The Industrial Practice of Chemical Process Engineering

The Industrial Practice of Chemical Process Engineering Samuel W. Bodman

The MJ.T. Press Massachusetts I nstitute of Technology Cambridge, Massachusetts, and London, England

Copyright

©

1968

by

The Massachusetts Institute of Technology Printed and bound in the United States of America by The Maple Press Company, York, Pennsylvania. All rights reserved.

No part of this book may be

reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher 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 fundamentals to practical chemical processing problems.

In 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 encounters

difficulties in becoming acclimated to the environment 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 fully for the challenges of industrial work, the M.I. T. Chemical Engineering Department has developed two specific programs.

First, tht. department's School of Chcmical Engineer­

ing Practice exposes graduate students to actual problems in an industrial environment under 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 design course in his tell.i "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 required 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 the Practice School program with elements of Professor Sherwood's case-study approach to in­ struction in engineering design.

It is hoped that the result of this combination

will not only be useful as a text for academic process design instruction but will also serve as a reference book for the young engineer embarking on a career in industry. iii

iv

Preface An introductory chapter briefly describes the interaction and inter­

dependence of market research, process design, and financial evaluation functions in commercializing a chemical product.

Methods of planning and

analyzing laboratory experiments, of utilizing market and financial infor­ mation, and of preparing and presenting a chemical process design are discussed.

The introduction is followed by a series of six case studies in

engineering design, typical of those encountered by a young engineer in his initial industrial assignments.

The need for accurate analysis and correla­

tion of laboratory data is given great emphasis.

Secondly, the need for

devising creative and practical solutions to processing problems is given some discussion.

Finally, each case illustrates methods of combining tech­

nical and financial information to provide a realistic evaluation of a proposed process.

Each case is concluded with a recommended design as well as

suggestions for further 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 in the analysis and presentation of design problems; computer programs are presented for three of the cases discussed.

These programs

and others like them have proven to be particularly stimulating when used in a "computerized classroom, " wherein the student can communicate directly with the machine.

In preparing portions of the text, it has been assumed that

the reader has at least some knowledge of the FORTRAN coding language. The present volume was written while the author served as Director of the Bound Brook Station of M. I.T.'s School of Chemical Engineering Practice. This station is located with the Organic Chemical Division of the American Cyanamid Company,

Bound Brook, 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 Bound 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 substantially to the case studies summarized in this book.

First of all, thanks are due to Professor Thomas

K. Sherwood, who has continued his guidance in the instruction of chemical process design at M.I.T.

It was he who originally suggested the preparation

of the present text, and Chapters 2 and 4 are based almost entirely on design cases developed by him for his process design courses.

Sincere appreciation

is el\.'Pressed to Professor Robert York of Cornell University for his guidance in the economic evaluation of chemical projects and for his instructive re­ view of portions of the present text. Professors C. J. King and Scott Lynn of the University of California in Berkeley were kind enough to review the entire text of this book.

Their comments were particularly significant in giving

additional breadth and meaning to several of the case studies.

Dr. Howard

Kehde and his colleagues at Dow Chemical Company, Midland, Michigan, very kindly reviewed the cases concerned with styrene production; their comments provided an invaluable dimension to those cases by bringing to bear the knowl­ edge and sophistication of a major producer of styrene monomer. Professor Giles R. Cokelet of California Institute of Technology provided a valuable commentary on the sulfur transportation case. Professor R. G. Thorpe of Cornell University originally acquainted the author with some of the difficulties

Preface

v

associated with the vacuum fractionation of styrene-ethylbenzene mixtures; this information was most useful in the preparation of Chapter 6.

The data

used as a basis for Chapter 3 were gathered under the direction of Professor Michael Modell of M.I. T.; his cooperation in making this in­ formation available is greatly appreciated.

Several former teaching assist­

ants and students at M.I. T. contributed substantially to the development of computer programs for Chapters 5, 6, and 7.

In this respect, special

thanks are due to Robert L. Blumberg, Bruce Crocker, Avelino R. Rodriguez, and Robert L. Sandel. Machine computations for this book were performed at the M.I. T. Computation Center.

The author gratefully acknowledges the financial support

of a research initiation grant from the Rohm and Haas Company which was instrumental in the preparation of this text. The author wishes to express his thanks to Messrs. Georges F. Doriot, Henry W. Hoagland, William H. Congleton, John A. Shane, and Miss Dorothy E. Rowe of American Research and Development Corporation.

Their guidance

during the past four years in various aspects of financial analysis and business

j udgment

contributed substantially to the present work.

The greatest source of encouragement and assistance in the preparation of this manuscript has come from the author's wife Betsy.

Without her many

very thoughtful and skillful contributions, this work could not have been com­ pleted. S. W. Bodman Cambridge, Massachusetts March, 1968

Contents Preface 1.

Introduction

2.

Reactor Design, Optimization, and Control in the Production of Monochlorobcnzene

3.

iii 1

18

Process Improvement for Liquid-Liquid Extraction of Fcnway AcW

51

4.

Catalytic Reactor Design for Benzene Hydrogenation

75

5.

Evaluation of New Methods for Sulfur Transportation

100

6.

Vacuum Fractionator Design for the Purification of Styrene Monomer

7.

125

Process Design and Evaluation for the Production of Styrene Monomer

151

Appendix

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. 2. 3. 4. 5.

Recognition of an economic opportunity Conception of a plan or design Preliminary analysis of the design Completion of a final design 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. In 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 specific 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 performance 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 variations 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 difficult 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 difficulty 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 exploited 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 likely configuration 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 varied 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 design. 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 guide in making 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.

Establi�hment 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. DETERMINATION 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 thinking 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 design 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 example, 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 another, one can assess the need for making 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 expenditure 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 many years of experience, and the novice often finds it quite difficult 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 in establishing preliminary limits on the process variables but also in 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. If 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 MARKET CONDITIONS

Cooperation between the design 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. On the other hand, an accurate sales forecast for a product is usually quite critical to a realistic prediction of the ultimate financial per­ formance 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 discussions 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 informed on the status of the market estimates. Only in this way can a final design 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 design 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 in 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 Arr.erican 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 only clarify the situation for the student but would also hopefully lead to better market analysis techniques for the in­ dustry as a whole. EVALUATION OF

EXPEIUMENTAL

DATA

In the manufacture of a particular chemical, the required process steps generally follow the sequence shown below:

1. 2. 3. 4.

Preparation of reactants Carrying out of reaction ( s) Heating or cooling of reaction products Separation of reactants from products and purification 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 exposure 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 f!uality of the information to be used. For example, data reported many 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 experimental 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. In 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. If 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'ltive 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 with 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 m�ant 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 theore tical developments. The utility of the information summarized in Table 1-1 naturally varies significantly from one segment of the chemi.cal 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 experimental data is developed, the design calculations to op­ timize the number of extraction or leaching stages can proceed quite smoothly. When theoretical correlation of process information is impossible, the use of a factorially designed experimental technique may be of great value. The use of statistically designed experiments is parti­ cularly valuable in reducing the required amount of experimental and analytical effort to solve a problem for which there is little or no theoretical basis. The following references, arranged by Koehler, provide 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 experimentation (22), to the use of computers in data re­ duction (41), to the selection of appropriate production-line control charts ( 1 7), 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 expenditure of a large capital investment is required. ESTABLISHMENT OF THE CRITICAL DESIGN PARAMETERS­ SIMULATION AND OPTIMIZATION

In his text, Sherwood (41) states quite appropriately that the de­ signer must be willing to make assumptions. Once sufficient informa-

Useful

Bases for Evaluation or Correlation of Data

Literature References

Methods of Evaluating and Correlating Chemical Engineering Expe rimental Data

Required Design Parameter

Table 1-1.

Process Operation

=

=

=

F(Nl T2' and T3' each 70°C 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 reacto r s at 5 0°C . 60°C . One po ssible combination is Tl 70°, T2 = 6 0°, and T3 From Eqs. 2 . 5 and 2 . 6 a minimum annual cost of $ 750, 000 at a feed =

=

Reactor Design, Op timization, alld Control

41

rate of about 1 1 0 lb mole s/hr can be c alculated. Figure 2-5 compar e s the results o f this combination with tho se for isothe rmal operation. Pursuit of further combinations of reactor temperature s is clearly a job for a computer, since the possibilitie s are too numerous for slide- rule calculation. Formulation of the problem fo r a digital com­ pute r is re latively simple, s ince the cost expres sions are not com­ plicated. The rate constant s must be expre ssed as function s of te m­ pe rature, e ithe r by e mpirical polyno mials 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 se milogarithmic graph, and may be expre s se d by kl

=

5. 1

x

1 01 2 e-19600fRT

(2 . 2 3 )

1< 2

=

2. 9

x

1 0 2 0e-3 2 600/RT

(2 . 24)

where T is now in oK. An incomplete compute r search of tempe rature sequence s in­ dicate s that the re is little to be gained by temperature variations, even though total annual costs in the vic inity of $73 5, 000 to $740, 000 can be attained. The principle that the temperature should be reduced as the reaction proce eds is sound [ see Denbigh (4), Chapter 5], but the obj ect is now minimum cost rather than maximum yield. The cost data applicable to the pre sent analy sis are such as to give very flat optima in the economic evaluation curve s . T h e selection o f a n optimum te mpe rature sequence is a classic for the more advanced optimization technique s mentioned in Chapter 1 . A mor e sophisticated approach to the selection of reactor tempe ra­ tures would provide a wo rthwhile exposure to some of the se mo dern technique s . SUMMARY OF THE REACTOR DESIGN

The minimum total annual cost for MCn p roduction is approxi­ mately $ 740,000; this inc lude s pu rchased chemicals, recy cle and pro ­ duct re cove ry, ope rating labo r, and fixed charge s, and would re sult in an annual saving of $ 1 00, 000 as compared with use of monochloro­ benzene purchased at $0. 1 05/lb Optimum operation is at about 60°C with a benzene fee d rate (including recycle) of 80 lb moles/hr (14. 2 gallons/min) . Ope rating at this tempe rature and flow rate should make it possible to produce the requi red amount of mono­ chlorobenzene in only 1770 hr /year . A slight gain c ould be made by operating successive reacto r s at lowe r temperature s , but the po ssible cost reduction would not appear to j ustify the more complicated control syste m that would be needed. The total annual cost is relatively in sensitive to feed rate and would inc rease to only about $775, ODD/year if the fee d rate at 60°C were doubled and the yea rly operating hours cut fro m 1 770 to 1 290. This means that the eqUipment could be made available much of the year for some othe r use . .

42

Chapter 2

It should be pOinted out that the pi'oduction of dichlo robenzene r epresents a substantial cost for chemicals . Its recovery and puri ­ fication would not b e ve ry difficult, and the proposed process would look very much more attractive if a market could be found for a few hundred thousand pounds per year.

HEAT-TRANSFER AND REACTOR-STABILITY CONSIDERATIONS

With the basic nature of the proc e s s established, it is e s sential that the reactor cooling system be examined in some detail. The adequacy of the cooling medium and the available heat-transfe r area must be asce rtaine d before the reaction equipment can be recommen ­ ded for MC B production. Of equal importance , the po s s ibility that the reacto r ope ration might be unstable unde r the proposed operating conditions should be inve stigated. To compute the rate of heat release in each reactor, the following enthalpy c hange s of the chlorination re actions must be calculated from the the rmodynamic data on the reactants and products: C6H6 + C12= C6HsCI + HCI

(a)

1 1.70

+

0 = 2 . 50 -- 22 . 0 6

Ml =

(b)

C6HsCI + Cl2

2 . 5 0 +0

=

=

MI =

- MI

- 3 1 . 26 Kcal C6H4Cl2 + HCI

-4 . 90 - 22 . 0 6 -

MI

-29 . 46 Kcal

The rate of heat e volution in a single reactor may now be computed as Qn

=

=

3 1 , 2 6 0 ( 1 . 8) (NJl1xA) + 2 9 , 4 6 0 ( 1 . 8 ) (NRk1 xB ) 2, 1 2 0, 000 k1 xA +2, 000, 000 1

0.02

1\

'\

\.

0.0 I

o 140

160

"

180

"'-

200

I'-..

220

I'--.r-

240

260

280

r-'

300

320

Temperature I °C

Figure 5-1. Viscosity of liquid sulfur (low­ temperature range). (From Tuller, reference 9) 70

60

,

'"

�

h\

50 �

u .,

40

D ,..

;;; 30

§

:>

20

10

)

�oo

1\ \ \ ''\

400

'

,,---

500

600

Temperature 1°C

Figure 5 -2. Viscosity of liquid sulfur (high-temperature range). ( From Tuller, reference 9)

107

108

C/w/)/eY 5

Both figures show a sharp increase in liquid viscosity at a tempera­ ture of approximately 320°F. This increase has been attributed to the rupturing of S8 rings to form polymeric-type chains of various lengths. Above 370°F the chains are thought to break into smaller segments; the continued chain breakage would lead to the observed decrease in 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 specific 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 (3) and (9) contain all the information regarding the properties of liquid sulfur which will be necessary for the design calculations.

Assumptions to

be Used

in

the Design

In addition to the approximations and Simplifications previously discussed, the following assumptions will be employed to complete the preliminary design: 1. Saturated steam at 85 psig (Ts a t 327°F) will be used as heating medium, and water at 15 psig (Tsat 250°F) will be with­ drawn from the annular section of the line. By this means, the sulfur will be maintained above its melting point of 246°F but below the tem­ perature range that produces the sharp viscosity increase. 2. The rate of heat loss through the insulation on the steam line will be assumed to be constant along the length of the line. 3. The fluid mechanics within the steam annulus will not be in­ vestigated in the present study. A theoretical solution for the pres­ sure profile in the annulus would make a relevant and interesting pro­ ject; this analysis will be required in the final design stage. Because of the simplification 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 sufficient 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 investment, both the diameter and schedule number of the pipe are varied along the length of most commercial oil and gas pipelines. For Simplicity,this practice will not be followed in the present preliminary design, and both and t 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 t. =

=

D

D

Sulfur Trallsportation Tcclmiqucs

109

DESIGN COMPUTATIONS

The sulfur pipeline is shown schematically as follows:

To barQes for shipment

where the cross section of the pipe may be visualized as: _

.-

Insulation

02

Steam annulus

Ds

D,

Sulfur pipe

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 fixed costs for the system will be expressed as functions of P, and L. Making use of this expression, the optimum set of design parameters will be selected.

D,Di,

Fluid Mechanics Considerations

The sulfur flow rate has been fixed at x lb/year or lb/hr. Thus the velocity and Reynolds number are given, res­ pectively, by

7. 5 108 85, 500 85,500(144) U 3600(112) (0. 78 5D2) 38.9 D2 112 7.27 104 D 38.9 12 D2 0.005 D =

Nne

------

DUp

=

--

J1.

=-

x--

=

/

fps

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

=

) 41 (U2)(� 2gc D

11 0

Chapter

where f

5

0.055/NRcO.2. .2 (38.9)2 112 � (12) (5280) 4(0.055) DO 144 (7.27 104)0.2 2(32.2) D 2.72D�·8 104 i/ (5.1) Implementing this equation, we have

=

p L

D4

x

P

X

=

L

ps mile of pipe

The total power for pumping is given by

OP J°

o

W

0

=

�

vdP vP Hence, 85, 500 50 144 3600 112(550)(0. 75) 3. 70P/ 074P by 3.70P (0.745) (8750) (0.01) 241 P/ Weight of Steel Required for Pipeline DP DP 30,000 2S weight =

o

x-x Px L

w=

L

=

hp (�O.

------

hp per station )

The total annual cost for pumping Cpower

=

is

given

--

=

L

L

dollars/ year

Sulfur Pipe. From the standard hoop stress formula, the wall thick­ ness of the sulfur pipe is given by t

=

-

and the

W1

= ---

in.

of sulfur pipe is computed as

=

•

t (D + t) (7.7) (62. 3) (52 80) (50) 2000 (144)

EliD,minating 0.0460 D2p 0.2 in.,

the wall thickness in favor we have

t «

WI

=

=

of

1380t (D P,

+

t)

tons

and assuming that

tons

the

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) .(D

+

2.4) (7.7) (62.3) (5380) (50) 2000 (144)

=

275 (D + 2.4)

tons

Sulfur Transportation Teclmiques

111

Total Steel. The total amount of steel is thus written as W steel

=

0.046 D2p + 275 (D

+

2.4) tons

Heat-Transfer Considerations

We have the following heat-transfer problem:

" Earth surface at temperature TA

1 !

Dj

where steam is condensing on the inner wall of the steam pipe having a diameter Ds' For steady-state conduction from a buried pipe,McAdams (4) gives the relation Q=

1

4Z In-

Btu/hr

Di

where Z» Di (see reference 1 for a derivation of this equation). Note that kg is the thermal conductivity of the earth,given as 0. 40 Btu/hr-ft OF and that X 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

112

Cl[{[pter 5

Taking average values of Ts 0. 05, Z as 6 ft, we have

285°F, TA as 60°F, I--

g o

0.7 0.6

I �

0.5

:

0.4

g

0..3

ler

7

BZ

Mole fraction of benzene in gas phase

CATCST

Total capital investment required for catalyst in two reactors

CATDEN

Bulk density of catalyst bed, lb m/cu ft

CATDIA

Diameter of catalyst pellet, ft

CATPRC

Price of catalyst, dollars/lb

C PEB

Heat capacity of ethylbenzene, Btu/(lb) (OF)

CPFG

Heat capacity of flue gas, Btu/(lb) (OF)

DEN

Density of reacting gas mixture at any point in the reac­ tor, lb/cu ft

EB

Mole fraction of ethylbenzene in vapor phase

E BPRCE

Price of ethylbenzene, dollars/lb

EK1

Equilibrium constant for Reaction 6.1, atm

EPS

Bulk void fraction of packed bed, dimensionless

ETA

Effectiveness factor, dimensionless

FRATE

Feed rate of ethylbenzene, lb/(hr) (tube)

H

Over-all heat-transfer coefficient, Btu/(hr) (sq ft) (OF)

HYD

Mole fraction of hydrogen in vapor

I

Integer subscript controlling the numerical calculations; I advances by 1 for each half step in distance for which the calculations are completed.

J

Integer variable that controls the value of the temperature, pressure, and composition variables to be used in taking 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 = 3).

K

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 .

M

Integer equal to 1/2; when M exceeds MM, the temperature, pressure and composition at that point in the calculation are printed.

MDEL

Input integer variable which controls the frequency of printing composition, temperature, and pressure values

157

Process El'aluation for Styrene Production

as the calculation proceeds. MDEL was set equal to 1 0 for the example shown i n the appendix. MM

Integer used to control printing of output, see definition of M

NSLICE

Number of distance increments used for the numerical calculation. A value of 100 was used in the example shown in the appendix.

PRESS

Gauge pressure of reacting gas at any point in the reac­ tor tube, psig

PRESSG

Absolute pressure of reacting gas at any point in the reactor tube, pSia

PREXG

Gauge pressure of reacting gas at exit of reactor tube, psig

PREXIT

Absolute pressure of reacting gas at exit of reactor tube, psia

RATE FG

Flow rate of flue gas, lb/(hr) (tube)

RCCOST

Cost of recycling ethylbenzene, dollars/year. Used putations for first economic basis

RCPRICE

Price of recycling ethylbenzene, dollars/lb

RCTCST

Capital cost for two shell-and -tube reactors, dollars

RK1

Kinetic constant for Reaction 6. 1 , lb moles/(hr) (atm) (lb catalyst)

RK2

Kinetic constant for Reaction 6.2, lb moles/(hr) (atm) (lb catalyst)

RK3

Kinetic constant for Reaction 6. 3, lb moles/(hr) (atm2) (lb catalyst)

SLICE

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, of

STY TDEL TEMP TEMPG

in

com­

TID

Absolute temperature of reacting gas at any point in the reactor tube, oK Internal diameter of reactor tube, in

TLEN

Total tube length, ft

TEMPK

TOD

Outer diameter of reactor tube, in

TOL

Mole fraction of toluene in vapor

TOTCST

Total operating cost for first economic basis, equal to RCCOST + YLCOST, dollars/year

158

Chapter 7

TUBES

Number of tubes required for the reactor

VEL

Velocity of reacting gas at any position in reactor tube, ft/sec EPS)/EPS**3

VOID

(1. 0

X

Total convei'sion of ethylbenzene, equal to STY+ BZ + TOL

XI

The reciprocal of the fraction of TDE L 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.

XMOLE

1. 0 + BZ

Y

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 $ O. 1 2/lb ethylbenzene lost, dollars/year

Z

Distance along the reactor tube, ft

-

+

STY

DESIGN OF TIlE REACTOR FOR CONVERSION OF ETIIYLBENZENE TO STYRENE 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 co mpleted. As an integral part of the reactor design, it will be necessary to consider various techniques in the nu merical 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 benzene

of

the Catalyst Used

for

the Dehydrogenation

of

Ethyl­

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 determined reaction rate data. Both the effects of mass transfer to the catalyst pellet sur-

Process Elyliualioll foy Styrene Production

159

face and diffusion within the catalyst pellet must be considered . If mass transfer is found to have li mited 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 fro m the orig­ inal article: Bulk density: 6 1 lb m/cu ft Pellet size: 4-8 mesh, standard Tyler screen size. From Perry's Handbook (7), we find that O.

185 in .'" 3/16 in . 8 mesh'" 2 . 36 mm '" 0. 093 in. '" 3/ 2 in. 3 4 mesh'" 4 . 70 mm '"

As an average pellet diameter we may then take cl'j>

=

1 -

2

(

- ) = %4

3 3 -+ 16 32

in.

= O.

1 4 0 in.

=

0 . 0 1 1 7 ft

This information constitutes the only specifications 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 platinu m 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

=

240 m2 /g

e

=

0. 5 9

=

1 . 3 3 g/cm3

Pp PT

=

3 . 2 5 g/cm3

where S g is the surface area of the catalyst, 0 the porosity of a catalyst pellet, Pp the density of a catalyst pellet, and P T the intrinsic density of the solid material that comprises the pellet. For this catalyst, the pore volume per mass of catalyst, Vg, is given by Vg

=

e /pp

=

1 -

Pp

1 - -

PT

=

O.

5 9 / 1. 3 3

=

0 . 443 c m3/g catalyst

Now consider a container of volume VB' packed full with one gra m of catalyst pellets. If the pellets are assumed to be spheres, it is reason-

Chapter

160

7

able to take the void volume between pellets, Vy, as O. 38 of the total volume of the container. To find the total volume of the container:

v,

=

O. 443 cm3/g catalyst

Vy

=

0.38 VB

Vr

=

1/Pr

=

0.751

VB

=

0.308

+

0.38 VB cm3/g catalyst

Hence, VB

=

1. 21 cm3/g catalyst

(:. )

or the bulk density of the catalyst bed is "

=

(62.4)

=

51.5 lb/cu

ft

This number is somewhat less than the value of 61 lb/cu the original article. However, if the property values e

=

Pr

=

Pp

=

ft

given in

0.50

1.57

3.15

are taken, the container volume can be computed in a similar manner to be 1 . 024 cm;$ /g catalyst . A value of 61. 0 lb/cu ft then follows for P 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 (12).

General Examination of the Kinetic Data

Before proceeding to analysis relating to the mass transfer and diffusion reSistances, it is interesting to examine the kinetic data shown in Table 6- 1 for the possibility that chemical equilibrium may have limited some of the observed conversions. The equilibriu m ex­ pression for styrene formation is written as

Process Evaluation for Styrene Production

161

Values of J( 1 are shown in Fig . 6-1. If the side reactions are neglec­ ted for the moment, the conversion of ethylbenzene may be calculated as y'2 S

---

1 -

Ys

The equilibrium conversion of ethylbenzene, as computed with this equation, is shown in Table 7 1 From the comparison of (Ys) E(,I and (Ys) EXP' it is clear that the measured conversions in runs 2 2 7, 2 2 8, 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 (12) 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 k 1. For example, by doubling the length of the experimental reactor, the same equilib­ riu m conversion would be obtained; however, the computed value of k1 from this hypothetical experiment would be quite different from that obtained from runs 2 2 7 to 2 30. -

Table

7-l.

.

Comparison of Observed Conversions with Those Computed Assuming Chemical Equilib rium (atm )

(YS)EQ

(Ys) EXP

0. 082

0.2 4 6

0.2 2 6

5 52

0. 082

0. 2 4 6

0. 22 5

2 29

555

0.082

0.2 50

0. 2 2 7

230

556

0. 082

0.2 50

0. 2 27

237

598

0. 2 2

0.3 6 7

0.2 4 6

239

600

0. 22

0 . 3 72

0. 269

2 52

676

0. 9 5

0 . 5 80

0. 366

2 54

676

0. 95

0. 5 80

0. 3 60

242

5 98

0. 2 2

0. 3 72

0.2 60

246

650

0. 64

0. 500

0. 3 10 0. 3 5 5

T (OC)

[(1

227

555

22 8

Run Nu mber

251

671

0. 93

0. 5 7 5

255

6 68

0.93

0. 60 1

0. 3 72

232

555

0. 082

0. 2 50

0. 1 9 5

Following u p o n 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 1�1 shown in Fig. 6 -2

162

Cha/)ter 7

to a specific experimental run sum marized in Table 6-1. However, an attempt is made in Fig. 7- 1 to associate the kl values with specific experimental 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 2 2 7 to 2 3 0 i s seen in Fig. 7- 1. There i s great scatter i n the data a t 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 O. 1 4 to 0. 2 1 lb moles/hr led to the computation of kl values that were too high. There­ fore the data from runs 2 2 7, 22 8, 2 2 9, and 2 3 0 will be discarded in future computations. 3.lCrtr-----r----r--.

3 .1O·'':::-----l...:----.L,----...J 1.0 1.1 1.2 1.3 fOOOl T with T

In

·K

Figure 7-1. Kinetic constants for the dehydrogenation of ethylbenzene. ( Wermer and Dybdal, reference 12.) Identification of experimental runs has been inferred by examination of data tabulation in reference 12, Table 1. Before proceeding, it should be noted that a rather serious dis­ crepancy remains. A comparison of runs 2 2 7 and 2 3 2 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 increased conversion of ethylbenzene to styrene. Other than experimental error, no reasonable explanation for this behavior is apparent. General Considerations

in

the Analysis of Catalytic Reactions

In analyzing a catalytic reaction, seven possible resistances to the progress of the reaction must be considered:

1. 2.

Transfer of reactants from the bulk mediu m to the catalyst pellet surface Diffusion of reactants from the pellet surface to the interior

Process Em/l/atiol1 jm' Styrene Productioll

163

of the pellet where most of the catalytically active area is found Adsorption of reactants on the catalyst surface Reaction upon the catalyst surface Desorption of the products Diffusion of products from within the pellet to the pellet sur­ face Transfer of the products from the pellet surface to the bulk medium

3. 4. 5. 6. 7.

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 (10) 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 (12). However, com­ putations will be made to analyze for possible mass transfer or dif­ fusional resistances that may have influenced the experimental data of reference (12). Mass-Transfer Rates at the Surface

of

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 jD factor. Experimental data have shown that the best expression for JD is that given by (9) jD

=

k lIip L-1.. NO .66 7 Sc G

=

. 41 for .l"ite > 3 50 0 989 N-O Re

(7. 2a)

=

1. 82 �i�('·51 for .l"iic < 3 5 0

(7. 2b)

•

As the first step in the computation of jD it is necessary to estimate the Reynolds nu mber for the flow over the catalyst pellets. In this calculation, the lowest mass flow rate obtained will be used (run 2 42); 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 �tc is (a) Mass velocity,

G

For run 2 42, G

=

(0. 0064) (106) (0. 00307) (3 600)

=

0 . 0 6 1 3 lb m /sec-sq ft

Chaptey 7

164

(b)

Viscosity,

/.1.

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 ethylbenzene is probably quite similar to that of toluene, which is listed in Perry's Handbook (7 ) as 0.0202 cp at 598°C. 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) Tc = 346°C and Pc = 38.1 atm. For a Zc value of 0.27 (equal to that for toluene) , ZcRTc 0.27 (82.05)(619) cc =360 --Vc = -- = g mole Pc 38. 1

Therefore, /.1.

=

0.00333 (106

x

619)1/ 2 (1.137)

(360) 2 /3

= 0.0191 cp

which agrees quite well with the value for toluene. (c)

Reynolds number and jD N

-

Re -

dpG

-;;

(0.0117)(0.0613) ---,-� - ��-(0.02)(6.72 x 10-4)

- --

-

=

56

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): D1

2

=

0.001858 T3/2 (M ave ) 1/2 ...: ::...:.� Pa�2 ilD

-------

­

(7.4)

Process Evaluation for Styrene Production

165

The Lennard-Jones force constants are estimated by the conventional empirical method, whereby a1 2

E/k

=

0.833 Vl /3

=

O.

77Tc

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: a12

=

0.833 (360)1/3

E/k

=

0.77 (619)

llT/E

=

871/476

=

=

=

5.92 A

476°K and

1.83

Thus, from Reid and Sherwood (8) the collision integral is found to be QD

=

1. 116 - 0.6 (0.011)

=

1. 109

The diffusion coefficient can now be computed as

[

0.00 1858(871)3/2 =

DES

104+106 1 /2 104 X 106

J

(1.0)(5.92)2 (1.109)

:"-00.168 cm2/sec

And the Schmidt number is found as Ns c

=

Jl/p DES

=

0.0191 102 (1.47

x

10-3)(0.168)

=

0.772

The mass transfer coefficient can now be calculated:

kg

=

(0.234)(0.0613) (105)(0.9)(772)2 /3

=

1.73

X

10-4

lb moles -----

sq ft-atm-sec

To compute the possible diffusional resistance in the kinetic experiments, the superficial area of the catalyst pellets must be determined. If it assumed that the pellets are uniform spheres of diameter cJ" then the porosity of the bed is 0.38 and the total superficlal area may be computed: a =

(0.62)(71')(0.0117)2 1/671'(0.0117)361

=

5.12 sq ft/lb catalyst

Chapter 7

166

The observed reaction rate in run 242 is r =

0.00640 (0.26) 3600(0.749)

=

6. 19

X

10-7

lb moles EB reacted ------

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 (P,If -p,.) =

6.19 (1.73

x

X

10-7

10-4){5. 12)

= 7.00

X

10-4 atm

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 Within the Catalyst 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 1), the effective­ ness factor. The parameter 1) is equal to the ratio of the actual reac­ tion rate to the reaction rate that would have been observed if 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)

[1 IJ

3 -1)=¢ cp tanh cp --

(7.5)

where

(7.6) The variable ct> 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

1.0 0.9

Elytiualioll jor

Styrene Prodllction

167

\

\

0.8

1\ \

\

1\

\

0.4 0.3 0.2

2

o

4

"'"i'-

6

Thiele modulus t

8

cP

......

'-....

10

12

Figure 7-2. Effectiveness factor as a function 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

--

Deff

=

1

(D12)eff

+

1 (DKn)eft

(7.7)

The effective molecular diffusivity is defined as D12B/T, where T, 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 Pp may be thought of as the mean pore radius in the pellet, re' Equation 7.8 has been derived from the kinetic theory of gases for cases when re is less than the mean free path of the gas under consideration. From Eqs. 7.5 through 7. 8 the value of 1/ may be computed for each experimental point of Wenner and Dybdal . This trial-and-error computation is illustrated in the following section.

168

Chapter 7

Computation of Effectiveness Factor for the Experimental Data

From the definition of ( D12 ) eff there follows (D12)eff =

(0.16�(0.5� 7

0.084 = -- cm2 /sec 7

And from Eq. 7.8 we have (D Kn )

I( 6.L

> -2

Since the value of I( 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 -2 6.L < ­

(7 . 2 5)

I(

Equation 7 . 2 5 then provides the basic criterion for the stability of the numerical solution. To implement it, 1( is first calculated as I(

=( aa/x )

=

-1. 3 5 1l 1 P

[1

+� (2x 1( 1

[

- Z)]

The largest value of K would be found when Kmax

9!

-1. 3 5 (9

x

10-3) (3.79) 1

+

)J

7 3. 9 ( 1,1 8 x 10-2

And the maximum value of 6.L to be used in the numerical calculations is ( 6.L) m a x

=

-2 1. 14

=

1. 7 5 ft

Chapter 7

1 78

5

Since a 6.L value of ft was used in the calculations sum m arized in Table 7- 5, the app roxim ate analysis just carried out would indicate that this value is too high. Sm aller value s of M will therefore have to be used to m ake a stable and ac curate solution to the problem pos sibl e . It was probably intuitively obvious that 6.L was too large 5 in the cal culations shown in Table 7 ; however , the fo regoing ana­ lysis provide s at least a s e miquantitative basis for sele c ting an app ropriate value of the distance increm ent. It should be pointed out that there are m any other te chniques that might have been applied in the num e rical solution to this p roblem ; for example the m e thod of l\I ilne or that of Runge - Kutta m ight have been used. Frequently these techniqu e s are available in computation cente r s in the form of " canned" program s . If applied in the pre sent problem such program s might well reduce the requirem ents for computation tim e . -

Preparation of a FORTRAN Program for the Reactor Design

A computer program for the r eactor de sign is prepared by solving the finite difference equations , E q s . 7. 18 to 7 . 2 2 . One additional equation , a relation for the p r e s sure drop in the reacto r , is r equisite for an accurate solution to the problem . The equation to be used in the final design calculations is that given in reference {l) : dP

dL

=

3 5 .

p v2 dp

(�) 1 -

€

( 7 . 2 6)

In solving the system of finite difference equations, the " m arching" 5 technique used in Table 7 will be m odified so as to insur e a num e r i ­ c a l solution that i s n o t only stable but als o highly accurate . This improved m e thod of solution, som etim e s de scribed as the improved Euler te chnique , consists of the following : -

1. 2.

3

.

A s befo r e , divide the interval of solution into N equal inc re ­ m ents of length . Since this is an initial value problem , value s of all the inde ­ pendent var iabl e s are known at the inlet. Substituting these value s into Eqs. 7. 18 through 7 . 2 2 and E q . 7 . 2 6 , the varia­ tions in the dependent variable s over half the first inc rem ent are computed. Using the values of the dependent variabl es computed at the midpoint of the first distance increm ent, E q s . 7 . 1 8 through 7 . 22 and Eq. 7. 2 6 are used to compute the changes in the dep endent variables ove r the entire first increm ent. Steps ( 1) and ( 2) are rep eated for all the distance inc rement s , i . e . , until l = N.

This technique m ay be expressed m athem atically as Xr + l - Xr

---- =

6.L

f ( Tr+ 1 / 2 • x r+1 /2 ' " /+ 1 / 2 ' z / + 1 /2 )

( 7 . 27)

Process EI'([lllalioJl for Sty rene Producli(}//

179

Analogous equations m ay be written fo r the other dependent variabl e s . B y u s e o f m ethods sim ilar t o tho se previously illustrated, t h e growth factor for the num erical s olution to E q . 7 . 27 m ay be shown to be � = 1 +

J( t:.L

(

J( t:.

L

1 + -2

)

( 7 . 2 8)

With a value of -1 . 1 4 for K, the m aximum value of t:.L that will en­ sure stability i s found to be 1 . 75 ft . Thus , the stability c riterion has remained unchanged from that for the simple r mar ching technique ; however , use of the improved Euler m e thod should re sult in a m arked improvem ent in the ac curacy of the solution . A FORTRAN program m ay now be prepared so that design cal cu ­ lations can be carried out for wide range s of the param ete r s th:l

::::

� .....

� -.

"'(

g '0-

CI)

"'(

�

� 2i

"'( c

'"

..... -.

!2;:;

g

..... C:l e;,

36. 976

Inlet Press. psig

4 . 070

4. 2 78

E xit Press. p sig

0 . 840

0 . 740

O. 794

Ope rating Cost on First B� sis

95. 618

1 00

123

156

6 64 6 3 0

876098

1 1 75543

1 64650

1 1 7 702

1 03 1 9 6

1 1 813 7

Reacto r Number of Recyc l e Cost, Yield Tube s in E q. 7 . 3 3 . C o st, E q . Cost, E q . Reactor 7 . 3 1 , $ /yr 7 . 3 2, $ / y r Dollars

93 . 9 7 6

Fraction S t yr en e Yield

9 1 . 841

6 2 . 694

55. 436

4 . 2 62

6. 411

5 . 882

0 . 444

0. 494

7 7 . 886

80. 246

80. 2 93

55

59

60

2 1 9087

2 6 01 8 9

267946

3 3 5 601

729252

63 22 6 8

6304 1 3

501791

96864

1 00504

86003

91110

10

F ract io n Unconverted E thy l benzene

Results of the Design Calculations for Reactors With 2 - in. Tu be Diameter

1656

3 . 849

Table 7 - 8 b .

Inlet Temp . OF 1 148

3 8. 2 1 3

Flue - Gas Inlet T e mp . OF

L ft 842

3 9 . 544

E xit T e mp . OF

10 1 651

82765

9 1 723

1 64 7

3 1 4449

97145

1 04 2 9 4

1 1 71

3 0 7 5 20

1 1 94

5093 3 8

3 94807

932

5 1 00 5 7

ID22

84

41 4948

10

8 9 . 093

84

10

0. 6795

74

2281 79 3 . 63 7

89. 3 1 6

40. 993

86. 677

1 643

0. 680

1216

0. 627

1112

5 . 594

4 . 644

1 67 9

5 1 . 845

1 6 73

4 9 . 973

1236

1 6 63

6 3 . 884

15

1 699

15

1 22 0

83. 656

842

0. 5 6 7

932

6 . 02 9

1 2 74

1 693

5 3 . 664

1 668

1 2 82

1 02 2 1112

1298

15

1254

66

842

O. 501

15

932

1 3 8209

1 03 6 7 1 4

90530

93424

1 76697 49

52 7 1 . 244

74. 6 1 9

20

0 . 3 82 0. 3 1 5

873638

20

5 . 231 5 . 339

66. 012 68. 081

1 686 1 6 80

1315 1334

1 02 2 1112

20 20

..... 00 C)

(')

--:

�

;::;>:> """

"'l

Table 7-8b continued

0. 988

0. 991

X• • Column Feed Rate Xu in lb moles/yr in Feed Distillate

3 . 516

3 . 289

3 . 1 07

2. 960

1 . 35

1 . 35

1 . 40

1 . 40

1. 45

66

66

65

66

66

43965

43937

49891

56044

68701

79905

79820

1 01 1 92

126958

1 68137

4241

4236

53 70

6738

8924

1 281 1 1

1 27993

1 58454

191742

245762

Optimum Column CorrespondiDg to Each React�r

0. 846

0. 985

3 . 51 5

593383 0. 61 1

0. 660

0. 963

0. 970

0. 976

4. 3 74

4. 013

3 . 736

1 . 30

1 . 30

1 . 35

65

66

65

32504

36034

3 9521

48650

49547

57358

68626

2330

2582

2630

3044

3842

78450

83383

84681

98437

1 1 1 790

Total Operating Co� for Column, $ /)'1'

0. 804

0. 980

4733 51 8527

0. 556

43894

Heating Cooling Cost, Cost, $/yr $/yr

Catalyst Cost, Dollars 1 3 1 0925

0. 756

0. 980

15 4230 454702

321 50

Depreciation Cost, $ /yr

6672 1028430

0. 704

15 3823

30226

Trays

10 5263 828933

0. 704

(O/D) (O/D). 1n

10 4280

683109

682430

Minimum Reflux

10

5361

3574

L ft

15

10

15

65

67664

65

61 083

1 . 30

1 999

1 . 30

37666

1 768

4. 427

33308

4. 767

66

27998

0. 955

1 . 25

26007

1 . 25

66

0. 962

5. 984

5. 274

0. 506

0. 930

0. 945

0. 549

0. 452

447385

0. 392

408609 36861 9

471 0

332309

5024

4420

20

4183

20

20

20

'tl

"( I;) .(ll' 486B'I. '7 221362.44 16� �'1� .44

PIPf COST

BELOW AS FUNCTIONS OF

��t;'NEfqING

ARF SHOWN

DESIGN CALCULATlnNS

"rPF r:tfa,r".c ...

ANNUAL

�UIFUR PIPELINE

12 1475.81 lZ18��.S6 13419S.81 1405SS.81 146'HS.8 1

102394.69 II'R75S.RI l 1S 115.81

1695,.94 RBlS.94 8q67S.94 Q6!'35.94

642H.99 705q�.q4

� 15 1'.9q �1 8 n .0 4

4�1�5.9q

261'15.98 H435.9q 3 9 79 '.9 9

INSULATION

310392.00 318722.31 3268.. 8.81 1347'15.37 3425'3.56 35022'1.69 3577"'1.50 365157.44

292'1'14.87 301827.06

2(431)'1.75 21832'1.63 211105.61 242'122.44 253'178.'14 264418.31 27434'.'14 283.52.6'1

STEAM COST !T0551.CO 1 886 1 5.06

87'1'18'1.25 9 12734.69 945378.31 '177'127.11 101038'1.50

847142.69

6U'I91.1'I 681.13.62 7lU'I1.31 14803".'1" 781150.56 814 "".QIl

61 6.16.12

TOUL COST 280"70..0.00 1688370.00 6IZ54.. . 87 552278.87 5"1331.19 559080.50 58615Z.'14

FORTRAN IV Program and Sample Printout for Chapter 5 203

nF

PIPE

7"''',:\'''.('1' 7-;""''''.'''('1

t-..�'" 11.C'r 1'1."'1"1 1Q.I''' ,1'1.,.."

1�."!'" 14."'''' l'i."1"!

�."" CJ.'"'' I"'."" 11.1'1' 12.,)'"

715��".�� l,:\,u"In.".,.

715"'''('1.(,''' 7-;"""."'"

71)t""('.�n 7,,(,,:,1"'.(,"

7':'t:'''r.'''f'

H""".C'C'

7-;"0"."'"

751"('0.1''':-

'''QQ''CJ.44

388664.5r. 4('7'162.1Q

'77312.31 2"56"7.7� 314l!1.94 H2n6.'14 351341.8 I

'�9�6R.12

241�'1.7�

nHP.I'I'

71)�('IC'.':'('I 7';"t:'n.,,1'

7151"I'Ir.f"'!('

t.en

1�5�9�.44

1�6R2r.12 16277R.31 17482�.44 1 R"714. 75 7C6C'''4.�6

ft."'''

1t;t:'�C'."'''

15'1�C'."t;' 7c;n,..�.,..,..

CIl�T

29�qlli(,!,'I.n,t;o 4Rftt"�.37 22136?44

PIPE

FUNCTIONS OF SULFUR

I.�

Ij."'"

4.�'"

nc"('.�I' 7S('Il"'n.M"

�S

F'I(;''IHRINr.

SHOWN BelOW

1.0" '- . "''' 3.�(,

4RE

•

1.1'

I"CHF �

=

f1U.INCW

AH'lUAL cns TS

1'1

DIIMOJN� ST'Trn��

IN�Ul'T'ON 'WICKNFS�

NU��E�

�IJI FliP PIPf1.INF OF�Ir.N CAI.CULATlnN�

PUMP COST

OJAMFHR

24el�93�.",! '1"H81.12 143076.75 47"96.89 26�e9 3C7186.56 326266.37 345347.50

PU"P COST 248316H.OO '1721ZI.12 159C16.75 �,"36.89 42829.55 36443.8"

31991.56 31958.1 7

364428.25 3835nR.25 4C2588.25 4Z1667.50 4�OH7.50 45'1827.50

34C57.62 31(121.15 32531.76 '2271.n� 32128.75 3Zt)43.a2

31936.13 3 1'121.18 31HO.78 319('3.1'1 318'18.1)5 31"94.11

SfEAM COST 120580.06 135'ZO.25 H97IT.62 162531.9. 1 7.,87.00 1855U.9� 1'6051.31 206078.81 215615.13 22.900.116 213801.87 ZU.ZO.OO 250788.75 258935.6Z 266885.50 27.658." 2822U ••• 2897.5.50 2970".94 304315.62

fOUL COST 2662'232.00 15266.1.00 6U503.50 5'1815.00 60660•• 56 6U80'.1I 6".51.12 U0311.25 116152.06 822201.00 868303.1. ."'33 •.•• 960Z52.81 10060.0.12 1051690.00 10'7206.00 11425'•• 00 1187861.00 1233016.00 1278061.00

�

0)

�

:g:

(\) � �

�.

�J�PTN6 STAT["NS

•

PIPF

COSTS

7.n.""

17.r." Jf'.t'l1"I ,q.l"!'"

3.�":' 4.1"" "'''' '3. 6."" 7.t:''':' tl.t'\1"1 9.0" 1".l"n tt."'!'" I z.n" 11.06

1 79 8 12 .5 0

966'1.61

2 41 68 7 .8 7

V> 1:1

n91.75

tR.""

4 6�9q.� n

182185.56

91'15 .56

2 H5 5 1 . 1 9

�

17�C66. 31

4� . � f"\

tA."f"I

46961.�n

18455R. 6 9

992 1 . 5 1

H1U2.69

4 7Q 9 � . 4 �

3.4 �

4'.""

lit."'"

473n.6b

1 869 31.8 1

1 0 1) 4 7 . 4 7

lA.I'\(\

,.Ii"

4":'."''''

1,.,.1'\'"

476'17. qq

IR93�4.94

2 5 0 212. 3 1

415.':'1'\

,.lie;

.'."1'1

Ift.n,..

481'6�.3n

1 0In.�2

253\30.12

'\.'."

,.It''

4,.,.r

lit."''''

191678.CO

10 2 99. 3 7

255 9B6.2 5

� C")

:\.qCj

'\.�"

'\.71i

4'."1'

4'1,"'"

1".l"If'I

16.f'\'"

4'14'12.0 7

I 9 B7 q

0)

�

I\)

;:,-

'c-

....

C $::

....

;:!

....

�

I\)

-

;:!

�

;:!

1:1

�

�

�

,.�t:;

4'\."'"

4 42 'l . R 9

1'141'5 1. 1 2

10 4 25.3 3

2 5 88 4 0 .62

.....

�

,\.7'"

11,.1\1\

4.17 '.�?

196424.25

10551.28

4"."'"

?

,\.f)"

A

4,.rl'll

tft.f'I"

49136.71

3\

� .... 2 0 1l 7t' . 4 4

2645��.31

2616 9 3 . 37 4 9R45. 7I

?("I3�41.5n

1 1)8 1) 3 . 1 9

10 6 71 . 2 4

.h.C'"

5 " 19 7 . 64

4,\.""

1",."'n

3.q�

4,"''''

�

'"

214

Appendix

FORTRAN IV Program and t

P

1 28 1 . 10

1 2 5 4 . 39

1 2 2 b . 42

1 6 7 1> . 6 4

1670.83

1 6 1> 4 . 6 7

1658. 1 8

165 1 . 35

1 1>4 4 . 1 6

1 1> 3 6 . 5 1>

4.935

1 2 . 6 33

1 8 . 35 2

2 2 . 997

2 1> . 9 2 5

30 . 3 2 3

33 .305

8 4 . 596

85. 551

88. 288

90 . 00 3

9 1 . 898

9 3 . 895

0

o . O vO

0 . 6 0 1 49

9 5 . 864

1 . 0uO

0 . 5 5 '1 2 4

RE AC T E D

Y I EL D L a MOL E S STY ,

0 .0

0 . 0 480?

1 1 6 3. 0 5

1 65 8 . 1 0

L I MOL E

0 . 0 00 1 l

0 . 0 �4 1 3

1 191.08

1664 .57

� . 6U762 H1 81

0.0

0 . 0 0 1 26

0. 1 6�03

PS [ G

0 . 0 00 1 5

0 . 1 2 16 3

1227.H

1 6 7 0 . 1> 9

PR E S S J R E

0.0

31

0 .00240

0 . 1 6 3 80

1 25 5 . 5 3

GAS

0 . 00 006

0 . 005

0.20H2

1 30 9 . 44

1 2 8 2 . 49

TEMP

0. 00041

0 . 009 4 0

0 . 2 76 2 4

0 . 2 4 0 79

f lU E

E T HY L I H N l E NE

C .O

0.00125

0 .0 1429

0 . J Od68

T E �P

FT

I . OOOOC 0 . 04d 7 1

C . O I 'l l l

0 . 00288

0 . 0 1 9 46

R E AC T I O N

0.0

" . 9 80 7C 0 . 9 �O0 7 0 . 1 2 8 "4

0 . 08&52

0 .005�1

0 . J2435

H Y U � OG E N

1 . J v l" 0 . 90�d 3

0 . 0092 5

o.na �8

T OL UE N E

l . �OO

v . 862 8 1

0 . 1 7 3 20

0 . 0 2 006

6ENZ E �E

10 3 . 0 0�

0 . 2 1 74 1

O . C 2 6 1�

0.01413

S T YR E NE

20 4 . 11 00 0 . 8 1 1 90

0 . 2 & .1 2 5

J.

0 . (; 1 9 4 3

0 .005 5 1

MUL E S

30 5 . 0 (, (; O. 7�905

0 . 300�9

1 . 0 0 .1 0 0

0 . 087b3

0 . 04 9 4 '1

0 . 009 8 1

L8

40 & . (; 0 0 ;; . 6 5 � 0 0

0 . 70& 1 5

� . 900

.. . 9 00 3 1 0 . � 49 3 1

0 . 00555

o . O C 2 �O

9 . 0 0 (; 0 . ",

0 . 90859

0 . 1 7� 90

0 . 1 3036

L ENuTH

50

7.UOO

NUM d E ot

60 8 . �v O

I NCR E ME NT

80

70

1 . 00C

J . 6b 1 1 b

90

) . U I.I O

2 . 0 0 1)

J . 80968

0.0

10

4 . uJO

1> 0

� . � O"

8 . t O (;

3 5 . 94 5

20

, . (lUU

70

1 0. O L O

0.085 1 1

50

3(; 40

1 196. 1 5

80

86. 791

90 100

�

00

�

� CIiI

::! �

�.

249 . 1 8 1 N U M B � � ut T U B E � = "YR A � N U AL CU � T a t � E C YC L ' � G E T H Y L B E Nl E N t A T O . O I C ' / L � = 32 1 0 1 8 . 4676 73. A N N U A L C O S T U F U H YL B � N L E N E C ON V E R T E D Tel B Y- P R O DU C T S A T O o l 2 0 H L B = HYR T O T AL A N N U AL u P � R A T I N G C U S T = 78d6Q2 . S U � f AC E A R � A U F R E A C T U R = 85 1 . 04 4 SO F T C O S T Of T w O R E Ae T U K , Of A BO V . S I l E = 1 0 9 96 0 . 0 0 $ CO S T UF L A T A L V , ' A T 3.00 $ / L B = 2 7 30 . 7 9 8

D I S J I LL A J ION t ALtUL A J I ONS fOR R E At JDR DE SCR I BED A80VE

HYR

IN 0 I S J I L L AfE-0 . 9 .9 � I N I "U" R -

HUT con

IN F E E D- 0 •• 00 EB FRAt J I DN

DEPR COST

E_ FRAt J I ON

F E ED fRAY

50 4 7 10 .

TR AYS

f E E D RUE -

R R fAClLlit

" . 07 9

COOL

CUST

YEAR COST

1 2002 9 . 1 9

1 18 5 1 0 . 8 1

3715. 35

3 858. 7 8

7000 1 . 1 9

7 2 70 3 . 4 •

1 t 87 5 4 . 2 5

•• 1 1 2 • ••

• 1 �8. 6 2

4002 . 2 1

• • • 00

3 9 . 00

7S.0 S . U

9 •• 00

85 . 00

39 3.6. 2.

1. 05

1 . 10

3 5 . 00

1 2 1 38 1 . M

78 . 00

4289. 06

1233H .87

1.15

808 1 0 . 50

4"32.49

94966 . 8 1

1 1 9348 . 75

3 U 8 Z . 32

8 35 1 2 . 1 1

3050 . 6 1

9 6 95 0 . 7 5

4 1 45 . 6 3

3 1 . 00

3 5 1 96 . 5 7

57.76 . 7 7

3 1 46 . 2 3

78 108 . 1 2

.9. CO

2 9 . 00

34U9. U

59 2 7 8 . 29

3 7 09 5 . 05

1.25

••• 00

2 B . 00

3UU. 26

3 Z . 00

1 . 30

u . OO

2 6 . 00

7 2 . 00

I.J5

.2 . 110

1 . 2.,

1 . 40

100Z5 8 . 56

9 8 90 6 . 8 1

33 3 7 . 4 7

1 0 2 1 5 0 .7 5

324 1 . 85

6 2 8 8 1 . 38

343 3 . 0 9

. 1 0 7 9 . 116

34039. n

..6B 2 . 96

3 . 5 85 . 1 1

2 5 . 00

340' • • 7 .

2 6 . 00

5 9 . 0C

2 5 .00

6 1 . 00

1 . 50>

S 8 . CO

1• •5

1.55

'l1 0 �

�

;;5

�

.....

' S U M I: D E X I T T I: MP UF FLUE GAS ' l ItO O . OO O DEGREES F C A T A L Y S T P I: L l t r U l lMETER . 0 . 0 1 04 0 FT C A T A L Y S T 8 U L � U I: N , I TY. . 1 . 000GO Ld /C U f T E F F E C T I � E N I: 5 . F AC T OK • 0.44 V O l o) F R A C T l O " O F P l C KED B ED . 0 . )8000 O V E � - A L L ti l: A T T k A � S fER COEFF IC IE N T ' 8 . 00000 B TU l liR/SO f Tlf H l A T C A P A C I T Y OF " THYl BENZENE ' 0 •• 3500 BTU I LB l f H I: A T C A P A C I T Y O f F L Uk G AS . 0 . 30000 B TUIL BI F tl A l t . 1> 5 2 0 . 000 LII f l UE GU/HRI TUIIE P R t S � UR E ' 5 . 00000 P SIG 100 F L UE

A > S U M I: U

N U MB E k

�

C

�

�

'\:)'\:)(';) �

):;.

NUM8ER

I NCREIIE NT

0 10 20 30 40 50 60 70 80 90 100 0 10

�

60 50 60 70 80 90 98 0 10 20 10 60 50 60 70 80 90 100

D . ,) :I . UUO

1 . 00000

l ENvTH FT

1 . 50C;

0 . 12341 0 . 14165 0 . 1 5 940 0. 17719 0. 19519 0 . 2 U J6 O . 2 3 U" 0.0 0 . OU 88 C . 0 7 6 49 0 . 1 0 U' 0 . 1 22 U 0 . 14098 0. 1 59 1 2 0 . 1 7 106 0 . 1 94 . 8 0 . 2 1 2 22 0 . 2 2 4 90 0 .0 0 . 0"4 " 1 0 . 0 7 7 11 0 . 1 0 1 88 0 . 1 22 " 7 0 . 1 " 1 20 0 . 1 59 2 6 0 . 1 77 1 9 0 . 1 9lo 1 0 1i . 2 1 2 7J 0 . 12 9 1 2

11 . 1 0 3 4 4

0 .0 0 . 046 3 0 0 . 07920

' T YRENE t T HYL 8 t N l EN � L8 NUL E S FO RMED PER

J . OIiO

0 . 8 0!> 1 6 0 . 7 114 2 2 0 . 7 . 3 '! !> 0 . 7 .. lI.2

Ii . UUO

0 . '. " 7

0 . 8.914

0 . nU8 0 . 7 ) 744 1 . 0 000 0 0) . 9 :0:'08 0 . 9 20 . 1 0 . 8 9U 9 0 . 86997 0 . 8482 1 01 . 8 2 10 3 0 . 8 06 0 7 0 . 7 115 44 11 . 7 0 5 7 3 " . 7 :0 1 70 1 . 0 0000 0 . 9 l1 U l 11 . 9 1 990 0 . 8 92 7 1

II . U,),.

0 . 9 1 74 ) 0 . 89046 :1 . 8 6 7 ) 2 ,) . 8 4 n 4 0 . 8 2 3 99 11 . 8 0 2 2 7

4. :OOU 6 . ,)r.D 7. 500 9 . 0 011 1 ,) . 50(, 1 2 . 0 00 U. :.IiU ".0 1 . 500

u . ooo

) . oou

U . loOO

4. 500 6 . 000 7. 50 0 9 . t OO 1 0 . SOli 1 2 . 00 0 1 4. 700 0.0 1 . stll 3. 0110 4 . 5110 • • 11 00 7. :000 9 . 11110 1 ,) . loDe;; 1 2 . .. 00

U. :oOO U . OOli

8EN Z EN E E TH Y L E N E

II MOLE

0.0

O . O O O ) !> 0 . 0 0065 0 . 00098

O.OOlU

0.C0413

0 . 0 0 1 74 0 .0 0 2 2 5 0 . 002114 0 . 0 0 3 48

O . OO l n

0 . 0 0 063 0 . 0 0094

o.ooon

F L UE G AS TE M P DEG F

P R E S SU R E PSIG

HYDROGEN

RUCT I ON T E IIP DEG F

T OL UE N E OI U HANE

OF E T HY L BEN I E N E F E D

1600.0C 1 60 4 . 2 2 1 608 . 411 1 6 1 2 . 70

2 8 ] 1 79 .

3 1.281 26.625 21 . 2 3 5 14.709 4 8 . 9 91 4 6 . 3 32 43. 441 40 . 2 9 7 3 6 . 860

n . l 70

4 1 . 4 50

2 8 . 8 39 24 . 0 2 1 1 8 . 3 62 1 1 .]20 ] .�66 4 '1 . 9 9 ] 4 1 . 3 74 44.514

3 8 . 0 87 3 4 . 3 92 3 0 . 2 84 2 5 . � 39 2 0 . 2 42 1 3 .668 4 . 165

3 . 2 54

SlYR

1 66 0 . 1 9

1 62 0 . 95 1 6 2 6 . 95 1 6 2 8 . 88 1 6 32 . 7 3 ln6 . 5 1 1 660 . n 1600 . 0 0 1604. 2 1 1 60 8 . 4 5 1 6 1 2 . 67 1 6 1 6 . 1l 1 62 0 . 90 1 62 4 . 90 1 62 ' . 1 3 1 6 3 2 . 68 1 63 6 . 6 7 1 63 9 . 66 1600.00 1 606 . 2 1 1608. 66 161 2.67 1616. 83 1620.91 1 626 . 9 1 1 62 8 . 86 1 63 2 . 70 1 6 3 6...

1 6 1 6 .16

1 0 2 2 . 00 1 0 1 5 . 79 1020. 5 1

A 8 0V E

35.401

5 3 . 664 5 1 . 1 90 48.5 19 45.b36 42.5 18 19 . 1 2 6

0 . 02 0 5 4 0 .0 2 3 ) ) 0 . 02 5 6 0 0.0 0 . 0 007 1 0 . 002 2 7 0 .C 04 2 9 0 . 0066 0 0 . 00908 0.01161 0 . 0 1604 0.01621 0.01791 0 . 0 1 8 79 0.0

0.0121 5 0 . 0 16 7 5 0.01712 0 . 0 1 90 S 0 . 0 20 ]6

0 . 0 0967

0 . 0 02 1 6 0 . 0 0.. .. 6 0 . 0 06 8 7

0 . 0 0 461 0.0

o.onu

0.0 0 . 0 00 8 5 0 . 002 7 1 0 . 005 1 2 0 . C 07 9 0 0 . 0 1096 0.01418

1 0 3 1 . 89 1 0 46. 22 1 06 1 . 36 1 0 76 . 30 1 09 0 . 65 1 1 06. 30 1 1 1 7 . 34 l U O . 02 1 0 2 2 . 00 1 0 1 7 . 66 1 0 2 2 . 50 1 0 ) ) . 28 1 066. 9 3 1 06 1 . 55 1 0 76 . 18 1 09 0 . 6 2 1 106. 23 1 1 1 7. 89 1 1 2 9 . 60 1 0 2 2 . 00 1 0 1 7 . 25 1 0 2 2 . 0.. 1 0 32 . 9" 1 04 • • 72 1 0 6 1 . 66 1 0 76 . 1 5 1 0 9 0 . 60 1 10 ... 1 .. 1 1 1 7 . 59 1 1 3 1 . 51

0 . 00 0 7 ..

0 . 0 0 1 15 0 . 0 0 2 43 0.0031 1 0 . 00389 0 . 0 04 74 0 . 0 0561 0.0

0.0 0. 0 . s 45 0. 0 7649 0 . 0 98 3 3 0. 1 1551 0 . 1 )069 0 . 1 4522 0 . 1 5 9 76 0 . 1 7465 0 . 1 90 0 1 0 . 2 0 5 74 0.0 0.04)1 8 0 . 0 7422 0. 0 9709 0. 1 1 5!> 3 0 . 1 ) 1 90 0. 1 4751 0 . 1 6 301 0 . 1 71" 0 . 1 94 ) 1 0 . l061 1 0.0 0 . 0 .. 368 0 . 0 7.. 7 5 0 . 0 9762 0. 1 1 560 0 . 1 ) 1 72 0 . 1 .. 7 1 1 0 . 1 62 .... 0 . 1 7 79. 0 . 1 9368 0. 2 08 7 8

0 . 00 0 36

0 . 0 0 ..92

0 . 00 0 11 3 0 . 00095 0 . 0 0 U2 0 . o e l l6 0 . 0 0 229 0 . 00 289 0.00JS7 6 . 0 0 .. 2.

DE S C R I B E D

E d F R A C T I O N I N F E E O - 0 . 7 6 5 E B F R AC T I O N

I N D I S T I L l A T E = O . 9 86 M I N I M UM R -

27.5]1 NUN8EM �F T U t t i • I l YR ANNUAL c u S T OF II t C YCL I NG E T HYL .EN l t .1E AT 0 . 0 1 0 " L 8 • .9.5 5 ) . ANNUAL C O S T Of e T HYL '�NZ t Nt C ON V E M T E D TO BY-PRODUC T � AT O . I ZO ./ L ' • i l YR 9 79 7 32 . T U T AL AhNUAl IIPt M A T I NG CO� T • SQ F r 486 . 509 SUR FAt t A II E A at K E At T O R • C U � T OF T �U II tA' T D M S OF A 80 V � i l l E • • 79589. 69 • 6680 . 8 98 C O S T UF C AT A L Y S T A T ] . OO . / L I •

d ,9 J 4 8 .

D I S T 1 L L A T I �N C AlC� l A T I U N S F O R RE A C T O R

FEED RATE-

Y l E L 'l L B MOL E S STY I LB M O L E E B R E AC T E D

0.0

9 7 . 46 9

94 . 43 2

9 5 . 92 1

93 . 0 1 2 9 1 . 70 7

0.0 91.688

90 . 56 3 89 . b 1 1 8 8 . 816 88 . 312 88. 1 1 2

96 . 3 4 9

9 5 . 094

9 3 . 92 1 9 2 . 81 4 9 1 . 98 8 91 . 29 7 90 . 8 2 6

'1 0 . 5 9 1

9l . 51 3

0.0

90 . 1 2 2 90. 010

97. 642 96. 2bO 9 4 . 95 6 9 3 . 73 3 92 . b30 9 1 . 68 9 90 . 942 '1 0 . 4 1 4

�

@

�

�

....

�

�

;;!

I:l

�

I:l,

I:l ;::!

;;!

�

� .... �

"tl

� ... ;::! .... I:) $:: ....

�

�

::so

(")

.§.

.... � �

�

� �

R� f A C r U K TRAYS OE P R co s r H E Ar C O S T

COOL C O S r

YEAR COST

203373.06

2 4 7 78 5 . 0 0

7385 .93

Z 0 5 6� 2 . 94

8 86 7 . 8 9

1 39 1 5 8 . 6 2

7630. 93

2 1 002 7 . 94

1 6 7 08 0 . 3 7

56d2 8. 52

1 4 3 77 4 . 7 5

7 875. 9 3

Z 1 35 7 2 . 5 6

11 8 36. 7 5

78 . 0 0

5 4 2 H. 3 0

1 48 3 9 0 . 8 7

8 120. 9 3

1 6 1 69 Z . 3 1

1 05 . eo

1 . 2 1l

73 . 00

5 3 16 1 . 1 4

1 5 3 00 7 . 0 0

5577. 29

1 64 2 2 6 . 94

l.e5

1 .2 �

1 1 . 00

524H. 68

1 0 5 08 Z . 0 6

5740 . 6 3

1 6 7 5 0 '1 . 5 6

20044 7.06

1 . 30

6 8 . 00

5 1 033. 0 1

1 08 1 5 9 . 5 0

5 90 3 . 9 6

1 70 75 6 . 1 2

6 8 95 . q z

1 . 3,

6 5 . eo

50326. 85

1 1 1236.87

6 06 7 . 3 0

1 73 1 2 8 . 00

l Z99Z 6 . 44

1 . 40

6 3 . eo

50368. 7 8

1 14314.31

6230.64

1771�3.5a

63624. 1 1

1 . 10 5

6 2 . 00

5 0 3 7 4 . 5 10

1 1 7 39 1 . 7 5

6B3. 97

1 8 02 8 5 . 8 1

9 1 . 00

1 . !> Il

6 1 .eo

49505.64

1 2 0 4 6 '1 . 1 9

6557. 3 1

1 8 42 7 2 . 8 1

1 . l�

1 . 5;

59. CC

5 0 2 8 0 . 36

1 2 3 54 6 . 62

6 7 2 0 . 64

1 86 4 7 1 . 06

200229. 1 2

l.bu

59. 00

50 1 8 1 . 92

1 2 6 62 4 . 0 0

6883.98

1 9 0 4 1 8 . 75

7 140. 92

1 . 6 !>

5 8 . 00

50qz8. 22

1 2970 1 . 44

7047. 32

1 3454Z . 50

1 . 70

5 8 . CO

49885.68

1 32 7 7 8 . 8 7

58545. 11

1 .7)

56 . 00

50592. 57

8Z.00

l.dJ

56 . 0 0

1.15

1 .85

� t\)

:t>.

�.

'\:)'\:)(1) ;:$ >:\..

2 . 70

2.b�

2 . bJ

2.��

2 . �0

. 6 7469 0 . b 1"38 0 . 55938 a . B6 S 7 1 . 00000 0 . 9 7 706 0 . 94 1 8 8 0 . 8 98 7 7 0 . 8 49 3 7 0 . 79" 1 9 0 . 13448 0 . 6 72 4 9 0 . 6 1 1 09 :> . 5 5" 2 2 0 . 5 2 1 08 1 . 000CO 0 . 9768 I 0 . 9"1 " 3 0 . 8 98 2 " a . 8" 8 7 3 0 . 7 9325 0 . 7 3 3 06 0 . 6 70 3 5 0 . /0 0 794 0 . 5 4942 () . 5 0 5 � � 1 . 000 0 0 0 . 9 76 5 5 0 . 9" 1 00 0 . 8 9774 0 . 8 4� 1 0 0 . 7923� 0 . 7 3166 0 . 1> 6 8 2 6 0 . 60"89 0 . 54"89 0 . 49382 0 .0 0 . 02 2 4b 0 . 05b 1 3 0 . 0 95 b l 0 . 1 3S I9 (' . 1 8 3 0 3 0 . 2 2 9 32 0 . 2 7b02 0 . 32 1 8 0

0.0

0 . 3b433 0 . 38l53 0.0 0 .02271 0 . 05b54 0 . 09599 0 . I H49 O. I832b 0.2N54 0 . 2 7b 3 0 � .32229 0 . 3b�58 0 . 39 1 b 7

0 . 05 b Q"

a . 0 2 2 9b 0 . 09b 3 5 C . DB 1 5 0 . 1 834 b 0 . 22nl 0 . 2 16 5 1 0 . 322bb 0 . 3 66 5 2 0 . 40 0 b 5 0.0 0 . 02 3 2 1 0.05733 0 . 09669 0 . 1 3899 0 . 1 8 363 0 . 22985 0 . 2 7 b 60 0 . 3 2293 C . 36 7 2 0 0 .4 0 6 20

0.0 0 .00003 0 .00017 O. OOObl

0.0 0 . 0 00 1 9 0 . 0 0 l l6 0 . 0044 7 0 . 0 10 1 1 0 . 01820 0 . 02 7 9 5 0 . 0 38 0 8 0 . 04 7 1 3 0 . 0 53 6 6 0 . 055 5 3 0.0 0 . 000 2 0 0 . 00 1 4 1 0 . 00462 0 . 0 1 04 7 0. 01888 0 . 02 906 0 . 0 3 9 75 0 . 049 ' " 0 . 05/0 7 9 0 . 0 5 9 74 0.0 0 . 00020 0 . 0 0 1 46 0 . 004 7 8 0 .0 1 0 8 3 0 . 0 1 956 0 . 0 30 1 8 0 .04142 0 . 05 1 8 1 0 . 05 9 9 0 0 . 064 1 3 0.0 0 . 000 2 1 0 .0 0 1 5 1 0 . 00494 0 . 01 1 20 0 . 02024 0 .031 ) 1 0 . 043 1 0 0 . 054 1 4 0 . 0 1> 3 0 1 0 . 06 8 19

O . O I O S /L B • B Y- P RO DU C T S

0 . 00 1 b5 0 . 0 0 3b2 0 . 0 0 b 79 0 .O t t 22 0 .0 1b69 0. 022b3 0.C25H 0.0 0 . 00003 0 . 000 1 7 O.OOObl 0.001b7 0 . 003b8 0 . 0 0692 0 .O t t 47 0.01 714 0 . 0 2 341 0 . e 2 751 0.0 0 . 00003 0 . 0 00 1 7 0 . 0 00b2 0 . 0 01 b 9 0 . 0 0 3 14 0 . 0 0 10 5 0 . 0 1 1 72 0 . 0 1 7 59 0 . 0 24 1 6

AT TO

0.02977 0.0 0 . 00003 o.OOOl7 0 . 00003 0 . 0 0 1 11 0 .00379 0.00718 0 . 0 I I98 C . 0 1 8 0) 3 0 . 02"90 0 . 0 3 1 60 • 58 . 9 1 3 R E C Y C L I N G E T H Y L 8 E NZ E � E E T H Y L B E N Z E N E C ON V E R T E D

C A T A L YS T

AT

3 . 00 S / L 8

•

50 2 4 . 5 3 5

Sl YR s n 4 42 . T O T A L A N N U A L U P E R A T I N G CO S T · SO f T 7 3 2 . 60 4 S U R f A c e A K EA Of � E A C T O R • C O S T Of T w a K E AC T O R S O f A 8 0 v e S I Z E ' 1 0 0 50 5 . 7 5 COST

S /Y R

1600.00 16l l . 1 7 1 62 5 . 2 1 16 3 6 . 40 1 646 . 85 1656. 7 1 1666.06 1 6 74 . 97 1 68 1 . 50 1 69 1 . 6 5 1 695 . 5 9 1 60 0 . 0 0 1 61 3 . 1 7 1 6 2 5. 2 3 1636.41 1 6 4 6 . 87 1656. 72 1 666 . 0 8 1 6 7 4 . 99 1 68 3 . 52 1 69 1 . 6 9 1 6'1 7 . 1 7 1 600 . 00 161 3. 1 7 1 62 5 . 2 3 1 6 3 6 . 42 1 646 . 88 1 6 5 6 . 74 1 6 66 . 0 9 1615 . 0 1 1 61 3 . 5 4 1 69 1 . 7 1 1 6 9 8 . 74 1600 .00 161 3. 1 8 162 5 . 24 1636.43 1646 . 8'1 1 65 6 . 7 5 1666. 1 1 1 675.02 1 6 8 1 . 56 1 69 1 . 7 3 1 69 9 . 54

b322 3 1 .

8 42 . 00 92 5 . 1 9 990 . 8 7 1 045 . 19 1 09 1 . 53 1 1 3 1 . 47 1 1 66. J 1 1 1 9 7 . 60 1 22 6 . 59 1 2 55. 32 1 2 7 1 . 00 842 . 00 9 2 5 . 00 990 . 59 1 0 44. 9 7 1 09 1 . 4 1 1 1 3 1 . 41 1 1 66 . 3 3 1 197.53 1 22 6. 36 1 25 4. 56 1275.91 8 42 . 00 9 24 . 82 9 9 0 . 32 1 044. 7 7 1 0 9 1 . 30 1 1 3 1 . 37 1 166. 3 1 1 1 97 . 49 1 22 6 . 2 1 1 2 54. 00 1 28 0 . 82 8 42 . 00 92". 6 3 990 . 06 1 0 44. 58 10 9 1 . 2 2 1 1 3 1 . 35 1 1 6 6 . 32 1 1 97 . 48 1 2 26 . 1 2 1 2 5 3 . 62 1 28 2 . 52

S / YR 2602 1 1 . 0 . 1 20 S/LB •

0.0 0 .02227 0 . 0 54 77 0 . 0 9 1 14 0 . 1 2808 0 . 1 64 8 2 0.20137 0 . 2 ) 7 94 0 . 2 746/0 0. 3 1067 0 . 3 2 700 0.0 0.02251 0.05513 0.09137 0 . 1 2802 0 . 1 64 3 8 0 . 2 0048 0. H655 0. 2 1282 0 . 3 0 880 0 . 3 3 1 93 0.0 0 . 0 2 2 76 0 . 0 5548 0.09157 0 . 1 2792 0 . 1 6 390 0 . 1 99 5 3 0 . 2 3509 0. 2 7086 0 . 3 06 6 1 0 . 3 J6 5 2 0.0 0 . 0 2 3CO 0 . 0 5 5 82 0 . 0 9 1 75 0 . 1 l 7 80 0 . 1 6 338 0 . 1 98 5" 0 . 2 ) )56 0. 26879 0 . 3 0� 19 0. H781

AT

5 9 . 6 94 56 . 9 7 3 5 3 . 899 50.U8 4 6 . 4 60 41 .919 3 6 . 6 39 30 . 3 5 2 2 2 . 5 50 1 1 .939 4.2 1 7 60 . 6 94 58.0 1 0 54 . 9 8 1 51 . 5 5 7 47.670 4 3 . 2 24 38 .072 3 1 .977 2 4 . 496 1 4 . 5 89 4 .333 6 1 . 6 '14 59.046 56 . 0 6 1 52.691 48 . 8 73 4" . 5 1 6 3'1 . 4 8 7 n . 569 26 . ) 75 1 7 .05 1 4 . � 59 62 . 6 '14 60 . 0 82 5 7 . 1 )9 ".822 50 . 069 4 5 . 1 98 40 . 8 8 1 )5 . 1 )0 2 8 . 1 96 1 9 . 3 68 6.410

D . ..

Q q . 04 0 97 . 3 5 5 94. 958

92. 158 8�. 347

86 . 84 5 8 4 . 847 8 3 . 449 82 . b 8 6 8 2 . 5H 0.0

99 . 0 1 7 9 7 . 28 9

9 1. 9 42

9". 82 6

8 9 . 04 1 8 b . 44 9 84 . 3 b 2 8 2 . 872 82 . at t 8 1 . 78 ? 0.0 99 . 9 95

97 . 222 Q". b92

9 1 . 724

8 8 . 714 8b . 05 1 83 . 879 8 2 . 300 8 1 . 143 81 .012 0.0 98.912 97 . 15 5 94. 556 9 1 . 50 4

8 � . 42 5

8 5 . 65 � 8 3 . 3'" 8 1 . 732 80 . 68� 80 . 2 4 7

":j

�

�

�

�

..... 00:::: 't1

� �

�

�

\:) �

�

�

(1)

'1:;)0 .....

�

N.

�

0-

.....

::::

�

�

;:so \:)

("')

(1) �

'1:;)0 .....

"I

�

0,

FEE D

��740� .

� B F RA C T I ON F R AC T I ON

DE S C� I S E O A 8 0VE

I N F E E O - 0 . 5 � 9 E8

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

�ATf'

I � D I ST I L L A T E =0 .962

COST

� I N ' HU H R "

HEAT

3 8 . 00

35 329. 22

31 135. 78

68 196. 2 5

65 804 . 94

6 34 1 3. 62

3 74 6 . " 8

3& 19. 56

3 '0 9 2 . 6 "

3365 . 7 1

83386. 1 2

1 07 3 0 3 . 5 6

1 0601 0. 1 2

1 04 6 2 6 . 75

1 045 ' 5 . 06

� . �27

83.CO

3 3 . 00

3 '0 1 9 4 . 3 6

7058 7 . 6 3

2 5 82 . 2 7

8 4 6 8 '1 . 31

YEAR COST

1.10

71. . 0 0

3 1 .00

32 969. 46

" 8 65 2 . 66

2666. 8 8

8 5'1 3 6 . 6 2

COST

COOL

I.I�

?l . 0 0

2 9 . 00

32 1 5 1 . 2 "

5 0 2 '0 6 . 8 9

275 1 . 50

8 7 H 3 . 75

OEP R COST

1 . 20

6 8 . 00

2 8 . 00

3 1 7 70 . 6 6

H 8 1t • • 1 2

2 836. 1 1

8 8 8 3 9 . 44

T RA Y

1.25

6� . 0 0

2 6 . 00

3 1 3" " . 0 "

5 3 4 3 5 . )It

2920. 13

8 9 96 8 . 6 2

F EED

1 . 30

6 3 . 00

2 6 . 00

30872 . H

5 5 0 2 9 . 60

3 005 . 34

9 1 60 6 . 1 9

T�AYS

l o J5

6 1 .CO

25 .00

3 0 8 89 . 1 6

5662 3 . 8 3

3 0 8 9 . 96

'1 3 7 8 0 . 1 2

R R F ACrO�

1."0

5 '1 . 0 0

2 5 . 00

30 339. " 8

582 1 8 . 0 3

3 1 7'0 . 5 1

948 1 1 . 3 1

1 0 6 46 7 . 8 1

1.";

5 8 . 00

2 3.00

3 0 2 9 8 . 2 '1

598 1 2 . 2 3

3259. 1 8

'1 6'1 5 7 . 3 1

3 23 8. 79

1.50

5 6 . 0U

2 3 . 00

3 0 79 3 . 35

6 1 '00 6 . 4 8

3343 . 80

61022 .27

1 . 55

55.00

23.00

30 1 " 5 . 7 0

6 1 00 0 . H

42 206. 7 9

l.bO

5�.OO

2 l . 00

30b 1 2 . 84

"3.00

1 .65

5 3 . 00

2 2 . 00

9 5 . 00

1 . 70

53 .00

1 .�5

1.75

�

0\

;J:..

� �

;:! � �

... .

1 .9>

1 .90

1.8,

l . dO

51 .CO

51.GO

52.00

5 2 . CO

2 1 . 00

22 .00

2Z.00

2 2 . 00

3 1 2 1 Z . 05

30 779. 6 8

30938.26

3 0 4 ij 8 . 9 7

6 9 37 7 . 6 3

6 7 78 3 . 37

66189. 19

64594 . 94

3 6 82 . 2 6

3597.64

3513.03

H2 8 . H

1 0427 1 . 87

10Z160.69

1 00 6 4 0 . 44

9851 2 . 3 1

2.40

2.35

2 . 30

2.l5

l . 20

2.15

2.10

2 .v>

4 6 . 00

46 . 00

4 6 . 00

4 8 . 00

4 R . 00

4 8 . CO

4 8 . 00

4 9 . 00

4 9 . 00

1 9 . 00

1 9 . 00

1 9 .00

1 9 . 00

1 9 . 00

2 1 . 00

2 1 .00

Z I .00

Z I .00

2 1 .00

3258 1 . 33

32227.92

3 1 8 7 1 . 92

31 513.25

3 1 1 5 1 . 89

32 1 2 6 . 2 5

3 1 74 3 . 1 5

3 1 357. 00

3 0 96 7 . 5 6

31Z11.87

30807 . 5 4

9 0 1 0 2 . 62

8 8 50 8 . 3 1

8691 4 . 1 3

8 5 3 1 9 . 87

8372 5 . 62

8 2 1 3 1 .44

80537 .25

7894 3 . 0 0

7 7 34 8 . 8 1

7575 4 . 5 6

74 1 6 0 . 2 5

7 2 5 6 6 . 06

7097 1 . 8 1

4782 . 2 5

4 6 97 . � 3

4 6 1 3 . 02

452 8 . 4 0

4 ,.,. 3 . 7 9

4359. 1 7

4274.56

4 1 89.95

4105. 33

4020. 72

3936. 1 0

3 85 1 . 4 9

3 76 6 . 8 7

1 2710 1 . 12

1 25787. 3 1

1 2 3 7 5 5 . 06

1 21720. 19

1 19682 . 6 2

1 1 76 4 1 . 5 0

1 1 6 93 8 . 0)

1 1 4 87 6 . 06

1 1181 1.12

1 1 0 74 2 . 8 1

1 09 3 0 8 . 1 9

1 0722 5 . 06

1 0 5 1 3 8 . 25

3 0 399 . 6 2

2 . 4t ::»

40. CO

1 9 . 00

32 2 1 6 . 2 9

2 1 .00

2 . 5U

4 6 . 00

1 9.00

4 9 . 00

2.55

4 > . 00

2.�0

2 . 60

"rj

@

�

.....

�

"'(

-.:::: "tl

�

�

"'( \:l

\:l ;:! �

�

�

�

�

"'(

"tl

�. 0-

!:: ....

'c"'(

()

.",. \:l � '"'� "'(

'I

�

'I

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

Index Computer program flow sheet,147,

Absorption,9

180

Activation energy,170 Activity coefficient,61 Adsorption,kinetics of,163

Computer programming, 146,178 Consecutive reactions,22

see Economic Analysis,economic,-

Control,production,8

Approximations,86

Conversion of a chemical reaction,

quality,8

evaluation

85,97,152,192

Arrhenius correlation,9,41, 161

Conveyor belt, 103 ,194 Continuous stirred-tan k reactor

(CSTR ),11, 24 Cooling water, cost of, 58 temperature rise of,43

Batch reaction,35 Bench-scale experimentation,4, 19. 76

CorrOSion, 106 Cost,capital, 38,57,58,181,195

Bleed stream,87,89 Boiler design,88

conversion, 73 operating,73 optimization of,28,39,70, 116, 142,183

Calculations,iterative,66

raw-material,38, 178

Calculations,tabulation of reactor-

Cost accounting,12, 13 Cost correlations,57

design,80, 172 Calculus of variations, 11 Capital cost estimation, 131

Counter-current staged process,

kinetics on,9

69 Crystallization,9

porosity of, 159

Cyclohexane production,97

Catalyst, activity of, 97

properties of, 158-160 surface area of,159 Data,consistency of,6

Catalyst pellets,9,98,162,166

correlation of, 9, 60 extrapolation of, 9,60

Censorship,7 Chemical equilibrium,effect of

Data analysis,8, 22 Data processing,3

temperature on, 85 Chemical reaction, see Reaction Chilton-Colburn equations,9,163

Data reduction, 8

Chlorination,chemical kinetics of, 22

Depreciation,122 DeSign,industrial practice of, 1-17

Chlorobenzenes,production of, 22

DeSign parameters, 8

Clausius-Clapeyron equation,9, 132, 143

Design strategy,10,11

Computer,classroom use of,2

Desorption,162

Design results,extrapolation of,95

229

The Industrial P rac t ic e

230

of Clzemical Process

Engineering

Difference equations,174

Gibbs-Duhem equation, 7, 9

Differential equations, derivation of, 81,172

Gilliland correlation, 135

numerical integration of,173 Diffusion within catalyst pellets, 9, 98,162,166 Diffusivity,binary,164

Heat exchanger,141 Heat of solution, 7 Heat transfer,boiling, 9

effective, 166

fluids of, 88

Knudsen,167

indirect,132

Distillation, see Fractionation

Heat transfer to buried pipes,111

Distribution coe fficient, 9,61

Heat transfer coefficients,43

Dynamic programming,11

Heat transfer of condensation, 9 Heat transfer in reactors, 171 Henry's law, 9 Hoop stress, 110

Economic evaluation,2,38,71, 72, 96,98,115,181 terminology of,12

Hydraulics,distillation-tray, 140 Hydrogen, industrial sources of,87 Hydrogenation,catalytic, 77

Effectiveness factor, 9,166,169,174 Efficiency, tray, 9 mixer-settler, 69 Enthalpy 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

Incentive,financial, 10 Integration,numerical, 84 Investment,return on, 13,196 Isothermal reaction,39 Iterative calculation, 66

Erbar and Maddox correlation, 134 Error,experimental, 162 probable, 8 Error analysis, 8 Experimental variables, selection of,

Kinetic data,analysis of, 22, 80,161 Kinetic data,correction for diffusion effects on, 170

60 Experiments, in-plant, 8 statistically designed,8,54, 59 Extraction, liquid-liquid, 9 Extrapolation of design results, 95

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

Factorial experimental design, 54 Fanning equation, 9, 109

process-design for, 64 Literature,methods for searching, 6

Fenske equation,134 Financial evaluation,see Economic evaluation

Market demand, 13

Financial incentive, 10

Market information,5

Finite difference techniques, 173

Market price, 13 Market research,5

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

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

Index Nernst's law, 9, 61

23]

Reactor control,42-48 Reactor design, 24,41, 81, 86,171 Reactor flow sheet, 25,89,17 1

Optimization,3, 70,80, 1 13,116, 140,

Reactor stability,42-48

182,193,194 Oral presentations,14

Reflux ratio,134

Recycle, 38, 87, 97 Residence times,38,58 Resistances to catalytic reactions,

Partition coefficient,9,61,62

162 Results, presentation of,14

Parallel reactors,35,197 Petrochemical industry,196

Return on investment,13,196

Phase diagram,62

Risk analysis, 13

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

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

Pressure drop in pipelines, 109

Smoker equations, 147

Pressure-enthalpy diagram,141

Solubility, gas in liquid,9

Process economics, 11

liquid in liquid, 9, 54

Process engineering fundamentals 1-17

Solubility data,7, 9

Process flow sheet, see Flow sheet, process

Sparged reactor, 9

solid in liquid, 9 Solution,heat of, 7

Production control, 8

Stability, reactor,42-48

Programming,discussion of, 146 dynamic,11

Stability of finite difference equa­ tions,175

FORTRAN, 146 linear, 10

Staged process,economic evaluation of,72 Statistically designed experiments, 8,60

Quality control,8

Steam costs,58 Steepest ascent, method of,11 Stoichiometry,80, 172

Raoult's law,7,9

Stripper design,67

Raw material costs, 39 Reaction, adiabatic,87

Styrene,industrial manufacture of, 126,129,146,197

catalytic, 162 equilibrium,9 isothermal. 39, 87

Sulfur,mining and production of,101 properties of,106

kinetics of,9

Supersolubility, 9

temperature of,39

Surfactant,73

Suboptimization,69

yield of,40,152,192 Reaction conversion,80,85, 97, 152, 192 Reactor,bench-scale, 152 catalytic fixed-bed,9,38,80,171 continuous stirred-tank,11, 24

Temperature of reaction,39-41 Thermal expansion, 105 Thermodynamics, first-law balance of,82

parallel,35,197

Thiele modulus,9,166-168

pressure drop in, 98

Tortuosity,167

shell-and-tube,l71

Transportation costs,123

slurry,9

Transportation of liquids,10 1

sparged,9

Triangular phase diagram, 62

232

Tile Industrial Practice of Clzell1 ical Process Ellgineerillg

Underwood equations, 147

Viscosity of gases,estimation of,

Vacuum distillation,138

Visual aids,14 Volatility,relative, 9,134

164

Van Heerdon criterion,45 van't Hoff relationship,9 Vapor-liquid equilibrium, 9, 57, 132 Variables, critical, 10

Writing, technical, 14

selection of, 60 Venture analysis, 12

Yield,reaction, 40, 152, 192