8
2
"

0
.
0
9
b
3
5

0
.
0
0
0
b
2

0
.
0
0
4
7
8

0
.
0
9
1
5
7

1
0
4
4
.
7
7

1
6
3
6
.
4
2

5
2
.
6
9
1

Q
"
.
b
9
2

d
-
L
L
0

a
.
8
"
8
7
3

0

e
l

1
5

0
.
0
0
1
b
9

0
.
0
1
0
8
3

0
.
1
2
7
9
2

1
0
9
1
.
3
0

1
6
4
6
.
8
8

4
8
.
8
7
3

9
1
.
1
2
4

5
0

1
0
.
t
t
O

0
.
7
9
3
2
5

0
.
1
8
3
4
b

0
.
0
0
3
1
4

0
.
0
1
9
5
6

0
.
1
6
3
9
0

1
1
3
1
.
3
7

1
6
5
6
.
7
4

4
"
.
5
1
6

8
8
.
7
1
4

"

~

6
0

1
2
.
0
0
0

0
.
7
3
3
0
6

0
.
2
2
n
l

0
.
0
0
1
0
5

0
.
0
3
0
1
8

0
.
1
9
9
5
3

1
1
6
6
.
3
1

1
6
6
6
.
0
9

3
'
.
4
8
7

8
b
.
0
5
1

w

7
0

I
"
.
O
L
O

0
.
6
7
0
3
5

0
.
2
1
6
5
1

0
.
0
1
1
7
2

0
.
0
4
1
4
2

0
.
2
3
5
0
9

1
1
9
7
.
4
9

1
6
1
5
.
0
1

n
.
5
6
9

8
3
.
8
7
9

S
O

I
b
.
I
"
O

0
.
/
0
7
9
4

0
.
3
2
2
b
b

0
.
0
1
7
5
9

0
.
0
5
1
8
1

0
.
2
7
0
8
6

1
2
2
6
.
2
1

1
6
1
3
.
5
4

2
6
.
)
7
5

8
2
.
3
0
0

9
0

l
d
.
O
O
u

0
.
5
4
9
4
2

0
.
3
6
6
5
2

0
.
0
2
4
1
6

0
.
0
5
9
9
0

0
.
3
0
6
6
1

1
2
5
4
.
0
0

1
6
9
1
.
7
1

1
7
.
0
5
1

8
1
.
1
4
3

N
.

9
9

1
9
.
d
O

(
.
5
0
5

0
.
4
0
0
b
5

0
.
0
2
9
7
7

0
.
0
6
4
1
3

0
.
3
J
6
5
2

1
2
8
0
.
8
2

1
6
9
8
.
7
4

4
.

5
9

8
1
.
0
1
2

0
.
0

1
.
0
0
0
0
0

0
.
0

0
.
0

0
.
0

0
.
0

8
4
2
.
0
0

1
6
0
0
.
0
0

6
2
.
6
'

0
.
0

0
1
0

2
.
V
O
O

0
.
9
7
6
5
5

0
.
0
2
3
2
1

0
.
0
0
0
0
3

0
.
0
0
0
2
1

0
.
0
2
3
C
O

9
2
"
.
6
3

1
6
1
3
.
1
8

6
0
.
0
8
2

9
8
.
9
1
2

K

2
0

4
.
0
0
v

0
.
9
"
1
0
0

0
.
0
5
7
3
3

o
.
O
O
O
l
7

0
.
0
0
1
5
1

0
.
0
5
5
8
2

9
9
0
.
0
6

1
6
2
5
.
2
4

5
7
.
1
)
9

9
7
.
1
5
5

~

3
0

b
.
G
O
O

0
.
8
9
7
7
4

0
.
0
9
6
6
9

0
.
0
0
0
0
3

0
.
0
0
4
9
4

0
.
0
9
1
7
5

1
0
4
4
.
5
8

1
6
3
6
.
4
3

"
.
8
2
2

9
4
.
5
5
6

"
0

d
.
0
0
0

0
.
8
4

1
0

0
.
1
3
8
9
9

0
.
0
0
1
1
1

0
.
0
1
1
2
0

0
.
1
l
7
8
0

1
0
9
1
.
2
2

1
6
4
6
.
8
'

5
0
.
0
6
9

9
1
.
5
0
4

S
O

1
0
.
0
U
0

0
.
7
9
2
3

0
.
1
8
3
6
3

0
.
0
0
3
7
9

0
.
0
2
0
2
4

0
.
1
6
3
3
8

1
1
3
1
.
3
5

1
6
5
6
.
7
5

4
5
.
1
9
8

6
0

L

.
0
0

0
.
7
3
1
6
6

0
.
2
2
9
8
5

0
.
0
0
7
1
8

0
.
0
3
1
)
1

0
.
1
9
8
5
"

1
1
6
6
.
3
2

1
6
6
6
.
1
1

4
0
.
8
8
1

8
5
.

7

7
0

1
"
.
V
O
O

0
.
1
6
8
2
6

0
.
2
7
b
6
0

0
.
0
I
I

9
8

0
.
0
4
3
1
0

0
.
2
)
)
5
6

1
1
9
7
.
4
8

1
6
7
5
.
0
2

)
5
.
1
)
0

8
3
.
3
'
"

8
e

1
6
.
0
0
e
.
0
.
6
0
"
8
9

0
.
3
2
2
9
3

C
.
0
1
8
0
3

0
.
0
5
4
1
4

0
.
2
6
8
7
9

1
2
2
6
.
1
2

1
6
8
1
.
5
6

2
8
.
1
9
6

8
1
.
7
3
2

"

9
0

I
S
.
J
u
O

0
.
5
4
"
8
9

C
.
3
6
7
2
0

0
.
0
2
"
9
0

0
.
0
1
3
0
1

0
.
3
0

1
9

1
2
5
3
.
6
2

1
6
9
1
.
7
3

1
9
.
3
6
8

8
0
.
6
8

~

1
0
v

2
0
.
0
0
0

0
.
4
9
3
8
2

0
.
4
0
6
2
0

0
.
0
3
1
6
0

0
.
0
6
8
1
9

0
.
H
7
8
1

1
2
8
2
.
5
2

1
6
9
9
.
5
4

6
.
4
1
0

8
0
.
2
4
7

(

N
U
K
b
L
K

U
b
L
>

5
8
.
9
1
3

A
N
N
U
A
L

C
O
S
T

O
f

R
E
C
Y
C
L
I
N
G

E
T
H
Y
L

8
E
N
Z
E

E

A
T

O
.
O
I
O
S
/
L
B

2
6
0
2
1
1
.

S
/
Y
R

A
N
N
U
A
L

C
O
S
T

E
T
H
Y
L

B
E
N
Z
E
N
E

C
O
N
V
E
R
T
E
D

T
O

B
Y
-
P
R
O
D
U
C
T
S

A
T

0
.
1
2
0
S
/
L
B

b
3
2
2
3
1
.

S
/
Y
8

T
O
T
A
L

A
N
N
U
A
L

U
P
E
R
A
T
I
N
G

C
O
S
T

s
n
4
4
2
.

S
f
Y
R

S
U
R
f
A
c
e

K
E
A

U
r

E
A
C
T
O
R

7
3
2
.
6
0
4

S
O

f
T

C
O
S
T

O
f

T
w
a

K
E
A
C
T
O
R
S

O
f

A
8
U
V
E

S
I
Z
E
'

1
0
0
5
0
5
.
7
5

C
O
S
T

U
r

C
A
T
A
L
Y
S
T

A
T

3
.
0
0
S
/
L
8

5
0
2
4
.
5
3
5

W

D
I
S
T
I
L
L
A
T
I
U

C
A
L
C
U
L
A
T
I
O
N
S

f
O
R

R
E
A
C
T
O
R

D
E
S
C

I
S
E
O

A
8
0
V
E

F
E
E
D

A
T
f
'

7
4
0

F
R
A
C
T
I
O
N

I
N

F
L
L
D

0
.
5

9

E
8

F
R
A
C
T
I
O
N

I

D
I
S
T
I
L
L
A
T
E
=
0
.
9
6
2

N
l
K
U
H

R
R

F
A
C
r
O

1
.
1
0

I
.
I

1
.
2
0

.
3
0

l
o
J
5

1
.
"
0

1
.
"
;

t

l
.
b
O

1
.
6
5

1
.
7
0

1
.
7
5

T

A
Y
S

9
5
.
0
0

8
3
.
C
O

7
1

.
.
0
0

?
l
.
0
0

6
8
.
0
0

6

.
0
0

6
3
.
0
0

6
1
.
C
O

5
'
.
0
0

5
8
.
0
0

5
6
.
0
U

5
5
.
0
0

5

.
O
O

5
3
.
0
0

5
3
.
0
0

F
E
E
D

T
R
A
Y

"
3
.
0
0

3
8
.
0
0

3
3
.
0
0

3
1
.
0
0

2
9
.
0
0

2
8
.
0
0

2
6
.
0
0

2
6
.
0
0

2

.
0

2
5
.
0
0

2
3
.
0
0

2
3
.
0
0

2
3
.
0
0

2
l
.
0
0

2
2
.
0
0

O
E
P
R

C
O
S
T

H
E
A
T

C
O
S
T

4
2
2
0
6
.
7
9

6
1
0
2
2
.
2
7

3
1
1
3
5
.
7
8

6
3
4
1
3
.
6
2

3
5
3
2
9
.
2
2

6
5
8
0
4
.
9
4

3
'
1
9
4
.
3
6

6
8
1
9
6
.
2
5

3
2
9
6
9
.
4
6

7
0
5
8
7
.
6
3

3
2
1
5
1
.
2
"

"
8
6
5
2
.
6
6

3
1
7
7
0
.
6
6

5
0
2
'
6
.
8
9

3
1
3
"
"
.
0
"

H
8
1

1
2

3
0
8
7
2
.
H

5
3
4
3
5
.
3

3
0
8
8
9
.
1
6

5
5
0
2
9
.
6
0

3
0
3
3
9
.
"
8

5
6
6
2
3
.
8
J

3
0
2
9
8
.
2
'

5
8
2
1
8
.
0
3

3
0
7
9
3
.
3
5

5
9
8
1
2
.
2
3

3
0
1
"
5
.
7
0

6
1
'
0
6
.
4
8

3
0
b
1
2
.
8
4

6
1
0
0
0
.
H

2
7

C
O
O
L

L
J
5
f

Y
E
A
R

C
O
S
T

4

3
2
3
8
.
7
Y

1
0
6
4
6
7
.
8
1

<

~
.

3
3
6
5
.
7
1

1
0
4
5
'
5
.
0
6

9
2
=
6
"

1
0
4
6
2
6
.
7
5

3
&
1
9
.
5
6

1
0
6
0
1
0
.
1
2

3
7
9
6
.
"
8

1
0
7
3
0
3
.
5
6

2
5
8
2
.
2
7

8
3
3
8
6
.
1
2

2
6
6
6
.
8
8

8
4
6
8
'
.
3
1

2
7

1
.

0

8

9
3
6
e
6
2

2
8
3
6
.
1

8
7
H
3
.
7
5

2
9
2
0
.
1
3

8
8
8
3
9
.
4
4

3
0
0
5
.
3
9

8
9
9
6
8
.
6
2

3
0
8
9
.
9
6

9
1
6
0

9

3
1

.
5

'
3
7
8
0
.
1
2

3
2
5
9
.
1
8

9
4
8
l

e
3
1

3
3
4
3
.
8
0

9
6
9
5
7
.
3
T

o
d
J

5
2
.
C
O

2
2
.
0
0

3
0
4
i
8
.
9
7

6
4
5
9
4
.
9
4

H
2
8
.
H

9
8
5
1
2
.
3
1

1
.
8
,

5
2
.
0
0

2
Z
.
0
0

3
0
9
3
8
.
2
6

6
6
1
8
9
.
1
9

3
5
1
3
.
0
3

1
0
0
6
4
0
.
4
4

1
.
9
0

5
1
.
G
O

2
2
.
0
0

3
0
7
7
9
.
6
8

6
7
7
8
3
.
3
7

3
5
9
7
.
6
4

1
0
Z
1
6
0
.
6
9

@

1
.
9
>

5
1
.
C
O

2
1
.
0
0

3
1
2
1
Z
.
0
5

6
9
3
7
7
.
6
3

3
6
8
2
.
2
6

1
0
4
2
7
1
.
8
7

2
.

0

4
9
.
0
0

2
1
.
0
0

3
0
3
9
9
.
6
2

7
0
9
7
1
.
8
1

3
7
6
6
.
8
7

1
0
5
1
3
8
.
2
5

=
b
>

4
9
.
0
0

2
1
.
0
0

3
0
8
0
7
.
5
4

7
2
5
6
6
.
0
6

3
8
5
1
.
4
9

1
0
7
2
2
5
.
0
6

"

2
.
1
0

4
9
.
0
0

Z
I
.
0
0

3
1
Z
1
1
.
8
7

7
4
1
6
0
.
2
5

3
9
3
6
.
1
0

1
0
9
3
0
8
.
1
9

2
.
1
5

4
8
.
0
0

Z
I
.
0
0

3
0
9
6
7
.
5
6

7
5
7
5
4
.
5
6

4
0
2
0
.
7
2

1
1
0
7
4
2
.
8
1

l
.
2
0

4
8
.
C
O

2
1
.
0
0

3
1
3
5
7
.
0
0

7
7
3
4
8
.
8
1

4
1
0
5
.
3
3

1
1
1
8
1
1
.
1
2

2
.
l
5

4
R
.
0
0

2
1
.
0
0

3
1
7
4
3
.
1
5

7
8
9
4
3
.
0
0

4
1
8
9
.
9
5

1
1
4
8
7
6
.
0
6

2
.
3
0

4
8
.
0
0

1
9
.
0
0

3
2
1
2
6
.
2
5

8
0
5
3
7
.
2
5

4
2
7
4
.
5
6

1
1
6
9
3
8
.
0
)

.

2
.
3
5

4
6
.
0
0

1
9
.
0
0

3
1
1
5
1
.
8
9

8
2
1
3
1
.
4
4

4
3
5
9
.
1
7

1
1
7
6
4
1
.
5
0

0 ~

2
.
4
0

4
6
.
0
0

1
9
.
0
0

3
1
5
1
3
.
2
5

8
3
7
2
5
.
6
2

4
,
.
3
.
7
9

1
1
9
6
8
2
.
6
2

'
^

2
.
4
:

4
6
.
0
0

1
9
.
0
0

3
1
8
7
1
.
9
2

8
5
3
1
9
.
8
7

4
5
2
8
.
4
0

1
2
1
7
2
0
.
1
9

2
.
5
U

4
0
.
C
O

1
9
.
0
0

3
2
2
2
7
.
9
2

8
6
9
1
4
.
1
3

4
6
1
3
.
0
2

1
2
3
7
5
5
.
0
6

~

2
.
5
5

4
6
.
0
0

1
9
.
0
0

3
2
5
8
1
.
3
3

8
8
5
0
8
.
3
1

4
6
9
7
.

3

1
2
5
7
8
7
.
3
1

2
.
6
0

4
>
.
0
0

1
9
.
0
0

3
2
2
1
6
.
2
9

9
0
1
0
2
.
6
2

4
7
8
2
.
2
5

1
2
7
1
0
1
.
1
2

Appendix
FORTRAN listings and
TY/Jical Results for the
Com/JII/er Programs Used in
Chapters 5, 6, and 7

Index
Absorption, 9
Activation energy, 170
Activity coefficient, 61
Adsorption,kinetics of, 163
Analysis, economic, see Economic
evaluation
-
Approximations, 86
Arrhenius correlation, 9, 41, 161
Batch reaction, 35
Bench-scale experimentation, 4,
19. 76
Bleed strea, 87, 89
Boiler design, 88
Caculations, iterative, 66
Calculations, tabulation of reactor-
design, 80, 172
Calculus of variations, 11
Capital cost estimation, 131
Catalyst, activity of, 97
kinetics on, 9
porosity of, 159
properties of, 158-160
surface area of, 159
Catalyst pellets, 9, 98, 162, 166
Censorship, 7
Chemical equilibrium, effect of
temperature on, 85
Chemical reaction, see Reaction
Chilton-Colburn equations, 9, 163
Chlorination, chemical kinetics of,
22
Chorobenzenes, production of, 22
Clausius-Clapeyron equation, 9, 132,
143
Computer, classroom use of,2
229
Computer program flow sheet, 147,
180
Computer programming, 146, 178
Consecutive reactions, 22
Control, production, 8
quality, 8
Conversion of a chemica reaction,
85,97,152,192
Conveyor belt, 103, 194
Continuous stirred-tank reactor
(CSTR), 11, 24
Cooling water, cost of, 58
temperature rise of,43
CorrOSion, 106
Cost, capital, 38, 57, 58, 181, 195
conversion, 73
operating, 73
optimization of, 28, 39, 70, 116,
142,183
raw-material, 38, 178
Cost accounting, 12, 13
Cost correlations, 57
Counter-current staged process,
69
Crystallization, 9
Cyclohexane production, 97
Data, consistency of, 6
correlation of, 9, 60
extrapolation of, 9, 60
Data analysis, 8, 22
Data processing, 3
Data reduction, 8
Depreciation, 122
DeSign, industrial practice of, 1-17
DeSign parameters, 8
Design results, extrapolation of, 95
Design strategy, 10, 11
Desorption, 162
230 The Industrial Practice of Clemical Process Engineering
Difference equations, 174
Differential equations, derivation of,
81,172
numerical integration of, 173
Diffusion within catalyst pellets, 9,
98, 162, 166
Diffusivity, binary, 164
effective, 166
Knudsen, 167
Distillation, see Fractionation
Distribution cfficient, 9, 61
Dynamic programming, 11
Economic evaluation, 2, 38, 71, 72,
96,98,115, 181
terminology of, 12
Efectiveness factor, 9, 166, 169, 174
Efficiency, tray, 9
mixer-settler, 69
Enthapy of formation, 42, 77
Enthalpy of reaction, 42, 80, 82
Equilibrium, chemical, 77, 85, 86, 169
liquid-liquid, 9, 54,60
vapor-liquid, 9, 57, 132
Erbar and Maddox correlation, 134
Error, experimental, 162
probable, 8
Error analysis, 8
Experimental variables, selection of,
60
Experiments, in-plant, 8
statistically designed, 8, 54, 59
Extraction, liquid-liquid, 9
Extrapolation of design results, 95
Factorial experimental design, 54
Fanning equation, 9, 109
Fenske equation, 134
Financial evaluation, see Economic
evaluation
-
Financial incentive, 10
Finite difference techniques, 173
First law of thermodynamics, 82
Flooding velocity, 141
Flow sheet, process, 63, 69, 97, 131,
138, 147
Fluid mechanics, 110
FORTRAN, 146, 180
Fractionator, design of, 9, 67, 132, 140
optimization of, 142
tray design for, 146
vacuum, 138
Fractionator column, diameter of, 140
Fugacity, 61
Gibbs-Duhem equation, 7, 9
Gilliland correlation, 135
Heat exchanger, 141
Heat of solution, 7
Heat transfer, boiling, 9
fluids of, 88
indirect, 132
Heat transfer to buried pipes, 111
Heat transfer coefficients, 43
Heat transfer of condensation, 9
Heat transfer in reactors, 171
Henry's law, 9
Hoop stress, 110
Hydraulics, distillation-tray, 140
Hydrogen, industrial sources of, 87
Hydrogenation, catalyic, 77
Incentive, financia, 10
Integration, numerical, 84
Investment, return on, 13, 196
Isothermal reaction, 39
Iterative calculation, 66
Kinetic data, analysis of, 22, 80, 161
Kinetic data, correction for diffusion
effects on, 170
Lennard-Jones force constants, 165
Lewis-Matheson method, 136, 148
Limiting case analysis, 4, 95
Linearization, 93, 177
Liquid -liquid extraction, 9
equilibrium data for, 9, 54, 60
process-design for, 64
Literature, methods for searching, 6
Market demand, 13
Market information, 5
Market price, 13
Market research, 5
Mass transfer to catalyst pellets, 9,
98, 163
Mass transfer to gas bubbles, 9
Mass transfer in packed beds, 162,
165
Mass transfer in slurry reactors, 9
Mass transfer in sparged reactors, 9
McCabe-Thiele diagram, 136
Mean free path, 168
Mixer-settler design, 58, 64
Murphree efficiency, 9
Nernst's law, 9, 61
Optimization, 3, 70, 80, 113, 116, 140,
182,193,194
Oral presentations, 14
Partition coefficient, 9, 61, 62
Parallel reactors, 35, 197
Petrochemical industry, 196
Phase diagram, 62
Pipeline, pressure drop in, 109
thermal expansion in, 105
transportation by, 102
Plug-flow reactor, 38
Pneumatic conveyance, 102
Pore radius, 167
Pressure drop in fractionators,
137-139
Pressure drop in packed bed
reactors, 98
Pressure drop in pipelines, 109
Pressure-enthalpy diagram, 141
Process economics, 11
Process engineering fundamentals
1-17
Process flow sheet, see Flow sheet,
process
Production control, 8
Programming, discussion of, 146
dynamic, 11
FORTRAN, 146
linear, 10
Quality control, 8
Raoult's law, 7, 9
Raw material costs, 39
Reaction, adiabatic, 87
catalytic, 162
equilibrium, 9
isotherma. 39, 87
kinetics of, 9
temperature of, 39
yield of, 40,152,192
Reaction conversion, 80, 85, 97, 152,
192
Reactor, bench-scale, 152
cataytic fixed-bed, 9, 38, 80, 171
continuous stirred-tank, 11, 24
parallel, 35, 197
pressure drop in, 98
shell-ad-tube,l71
slurry, 9
sparged,9
Reactor control, 42-48
Index 23]
Reactor design, 24,41, 81, 86, 171
Reactor flow sheet, 25, 89, 171
Reactor stability, 42-48
Recycle, 38, 87, 97
Reflux ratio, 134
Residence times, 38, 58
Resistances to catalytic reactions,
162
Results, presentation of, 14
Return on investment, 13, 196
Risk analysis, 13
Safety, 43
Sales forecast, 5
Scale-up, 87, 95
Selling price, 13
Simpson's rule, 84
Simulation, 3, 8
Slurry, pumping of, 102
Slurry reactor, 9
Smoker equations, 147
Solubility, gas in liquid, 9
liquid in liquid, 9, 54
solid in liquid, 9
Solubility data, 7, 9
Solution, heat of, 7
Sparged reactor, 9
Stability, reactor, 42-48
Stability of finite difference equa
tions, 175
Staged process, economic evaluation
of,72
Statistically designed experiments,
8,60
Steam costs, 58
Steepest ascent, method of, 11
Stoichiometry, 80, 172
Stripper design, 67
Syrene, industrial manufacture of,
126,129,146,197
Suboptimization, 69
Sulfur, mining and production of, 101
properties of, 106
Supersolubility, 9
Surfactant, 73
Temperature of reaction, 39-41
Thermal expansion, 105
Thermodynamics, first-law balance
of,82
Thiele modulus, 9,166-168
Tortuosity, 167
Transportation costs, 123
Transportation of liquids, 101
Triangular phase diagram, 62
232 Tie Inustrial Practice of Clell1 ical Process Ellgineerill
Underwood equations, 147
Vacuum distillation, 138
Van Heerdon criterion, 45
van't Hoff relationship, 9
Vapor-liquid equilibrium, 9, 57, 132
Variables, critical, 10
selection of, 60
Venture analysis, 12
Viscosity of gases, estimation of,
164
Visual aids, 14
Volatility, relative, 9, 134
Writing, technical, 14
Yield, reaction, 40, 152, 192