CHEMICAL ENGINEERING

A Review for the P.E. Exam

WILLIAM E. CROCKETT
West Virginia College of Graduate Studies
Institute, West Virginia

A Wiley-Interscience Publication
JOHN WILEY & SONS
New York • Chichester • Brisbane • Toronto • Singapore
Copyright ",) 1986 by John Wiley & Sons, Inc.

All rights reserved. Published simultaneously in Canada.

Reproduction or translation of any part of this work
beyond that permitted by Section 107 or 108 of the
1976 United States Copyright Act without the permission
of the copyright owner is unlawful. Requests for
permission or further information should be addressed to
the Permissions Department, John Wiley & Sons, Inc.

Library of Congress Cataloging in Publication Data:

Crockett, William E.
Chemical engineering.

"A Wiley-Interscience publication."

Includes index.
1. Chemical engineering. I. Title.
TP155.C74 1985 660.2 85-5382

Printed in Singapore

10 9 8 7 6 5 4 3 2 1
 PREFACE

The West Virginia College of Graduate Studies offers the Master's degree in
Chemical Engineering for the employees of the many chemical companies near
Charleston, West Virginia. As a public service, the College offers a review course
for chemical engineers who wish to become registered. I have taught this course
eleven times.
Like many college professors, I could not find a suitable text. So I taught the
course from notes that I changed (improved?) each time that I taught. With
yearly after-the-test comments from most course participants, the notes have
reached the present form. I am satisfied with them and the students are satisfied
with them, so it is time to have them published.
I am indebted to Mrs. Reva Dunn for her many months of typing, even
though she now hates equations; to Mrs. Margery Edington for many hours
over a hot photocopy machine; to Mr. Robert Oneacre for his weeks of
hunching over a drawing table doing the figures; to Ms. Jeannette Stiefel and
Ms. Margaret Comaskey for making me look good with their editing; and to the
West Virginia College of Graduate Studies for the timely financial grant. I thank
the students, for it is they who make it a joy rather than a job. Above all, I am
grateful to Bunny, who makes it all worthwhile.

William E. Crockett
P.E. #6853 (WV)
Institute, West Virginia
August 1985

,j i
 CONTENTS

Introduction 1

Thermodynamics 5

Process Design 49

Mass Transfer 79

Heat Transfer 137

Chemical Kinetics 173

Fluid Flow 197

Engineering Economics 233

Index 263

ix
 INTRODUCTION

This book will help you prepare for the Chemical Engineering Principles and
Practice of Engineering Examination (P. E. Exam). Although it will increase
your chances of passing the exam, I hope you will also relearn neglected parts of
your undergraduate chemical engineering. I have included only those subjects
and concepts likely to occur on the test.
To make the book useful during the exam, I have:

• used the table of contents for each chapter as a topic index. I assume that
upon reading a problem, you will be able to identify the type of problem.
Scanning the table of contents of the appropriate subject chapter should
lead you to the proper section of the chapter.
• defined the nomenclature in the text rather than in a separate section, which
should be more convenient during the test.
• followed the same topic sequence as used in McCabe and Smith's Unit Oper-
ations of Chemical Engineering for the chapters on mass transfer, heat
transfer, and fluids. The thermodynamics topic sequence follows that of
Smith and Van Ness' Introduction to Chemical Engineering Thermodynam-
ics. Both of these texts are standards; many of you used them when you were
undergraduates. The process design, kinetics, and economics chapters fol-
low schemes that I have found useful.

THE TEST

The P. E. Exam is an 8-hr test, administered in two 4-hr sessions. In each session
you are to work any 4 of 10 problems. The approximate number of problems in
each subject is

Thermod ynamics 3 Chemical kinetics 4

Process design 4 Fluid flow 2
M ass transfer 4 Engineering economics
Heat transfer 2
2 INTRODUCTION

A test booklet contains all the problems. You must put all the solution steps
and calculations on the graph-paper solution pages provided in the booklet.
You should take pens or pencils with you to the test. You may use a hand-
held calculator, but it must be battery operated. Take extra batteric' to be safe.
Since the test is open book, reference texts, handbooks, and math tables are
permissible. But loose pages are not permitted unless they are kept in a binder or
notebook.
People grade the test. Partially correct solutions receive partial credit. Do not
give the graders any opportunities to misunderstand your solutions. If any part
of a problem is stated ambiguously, state (and underline) your interpretation of
the meaning.

REFERENCE TEXTS

I shall suggest reference books that I prefer, but it would be better if you choose
books with which you are familiar. My reference texts are

W. L. McCabe and J. C. Smith, Unit Operations oJChemical Engineering, 3rd ed.,

McGraw-Hill, New York, 1976. An excellent reference with a great index.
J. M. Smith and H. C. Van Ness, Introduction to Chemical Engineering Thermo-
dynamics, 3rd ed. McGraw-Hill, New York, 1975. Comprehensive and easy
to understand.
F. Kreith, Principles of Heat TransJer, 3rd ed., Harper & Row, New York, 1973.
Comprehensive treatment of all types of heat transfer.
O. Levenspiel, Chemical Reaction Engineering, 2nd ed. Wiley, New York, 1972.
Widely used with excellent example problems.
E. L. Grant, W. G. Ireson, and R. S. Leavenworth, Principles of Engineering
Economy, 7th ed. Wiley, New York, 1982. The standard text.
R. H. Perry, Chemical Engineers' Handbook, McGraw-Hill, New York, 1973,
latest edition. Excellent source for data but written by a committee.
J. H. Keenan and F. G. Keyes, Thermodynamic Properties oJSteam, Wiley, New
York,1954.
CRC Standard Mathematical Tables, 20th ed., The Chemical Rubber Company,
Cleveland, Ohio, 1972.

I know several engineers who took only Perry and this book to the test. They
did fine, but several reference books may provide a psychological benefit.

PREP ARING FOR THE TEST

The test is easier and less frightening if you have prepared properly. Here are
some suggestions:

• Start preparing at least several months before the test. Tests are usually
given in mid-April and at the end of October of each year.
 PREPARING FOR THE TEST

Decide upon and get a copy of the reference books that you prefer.
• Read a chapter in this book, stopping before the problems. Each chapter
deals with one subject; they do not have to be studied in sequence. You may
wish to do your most difficult subject first.
• Page through the reference book you have chosen for the subject and be-
come familiar with the contents.
• Returning to this book, read all the problems provided at the end of the
chapter, and find the easiest. You should do this on the test too. The secret
to success is finding and solving the easiest problem first. It builds confi-
dence and saves test time for more difficult problems.
• Determine what problem-solving strategy you will use. This is usually done
mentally as you read and think about each problem. Place the problem in a
broad category and then narrow the category again and again; for instance,
heat transfer, condensing vapor to heat a liquid, double-pipe exchanger, log
mean temperature difference, and dirt factor. By using this strategy you start
to "see" the needed equations or, at least, to know which areas to look for in
the table of contents of the subject chapter.
• Compare your strategy with the strategy given in the chapter. If you are not
used to solving problems using this step, you may find it beneficial.
• Sblve the chosen problem. Treat it just as if it were an actual test problem.
While both preparing for and taking the test:
Time yourself.
Sketch the proc.:ss, if appropriate, to make the problem clearer.
Write down the basis for your calculation; you will be less likely to
change bases inadvertently.
Write the units with each number. You will prevent some simple mistakes
this way.
List the reference for any data or equations used, so the grader can check
your source.
• Compare your solution with the one given at the end of the chapter.
• Continue with the other problems, easiest to hardest.
• Continue with the other chapters, possibly hardest to easiest. If you should
run out of preparation time, at least you have studied those subjects on
which you most need refreshing.

Remember that you do not have to be proficient in all subjects to be able to

work 8 problems out of 20. Eight correct solutions would give you a perfect
score, and you do not need a perfect score to pass.
You may wish to finish your test preparation with a sample examination
markekd by the P.E. Exam makers-the NCEE, National Council of Engineer-
ing Examiners.t The cost is about $20.

t P. O. Box 16H6, Clemson, South Carolina 29633-1686.

 THERMODYNAMICS

I. Definitions 7
A. Fundamental Quantities 7
1. Time 7
2. Length 7
3. Mass 7
4. Force 7
5. Temperature 7
B. Secondary Quantities 8
l. Volume 8
2. Pressure 8
3. Work 8
4. Energy 9
5. Heat 10
IT. The First Law and Some Basic Principles 10
A Internal Energy 10
B. First Law II
C. State Functions 11
D. Enthalpy 11
E. The Steady-State Flow Process 11
F. Equilibrium 12
G. The Phase Rule 12
H. Reversible Process 13
I. Specific Heat and Heat Capacity 13
III. Pure Fluid Volumetric Properties 13
A PVT Behavior 13
B. The Virial Equation 14
C. An Ideal Gas 15
1. Constant-Volume (Isometric) Processes 15
2. Constant-Pressure (Isobaric) Processes 15
3. Constant-Temperature (Isothermal) Processes 15
4. Adiabatic Processes (dQ = 0) 16
5. Polytropic Processes 16

J. M. Smith, and H. C. Van Ness, Introduction to Chemical Engineering Thermodynamics, 3rd ed ..

McGraw-Hill, New York, 1975.

5
6 THERMODYNAMICS

D. Cubic Equations of State 16

1. van der Waals Equation 16
2. Redlich-Kwong Equation 17
E. Generalized Correlations and Acentric Factor 17
F. Liquid Behavior 18
IV. The Second Law of Thermodynamics 18
A. Statements of the Second Law 18
B. The Heat Engine 18
C. The Ideal-Gas Temperature Scale 19
D. Entropy 20
E. Real Processes and the Second Law 21
F. Irreversibility and Entropy Change 21
G. The Third Law of Thermodynamics 22
V. Thermodynamic Properties of Fluids 22
A. Some Common Thermodynamic Definitions 22
B. Thermodynamic Relationships 22
VI. Properties of Homogeneous Mixtures 24
A. Fugacity 24
B. Fugacity Coefficient 25
C. Activity 26
D. Activity Coefficients and Excess Properties 26
VII. Phase Equilibria 27
A. Criteria 27
B. Vapor-Liquid Equilibrium 27
C. Activity Coefficients from Experimental Data 27
VIII. Chemical Reaction Equilibrium 28
A. The Reaction Coordinate 28
B. Equilibrium and Gibbs Free Energy 29
C. Temperature EtTect 30
D. Equilibrium Constant and Composition 30
IX. Heat into Work by Power Cycles 31
A. Behavior of Some Heat Engines 31
B. T-S and P- V Diagrams for Representing Heat Engine Performance 32
X. Refrigeration and Liquefaction 34
A. Carnot Refrigeration Cycle 34
B. Air-Refrigeration Cycle 34
C. Vapor-Compression Cycle 35
D. Liquefaction Processes 36
Problems 36
Problem-Solving Strategies 38
Solutions 40
 DEFINITIONS 7

I. DEFINITIONS

A. Fundamental Quantities

1. Time

The fundamental unit of time is the second.

2. Length

The fundamental unit of length is the meter. A foot is 0.304 meter (m).

3. Mass

The fundamental unit of mass is the kilogram. The pound mass (Ibm) is 0.454
kilogram (kg). Mass is the measure of quantity of matter, while weight refers to
the force exerted by gravity on mass. An object's weight may vary from place to
place but its mass remains constant.

4. Force

The fundamental unit of force is the newton (N), which is the force which, when
applied to a one kilogram mass, will cause an acceleration of one meter per
second squared. This unit comes from Newton's second law of motion
F = ma/g c

where gc is a proportionality constant. Note that this definition provides a rela-

tionship among four fundamental quantities-time, length, mass, and force.
Thus the units of these quantities determine the value of gc. The newton is de-
fined such that
(kg)(m)
(N)(sec 2 )
For the English system, using pounds force (lbf ) and pounds mass (Ibm),

lbmft
gc = 32.174 - l b2
f sec

5. Temperature

The fundamental unit of temperature is the degree Celsius Cc). It is related to

the absolute temperature K on the Kelvin scale, by
teC) = T(K) - 273.15
8 THERMODYNAMICS

The English unit is the degree Fahrenheit CF), which is related to the absolute
temperature OR on the Rankine scale by
tCF) = TCR) - 459.67
The relationship between the two temperature scales is
tCF) = 1.8tCC) + 32
B. Secondary Quantities

1. Volume

The amount of material being considered affects the volume. But specific volume
is independent of the total amount of material since it is defined as volume per
unit mass or volume per unit mole. The density p is just the reciprocal of specific
volume.

2. Pressure

The pressure P on a surface is the force exerted normal to the surface per unit
area of surface. The pascal, newtons per square meter, is used in the SI system
while psi or pounds force per square inch is used in the English system. Pressures
are usually measured in terms of a height of a column offluid under the influence
of gravity. Since
F = rng/g o
if the mass of the column of fluid is rn,
rn = Ahp
then
F/A = P = hpg/gc
Remember that many of the methods of measuring pressure give the difference
between the measured pressure and the surrounding atmospheric pressure, but
all thermodynamic calculations require the use of absolute pressure.

3. Work

When a force acts through a distance, we have work being done. The quantity of
work is defined by
dW = F dl

where F is the force component acting in the displacement dl direction. With this
definition, work accompanies a change in fluid volume. For the expansion or
compression of a fluid in a cylinder caused by a piston movement, the force
exerted by the piston on the fluid is given by the product of the piston area and
 DEFINI TIONS

FIGURE TI. PV work.

the fluid pressure. The piston displac ement is given by the fluid volume
divided
by the piston area. So,

dW=PAd~=PdV
for constan t A. Then

W= I V2

VI
PdV

as shown in Figure Tl. In the SI system, the unit of work is the newton
meter,
which is called a joule (1). In the English system, the foot-po und force
(ft lb r) is
often used.

4. Energy

For a mass m acted upon by a force F, the relation ship was given by
F = ma/ge'
so

Now acceleration is a = du/dt, where u is the body's velocity. Thus

m du m dl
dW =_ .. dl =- - du
ge dt ge dt
But velocity is u = dl/dt, so
m
dW = -u du
ge
Integra ting for a change in velocity from U1 to U2,

or
10 THERMODYNAMICS

where (mu 2)/(2gc) is called kinetic energy, or

mu 2
EK = - - = ft-Ib e
2gc
Raising a mass m from an elevation Z 1 to an elevation Z 2 requires an upward
force at least equal to the weight of the body and this force must move through
the distance Z 2 - Z l' or

Then
W = F(Z2 - Zl) = m(g/gc)(Z2 - Zl)

= mZ 2 !L - mZ 1 rL = ~(mZ{j) = ~Ep
gc gc gc
where Ep is called potential energy.
If a body is given energy when it is elevated, it is reasonable to assume that
the body will retain this energy until it performs the work of which it is capable.
So if an elevated body falls freely, it should gain in kinetic energy what it loses in
potential energy, or
~EK + ~Ep = 0
This has been confirmed experimentally. So work is energy in transit.

5. Heat

When a hot body is brought into contact with a cold body, the hot body be-
comes cooler, the cooler body hotter. This "something" that is transferred from
the hot body to the cold is called heat Q. Heat is a form of energy, measured as
the calorie or the British thermal unit. These units are related to the primary
energy unit, the joule, by
1 calorie = 4.1840 J
1 Btu = 1055.04 J
and
1 Btu ~ 1055 J ~ 252 cal ~ 778 ft-lb r

II. THE FIRST LAW AND SOME BASIC PRINCIPLES

A. Internal Energy

Energy added to a body as work exists as internal energy until it is extracted as

heat. This internal energy U increases the energy of the molecules of the body.
 THE FIRST LAW AND SOME BASIC PRINCIPLES ))

B. First Law

The total quantity of energy is constant. When it disappears in one form, it

appears simultaneously in other forms. To use this concept, it is convenient to
divide the world into two parts, the system and its surroundings. Everything not
Included in the system is considered to constitute the surroundings. And so the
trst law applies to the system and the surroundings, or
fl(energy of the system) + fl(energy of surroundings) = 0
If a system boundary does not allow the transfer of matter between the system
and its surroundings, the system is closed and its mass is constant. Therefore all
energy passing across the boundary between the system and its surroundings is
transferred as heat and work. So for a closed system,
fl(energy of surroundings) = ± Q ± W,
with the sign choices depending upon which transfer direction is called positive.
F or the closed system,
fl( energy of system) = flU + flEK + flEp
If heat is positire when transferred to the system from the surroundings and if
work is positive when transferredfrom the system to the surroundings,
flU + flEK + flEp = Q- W
or the total energy change of the system is equal to the heat added to the system
less the work done by the system.

C. State Functions

Thermodynamics is made up of two types of quantities: those dependent on

path and those not. Those quantities not depending upon path are called state
or point functions. Internal energy is a state function; work and heat, being path
dependent, are not state functions.

D. Enthalpy

Enthalpy is defined by the equation

H = U + PV
It is also a state function.

E. The Steady-State Flow Process

Consider the process shown in Figure T2, with a heat exchanger and a turbine,
operating at steady state (as opposed to the prior closed system). Our first-law
equation for this system is
flu 2 g flZ
flU + -- + - = Q- W
2ge ge
12 THERMODYNAMICS

I Qt-,
U1--

Zl

l1r
U2

j
FIGURE T2. Steady-state flow process.

but the W includes all work interchanged between the system and the surround-
mgs, or

Then

or

since
AU + A(PV) = AH
Of course, for systems where kinetic and potential energy changes are negligible,

AH=Q-Ws

F. Equilibrium

An equilibrium system is one in which there is no tendency for a change in state

to occur. There is no driving force present.

G. The Phase Rule

The state of a pure homogeneous substance is fixed when two state functions are
set. For more complex systems, this number changes. For nonreacting systems,
J. W. Gibbs showed that the number of degrees offreedom F is given by
F=2-n+N
where n is the number of phases and N is the number of species. (This is some-
times remembered by P + F = C + 2, Police + Force = Chief + Two, P =
number of phases.)
 PURE FI~UID VOLUMETRIC PROPERTIES

H. Reversible Process

A frictionless process operating in a differential manner producing the ma.\J-

mum attainable work is a reversible process.

I. Specific Heat and Heat Capacity

A definite amount of heat is required to raise the temperature of a given mass of

any material by one degree. The quantity is called the heat capacity of the sub-
stance. This relationship may be expressed as
dQ = nC dT
where n is the number of moles, C is the molar heat capacity, and dT is the rise in
temperature caused by the dQ quantity of heat. Had we expressed C as the heat
capacity per unit mass, we would have used the specific heat.
For a constant volume process

or

~u = Q= f CvdT

while for a constant pressure process

or

III. PCRE FLUID VOLUMETRIC PROPERTIES

A. PVTBehavior

Homogeneous fluids are either gases or liquids, as shown in Figure T3. But the
distinction becomes blurred at the critical point, where there is but one phase.
Figure T3 provides no information about volume. If we were to draw isotherms
on Figure T3 to the right of the solid region and plot pressure versus molar or
specific volume for each isotherm, we would obtain Figure T4. The T1 and T2
lines are isotherms at temperatures greater than the critical temperature; they
are smooth. This figure shows that where a single phase exists, there is a relation
connecting P, V, and T, or f(P, V, T) = 0, an equation of state. Of course, the
simplest is the ideal-gas law
PV=RT
14 THERMODYNAMICS

Solid

Triple
Point

Temperature FIGURE T3. Pure fluid PT diagram.

B. The Virial Equation

Experiment has shown that we can write

PV = a(1 + B'P + Cp2 + ... )
and when data for pure substances are plotted in this way and extrapolated back
to P = 0, all substances intercept at the same point. So we fix the absolute tem-
perature scale so that a is directly proportional to T, that is,
a = RT
where R = 82.05 cm 3 atm/g mol K, the universal gas constant. We can now
rewrite our eq uation

PV ' C ' P2
Z=-=l+BP+ + ...
RT
where Z is called the compressibility factor.

v
FIGURE T4. Pure fluid PV diagram.
 PURE FLUID VOLUMETRIC PROPERTIES l~

Alternatively we could write

B C
Z=I+-+-+···
2 V V

Both equations are virial expansions; Band B' are called second vi rial coeffi-
cients, C and C' third vi rial coefficients, and so forth. Note that the two sets of
coefficients are related by

B'=~ D' = D_-_3_B_C---,+:--2_B_3

etc.
RT' (RT)3

C. An Ideal Gas

The definition of an ideal gas is one whose equation of state is Z = 1 and whose
mternal energy is independent of pressure and volume, being a function of tem-
perature only. If we apply the first law to reversible non Bow processes involving
1 mol of gas, then for

1. Constant- Volume (Isometric) Processes

dU = dQ = CvdT

2. Constant-Pressure (Isobaric) Processes

dH = dU + RdT
or

or

I
3. Constant-Temperature (Isothermal) Processes

dU = dQ - dW =0 or Q=W

or
V2 Pl
Q= W = RTln~ = RTln-
Vl P2
16 THERMODYNAMICS

4. Adiabatic Processes (dQ = 0)

dU = -dW = -PdV= -(RT/V)dV= CvdT

or
dT R dV
T C,. V
If we define " = Cp/C v, then

or

The expression for work W becomes

W = P1 [1 _(P2)(Y-l)/YJ =
V1
y-l P1
RT~
y-l
[1 _(P2)(Y-l)/YJ
P1

These expressions yield fairly good results for real gases. For monatomic gases,
y = 1.67; for diatomic gases, y = 1.40; and for simple polyatomic gases (C0 2 ,
S02' NH 3 , and CH 4 ), y = 1.3.

5. Polytropic Processes

F or this general case, when only the reversibility of an ideal gas is assumed:
dU=dQ-dW
dW = PdV
dU = CvdT
dH = CpdT
and so
dQ = Cp dT + P dV

D. Cubic Equations of State

1. van der Waals Equation

(p + :i)(V - b) = RT,
 Pt;RE FLUID VOLUMETRIC PROPERTIES 17

using

2. Redlich-Kwong Equation

RT
P= - --
V - b
~-c-c----
Tl/2(V
a
+ b)V
or

z=l~h-~C :h}
b A
B = RT' B

0.4278R 2 TZ· 5 b = 0.0867 R T.:

a=
Pc

E. Generalized Correlations and Acentric Factor

The principle of corresponding states (using reduced temperatures and reduced

pressures) is much better than the ideal-gas law. To make the results even better.
Pitzer developed a third factor, the acentric factor w, determined from the vapor
pressure measured at a reduced temperature of 0.7,
w = -loglo(p~athr=0,,7 - 1.000

where P r = PIPe' ~ = TIT.:·

Based upon this idea
z = 1 + (BP~)(Pr)
RT.: ~

BPe= B O + wBl
RT.:
o 0.422
B = 0.083 - T1.6
r

This equation is accurate for ~ > 0.486 + 0.457 Pr.

For conditions below this line, use
Z = zO + wZ 1
where ZO and ZI are obtained from generalized curves such as those given by
Smith and Van Ness, pp. 89-90.
18 THERMODYNAMICS

F. Liquid Behavior

The Tait equation

v= Vo - D In( P
Po
+ E)
+E
gives excellent results for liquids. Here Vo, Po are a set of reference state volume
and pressure, and D, E are correlated from experimental data at a given temper-
ature

IV. THE SECOND LAW OF THERMODYNAMICS

A. Statements of the Second Law

There are many different ways of stating the second law of thermodynamics.
Some of them are

1. Any process that consists solely in the transfer of heat from one tempera-
ture to a higher one is impossible.
2. Nothing can operate in such a way that its only effect (in both system and
surroundings) is to convert heat absorbed by a system completely into
work.
3. It is impossible to convert the heat absorbed completely into work in a
cyclical process.

Thus the second law-does not prevent the production of work from heat, but it
does limit the efficiency of any cyclical process.

B. The Heat Engine

The heat engine, a machine that produces work from heat in a cyclical process,
illustrates the second law. For instance, a steam plant operates as follows:

1. Part of the heat from the fuel is transferred to liquid water in the boiler,
converting it into steam at a high pressure and temperature. Let the heat
absorbed by the steam be Q1.
2. Energy is transferred from the steam as shaft work by a turbine. This is
adiabatic.
3. Exhaust steam from the turbine is condensed at a low temperature by the
transfer of heat to cooling water. Call this Q2. Note that, for the time
being, the numerical values of Q 1 and Q2 are regarded as positive.
4. Liquid water is pumped back to the boiler, completing the cycle. This is
adiabatic.
 THE SECOl'iD LAW OF THERMODYNAMICS 19

Now the system energy change must be zero for the complete cycle, or
~Ucycle = 0 = Ql - Q2 - W
or
Ql - Q2 = W
The ratio of W to Ql ' the efficiency of the engine for converting heat into work,
would be expected to depend upon the degree of reversibility of processes 1-4. In
fact, an irreversible heat engine cannot have an efficiency greater than a reversi-
ble one.

C. The Ideal-Gas Temperature Scale

Consider an ideal gas as the working fluid in a reversible cyclical heat engine or
Carnot engine as illustrated in Figure T5. The cycle consists of an isothermal
expansion from A to B, during which an amount of heat Q1 is absorbed at TI .
Then we have an adiabatic expansion BC to a lower temperature T2 • Following
this, we discard an amount of heat Q2 at the constant temperature T2 while
undergoing the isothermal compression from C to D.
Finally the gas returns adiabatically to its original state by compression D to
A.
Now the net work done by the gas is the algebraic sum of the work effects for
each of the steps:

The BC and DA processes are reversible and adiabatic, so the first law says that
the work is equal to -~U,

Wbc= -~Ubc= _fT2 CvdT

TI

Wda= -~Uda= _fT' CvdT

T2

Volume FIGURE TS. Carnot cycle PV diagram.

20 THERMODYNAMICS

for each mole. Thus these two amounts of work cancel each other and

But these are isothermal processes, so

and for isothermal processes AU = 0 so

The efficiency, or the ratio of work obtained to the heat absorbed, WoeJQ1' is
given by

Because PV = RT for all the processes in this ideal cycle, and PVY = constant
for adiabatic reversible processes,

Pb Pc
Pa Pd
Then
Woet Tl - Tz
Ql Tl

D. Entropy

We now return to the usual sign convention that heat is positive when absorbed
by the system and negative when rejected by the system. The prior equation then
is really

or
 THE SECOND LAW OF THERMODYNA:\IICS

This suggests that the quantities QdT1 and Q2/T2 represent property (h.lr~~'
the working fluid because their sum is zero for the cycle, a characteristIC: (If "

property or state function. Expressed differentially,

dQl + dQ2 = 0
T1 T2
The integral of dQR/T (R emphasizes the necessity of reversibility) is the same for
any reversible path and so depends only upon the initial and final states. This
quantity has been given the name change in entropy ,1.S. The calculation of en-
tropy changes must always be for reversible conditions.
Entropy ideas may be summarized:

1. The entropy change of a system undergoing a reversible process is equal

to dS = dQJT or
,1.S = f d;R
2. To calculate the ,1.S for a system undergoing an irreversible change in
state, we must devise a reversible process for accomplishing the same
change and then evaluate SdQ/T over the reversible process. This will
give us the entropy change for the original irreversible process.
3. If a process involves only the transfer of energy as heat to a system, the
entropy change is correctly evaluated from SdQ/T for the actual process,
even if it is irreversible.

E. Real Processes and the Second Law

All processes proceed in a direction such that the total entropy change is posi-
tive, approaching a limit of zero, as the process approaches reversibility, or

No process can be invented for which the total entropy decreases.

F. Irreversibility and Entropy Change

We develop the relationship between the total entropy change and the degree of
irreversibility with the following example.
When an amount of heat Q is transferred from a system at a temperature T[
to surroundings at the lower temperature T2 , the total entropy change is given
by

,1.Stotal = ,1.Ssurr + ,1.Ssys = ~ ~ ~ = Q( T~~p )

On the other hand, suppose that the heat Q is transferred to a reversible heat
engine that absorbs this amount at T1 , converting part to work and rejecting the
22 THERMODYNAMICS

remainder at T2 . The quantity of work that could be obtained, but which was
lost because of the irreversible nature of the actual process is given by

W = ---y:--
Q( Tl - T2) = TzQ (TlT-1 T2T2)
-

So if we eliminate Q between these equations,

G. The Third I~aw of Thermodynamics

The absolute entropy is zero for all perfect crystalline substances at absolute
zero temperature. Thus, the absolute entropy of a gas at temperature T is given
by
S= fTr (C 1 dT + ~Hf + fTV (C )1 dT + "!l!v + fT ~Cp)g dT
p p

o T 7;- Tr T 1:. Tv T
where 7;-. T., = temperature of fusion, vaporization and
AH r , AHy = heat of fusion, vaporization.

V. THERMODYNAMIC PROPERTIES OF FLUIDS

A. Some Common Thermodynamic Definitions

U is the internal energy of a system; it includes all forms of energy contained in a

system except the energy due to position (potential) or motion (kinetic) of the
system.
H = U + PV is the enthalpy. It is purely a convenient thermodynamic func-
tion relating internal energy and the product of pressure and volume.
A = U - TS, the Helmholtz free energy, is a convenient thermodynamic
function relating internal energy and the product of the two fundamental prop-
erties, temperature and entropy.
G = H - TS is another convenient thermodynamic function like A. It is most
frequently referred to as the Gibbs free energy, although there are those that say
this is a misnomer.
Anyone who has studied thermod;.namics is appalled at trying to learn all the
interrelationships between the various thermodynamic terms. We shall usc very
few of them, but for general information a memorization method is presented.

B. Thermodynamic Relationships

The method consists of eight thermodynamic functions arranged in a square:

The order of the symbols of the functions can be remembered by means of
t :\1ethod of Shih-Ching Suo modified by Ch111- Yung Wen.
 THERMODYNAMIC PROPERTIES OF FLUIDS 23

W
AOH
+'r G

U
(a)
e-

~
+T~ l~

W
~
U
(b)
~
+T~

A~
W
U
(c)
~

+T

+T~e- A

+V s-
(d)

FIGURE T6. Thermodynamic relationships.

the coded sentence; "The Gods Punish Humble Souls Until Vanquislled
Altogether," with the initial1etters placed as shown in Figure T6a.

1. To obtain dU as a function of the appropriate independent variables,

draw the two diagonals toward U, using the signs corresponding to the
symbols at the base of the vectors, as shown in Figure T6b, to get

dU = TdS - PdV

In a like manner,

dA = - S dT - P dV
dG = V dP - S dT
dH = V dP + T dS

2. The partial derivatives may be obtained from triangular relationships, as

shown in Figure T6c,

caul = T
(as)y

where the variable outside the parentheses, V in this case, is always at the
right angle. From this same triangle, we obtain

(aA) = -s
(aT)y
 24 THERMODYNAMICS

Using the other triangles, we obtain

(aU) = _p (oH) = v
(oVh ' (oP)s

(a H) = T (oGl = -s
(oS)p , (cT)p

(oG) = V (oA) = _p
(oPh ' (avh
3. The "Maxwell relationships can be obtained by considering the right
triangle with the same base, as shown in Figure T6d,
(oT) (OP)
(aVh (as)v
Likewise with the others
(av) (aT) (as) (oV) (ap) (as)
---- ---
(oS)p caph' (oPh caT)p' CaT)v (av)y
4. Four other relationships may be derived from the corners of Figure T6a:
(aT) P (av) T
-
(aV)A s' (as)u P
(as) v (oP) s
(OP)H T' (aT)G v

5. Another four relationships may be obtained from the trapezoids, as

shown in Figure T6e,
U - A = TS
Likewise,
A - G = -PV, G-H=-ST, H- U = VP

6. Many other relationships, such as chemical potentials, also can be de-

rived. for example, by considering each side of Figure T6a,
(aU) (aA) (oG) (oH)
/1=- - =---=---= -
Can)s.v (an)y,v (cn)p,T (an)p,s

VI. PROPERTIES OF HOMOGENEOUS MIXTURES

A. Fugacity

Gibbs free energy G is defined by

G = H - TS

.' ,
 PROPERTIES OF HOMOGENEOUS MIXTURES

It has a unique relation to temperature and pressure since dH = T dS + ~ dP.

for if we differentiate Gibbs free energy
dG = dH - T dS - S dT
= -SdT+ VdP
-\pplying this to 1 mol of a pure fluid i at constant T,
dG i = ~dP

But for an ideal gas ~ = RTjP, so

dP
dG·I = RT-
P

or
dG i = RTdin P

constant n. This suggests replacing the pressure P with a new function,.t;, or

I at
the fugacity of pure i,
dG i = RT dint;
with.t: having the units of pressure. To complete the definition,

II· m.t:- = 1
p-o P
The fugacity of a component in solution,i, is defined likewise:

lim j~ = 1
p-o xiP
(at constant T). Thus for an ideal gas
7'. =
.Ii x.p
l

and this is the definition of partial pressure.

B. Fugacity Coefficient

The fugacity coefficient is the ratio of the fugacity of a material to its pressure.
For a pure substance, ¢i = NP. For a component in solution, ¢i = .t:!(XiP),
These values may be calculated from P VT data.
dG i = ~ dP = RT dln.t:

(at com'ant n. Now since

dP
In.t: = In P + In ¢i' and d In.f; = p- + din ¢i
26 THERMODYNAMICS

and

or

But Z = PV;/(RT), so
d In <Pi = (Zi - 1) dP/P
or

(at constant T). Equations and generalized plots have been developed from this
equation (see Smith and Van Ness, pp. 236-237).

C. Activity

The activity di of component i in solution is defined to be

where Ii is the fugacity of component i in a standard state at the same tempera-

ture and pressure as that of the mixture. Now if

-AG = - 1 L x.(G.- - G.O) = L (x. In d·)

RT RT I I I "

For an ideal solution,

( AG) = L Xi In Xi
RT ideal

D. Activity Coefficients and Excess Properties

An excess property is the difference between an actual property and the property
that would be calculated for the same conditions of T, P, and X by the equations
of an ideal solution. Superscript E denotes the excess property. Using this con-
cept, it is possible to obtain
 PHASE EQUILIBRIA

where the activity coefficient f'i is given by

Based upon the partial molar excess property,

GE
R~=ln')'i

VII. PHASE EQUILIBRIA

A. Criteria

The equilibrium state of a closed system is that state for which the total Gibbs
free energy is a minimum with respect to all possible changes at the given T and
P, that is, (dGh,p = 0,

B. Vapor-Liquid Equilibrium

Using the fugacity coefficient for each phase and equating them at equilibrium,

Yi(foi = xi(fol
but the (fo) and ¢~ are composition dependent. When the vapor phase is an ideal
gas and the liquid phase is an ideal solution, we can reduce this equation to
Raoulfs law:

,_
Yi - Xi
(p~a[)
P

C. Activity Coefficients from Experimental Data

Activity coefficients for a given system are related to GEjRT, so most correla-
tions are based upon this function, The excess Gibbs free energy GE is a function
of T p, and the :<s but for liquids the P effect is weak. So for low pressure at
constant T

A power ,cries. the Redlich- Kister expansion,

GI::RT. ,
= B + ((XI - x 2) + D(x 1 - x 2 )" + ...
X I .\2
28 THERMODYNAMICS

is effective for binary mixtures. For instance, if we use

the corresponding equations for In 1'1 and In 1'2 are

In 1'1 = BxL In Yz = Bxf

If we use

then
In ~'l = X~[A!2 + 2(AZ! - A 12 )x l ]
In ~2 = xi[A 2 ! + 2(A12 - AZ1 )x Z ]

the Margules equations. If we use

then we obtain

In 1'1 = x 2A' '[ (A'All,12_1 ) J2

2 12//1+ Xl --A'lZ !( 1+ A' )Z
A 21 xX 2
,121

In h = XiA'12![1 + (~:: - l)X z J A~)(l

= + ~::::Y
the van Laar equations.

VIII. CHEMICAL REACTION EQUILIBRIUM

A. The Reaction Coordinate

Following Aris and Petersen, if we write the stoichiometric equation so that the
stoichiometric numbers Vi are positive for a product and negative for a reactant,
then we may write

dlJ) = dn z = ... = de
v! V2

or

where ni = moles of species i and e = reaction coordinate or extent of reaction.

 CHDHCAL REACTION EQUILIBRIUM 29

When two or more reactions proceed simultaneously, a separate reaction

coordinate I'.j is associated with each reaction j. If we have r independent reac-
tions, and the stoichiometric numbers of the species Ai are denoted for each
species and each reaction by Vi.) (i = 1,2, ... , N are species and j = 1, 2, ... , r
are the reactions), then

(j = 1. 2, ... , r)

and
(i = 1, 2, ... , N)

B. Equilibrium and Gibbs Free Energy

The total differential of the Gibbs free energy of a single-phase, multicomponent

system is

or
dC t = _st dT + V t dP + L (fliVJ de
where fli is the chemical potential of component i. Note that from this we obtain

and, at equilibrium,

Now
fli = G~ + RT In Iii
So

or
In TI ai" = - L viG;/RT
where TI means the product over all i. It is usual to define
K = TI at'
So, finally,
-RT In K = L viGi= ~Go
30 THERMODYNAMICS

where K is the equilibrium constant and AGo is the standard Gibbs free energy
change of reaction. The activities iii = ];1 ffare the connection between the equi-
librium state and the standard states of the constituents at the equilibrium tem-
perature T.
Sometimes data are given such that one must use

where AH and ASo are the standard heat of reaction and the standard entropy
change of reaction.

C. Temperature Effect

The van't Hoff equation relates the equilibrium constant to the temperature

dinK AH'
dT RT2

D. Equilibrium Constant and Composition

For gas-phase reactions

K=flft
where the fugacities are functions of T, P, and composition. But K is a function
of temperature only. So for a fixed temperature, the composition at equilibrium
must change with pressure in a way that fl .ni
remains constant. Now

with (Pi the fugacity coefficient, Yi the mole fraction of component i in the equili-
brium mixture, and P the equilibrium pressure. So

where v = L Vi and P is in atmospheres. The y;'s may be expressed in terms of

the reaction coordinate at equilibrium ceo The ;Pi may depend on composition
and thus upon 8 e .
lf the equilibrium mixture is an ideal solution. then (Pi = ¢i' the fugacity coef-
ficient of pure i at T and P. For this case

and the ¢i may be evaluated from generalized correlations if T and P are known.
If the temperature is sufficiently high or the pressure is sufficiently low, the
equilibrium mixture will be an ideal gas and (Pi = 1 or
r
I HEAT INTO WORK BY POWER CYCLES ~ 1

IX. HEAT INTO WORK BY POWER CYCLES

A. Behavior of Some Heat Engines

Several of the more important heat engines will be discussed here. A much larger
selection may be found in Reynolds and Perkins, Engineering Thermodynamics,
McGraw-Hill, New York, 1970.
The steps in the cycle of each engine are shown in Table T1. The engine
performance can be obtained by solving for the heat added and the work done
during each step. The overall performance can be calculated using the given
efficiency equations, based upon the nomenclature of Figure T7. To illustrate

TABLE Tl a
Otto Cycle Diesel Cycle Gas Turbine
Carnot Cycle (air standard) (air standard) (air standard)

Tj - T2
1]=1-
CY-
- 1
'1=1-
CJ -(~Y t/=I- (P-Y-1I!Y
-
1 C
1]=7'-
r Pb
-; -. -
1)
re r e

Adiabatic Adiabatic Adiabatic Isentropic

compression compression compression compression
Isothermal Const. volume Const. pressure Const. pressure
expansion heat absorp. expansion heat absorption
Adiabatic Adiabatic Adiabatic Isentropic
expansion expansion expansion expansion
Isotherm Canst. volume Const. volume Const. pressure
compression heat rejection heat rejection heat rejection

a Notes r = compression r, = compression Ph> Pa

ratio ratio
V before camp. expansion
Tl > 7~
V after compo
r0 = ratio
V after expan.
------
y = Cp/C
V before expan.

FIGURE T7. Nomenclature for heat engines of

w Table TL

=
32 THERMODYNAMICS

the use of Table Tl, consider this problem: A Carnot engine and air-standard
Otto, diesel, and gas turbine engines are each operating with a heat addition rate
of 10,000 Btu/hr. The Carnot engine is operating between heat reservoirs of 1200
and 300 F. The other engines have compression ratios of eight '.0 one, and the
J

diesel engine has an expansion ratio of two to one. Find the net horsepower
produced by each engine.

Solution: For the Car not engine

= Tl ~ T2 = 1200 ~ 300 = 0.5422

I] T1 1200 + 460
= (10,000 Btu/hr)(0.5422)
= 5422 Btu/hr
Converting to horsepower,
(5422 Btu/hr)(778 ft lbr/Btu)
W= --=2.13hp
(550 ft lbr/sec hp)(3600 sec/hr)
For the Dtto engine, r = 8 and the efficiency is
I] = 1 - (l/ry-1 = 1 - (i)O.4 = 0.5647 or 56.6%
W = Qll] = (10,000 Btu/hr)(0.5647) = 5647 Btu/hr = 2.22 hp
For the diesel engine, r = 8, re = 2,

I]
r r - Y ~ r -;'
= 1- - e = 1~-
8[2 - 1.4 - 8- 1.4J
= 0.3819
}' (r/re) - 1 1.4 (8/2) - 1
W = Qll] = (10.000 Btu/hr)(0.38 19) = 3819 Btu/hr = 1.50 hp
For the gas turhine engine, r = 8,
I] = 1 - (I/r)(r ll/y = 1 ~ (DO 41.4 = 0.4480
W = QlIJ = 4480 Btu/hr = 1.76 hp

B. T-S and P- V Diagrams for Representing Heat Engine Performance

In understanding the operation of a heat engine, it is often useful to plot the

operation on a diagram of temperature versus entropy (T -S) or pressure versus
volume (P - V). On a plot of pressure versus volume. each of the heat engine
processes can be represented as a line as shown in Figure T8. The lines are

1. V = constant, constant volume process;

2. P = constant, constant pressure process;
3. PV = constant, a process at constant temperature and enthalpy; and
4. PVY = constant, a process at constant entropy (adiabatic and reversible).
r HEAT INTO WORK BY POWER CYCLES

1
p

v ..
FIGURE T8. General PV diagram.

Now let us use the previous description of heat engines to plot a P- V dia-
gram. For the Carnot cycle, as shown in Figure T9, we have:
a -+ b: adiabatic compression from P a , V. to P b , Vb' Temperature rises from
Tz to T1 . Entropy remains the same.
b -+ c: heat adsorption at constant temperature T1 • Pressure decreases from
Pb to Pc while volume increases from Vb to v;,. Enthalpy remains the same, while
entropy increases.
c -+ d: adiabatic expansion from Pc, v;, to P d , J;d. Temperature falls from Tl
to T2 . Entropy remains the same.
d -+ a: heat rejection at constant temperature Tz . Pressure increases from
Pd to P a while volume decreases from Vd to Va' Enthalpy remains the same, and
entropy decreases.

-1---r-
I I
I I
___ L
I
p I
I
_L
I

FIGURE T9. PV diagram for Carnol engine.

34 THERMODYNAMICS

X. REFRIGERATION AND LIQUEFACTION

A. Carnot Refrigeration Cycle

For a continuous refrigeration process, since heat must be absorbed at a low

temperature, there must be continuous rejection of heat to the surroundings at a
higher temperature. So a refrigeration cycle is simply a reversed heat-engine
cycle. Since heat is transferred from a low level to a higher one, external energy
must be used so as not to violate the second law. The ideal refrigeration cycle is
the Carnot: two isothermal processes where the heat Q2 is absorbed at the lower
T2 temperature and heat Q1 is rejected at the higher Tl temperature; and two
adiabatic processes whose result is the addition of the net work W to the system.
Note that here we use both Ql and Q2 positive so W is positive and equal to the
work added to the system.
From the first law, since all f\.U of the fluid is zero for the whole cycle
W = Ql - Q2
From the Carnot cycle of Figure TIO,

so

Q2 T2
We use this to calculate the required work W for a given quantity of refrigera-
tion Qz using the ideal Carnot cycle.

B. Air-Refrigeration Cycle

The air-refrigeration cycle is not ideal; it can be built, but usually is not. In this
cycle, the refrigerator air absorbs heat at an essentially constant pressure P 1 in
the cold space, and, in the cooler, rejects heat to the surroundings at a higher,
constant P 2' The gas is compressed (ideally) at constant entropy, A to B, using

FIGURE no. Idealized Carnot T -S diagram.

r REFRIGERATION AND LlQLEFACTlO:\ .....:.

c A

s FIGURE TIL Air-refrigeration T S diagram.

as part of the energy required, the work obtained from the expansion process
CD. This process is illustrated in the T -5 diagram of Figure Tt 1.
The thermodynamic analysis of this cycle will illustrate the type of calcula-
tions needed for other, more practical cycles. If the specific heat of the gas is
constant, the heat absorbed in the refrigerator, Q2' and rejected in the cooler, Ql'
are

Q2 = mCp(TA - TD )

Ql = me p(TB - Te·)

For the reversible adiabatic processes CD and AB, the P- T relationship is

Since refrigeration performance is defined in terms of a coefficient of perfor-

mance (COP),

COP = -refrigeration
---- obtained
-- Q2
work required W
for this air cycle

COP = Qz = T"
W I;l-TA

C. Vapor-Compression Cycle

Constant-pressure evaporation of a liquid is a method of absorbing heat at

constant temperature; constant-temperature condensation of a gas at a higher
36 THERMODYNAMICS

pressure is a method of rejecting heat. This scheme is the vapor-compression

cycle:
constant pressure evaporation:
isentropic compression:
constant pressure condensation:
adiabatic expansion:

This leads to
COP HA -HD
= -------
HB - HA
12,000
circulation rate = m = -- ~­
HA -HD
= Ib per hr per ton of refrigeration
N ate: A ton of refrigeration = 12,000 Btu/hr.

D. Liquefaction Processes

The basis of allliq ucfaction processes is to cool the gas until it enters the two-
phase region. It is done by three main methods:

1. Cooling at constant pressure (heat exchanger).

2. Cooling by expansion in an engine from which work is obtained.
3. Cooling by an expansion-valve or throttling process.

Of these, the constant-enthalpy or Joule- Thomson expansion (3) is the one most
used.

PROBLEMS

A. Steam comes from a boiler at 650 psi a and 760°F. After expansion to 150
psia, the steam is reheated to 760°F. Now expansion occurs to 3 in. Hg abs.
For the Rankine cycle with reheating (ideal reheat cycle), calculate, for
1 lb of steam, (a) the heat added, (b) the heat rejected, (c) net work, (d)
thermodynamic efficiency, and (e) quality of exhaust steam. Consider a
Rankine cycle operating between the upper limit of 650 psia, 760c F, and the
lower limit of 3 in. Hg abs. Compute these quantities. Compare the thermo-
dynamic efficiencies of the two cycles. Also compare the thermodynamic
efficiencies with the Carnot efficiency.
B. (a) Calculate the change in entropy when 1 lb of NH3 is cooled from
400 to - 17 5-F at a constant pressure of 1 atm. Express results as
Btu/lbR. (b) Calculate the absolute entropy of solid NH3 at its melting
point at I-atm pressure, in Btu/lboR. Boiling point = - 33.4°C; melting
 PROBLEMS

point = - 77.7°C; Ciliq) = 1.06 cal/g K; C isolid) = u.5u: ;:"L ~ r....

)'v = 5581 cal/g mol; I.f = 1352 cal/g mol; absolute entropy of gas c: [
25°C = 46.03 cal/g mol K.
C. A hydrocarbon mixture (15 mol % methane, 15 % ethane, 25 % propane,
20 % isobutane, 20 % n-butane, and 5 % n-pentane) is flashed into a separa-
tor at 80°F and 150 psia. Fifty two percent of the mixture leaves the separa-
tor as liquid. What is the composition of this liquid? K values: methane,
17.0; ethane, 3.1; propane, 1.0; isobutane, 0.44; n-butane, 0.32; n-pentane,
0.096.
D. Iron oxide (FeO) is reduced to Fe by passing a mixture of 30 mol % CO and
70 mol N 2 over the oxide at 100 atm and 100°C. If Ka = 35 (at the stated
%

conditions) for the reaction

FeO(s) + CO(g) ( ) Fe(s) + CO 2 (g)
what is the weight of metallic Fe produced per 1000 ft 3 of the entering gas
measured at 1 atm and 25°C?
Calculate the above if pure CO had been used instead of the mixture.
Assume chemical equilibrium is reached in both cases.
E. A mixture (58.3 mol ~~ n-hexane, 8.33 % n-heptane, 33.3 % steam) is cooled
at a constant pressure (l atm) from an initial temperature of 350'F. If we
assume that the hydrocarbons and water are immiscible, calculate:

1. Temperature at which condensation first occurs.

2. Composition of the first liquid phase as it first appears.
3. Temperature at which the second liquid phase appears.
4. Composition of the second liquid phase as it first appears.

In P = A + BIT: A = 17.7109, B = -6816.4 for n-hexane; A = 17.9184,

B = -7547.4 for n-heptane; P = vapor pressure (mm Hg), T = OR.
F. It is claimed that pure ethane at 0.10 atm abs and 1000 K can be converted
to ethylene and hydrogen and 93 % conversion is claimed. Is this possible?
~G~98 = - 7860 cal/g mol
~H~98 = -20,236 cal/g mol
C; = 2.3 -I- 0.02T(cal/g mol K)
~G~98 = 16,282 cal/g mol
!lH~98 = 12,500 cal/g mol
C~ = 2.8 + 0.03T(cal/g mol K)
Hydrogen ~G~98 = ~H~98 = 0
(H 2) C~ = 6.9 +
0.004 T(cal/g mol K)
G = H - TS = Gibbs free energy; ~G~98' ~H~98 = Gibbs free energy,
enthalpy of formation at these conditions
38 THERMODYNAMICS

Feed
Compressor

Liquid

FIGURE Tt2. Schematic for problem G.

G. Find the pressure needed to obtain maximum liquefaction in the process

shown in Figure T12 with the following data on compressibility factors.

Temperature
Pressure
(psia) 70°F 80°F 90°F 100°F 110°F
700 0.166 0.177 0.415 0.535 0.590
1000 0.223 0.228 0.227 0.253 0.277
1500 0.316 0.319 0.324 0.330 0.338
2000 0.407 0.408 0.410 0.412 0.417
2500 0.494 0.493 0.493 0.495 0.498
3000 0.579 0.577 0.576 0.575 0.577

PROBLEM-SOLVING STRATEGIES

A. To do this problem, I need to know the steps involved in a Rankine cycle

with reheating, a plain Rankine cycle, and the Carnot cycle. If I can find the
steps defined-possibly in Smith and Van Ness-then I can solve the prob-
lem using a set of steam tables. Because of all the parts to this question, I
would, in the interests of time, try to find shorter, possibly easier, problems.
B. With the equation for the definition of entropy, part (a) of this problem is
simple. Take care with the differing units. For part (b) I must work back-
wards from the given gaseous entropy at 25°C since I do not know the solid
entropy at absolute zero. An easy problem if I am careful with the unit
conversions.
 PROBLEM-SOLVING STRATEGIES 39

C. This equilibrium flash should be simple, because the temperature is speCI-

fied and the values for K are given. (It would be trial and error if tempera-
ture were not specified.) I shall just use component balances
and
so
FZ i = LXi + VKiX i = (L + VK;)Xi
with LiF = 0.52, V/F = 0.48. J can calculate each Xi in turn.
D. This is a chemical equilibrium problem. The equilibrium constant is given
and it can be expressed in terms of total pressure, mole fractions, and fuga-
city coefficients. To get the mole fractions present at any time, do a material
balance with the stoichiometric equation. To get the fugacity coefficients,
use the reduced temperatures and pressures of the components with fuga-
city coefficient graphs. Thus J can solve for conversion if I have expressed
mole fractions in terms of conversion. With conversion available, all other
calculations are easy. Keep in mind that inerts affect the mole fractions of
reactants and products.
E. This problem is a little different. It could be trial and error. I know that at
the time that the first drop of liquid forms, the vapor mole fractions are
those given. Because of the immiscibility, if water condenses first, its tem-
perature will be at the point that gives a vapor pressure equal to the total
pressure times the mole fraction of water vapor, 0.333. If I make this as-
sumption, I must check. using vapor pressures and mole fractions, to see if
the hydrocarbon phase could have condensed at a higher temperature. Un-
less I can find a discussion of this type of problem (Smith and Van Ness or
Perry), I would pass over it-with relief.
F. At first glance this looks to be a simple chemical equilibrium problem. All
the equations are given to calculate the equilibrium constant. Set up the
stoichiometry in terms of conversion and then calculate mole fractions in
terms of con version. Using the equilibrium constant in terms of total pres-
sure, mole fractions, and fugacity coefficients, calculate the conversion. If
the equilibrium conversion calculated is greater than 93.%, the claim is pos-
sible. Fugacity coefficients can be obtained from a graph in Smith and Van
Ness, in terms of reduced temperature and pressures. This should be an
easy problem.
G. Liquefaction brings to mind the Joule-Thomson effect, a constant-enthalpy
expansion. But to get from enthalpy to compressibility will require using
various thermodynamic relationships. I would start with an enthalpy
balance around the valve. On the inlet side is all vapor; on the outlet side is
vapor and liquid. So enthalpy could be related to the fraction of liquid. I
would start this problem, if I felt I had time to think about what thermo-
dynamic relationships are needed.
 40 THERMODYNAMICS

SOLUTIONS

A. See Figure T13.

Basis: 1-lb circulating steam
Want: Qboilep Qreheat' Qcond' W,l' W,2' Wp , '1, and quality.
Compare to Rankine without reheating and with Carnol.

1 t - - - - wsl
650 psia
760 of

Boiler Reheater

Ws2 - - - - I
6
4

5
Condenser

FIGURE Tl3. Schematic for solution A(a).

At point (1), HI = 1382.8, S 1 = 1.6060, T = 760°F, P = 650 psia, W,I =

HI - H 2 · For turbine, assume isentropic operation, Sl = S2. So at point
(2), P 2 = 150, S2 = 1.6060, T2 = 410°F, H2 = 1225, so

W,l = 1383 - 1225 = 158 Btu/lb

At point (3), P 3 = 150 psi a, 760°F = T3 , S3 = 1.780, H3 = 1407,

Qr = H3 - H2 = 1407 - 1225 = 182 Btu/lb

At point (4), P 4 = 3 in. Hg, S4 = 1.7820, two-phase region T4 = 115°F.

Let X = fraction liquid
S = 1.7820 = X(0.1560) + (1 - X)1.9456 quality = 1 - X = 90.9 ~I"
H = ? = X(82.99) + (1 - X)(1111.6) X = 0.091
H4 = 1018 Btu/lb; - W,2 = H4 - H3 = 1018 - 1407 = -389 Btu/lb;
H 5 = 82.99 Btu/lb.

,
t ~

~~L,
 SOLUTIONS 41

At point (5), P5=P4=3in.Hg=1.473psia, T5=T4=115 c F. S,=

0.1560,
Qe = Hs - H4 = 83 - 1018 = -935 Btu/lb

For pump, assume isentropic compression

Wp = f V dp ~ V ~p = 0.01618(650 - 1.47) = 10.49C~:)C:~)

= 10 49(~)(~)(144in.2)(. ~__~~.tu_)
. Ibm in. 2 ft2 778 [t Ibr
Btu
= 1.94 lli

At point (6), H 6 = H 5 + Wp = 83 + 1.94 ~ 85 Btu/lb

Qb = HI - H6 ~ 1383 - 85 = 1298 Btu/lb

Check
Qb + Q, + Qc = 1298 + 182 - 935 = 545 Btu/lb
W,1 + W,2 - Wp = 158 + 389 - 2 = 545 Btu/lb
work 545 545
11 = ~+ Qb = 1298+ 182 = 1480 = 37 ~o

Conventional Rankine cycle (see Figure TI4).

HA = HI = 1382.8 Btu/lb, S", = S 1 = 1.6060, 1.6060 = X(O.l560)

+ (l - X)1.9456 or X = 0.190, quality = 81 %
HH = X(82.99) + (1 - X)1111.6 = 916.1 Btu/lb
W, = HA - HB = 1382.8 - 916.1 = 466.7 Btu/lb
Qcond = He - HH, He = H 5, Qc = 83 - 916.1 = -833.1 Btu/lb

3"Hg

c
Boiler Condenser

Wp

FIGURE T14. Schematic for solution A(b).

42 THERMODYNAMICS

As before

f v dP = 1.94 Btu/lb and HD = H6 = 85 Btu/lb

Qboiler = HA - HD = 1382.8 - 85 = 1297.8 Btu/lb

'1 = _~~et_ = 466.7 -_1.9 = 35.8 %
Qboiler 1297.8

Carnot efficiency
11 = (760 - 115)/(760 + 460) = 52.9%
204.4°C -115)C
B. !1.S for 1 Ib NH3 ~
400°F gas -175°F solid

Gas 400°F ~ - 33.4°C; !1.S = f(ep/T) dT;

C p = ex + f3T + yT2, rx = 6.086,

fj = 8.813 X 10- 3 , y = -1.506 X 10- 6

Tin K, C p in Btu/lb mol OF. (Smith and Van Ness, 2nd ed., p. 122)

!1.S = IT2 [tl. + f3 + )'T] dT = (ex In T + f3T + Y T2)T2

T, T 2 T,

TJ = 400°F = 477 K, T2 = - 33.4°C = 239.6 K, so

Btu 17lb Btu

ilS 1 = - 6.l645 Thm-;;J0R -:- Ib mol = - 0.3626 1b ci

Gas ~ liquid at-33.4°C

-ilHvap -5581 cal cal
ilS 2 = - - - = - - -- -- = - 23.29-- -- .
Tboi1 g mol 239.6 K g mol K

Btu 17lb Btu

= -23.29 -- - - -:- --- -- = -1.370 -
lb mol OR lb mol IbcF

Liquid -33.9°C ~ -77.TC

ilS 3 = f l95.3 C
---.I'dT;
239.6 T
C = 1.06-
p
cal
gK

195.3 cal Btu

= C In - - = -0.217 - - = -0.217 --
r 239.6 gK Ib"F
--
SOLUTIONS 43

Liquid ~ solid at -77.7°C

-1352 cal = -6.923 ~

g mol 195.3 K g moloC

=
Btu
-6.923 - - - 17 -
lb IBtu
= -0.407-
lb moloF I mol IboF
Solid -nrc ~ -115°C

~S5 = f 158

195.3
C
~dT,
T
C
P
= 0.502 cal
gK

= 0.502 In _158 = _ 0.1064 Btu

195.3 IboF
~S = 0.3626 - 1.370 - 0.217 - 0.407 - 0.1 064
-
Btu
= -2.463- answer to (a)
IbcF
Absolute entropy of solid NH3 at melting point - 77.7°C. Note that we
cannot use g=elt(Cp/T) dT with C p = 0.502 cal/g K, because we get an
indeterminate term at T = O. So we use the summation scheme shown in
Figure T15.
So S - ~S4 - ~S3 - ~S2 + ~Sn = ~S or S = ~S - ~Sn + ~S2 +
~S3 + ~S4

~Sn = 'Y. In 2~~6 + f3(298 - 239.6) + ~ (298 2 - 239.6 2 )

Btu 1.8185 cal Btu

= 1.8185· - - = -- = 0.1070 - or
lb mo\OF 17 gK Ib0R
46.03 Btu
S = - - - - - 0.107 - 1.370 - 0.217 - 0.407
17 Ib)F

= 0.6066 l~~~ answer (b)

T o K
State Solid Liquid Gas

(0 S = 46.03 call g mol K given

-I "I -I
FIGURE TIS. Schematic for solution B.
 44 THERMODYNAMICS

C. Basis: I-mol feed; Yi V + xiL = Zi with Yi = KiXi, so

Zi
Kix i V + xiL = Zi x·
,
=
KiV+L

and
L = 0.52 V = 0.48

Compound Zi Ki 0.48 Ki + 0.52 Xi Yi

Methane 0.15 17.0 8.68 0.0173 0.2938

Ethane 0.15 3.1 2.008 0.0747 0.2316
Propane 0.25 1.0 1.00 0.25 0.25
Isobutane 0.20 0.44 0.7312 0.2735 0.1204
n-Butane 0.20 0.32 0.6736 0.2969 0.0950
n-Pentane 0.05 0.096 0.5661 0.0883 0.0085
1.0007 0.9992

D. Basis: I-mol entering gas or 0.3 mol CO, 0.7 mol N 2'
FeO(s) + CO(g)
---
------+

Att=O,CO 0.3; CO 2 =0.

Fe(s) + CO 2 (g) T = 100°C; P = 100 atm.

At t = t, CO = Ceo; CO 2 = C eo2 ; 0.3 - Ceo = C eo2 = x, then Ceo =

0.3 -x; total moles = Ceo + C eo2 + 0.7 = 0.3 - x + x + 0.7 = 1; mole
fraction CO = 0.3 - x, mf CO 2 = x,

K = aFeaeo2 = aeo2 = Yeo2 ¢eo2 = 35

a aFeOaeO aeo Yeo ¢eo
Get ¢s from Smith and Van Ness, p. 354, 2nd ed.

TdK)

304.1 73 1.23 1.37 0.80

134 35 2.78 2.86 0.97

So

Yeo2 = 35 0.97 = 42.44

Yeo 0.80
Then
42.44 = (x)/(0.3 - x) or x = 0.293 (also moles of Fe produced)

" \\,
~;~~
 SOLL'TIONS

For pure CO entering

mfCO = 1 - x, mfCO z = x
x
-- = 42.44 or x = 0.977 (also moles Fe produced)
1-x

(1000 ft3) . /1[359. ~ (298)J

c
moles gas( 25 C );1000 ft 3 = = 2.55
1 atm I mol 273
For 30% CO: Fe produced = 0.293(2.55)(55.85) = 41.73 lb.
For 100,!~ CO: Fe produced = 0.977(2.55)(55.85) = 139.14 lb.
E. Composition: 0.583C 6 , 0.083C 7 , 0.333H zO; H 2 0 and hydrocarbons are
immiscible as liquids; cooling down from 350°F at moment of first drop of
liquid forming Y6 = 0.583, Y7 = 0.083, Yw = 0.333. So
P6 = 0.583 atm P7 = 0.083 atm P w = 0.333
= 443.33 mm Hg = 63.33 mm Hg = 253.33 mm Hg
= 4.90 psia

If H 2 0 condenses first, T will be that at which VP w = 4.90 psia or T =

161'F = 621R. Check for hc condensation:

X6 -
- ~). x-
Y'6( VP 6 ' f
= .Y7
" (!J_) ..
VP 7 '6
x. + x 7 -- 1

443.33 63.33
----
VP 6 ' VP 7
6816.4
In VP 6 = 7109
17. .
- - .T- ' In VP 7 = 17.9184 _ 7~~7.~
T
At 621R
VP 6 = 840.8, X6 = 0.5272
At 621 OR
VP~ = 318.9. x7 = 0.1986,
So therc is no condensation of hydrocarbons. First condensation occurs at
161 F (answer 1) and it is pure H 2 0 (answer 2).
Assume T = 150°F (610 R) for condensation of 2nd phase. At this T,
C

VP w = 3.718 psia (192.22 mm Hg),

Phc = 760 - 192.22 = 567.78 mm Hg,
567.78
Yhc = . 760 = 0.747.
46 THERMODYNAMICS

Thus

Y6 0.583
~=--=7 and Y6 + Y7 = 0.747
Y7 0.083

So

7 0.747
Y6 = 8" (0.747) = 0.6536, Y7 = -8- = 0.0934

VP 6 = 689.85, VP 7 = 256.11
X6 = 0.7201, X7 = 0.2772 I = 0.9972

Assume T = 149°F(609°R). At this T, VP w = 3.627 psia

(187.52 mm Hg),

Phe = 760 - 187.52 = 572.48, Yhe = 572.78/760 = 0.753.

So

7 0.753
Y6 = 8" (0.753) = 0.6591, Y7 = -8- = 0.0942

VP 6 = 677.31, VP 7 = 250.96
X6 = 0.7396, X7 = 0.2853 I = 1.0248
So to the nearest degree, the second phase condenses at 150°F (answer 3)
and its composition is 72.01 % C 6 and 27.72 % C 7 (answer 4).

o o Cone. at t = 0;
I-x x x t = t:
Total moles = 1 + x

AG~98 = (AG~98H4 + AG~98H,)f - (AG~98H.)f

= 16,282 + 0 - ( - 7860) = + 24,142 cal/mol
AH~98 = (AH~98H4 + AH~98H2)f - (AH~98H.)f
= 12,500 + 0 - ( - 20,236) = + 32,736 cal/mol

AH~ = AH~98 + fT (I ViC Pi ) dT

298

= 32,736 + fT (2.8 + 0.03 T + 6.9 + 0.004 T - 2.3 - 0.02 T) dT

298

= 32,736 + fT (7.4 + 0.014 T) dT

298
= 32,736 + 7.4(T - 298) + 0.007(T2 - 298 2 )
..
SOLUTIONS 47

Now ~Go = -RT In K., so at T = 298K, In Ka = -24,142/1.987(298) =

-40.77. Now
din K ~Ho
d( T) a = + R T2 (van't Hoff)
~H~ = 29,909.17 + 7.4 T + 0.007 T2
~H~/T2 = 29,909.17 T- 2 + (7.4/T) + 0.007

R i lnKaT

InKa 298
din Ka =
IT
298
(29,909.17 T- 2
74
+ _. + 0.007) dT
T

R(ln KaT - In K a298 ) = (

-29,909.17
+ 7.41n T + 0.007 T
)T
T 298

R(ln KaT + 40.77) = [ -29,909.17(~ - 2~8) + 7.41n 2~8

+ 0.007(T - 298)]
T = 1000K
R(ln KaT + 40.77) = 84.33, In KaT + 40.77 = 42.44

In KaT = 1.67, K al000 - 5317 _ YH , YH4 P2

- . - P
YHo
or

YH , YH4 = 53.17, x x 1-x

mfHl = -1-' mfH. = -1-' mfH6 = -1-
YHo -x -x +x
53.17 = + x)
x 2 (l _ _ X_2_
x2 = 0.9815,
(1 + X)2(1 - x) 1 - x 2'
x = 0.9907 equi1 conv.
So a conversion of 93 % is possible.
G. See Figure T16. This is a Joule-Thomson liquefaction problem.
Basis: 1-lb gas at 90°F and P = ? entering cracked valve. H in = H out ;
H in = xHsatliq + (1 - X)Hsalvap' or

X=--=--~~-
H in - H sal yap
H sat liq - H sat yap

~ ~
VAroR

1 lb '-_ _ _.... i-x lb

LlQUIlJ
x lb FIGURE T16. Schematic for solution G.
 48 THERMODYNAMICS

We want

ox) = (ox) = 0 = (OHin)

( oP (_ 1 )
90°F oP T ap T=90°F Hsatliq - Hsatvap
so we want

NowdH= TdS+ VdP.So

( OH)
oP
_ V + T(OS)
T=90° - ap T

from Maxwell relations

-(!!)T = G~)p
( OH)
ap
_ V _ T(OV)
T=900 - oT p
But
v = (ZRT)/P
So

( OV) = ~ [O(ZT)] = ~
oT p P oT p P
[Z + T(oZ)
aT
]
p

So

( OH)
oP T=900
= v _ RT
P
[z + T(OZ)
aT
] = ZRT _ ZRT _ RT2 (OZ)
Pp P P aT p

Then

_ RT2 (OZ) = 0 or OZ) =

( aT 0
P oT p p

From data
at P= 3000 psi a, oZ/oT = 0 at T = lOO°F
at P= 2500 psia, oZ/oT = 0 at T = 85°F
at P= 2000 psia, oZ/oT = 0 at T = 70°F
So at T = 90°F use 155 of (3000 - 25(0) added to 2500 or P = 2667 psia

·,
\

Ii..:
 PROCESS DESIGN

1. Introduction 50
II. Stoichiometry 50
III. Material and Energy Balances 52
A. Introduction 52
B. A Typical Problem 53
C. Heats of Formation and Combustion 54
D. Heat of Reaction 55
1. At 25T 55
2. At Other Temperatures 55
IV. Optimization 56
A. Introduction 56
B. Graphical Method 56
C. Analytical Method 57
1. General Conditions 57
2. Unconstrained Optimization 57
3. Constrained Optimization 59
a. Substitution 59
b. Lagrangian Multiplier 59
V. Process Control 60
A. Introduction 60
B. The Laplace Transform 60
C. Transfer Functions 61
D. Block Diagrams 62
E. Closing the Loop 63
F. Controllers 64
1. Proportional Controllers 64
2. Proportional-Derivative Controllers 65
3. Proportional-Integral Controllers 65
G. Stability 65
1. Introduction 65
2. Partial Fraction Expansion 65
3. Characteristic Roots 66
4. Routh's Method 67
Problems 67
Problem-Solving Strategies 70
Solutions 70

49
50 PROCESS DESIGN

I. INTRODUCTION

Most of the process design problems found on the test are really just unit opera-
tion problems covered under other sections of this book. Theref( ,re this chapter
will cover areas of process design not addressed elsewhere. These topics include
stoichiometry, heat and material balances, optimization, and control.

II. STOICHIOMETRY

Stoichiometry problems involve chemical balances. Combustion problems are

based on stoichiometry principles-the relationships of gas volume, mass, and
composition in a chemical reaction. So we use combustion problems to get you
involved, once again, with stoichiometry. Figure D1 represents most combus-
tion problems.

Flue Gas

I
Fuel

Air
f FIGURE Dl. Combustion problem schematic.

All combustions use oxygen, but costs usually make pure oxygen unattrac-
tive. Thus air is used. For calculation purposes, air is a mixture of 79 volume %
N2 and 21 volume % O 2 , giving a nitrogen to oxygen ratio of 79/21 or 3.76.
The flue gas is most often analyzed by the Orsat technique. This is a simple
volumetric method, but it does not account for water or nitrogen oxides. Thus,
an Orsat analysis is given on a moisture free basis. Because the Orsat cannot
account for nitrogen oxides, we assume that nitrogen oxides are negligible. For
the analysis, we say that none of the nitrogen in the air reacts. Therefore, for
calculation purposes, all the nitrogen entering the system in the air stream leaves
as part of the flue gas.
To insure complete combustion of all the carbon and available hydrogen in
the fuel, air is always fed in excess of the amount needed. So with complete
combustion, all carbon in the fuel shows up in the flue gas as either carbon
monoxide or carbon dioxide. The theoretical oxygen is that needed to burn all
carbon in the fuel to carbon dioxide. We calculate the percentage of excess air
from
. O 2 supplied - O 2 theoretical
percentage excess aIr = 0 h - . I 100
2 t eoretJca
 STOICHIOMETRY 51

Example Problem:

Fuel Air (vol) Flue gas (Moisture Free Basis)(mol)

Carbon 79 ~.~ N2 CO 2 X
Hydrogen 21 ~·o O 2 CO Y
Moisture O2 Z
Ash Nz 100 - X - Y - Z
Other

Combustion equation:
CAH B + M0 2 + PN z ----;. XC0 2 + YCO + Z02
+ (laO - x - Y - Z)N2 + RH 2 0
Relationships from the combustion equation:

Carbon A=X+Y
Hydrogen B= 2R
Oxygen M = X +!Y + Z + ~R
Nitrogen P = 100 - X - Y - Z

Calculation: Always choose and write down the basis

Basis: 100 mol of dry flue gas

1. Air feed rate.

N2 in flue gas = 100 - X - Y - Z = P
air feed 100 - X - Y - Z P
lOO~~ol flu~ gas = -. --0.-79 - = 0.79

O2 . au. feed = 0.21 (0.79

III -P)
- = 3P
.7
-6

2. Calculate composition of feed.

(a) Carbon in flue gas = carbon in fuel.
moles C = A = X + Y
C = CO 2 + CO
lb '
pounds of C = (X + Y) 12 ( . ~---)
lb mol
MW
(b) Hydrogen in fuel = hydrogen in water.
moles H 2 = ~B = R
52 PROCESS DESIGN

(c) Oxygen balance gives the moles of water. Moles O 2 in air = moles
O 2 in flue gas + moles O 2 used to form CO 2 , CO, and H 2 0. So

M = 0.21 (
100 - Z- Z) Y -
= X + ~ y + L. + ~R
0.79
or
1R = (0.21/0.79)(100 - X - Y - Z) - (X + ~y + Z)
R = 2[(0.21/0.79)(100 - X - Y - Z) - (X +}y + Z)]

pounds H, in feed ~ R (2 l~~l)

Percentage excess air:

O 2 supplied = (0.21/0.79)(100 - X - Y - Z)
O 2 theoretical = X + y + 1R
Note that we do not use! Y because theoretical oxygen is based upon complete
combustion.
supplied - theoretical
percentage excess = h . I 100
t eoretIca
Some Words on Gas Volumes. At standard conditions (STP) of O°C and
1 atm, 1 lb mol of dry gas occupies 359 ft3. For other than standard conditions
we use the ideal gas law PV = nRT
where P = absolute pressure,
V = volume,
n = number of moles,
R = gas law constant, and
T = absolute temperature.
So
PP r,RT
P sld V,ld nR Y,td
or V =
P (~)
VsId (r:ld) T
sId

III. MATERIAL AND ENERGY BALANCES

A. Introduction

All balances, whether material or energy, start with the fundamental balance
equation,
input - output + production = accumulation
 MATERIAL AND ENERGY BALANCES 53

Since unsteady-state problems on the test are rare, we shall assume steady state,
or
input - output + production = 0
The production term is of importance only for problems involving chemical
reaction. Without reactions, both material and energy balances are determined
from
input - output = 0

B. A Typical Problem

Consider the process, shown in Figure D2, in which streams A and B react to
form stream C. If the temperatures of streams A and B are given, what is the
temperature of stream C?

Heat Loss

A
c
B
FIGURE 02. Typical process energy balance.

We must first do a material balance, using the reaction stoichiometry. This

will determine the amounts of each of the components in each of the streams. To
do an energy balance, we need the enthalpy of each stream. Actually, the en-
thalpy of component x is given by

where Hx = enthalpy (Btu),

X = amount of x (lb),
C p = specific heat of x (BtUflbO F),
T = temperature of x CF), and
To = base temperature (chosen) OF.
Practically, most design work uses a mean C p , which is defined by

CpmcanCT - To) = fT C p dT
To

Plots of r1ean C p versus temperature T can be found in most thermodynamics

texts. So the enthalpy of stream A is
HA = cI lt~ C p mean)( TA - To)
54 PROCESS DESIGN

Likewise
Hs = (I WBCpmean)(Ts To)
He = (I WcCpmean)(Te - To)
where the W; are the weight fractions of the components in the stream.
If we know the heat loss HL and the heat of reaction t1HR at temperature T,
then the energy balance becomes t
HA + HB - He - H L - t1H R = 0
This equation is solved for the temperature T.

C. Heats of Formation and Combustion

The heat of formation t1H r is the heat used (+) or evolved (-) in forming one
mole of a compound from the elements that make it up. The heat of combustion
t1H c is the heat used ( + ) or evolved ( - ) in burning with oxygen one mole of a
compound to form specified combustion products. Standard tables of these
data, usually at 25°C, are available.
For example,

! N 2(g) + ! Gig) ------4 NO(g), t1H r = +21.6 kcal/g mol

!N 2 (g) + 02(g) ------4 N0 2(g), t1Hr = + 8.04 kcal/g mol
The first equation indicates that if and when 1 g mol of nitric oxide (gas) at 1
atm and 25C is formed from! g mol each of nitrogen (gas) and oxygen (gas),
both at 25C and 1 atm, 21.6 kcal must be added to the reaction in order to form
the nitric oxide (or 21.6 kcal must be consumed by the reaction). When the
equations are written as above, the + sign means that heat is used in the reac-
tion (called an endothermic reaction, that is, heat is put "endo" the reaction)
and a - sign means that heat is given off by the reaction (called an exothermic
reaction, that is, the heat "exits" the reaction).
An easy way of working with heats of reaction is to write the two preceding
equations in the following way:

1N 2(g)+ !02(g) + 21.6 kcal/g mol NO ------4 NO(g)

!N 2(g) + 0z(g) + 8.04 kcal/g mol NO z ------4 N0 2 (g)

Since most heats of formation (and heats of combustion) are tabulated at 25C
and 1 atm pressure. we would seem to be in trouble. In fact. it is doubtful if the
preceding reactions would even work at 25C and I atm. We shall resolve this
shortly.

t Note that thc production tcrm. the heat of reaction. has a negative sign because the convention is
that an exothermic reaction (one losing heat to the surroundings) carries a negative sign. So for a
heat-producing reaction. the term- AIiR is a positi\c amount of heat.
 MATERIAL AND ENERGY BALANCES 55

D. Heat of Reaction

The standard heat of reaction, tlH R , is the difference between the enthalpy of
the products and the enthalpy of the reactants when the reaction both begins
and ends at 25'C and 1 atm. If the standard heat of reaction cannot be found, it
can be calculated from standard heats of formation and/or standard heats of
combustion (if you can find them).
To illustrate, what is the standard heat of reaction for the reaction:

Using the previous two equations

tN 2(g) + 02(g) + 8.04 kcal - - N0 2(g)

tN 2 (g) + t02(g) + 21.6 kcal - - NO(g)
and subtracting the second from the first, using the -+ as an equal sign, and using
algebraic principles:
t02(g) - 13.56 kcal -+ N0 2 (g) - NO(g)
or
NO(g) + t02(g) - 13.56 kcal-+ N0 2(g)

So the standard heat of reaction is

tlH R = - 13.56 kcal/g mol NO

or 13.56 kcal are evolved for each gram mole of N0 2 formed.

For those who prefer to memorize formulas,
tlHR = tlHr(products) - tlHr(reactants)
or

2. At Other Temperatures

The standard heats are tabulated at 1 atm and 2YC (sometimes 18 C). We shall
D

need the heat of reaction at other conditions. Now most thermodynamic func-
tions are point functions; they are functions of the final and initial states only.
So, in Figure D3, the enthalpy change in going from point 1 directly to point 4 is
equivalent to that going from 1 to 2 to 3 to 4.
Equating the enthalpy balances for these two schemes gives
56 PROCESS DESIGN

FIGURE D3. Heat of reaction at different temperatures.

or

or

IV. OPTIMIZATION

A. Introduction

All chemical engineering designs are optimized, whether formally or informally.

The main difficulty is determining the objective function-the quantity that is to
be minimized or maximized. Fortunately, for testing purposes this is given. Most
practical problems have several variables that affect the optimum. In the interest
of practical test taking, single-variable optimization is reasonable. Most of these
problems can be solved either graphically or analytically.

B. Graphical :v1ethod

A graphical solution is straightforward if the objective function, such as total

cost, can be calculated as a function of an independent variable, such as insula-
tion thickness, which can be varied to affect the total cost. Consider the cost of
insulating a pipe. The more insulation you use the higher the cost. But the more
insulation you use the less heat is lost from the pipe. For example, if the cost of
lagging as a function of thickness X is given by

$(' = X
and the cost of the heat loss is given by

$L = 1 + 4/X

we can plot both curves and add them together for the total cost. We see from
Figure D4 that the minimum total cost is $5 at an insulation thickness of 2.
 OPTIMIZATION

II CP~~
6 '3-"'-
5
'\. ~~/ V
\ /
4
/
" -
$
3

2 ~ I'-...
1 V $L

1 2 3 4 5
X FIGURE D4. Cost of insulating a pipe.

C. Analytical Method

1. General Conditions

When we plot our optimization problem, we see how the objective function
varies as the independent variable changes. We can see if anything unusual is
occurring. When we solve the problem analytically, we must consider whether
the objective function has any discontinuities or if the independent variable is
bounded.
Solving optimization problems analytically we must be aware that the opti-
mum, if one exists, can occur only at:

• a boundary of the independent variable,

• a discontinuity in the objective function, or
• a stationary point of the objective function.

All possibilities must be evaluated for each problem. Of these possibilities, only
the methods for finding stationary points will be discussed.

2. Unconstrained Optimization

For continuous functions such as those shown in Figure D5, the slopes, or deri-
vatives, are also continuous. For curve a, to the left of the maximum the deriva-
tive is positive; to the right of the maximum the derivative is negative. At the

OL-----------------_
x FIGURE D5. Optima.
58 PROCESS DESIGN

b
+

yI 0 I----Ij~----"~-... X

FIGURE D6. Derivatives of optima.

maximum point, the derivative is zero. The opposite is true for curve b with the
mlmmum.
So to find the stationary points analytically, we set the first derivative of the
objective function to zero. The resulting independent-variable values are max-
ima, minima, or inflection points. If we plot the first derivatives of curves a and b
(Figure D6), we find that the curve with the maximum has a negative second
derivative at the stationary point and the minimum curve has a positive second
derivative at the stationary point. So analytically, we evaluate the second deriva-
tives at the stationary points.
The total cost of insulation was

4
C=l+-+X
X

The first derivative with respect to X is

, 4
C = - - -- +1
X2
Setting C = 0 and solving for X

or
X = ±2
Since we cannot have negative thicknesses of insulation.
X = 2

is a stationary point. Taking the second derivative of the total cost gives
C" = 8/X 3
 OPTIMIZATION 59

The second derivative, evaluated at the stationary point, is

C'(2) = 1
a positive number. Thus the stationary point, X = 2, gives the minimum total
cost.

3. Constrained Optimization

It often happens in engineering problems that an objective function is to be

minimized,
minimize C(x, y)
but at the same time, the independent variables are constrained by an equation
such as
f(x, y) = 0
There are several methods of solving this type of problem.

a. Substitution

If easily done, it is always best to solve the constraint equation for one variable
in terms of the others. For our case, we might be able to get
y = g(x)
easily from
f(x, y) = 0
We then substitute this equation for y into the function to be minimizcd,
minimize C{x, g(x)}
Now the cost function depends only upon x and we find the minimum as before.

b. Lagrangian Multiplier

For situations in which it is difficult to solve the constraint equation for one
variable in terms of the others, we use the Lagrangian multiplier method. With
this method we add to the original objective function, the constraint cq uation
multiplied by a constant j,
C = C(x, y) + ).f(x, y)
To find the optimum, we calculatc the first derivative of C with respect to each
independent variable and set it to zero, or
oC
--- =
oC
~ + I.,(If- = 0
tx rx -ax
oC ac ,o{
= - +) . = 0
cy oy , iJl'
60 PROCESS DESIGN

This gives two equations for the three unknowns x, y, and X The third equation
is
f(x, y) = 0
There is no easy method of telling whether the solution is a maximum, mini-
mum, or inflection point. The easiest method is to evaluate the objective func-
tion at locations near the solution to confirm that the solution is indeed a
minimum (if you were hoping to minimize an objective function).

V. PROCESS CONTROL

A. Introduction

Processes are designed to operate at steady-state conditions. They never do. All
processes change with time for many reasons: the feed rate, composition, or
temperature is usually variable; the steam temperature or rate varies; or the
atmospheric conditions change.
To introduce process control, we consider the well-stirred vessel in Figure D7
of volume V into which is flowing a fluid of concentration x(t) at a rate of q.
Leaving the vessel at rate q is the fluid of concentration yet). An unsteady-state
material balance gives
dy
V dt = qx - qy

If the input concentration suddenly changes from x = 0 to x = a (a step func-

tion), this first-order differential equation can be solved to give
yet) = a[l - e- qt / V ]

Note that the two constants that define the system, V and q, appear only as a
ratio. This ratio is called the time constant r since it has units of time
r = V/q
With the time constant notation, the concentration equation is
yet) = a[l - e- r,,]

B. The Laplace Transform

Most chemical engineering control problems involve keeping a process operat-

ing near a steady-state point. If the control scheme is a good one, large de via-

x(t) yet)
q q
V,y(t) FIGURE D7. A well-stirred vessel.
 PROCESS CONTROL 61

tions from the operating point are unlikely. Small deviations from the setpoint
allow one to assume that most control problems are linear systems. The time-
honored method for solving linear control problems is the Laplace transforma-
tion.
The Laplace transform of the function f(t) is defined by

L[f(t)] = io" .{(t)e- st dt

So we convert a function of time t to a function of s, or

L[f(t)] = F(s)
Taking the Laplace transform of
f(t) = e- at
we obtain, through a formal integration,
1
L[f(t)] = L[e- at ] = . -
s+a
Note that when the denominator is zero,
s = -a
the coefficient of t in the exponential term. Using integration by parts, we find
that

L[ r( ir')] = res Yes) - yeO)]

where Yes) is the Laplace transform of yet). Many texts contain the pairs of
Laplace transforms and operations.

C. Transfer Functions

Returning to our original stirred tank, by defining

L[y(t)] = Yes), L[x(t)] = Xes)
our differential equation in time becomes an algebraic equation in s,
rsY (s) + yes) = Xes)
or
Y (s) = ( __1 ) Xes)
rs +1
or, in general,
yes) = G(s)X(s)
The bracketed term that defines G(s) is called the transfer function of the system.
62 PROCESS DESIGN

Xes) yes)
G(s)
input output
FIGlJRE D8. General block diagram.

For any given input x(t) to our stirred tank, we could look up its transform
Xes). The transform of the output yes) is then known; the time response can be
found from Laplace transform tables. For a step input in x(t)
x(t < 0) = 0, x(t > 0) =a
the transform is
Xes) = als
Thus,

1 ) a air
Yes) = ( rs + 1 ~ = ~+ l/r)
a a
s s + 1/r
The inverse transform of Yes),
L ~l[y(S)] = y(t)

is
yCt) = a[l - e~t!rJ

D. Block Diagrams

One of the advantages of the transfer function is that it leads to block diagrams.
The stirred tank, in block diagram form, is
YCs) = G(s)X(s)

as shown in Figure 08, or

Yes) = (_1+
rs
-)X(S)
1
as shown in Figure 09.

Xes) ~i ,":,11----- Yes)

FIGl'RE 09. Stirred tank block diagram.

 PROCESS CONTROL 63

z(t) wet) yet)

q q q
V,z V,w V,y

FIGURE 010. Three stirred tanks in series.

We can extend this concept. Consider the three stirred tanks arranged in
series in Figure D 1O.
This cascade is represented in block-diagram notation by Figure D11 or by
Z(s) = Gj(s)X(s)
W(s) = G2 (s)Z(s)
Yes) = G3 (s)W(s)
which, when combined algebraically, gives
Yes) = G3 W = G 3 G 2 Z = G 3 G 2 G J X(s)

E. Closing the Loop

Consider the closed-loop (feedback) control system shown in Figure D12 where
the Laplace-transformed variables are

R(s) = set pomt,

E(s) = error between set point and Y,
Gc = controller transfer function,
pes) = output from controller (usually pressure),
Gv = valve transfer function,
Xes) = output of valve, same units as Y,
U(s) = outside disturbance to process,
G = process transfer function,
Yes) = process output variable,
H = transducer transfer function, and
B(s) = output of transducer.

Xes) Z(s) W(s) Yes)

Gl (s) G2(s) G3(S)

FIGURE 011. Block diagram for Figure D10.

64 PROCESS DESIGN

U(s)

R(s) +

FIGURE D12. Closed-loop feedback control.

Using step-by-step block-diagram algebra,

Yes) = GZ(s)
Z(s) = U(s) + Xes)
Yes) = GU(s) + GX(s)
Yes) = GU(s) + GGvP(s)
Yes) = GU(s) + GGvGcE(s)
Yes) = GU(s) + GGvGc[R(s) - B(s)]
Yes) = GU(s) + GGvGcR(s) - GGvGcB(s)
Yes) = GU(s) + GG,GcR(s) - GGvGcHY(s)
or

or

These could have been written by inspection. Note that the numerator is the
product of all transfer functions between the input and the output of the open-
loop system. The denominator of all closed-loop systems is one plus the product
of all transfer functions around the closed loop.

F. Controllers

The usual industrial controller is pneumatic. They are sold with up to three
different control modes.

1. Proportional Controllers

The idealized transfer function for the proportional controller is

pes)
G =-=K
c E(s) C

where Kc is a constant, usually called the gain.

-------- ---- - - -

PROCESS CONTROL 65

2. Proportional-Derivative Controllers

The idealized transfer function for the proportional-derivative controller is

where'D is the derivative time.

3. Proportional-Integral Controllers

The idealized transfer function for the proportional-integral controller is

G= ['(s)
C E(s)
= K (1 + _1 )
C 'IS

where " is the integral time.

G. Stability

I. Introduction

If the output of a controlled system is bounded for all bounded inputs, the sys-
tem is stable. Since most linear systems are exponential in time, that is,

if the exponent a is positive, the system is stable since yet) damps out as time gets
large. If the exponent a were negative the system would be unstable since yet)
would approach infinity as time increased. Only the simpler methods for deter-
mining stability will be discussed.

2. Partial Fraction Expansion

We have already seen that the Laplace transform function

A
Yes) = - -
s+a
gives the time function
yet) = Ae- at

If we set the denominator of Yes) to zero,

s+a=O
or
s = -a

1
 PROCESS DESIGN

we obtain the exponential term in yet). Were we to have a transfer function such
as
A
y(s)=-----
(s + sJ(s + S2)
we would write it as

The time solution would be

and it would be stable if both Sl and S2 were positive. To find K1 and K 2, we

write Yes) with the common denominator

Equate this numerator with the original numerator

or

Equating coefficients of like powers of s on both sides of the equation gives

K1 + K2 = 0
K 1 s2 + K 2 s1 = A
or

Thus the time solution is

3. Characteristic Roots

Suppose a closed-loop system has the transfer function form

G1
Yes) = 1 G R(s)
+ G1 2
 PROBLEMS

We would manipulate the equation so that we have both a numerator anl1

denominator polynomial in the Laplace variable, s:

Yes) = N(s)
D(s)
The denominator polynomial D(s) is called the characteristic equation. Its roots
[the values of s that make D(s) = 0] are called the characteristic roots. If we
have any real, positive roots, we know that the system is unstable. Knowing the
roots enables us to use the partial fraction expansion to obtain the time solution.

4. Routh's Method

When D(s) is higher than a second-order polynomial in s, numerical root-finding

procedures are necessary. But Routh's criterion allows us to determine stability
without determining the roots.
We write the characteristic equation in the form
ao sn + a 1 sn - 1 + a 2 sn - 2 + ... + an _ 1S + an = 0
where a o is positive. If any of the a i are negative, the system is unstable. If all the
ai are positive, develop the Routh array:
Row
ao a2 a4 a6
2 a1 a3 as a7
.' hl h2 b3

11 II
11+1 !l1
The elements in the third and succeeding rows are found from

The clements for the other rows are found from formulas which correspond to
those given above. All elements of the first column of the Routh array must be
positive and nonzero for a stable system.

PROBLEMS

A. A 70°; NaOH liquor at 160u F with a heat content of 325 Btu/lb is diluted
with a 20 % NaOH liquor at 80°F with a heat content of 40 Btu/lb to give a
final solution containing 40% NaOH. For every 100Ib of product, how
68 PROCESS DESIGN

much heat must be removed to maintain a final product temperature of

lOO°F? (Heat content of 40 % solution is 94 Btu/lb at lOO°F.)
B. A kiln is operated countercurrently to produce lime by burninQ: a calcium
carbonate sludge. The compositions of the flue gas leaving the cold end of
the kiln and the sludge entering the same end are

Sludge wt% Flue Gas Vol%

CaC0 3 45 CO 2 20
H 2O 50 CO 1
Inerts 5 N2 76
O2 3

Methane, at a rate of 30,000 ft3/hr (measured dry at 60°F, 1 atm), is used to

fire the kiln. If the lime conversion is 90 % complete, how many pounds per
hour of CaO is produced?
C. Benzene reacts with chlorine in a batch reactor to produce both mono- and
dichlorobenzene according to
C6H6 + Cl 2 ~ C 6 H sCI + HCI
C 6 H sCI + Cl 2 ~ C 6 H 4 Cl z + HCI
A l000-lb charge of benzene produced a product of 50 wt % monochloro-
benzene, 30 wt % dichlorobenzene, and 20 % unreacted benzene. The gas
produced consisted of 90 vol % HCI and 10 vol % Cl z at 1 atm and 200°F.
We wish to determine

(a) The volume of gas produced.

(b) The final weight of liquid product.
(c) The percentage of excess chlorine added, assuming the theoretical
amount of chlorine needed would produce pure monochloroben-
zene.
D. Air and gaseous HCI at 25°C are mixed and fed to a reactor in the Deacon
process. The gaseous product leaves the reactor at 500°C with the composi-
tion:
Product Vol%
HCI 12
N2 53
O2 6
Cl 2 14.5
H2O 14.5
 PROBLEMS 69

If the reaction stoichiometry is

4HCI + O 2 ~ 2Cl 2 + 2H 2 0
how much heat must be supplied to or removed from the reactor for each
100 g mol of product gas? Assume no heat losses.
E. Given the control system of Figure D13 with 'II = 1, '2 = 1, '!3 = t deter-
mine the range of Kc for which the system is stable.

+ y

FIGURE 013. Schematic for problem E.

F. Given the block diagram of the control system shown in Figure D14, if Y is
fed back positively, what K range is needed for stability? If Y is fed back
negatively, what K range is needed for stability?

R + E K Y
"I" (s+2)(s+3) (s+4)

FIGURE 014. Schematic for Problem F.

G. What reaction temperature T maximizes the reversible reaction rate for

if the concentrations are fixed and

with
70 PROCESS DESIGN

PROBLEM-SOLVING STRATEGIES

A. Using a basis of 100lb of product, do a total material balance and an

NaOH balance to determine the stream amounts. Then a h~at balance
should provide the answer.
B. This problem looks straightforward but involved. I need to deal with two
reactions: the burning of methane with air and the decomposition of
calcium carbonate. I must allow for incomplete combustion of methane, so
we shall have two methane reactions. The flue gas CO 2 will come from the
decomposition of lime and from the complete combustion of methane reac-
tions. The N 2 comes directly from the air; the O 2 comes from excess air. If I
had the time, I would work out the problem. I suspect that I would have to
think carefully as I worked it out.
C. With this material balance problem, start with the stoichiometric equa-
tions, relating products to reactants. The problem would require logical
thinking along the way to a solution. But I would not hesitate to attempt it.
D. With a basis of 100 gmol of product gas, determine the enthalpy of the
product at 500°C. If I'assume a base temperature of 25°C, I do not need to
worry about the reactant enthalpy. Find the heats of combustion or forma-
tion of the components so that I can calculate the heat of reaction. Once I
find the heat ofreaction for each 2 mol of C1 2 , multiply by 14.5/2 to obtain
the total heat. Finally, a heat balance provides the heat that must be added
to the reactor. To solve this problem I need Cpdata. If I can find it, mean C p
data are the most convenient to use.
E. Perform the block diagram algebra, step by step, to get Y as a function of R.
Find the denominator polynomial in s. Use th~ Routh criterion.
F. Perform the block diagram with a in the 1 block. Find Y as a function of R
and a. Get the denominator polynomial in s (and a). Use the Routh criter-
ion for stability. When a = + 1, the answer is for the first question; when
a = -1, the second question is answered.
G. Set up the equation for the rate of formation of B in terms of k l ' k2' Ca , and
Cb . Find the stationary point by differentiating rb with respect to T. Sub-
stitute this value of T into the rb equation for the solution. Make sure that
the second derivative of rb with respect to T is negative; this assures a
maximum.

SOLUTIONS

A. See Figure D15.

Basis: 100 lb of 40 % NaOH solution at 100 c F.
Total material balance: x +Y= 100
NaOH balance: 0.70X + 0.20Y = 40
Enthalpy balance: 325X + 40Y = Q + 94(100)
 SOLUTIONS 71
X Ib
7O%NaDH lOO°F
100 Ib
Y Ib 4O',I,NaOH
20%NaDH

Q FIGURE DIS. Schematic for solution A.

So X = 100 - Y.
Then 0.70(100 - Y) + 0.20Y = 40 or 70 - 0.50Y = 40
or 0.50Y = 30 or Y = 60
Then X = 40

Using enthalp y balance,

325(40) + 40(60) = Q + 9400

or

Q = 6000 Btu/100 lb of produc t solutio n

B. See Figure D16.

Basis: lOO-lb sludge

CaC0 3 CaO + CO 2
MW (100.0) (56) (44)

Sludge Wt% Pounds Moles

CaC0 3 45 45 0.45
H 2O 50 50
Tnerts 5 5
100 100
moles CaO made = 0.90(0.45) = 0.405

30,000 ft3fhr
dry ,60°F
Cf4 --"':""---.4
air
90% Conv.
Cao , CaC0:3
FIGURE D16. Schematic for solution
inerts B.
72 PROCESS DESIGN

The combustion reactions:

CH 4 + 20 2 = CO 2 + 2H 2 0
CH 4 + !02 = CO + 2H 2 0
So the flue gas CO 2 comes from both the CaC0 3 reaction and the CH 4
reaction.
Assume Z moles of flue gas per 100-lb sludge:

Flue Gas Mole % Moles

CO 2 20 0.20Z
CO 1 0.01 Z
N2 76 0.76Z
O2 3 0.03Z
100 Z

Moles O 2 in air used for burning = (0.76Z/0.79)0.21 = 0.202Z.

Moles O 2 in combustion = 0.202 Z - 0.03 Z = 0.172 Z.
Moles CO 2 from combustion = 0.20 Z - 0.405.
Moles O 2 burning CH 4 to CO 2 = 2(0.20Z - 0.405) = O.4Z - 0.81.
Moles O 2 burning CH 4 to CO = 0.172Z - (O.4Z - 0.81)
= 0.81 - 0.228 Z.

Moles CO formed = «0.81 - 0.228 Z) = 0.54 - 0.152Z = 0.01 Z

0.162Z = 0.54 Z = 3.33 mol
Moles C from CH 4 = (0.20 Z + 0.01 Z) - 0.405
= 0.295mol
To get the pounds per hour of CaO produced,
ft3 492°R mol mol
CH 4 = 30,000 hr 520 R 359f13
0 = 79.07 hr
Then
0.405 mol CaO 56 Ib CaO 100 lb sludge
CaO production = 100 Ib sludge mol CaO 0.295 mol CH 4 x

79.07 mol CH 4 _ 7 lb
hr - 60 9 hr

C. Basis: 100-lb liquid product.

+ Cl 2 ----+' C 6 H 5 CI + HCI
(10.9) (112.6) (36.5)
C 6 H 5 CI + Cl 2 HCI
MW (112.6) (10.9) (36.5)
 SOLUTIONS

Liquid Wt% Pounds Moles H

C 6 H s CI 50 50 0.444 x 5 = 2.220 0.222

C 6 H 4 Ci 2 30 30 0.204 x 4 = 0.816 0.204
C6 H6 20 20 0.256 x 6 = 1.536
0.904 4.572 0.426

moles C6H6/100-lb liquid product = 0.904 mol

Ib atoms H in feed = 6(0.904) = 5.424
lb atoms H in gas = 5.424 - 4.572 = 0.852

Gas Mole % Moles Moles el 2

HCI 90 0.852 0.426

Cl 2 10 0.095 0.095
100 0.947 0.521

0.947 mol gas 359 ft 3

(a) Volume of gas produced = - --- - - -- - -
100 Ib liq prod mol gas STP

100 lb liq pro~ 1000 Ib~H6 feed _ = 4815 ft 3 at STP

0.904 mol feed 78.11b C6H6/mol C6H6
100 Ib liq prod 1000 lb feed
(b) Weight ofliq prod = -c- X - - - .- = 1416lb
0.904 mol.eed 78.1 lb/mol
(c) Theoretical Cl z = 0.904; total Cl z feed = 0.426 + 0.521 =
0.947
0.947 - 0.904
percentage excess = - 0.904- 100 = 4.8 %

D. See Figure D17.

Basis: 100-g mol product gas
Base temperature 25°C.

Product Gas
hr 500°C
-~
25 °C ~_I-----'

Q FIGURE 017. Schematic for solution D.

74 PROCESS DESIGN

Vol %or Mean

Product Mol % Moles Cpat 25-500°C Moles x C p(500-25)

HCI 12 12 t7.l cal/g moloC 40,470

N2 53 53 7.2 181,260
O2 6 6 7.5 21,375
Ci l 14.5 14.5 8.6 59,233
Hl 14.5 14.5 8.5 58,544
360,882 cal

Heats of formation at 25°C.

HCI(g) = - 22,063 cal/g moP

01= 0
Ci l = 0
H10(g) = -57,798 cal/g mol

AHR25°C = AHc(prod) - AHc(react)

= 2( - 57,798) - 4( - 22,060) = - 27,344 cal/2 g mol Ci l
-27,344 cal
- - ' - - - x 14.5 g mol Ci l = -198,244 cal
2 g mol Ci l
Q - ( - 198,244) = 360,882 or
Q = 162,638 cal supplied/lOO g mol product
1
G --------:---
E. 1 - (s + 1)(is + 1)

1
E= R -HY H=--
ts + 1
Y = KcGJ(R - HY)
Y(l + KcGJH) = KCGJ R
Characteristic equation is numerator of

t Smith, and Van Ness, Introduction to Chemical Engineering Thermodynamics, McGraw-Hill, New
York,19 ,2nd ed., p. 126.
I Same reference, p. 138.
 SOLUTIONS

or

I + Ke -----
(s + l)(~s + 1)(t5 + 1)
The numerator is

(s + 1)(~s + l)(-!s + 1) + Kc
(!s2 + is + 1)(ts + 1) + Kc
~S3 + ~S2 + ~S2 +is + ts + 1 + Kc
~53 + 52 + II s + 1 + Kc

or

Routh array

6
6(11) - 6(1 + Kc)
6
6(1 + Kc)
All offirst column must be positive: 1 is (+),6 is (+),11 - 1 - Kc must be
( +), 1 + Kc must be ( + ).
and

for stability

F. Y= GE positive feedback a = 1
E = R + aY negative feedback a = - 1
Y= G(R + aY) or Y(I- aG) = GR

Characteristic equation is numerator of 1 - aG

aK
1 - aG = 1 - - - - - --------
(s + 2)(s + 3)(s + 4)
numerator = (s + 2)(s + 3)(s + 4) - aK
= (S2 + 5s + 6)(s + 4) - aK
= S3 + 4s 2 + 5s 2 + 20s + 6s + 24 - aK
= 53 + 9s 2 + 26s + 24 - aK
 76 PROCESS DESIGN

Routh array

26
9 24 - aK
(9 . 26 - 24 + aK)/9
24 - aK

First column 1,9, (210 + aK)/9, 24 - aK must all be positive:

For positive feedback a = 1 and 210 + K > 0
and

24 - K > 0 or for stability - 210 < K < 24

For negative feedback a = -1 and 210 - K > 0

and

24 +K> 0 or for stability - 24 < K < 210

G.

max (r b) = max (k 1 C a - k2 C b )

drb dk[ dk2

dT = Ca dT - C b dT

........
 SOLUTIONS

Stationary point

k- C -E2- -Ez/RT
2 b RT2 e

kCE
1 a 1 = e-(Ez-Ed/RT
k 2 Cb E2

~2rb = [lEI e-E1IRT(E _ 2RT)C _ l0. E2 e-E2/RT(E 2 _ 2RT)C

dT2 R2T4 1 a R2T4 b

= k 2C b E 2
R2T4
e-Ez/RT(E
1
_ 2RT - E
2
+ 2RT)

= k-
2 C b E 2(E 1 -
-R2T4 E 2)
e
-Ez/RT

This is negative so the stationary point gives a maximum.

 MASS TRANSFER

1. Introduction 81
A. Topics Covered 81
B. Symbols 81
C. Stage Principles 81
1. Terminology 81
2. Material Balances 81
3. Enthalpy Balances 82
4. Graphical Methods 82
a. The Operating Line 82
b. Ideal Stage (or Perfect Plate) 83
5. Absorption Factor Method 83
6. Enthalpy-Concentration Diagram Method 84
II. Distillation 85
A. Flash Distillation 85
B. Differential Distillation 85
C. Steam Distillation 86
D. Continuous Distillation 87
1. Principles 87
2. McCabe-Thiele Graphical Method 88
3. Ponchon-Savarit Enthalpy-Concentration Method 90
III. Leaching and Extraction 90
A. Leaching 90
1. Terminology 90
2. Equilibrium 91
3. Operating Line 91
4. Constant or Variable Underflow 91
5. Number of Ideal Stages Constant Underflow 91
6. Number of Ideal Stages- Variable Underflow 92
a. Overall Balances and Point J 92
b. Operating Line and Point P 93
B. Liquid Extraction 94
1. 3asic Data 94

W. L. McCabe and J. C. Smith, Unit Operations of Chemical Engineering, 3rd ed. McGraw-Hill, New
York. 1976.
79
 80 MASS TRANSFER

2. Principles of Extraction 95
a. Ponchon-Savarit Method 95
b. Rectangular Diagram Method for Extraction 96
c. Countercurrent Construction of Stages 97
d. Minimum Solvent Ratio 97
IV. Gas Absorption 98
A. Packed Towers 9g
./ 1. Principles 98
a. Material Balances 9g
b. Limiting Gas- Liquid Ratio 99
c. Differential Material Balance 100
2. Absorption Rate 100
a. Theory of Double Resistance 100
b. General Case-~x ~y Triangle 101
c. Overall Coefficients 102
d. Height of Transfer Unit Method 103
e. Controlling Resistance 104
f. Lean Gases 104
3. Coefficients and HTU's 104
B. Plate Columns 105
r V. Humidification Processes 105
A. Definitions 105
B. The Humidity Chart 106
C. Theory of Wet-Bulb Temperature 106
D. Humidification Processes 107
1. Introduction 107
2. Gas-Liquid Interaction 109
3. Packed Gas-Liquid Contactors 108
4. Adiabatic Humidification 111
/ VI. Drying of Solids 112
A. Introduction 112
B. Equilibria 112
C. Rate of Drying 112
D. Rate-of-Drying Curves 112
1. Constant-Rate Period 113
2. Critical Moisture Content and Falling-Rate Period 113
E. Calculation of Drying Time 113
1. Equations for Constant-Drying Conditions 114
Problems 115
Problem-Solving Strategies 118
Solutions 120
---
INTRODUCTION 81

I. INTRODUCTION

A. Topics Covered

We shall discuss the general stage principles and then some of the principal
methods of mass transfer: distillation, leaching and extraction, absorption, hu-
midification, and drying. We shall discuss only the material that is likely to
appear on the P.E. exam. Therefore, little on multicomponent separations will
be found here. The theoretical diffusion and mass transfer between phases will
not be discussed; the topic is amply covered by McCabe and Smith.

B. Symbols

\1cCabe and Smith's convention for identifying the vapor V phase and the liq-
uid L phase will be used:

Process V Phase L Phase

Distillation Vapor Liquid

Gas absorption Gas Liquid
Liquid extraction Extract Raffinate
Leaching Liquid Solid
Humidification Gas Liquid
Drying Gas Wet solid

The subscript a will refer to the end of the unit where the L phase enters; b to the
end where it leaves.

C. Stage Principles

1. Terminology

Stages are numbered in the direction of flow of the L phase, with the nth stage
being a general stage. The total number of stages is N.

2. Material Balances

Doing a total material balance into and out of envelope 1 of Figure M 1 gives
La + v" + 1 = Ln + Va
for steady-sta,e operation. A balance on component A of a two-component sys-
tem gives
r
81 MASS TRANSFER

H t - - - - - + - t n-1

1--+---_ _-1 n

Yn +1
1-_ _ _ _..1.--1 n+ 1

Vb
Yb FIGURE Ml. Terminology for a staged process.

Had we chosen to write the balances over the whole unit, we would obtain
La + Vb = Lb + v"
and

3. Enthalpy Balances

If the enthalpies are denoted by H, the envelope 1 enthalpy balance is

LaHL.a + Vn+1 H y,n+l = LnHL,n + v"HY,a
while the enthalpy balance around the total unit is
LaHL,a + VbHv,b = LbHL,b + v"Hy,a
4. Graphical Method

Many two-component problems can be solved graphically, If either the operat-

ing line or the equilibrium is straight, graphical methods are ideal. If both are i
straight, the various equations are quicker and easier.

a. The Operating Line

The operating line is simply the component material balance expressed as y =

function of x:
r
INTRODUCTION 83

I
X2 FIGURE M2. Graph of an ideal equilibrium stage.

",ote that the two end points, (X a' Ya) and (Xb' .Vb), satisfy this material balance
':'1 uation. In fact, all stages in between must also be represented by this material
jalance equation. Of course, it is much easier to use when it is a straight line;
:hat is, when L and V are constant or when they can be redefined so as to be
-::onstant.

b. Ideal Stage (or Perfect Plate)

We live with actual stages, but it is easier to calculate the theoretical or ideal
stages. With the number of ideal stages in hand, we can apply a stage efficiency
factor to obtain the actual stages.
The ideal stage is defined such that the V phase leaving a stage is in equili-
brium with the L phase leaving the same stage. Thus we need the equilibrium
-::urve relating y to x. As shown in Figure M2, the two phases leaving the same
stage are related by the equilibrium curve (point 1), and the phases from one
stage to the next are related by the operating line (material balance), or point 2.
Continuing up or down the unit in this manner allows one to determine the
number of stages needed to effect the separation.

5. Absorption Factor Method

If both the equilibrium line and the operating line are straight, the graphical
stage-to-stage construction can be expressed as a formula. We assume the
straight equilibrium line to be
Ye = mXe +B
or, for the ideal, nth stage,
Yn = mXn +B
Substituting this into the operating line (straight when we assume LjV constant),
L(y" - B) Lx.
Yn + 1 = - m V· + Ya - V
Define the absorption factor A = L/(m V), so
Yn+l = AYn - A(mxa + B) + Ya
84 MASS TRANSFER

Now note that the (mxa + B) quantity is the vapor concentration that is in equi-
librium with the inlet L phase concentration Xa' Denote it by Y: = mXa + B, so
that

Now do the stage-by-stage construction with n = 1, or

Y2 = AYi - Ay: + Ya = AYa - Ay: + Ya

= Ya(1 + A) - Ay:
since Ya = Yi' Then for n = 2

Y3 = AY2 - Ay: + Ya = A[Ya(1 + A) - Ay:J - Ay: + Ya

= Ya(l + A + A2) - y:(A + A2)

This eventually becomes

This is the Kremser equation. It can be solved for N:

N = In{Yb
Ya
= Y:}/ln A
Ya
Our equation can also be expressed in terms of L-phase concentrations x, and
the stripping factor S = 1jA = mLjV, or

Xa -
N=ln { - - * X:}/ InS
Xb - Xb

6. Enthalpy-Concentration Diagram Method

We have an A stream at a rate rna mixed with a B stream of rate rnb to give a C
stream. With concentrations Xa, Xb, Xc and enthalpies H a' H b' He, we can show
that the enthalpy balance is

and the component material balance is

Thus
rna Hb- Hc Xb - Xc
rnb He - H a Xc - Xa
the lever principle. McCabe and Smith (pp. 528-530) show how this may be
modified for nonadiabatic processes.
 DISTILLA TION 85

II. DISTILLATION

A. Flash Distillation

Consider a liquid mixture in which the pressure is reduced (usually through a

valve). A vapor-liquid equilibrium results. and the two phases are separated. If
we have I mol of a binary mixture with X f mole fraction of the more volatile
component and we let / be the fraction of the feed vaporized, then 1 - / is the
fraction of feed that lea ves as liq uid. If Yd and Xb arc the vapor and liquid concen-
trations, then a more volatile component balance becomes

XI = /Yd + (1 - f)Xb

We know neither Yd nor X b , but we have assumed that equilibrium exists. So,
once again, we need the equilibrium line to go with our material balance line.
With the equilibrium assumption, the point (Yd' x b ) falls on the equilibrium
curve, Y = function of x, so we shall write the material balance equation without
the subscripts
XI = y{ + (1 - f)x

or

which is a straight line with slope - (1 - f)1f = - L/V passing through the
Y= x diagonal at x = XI'

B. Differential Distillation

Now consider a liquid charge of no moles initially in a still. At any time t, there
will be n moles of liquid remaining in the still with a liquid composition x. At this
same time t, the vapor composition is y, and the amount of component A will be
l1a = xn.
At this instant, we assume that a small amount of liquid is vaporized, dn
moles. The amount of A in this vapor will be -dna or - Y dn, since na is decreas-
ing. Thus

dna = d(xn) = X dn + n dx
so
Y dn = x dn + n dx
or
dn dx
n y - x
86 MASS TRANSFER

This equation will be integrated from the initial charge no to the final charge n l ,
while the x goes from x o , the initial concentration to Xl' the final concentration,
giving the Rayleigh equation:

This equation may be integrated graphically using the equilibrium curve, Y =

function of x, or if the relative volatility
Ya Xb Y 1- X
aab = - -- = - - -
Xa Yb 1- Y X
is constant, the equation may be integrated to

where subscripts a and b refer to these components and the subscript 0 refers to
the initial charge.

C. Steam Distillation

Many organics are temperature sensitive; conventional distillation procedures

would be inappropriate. For these cases, steam distillation may deserve consid-
eration. Here, volatile organic liquids are separated from nonvolatile impurities
by vaporization of the volatiles by passing live steam through them. The general
equation for this scheme is
rns Ms p,
rna llA.a Pa
where lil s ' rna = mass flow rates of vapor, steam and volatile,
Ms ' M a = molecular weights, steam and volatile. and
Ps. Po = partial pressures, steam and volatile.
This equation reduces, for specific cases, to:
If the steam does not condense,

where 1[ = total pressure and

VP a = vapor pressure of volatile.
If the steam condenses and is immiscible (forming two liquid layers),
rns i\1 s VP s Ms 1[ -VP"
rna /vi a V PallA. a VPa
 DISTILLATION

Tfthe steam condenses and is miscible (forming but one liquid layer),

rn, = _Ms X,' YPs

rna Maxa·YPa

where x,, Xa = mole fraction in liquid, steam and volatile

D. Continuous Distillation

1. Principles

In the stage principles section we obtained the material balance or operating

line. For distillation, we have the same operating line form, but with the distilla-
tion notation shown in Figure M3. Now we have two, one above the feed and
one below the feed:

Ln DXd
Yn+ 1 = -v.- Xn + -- (above feed)
n+1 v" + 1

Ym+ 1 = v--
Lm
m+l
Xm -
BXb
V. -
m+l
(below feed)

Using a total balance around the top of the column, we find

D = Vn+ 1 - Ln

and around the bottom

F
Xf

XlJ FIGURE M3. A distillation column.

86 MASS TRANSFER

This equation will be integrated from the initial charge no n

to the final charge ~,
while the x goes from x o , the initial concentration to Xl' the final concentration.
giving the Rayleigh equation:

f dn
"O
"1 n
= In (no) =
nl
fXO ~
X, Y ~ X

This equation may be integrated graphically using the equilibrium curve. y

function of x, or if the relative volatility
Y 1~ X

1~ Y X

is constant, the equation may be integrated to

where subscripts a and b refer to these components and the subscript 0 refers to
the initial charge.

C. Steam Distillation

Many organics are temperature sensitive; conventional distillation procedures

would be inappropriate. For these cases, steam distillation may deserve consid-
eration. Here, volatile organic liquids are separated from nonvolatile impurities
by vaporization of the volatiles by passing live steam through them. The general
equation for this scheme is
ms Ms Ps
ma lvl a Pa
where ms ' ma = mass flow rates of vapor, steam and volatile,
M s' Ma = molecular weights, steam and volatile, and
Ps' Pa = partial pressures, steam and volatile.
This equation reduces, for specific cases, to:
If the stcam does not condense,

where rr = total pressure and

VP a = vapor pressure of volatile.
If the steam condenses and is immiscible (forming two liquid layers),
ms Ms VP, AI, rr ~ VP a
ma Ma VP" Aia VI\
 DISTILLA TION

If the steam condenses and is miscible (forming but one liquid layer).

lils = ~~xs' VP s
rna Maxa·VPa

where x s , Xa = mole fraction in liquid, steam and volatile

D. Continuous Distillation

1. Principles

In the stage principles section we obtained the material balance or operating

line. For distillation, we have the same operating line form, but with the distilla-
tion notation shown in Figure M3. Now we have two, one above the feed and
one below the feed:

Ln DXd
Yn+l = V- Xn + -V (above feed)
n+1 n+1

Lm BXb
Ym + 1 = - Xm - -- (below feed)
Vm+ 1 Vm+ 1

Using a total balance around the top of the column, we find

and around the bottom

F
--101
Xf

B
Xb FIGURE M3. A distillation column.
 88 MASS TRANSFER

Eliminating Vn + 1 and Vm + 1 we obtain for the material balances or operating

lines

and

2. McCabe- Thiele Graphical Method

If we can assume constant molal overflow, that is L. and Lm are constant, the
operating lines are straight. Assuming that we have available the equilibrium
data for our binary, we may use th~McCabe- Thiele graphical method.
Most analyses use the concept of reflux ratio or RD'
L V-·D
RD = -- =
D D
With this definition, the above feed operating line is
RD Xd
Yn+ 1 = RD +i Xn + RD + 1
Note that the operating line passes through the Y = x diagonal at x = X d . This is
true regardless of whether we are using a total or a partial condenser. But keep
in mind that if a partial condenser is being used, we subtract one theoretical
stage when counting stages. Thus we see that the operating line for the rectifying
section (above the feed) goes through the point y = x = Xd with a slopc of
RD/(R D + 1).
The operating line for the stripping section (below the feed) was
[ BXb
y=;:;:- x-=--
L-B L-B
with I thc liquid flow down the column beneath the feed point. You will notice
that this equation goes through the point Y = x = X b , regardless of the type of
reboiler. Bear in mind though that a partial reboiler is a theoretical stage. Thus.
in counting the theoretical trays or plates required. you must subtract one for a
partial reboiler.
The feed plate separates the stripping section from the rectifying section. The
addition of the feed increases the reflux in the stripping section, increases the
vapor in the rectifying section, or does both.
For example, if the feed is cold, the entire feed stream becomes a part of the
stripping section retlux. But also. to heat the feed to its bubble point. some vapor
must condense. and this condensate also becomes part of the stripping section

.tw~1~'__________-===================================~======~__~_________'~
 DISTILLA TION 89

reflux. Of course the vapor to the rectifying section is less, by the amount of the
.::ondensate.
To determine numerically the effect of the feed on the column internal flows,
voe use the q factor, which is defined as the moles of liquid flow in the stripping
5Cction that result from each mole of feed. Equivalently, it could be defined as
the heat required to convert one mole of feed to saturated vapor divided by the
molal heat of vaporization of the feed. The various possibilities are
q> 1, cold feed,
q= 1, feed at bubble point (saturated liquid),
o< q < 1, feed partially vapor,
q= 0, feed at dewpoint (saturated vapor), and
q< 0, feed superheated vapor.
,",ote that if the feed is a mixture of liquid and vapor, q is the fraction that is
liquid.
With the definition of q, the feed line can be derived as
q x
y = - - - x - -J-
q-l q-l

Once again, the line passes through the diagonal at x = xJ' that is, y = x = x J
and the slope of the line is q/( q - 1). All three lines, the rectifying, the feed, and
,he stripping pass through a common point.
Once the rectifying and stripping operating lines have been plotted, we start
stepping in the stages. But when are we at the feed stage? When we change from
stripping to rectifying or from rectifying to stripping. Note that this is not
unique. But we get the smallest number of stages by changing as soon as it is
possible.
What is the minimum number of stages that we could possibly use to obtain a
desired separation? The more reflux we use, the better the separation. But if the
separation is fixed, the more reflux we use, the fewer stages we need. So, in the
limit, an infinite reflux ratio should give us the minimum stages. And an infinite
reflux ratio gives us a slope of one, or the diagonal line. Thus, using the y = x
diagonal as the operating line and stepping off the stages determines the mini-
mum number of stages.
At the other extreme, as the reflux ratio gets smaller, the number of stages
required increases, to a point where we require an infinite number of stages.
Graphical analysis will show that if the operating line passes through the inter-
section of the feed line and the equilibrium line, an infinite number of stages
would occur at the pinch point.
The slope of this operating line is Rmin/(Rmin + 1), from which we can calcu-
late the minimum reflux ratio R min . Should the equilibrium line be so nonideal
that the operating line passes through it, we increase the slope of the operating
line until it just becomes tangent to the equilibrium line. The reflux ratio calcu-
lated for this slope is then the minimum.
90 MASS TRANSFER

3. Ponchon-Savarit Enthalpy-Concentration Method

This method is fully discussed by McCabe and Smith, pp. 571-585.

III. LEACHING AND EXTRACTION

We now discuss methods of removing one constituent from a solid or liquid by

means of a liquid solvent. There are two categories. Liquid extraction is used to
separate two miscible liquids with the use of a solvent which preferentially dis-
solves one ofthem. Leaching (solid extraction) is used to dissolve soluble matter
from its mixture with an insoluble solid.

A. Leaching

1. Terminology

The most important method is the continuous countercurrent method using

stages. Consider the diagram in Figure M4 where the stages are numbered in the
direction of the solids flow. The V phase is liquid that overflows from stage to
stage counter to that of the solid, dissolving solute as it moves from stage N to
stage 1. The L phase is the solid. Exhausted solids leave stage N and concen-
trated solution overflows from stage 1.
We assume that the solute-free solid is insoluble in the solvent and that the
flow rate of this solid is constant throughout the cascade. We let V refer to the
overflow solution and L to the liquid retained by the solid, both based on a
definite flow of dry solute-free solid. The terminal concentrations are Xa =
solution on entering solid, Xb = solution on leaving solid, Yb = fresh solvent en-
tering the system and Ya = concentrated solution leaving the system. Just as
with distillation or absorption, the calculations for a countercurrent system can
be analyzed by using an equilibrium line and an operating line. The method
used depends upon whether these lines are straight or curved.

FIGURE M4. A staged leaching system.

 LEACHING AND EXTRACTION 91

r 2. Equilibrium

If sufficient solvent is present to dissolve all the solute in the entering solid,
equilibrium is attained. And the concentration of the liquid retained by the solid
;eaving any stage is the same as that of the liquid overflow from the stage. The
equilibrium relationship is Xc = Ye'

3. Operating Line

The operating line equation is obtained from material balances around the first
stages:
'1

total solution:
solute:
Then
1 v..Ya - Laxa
Yn+l = -l-+-[(-V-.---La)/Lj xn + Ln + v.. - L~

This operating line passes through the points (xa, Ya), (Xb' Yb)'

4. Constant or Variable Underflow

Two cases are considered. If solution density and viscosity change considerably
with solute concentration, the lower stages may have solids with more liquid
retention than the higher stages. The slope of the operating line would thus vary
from stage to stage. On the other hand, if the mass of solution retained by the
solid is independent of concentration, Ln is constant and the operating line is
straight. This is called constant solution underflow. Of course if the underflow is
constant, so is the overflow.

5. Number of Ideal Stages-Constant Underflow

When the operating line is straight, a McCabe-Thiele diagram can be used to

determine the number of ideal stages; but with leaching, the equilibrium line is
always straight. so the Kremser equation

IV = In
Yo -
- y:
(~'b Y~)I!ln (Yb -Ya)
I Y~ - Y:
can be used with y: = xa and y~ = xb •

Note that this equation cannot be used for the entire cascade if La' the solu-
tion entering with the unextracted solids, differs from L, the underflow within
the system. This is illustrated later in the problems.
92 MASS TRANSFER

6. Number of Ideal Stages-Variable Underflow

When the underflow and overflow vary stage to stage, a modification of

Ponchon-Savarit can be used. The modifications are (1) consider each stream a
mixture of solid and solution and (2) use the ratio of solid to solution in place of
enthalpy. The solution is a mixture of solute and solvent. Let a represent solute.
b the solid, and S the solvent. Then the abscissa is X = a/(a + s). The ordinate is
Y = b/(a + s). The curved line of Figure M5 must usually be determined experi-
mentally. Since the solid is assumed insoluble in the solvent, all points represent-
ing overflow streams lie upon the X axis (Y = 0). Also since at equilibrium, the
solution concentrations for both under- and over-flow for the same stage are
equal, all tie lines are vertical. As with Ponchon-Savarit, a point for the mixture
of two streams lies on the straight line connecting the mixed streams and the
level rule applies. Bear in mind that lines on the X - Y diagram represent not the
masses of the total streams but only the masses of the solution in the streams.

a. Overall Balances and Point J

The overall balance for leaching has four streams: the inlet solvent Vb' the outlet
concentrated liquid Va, the inlet feed La' and the outlet exhausted solid Lb' We
assume the overall process replaced by two fictitious processes in series; in the
first, streams Vb and La are mixed to form stream J; in the second, stream J is
divided into streams Va and Lb' These processes are represented in Figure M6
by
Vb + La = J = Lb + Va
Five points, Vb' La, J, Y,,, L b , each representing a corresponding stream, are
shown. A straight line through points V. and Lb intersects a straight line through
points Vb and La at point 1. By the lever rule, the mass of solution in stream La is
inversely proportional to the line J La; the solution in stream Va is inversely

o Overflow 1.0
a
X = a+s
FIGURE MS. X - Y diagram for variable underflow leaching.
 LEACHING AND EXTRACTION 93

FIGURE M6. Ponchon-Savarit diagram for leaching.

proportional to line JVa; and so forth. If, for example, the mass ratio of the
solution in the incoming solvent to that in the feed is known, this ratio equals
--
J La/J Va and point J can be plotted. Then if the analysis of one other stream, for
example, v., is known, the line LbJVa can be plotted. Since point Lb is on the
underflow line, it is also the intersection of the underflow line with line LbJV.,
and Lb is thus determined.

h. Operating Line and Point P

Overall balances are treated by assuming a fictitious mixer added to each end of
the cascade. Imaginary stream P is assumed to be formed by subtracting La from
Va or by subtracting Lb from Vb:

or P = Va - La = Vb - Lb

Then a materiJ 1 balance over the first n stages and the fictitious mixer ahead of
stage 1 is
94 MASS TRANSFER

The operating line represented is a straight line passing through points P, V,d l'
and Ln" All other individual operating lines also pass through point P. These
lines, used alternately with the vertical equilibrium lines, give the Ponchon-
Savarit construction shown.

B. Liquid Extraction

When distillation is difficult or ineffective, liquid extraction is used. Close-boil-

ing mixtures are prime candidates for extraction. This method uses differences in
the solubilities of the components, thus exploiting chemical differences.

1. Basic Data

(a) A and B are substantially insoluble liquids, C is a distributed solute.

(b) E = lb/hr ft2 extract (solvent rich)
R = Ib/hr ftl raffinate (residual liquid)
B = lbjhr ftl solvent
(c) x = wt fraction of C in raffinate
y = wt fraction of C in extract
X = x/(l - x) lb Cjlb non-C in raffinate
Y = y/(l - y) lb C/lb non-C in extract
(d) Triangular coordinates. The sum of the perpendicular distance from any
point within the triangle to the three sides equals the altitude of the
triangle, as shown in Figure M7. Let altitude = 100 %. Point D: binary
20 ~<, B, SO % A. All points on DC have same ratio of A to B.
Adding R pounds of mixture at point R to E pounds of mixture at E
gives the new mixture shown on the straight line at point M, such that
R
- = =ME = XE
--- -
- XM .
mIxture rule
E RM XM - XR

(e) Kinds of Systems

(l) Type I (see Figure MS). Three liquids, one pair partially soluble; C
dissolves completely in A and B; A and B soluble to a limited ex-
c

7
I
I
A
~~
Iy?
__________ ~ B
[) FIGURE M7. Triangular coordinates.
 LEACHI"IG AND EXTRACTION

c c

1 phase

A _ _...;;..._ _ _ _..........AB A~ ___ ~~ __ ~

FIGURE MS. Extraction equilibrium FIGURE M9. Extraction equilibrium diagram-

diagram-type 1. Type II.

tent. (More C in E than in R -distribution coefficient greater than

1: large distribution Coefficient-less extracting solvent required.)
(2) Type II (see Figure M9). A and C completely soluble; A and B, B
and C limited solubility.

2. Principles of Extraction

Since most continuous extraction methods use countercurrent contacts between

two phases, one a light liquid, the other a heavy one, many of the fundamentals
of countercurrent gas absorption and distillation carryover to extraction. But
the equilibrium relationships in liquid extraction are more difficult.

a. Ponchon-Savarit Method

Since the graphical mixing construction and the lever rule both apply to triangu-
lar diagrams, the complete Ponchon-Savarit method carries over unchanged to
these coordinates. Typical constructions for type I and type II systems are
shown in Figures MIO and MIl, respectively. Points P and] have the same
significance as in leaching. With triangular coordinates. remember that stream
magnitudes and concentrations are based upon total streams.

Raffinate

p B

FIGURE MJO. Ponchon-Savarit for type [ extraction.

96 MASS TRANSFER

B A FIGURE MH. Ponchon-Savarit for type II extraction.

b. Rectangular Diagram Methodfor Extraction

Given a triangular diagram as in Figure M12a; we replot it as in Figure M12b,

where F and 0 are the two components to be separated, and S is the solvent.
Using Figure M 12b, we obtain the rectangular diagram of Figure M13. Notice
that all points on the x axis are solvent free (the Xo = 0% is the same as

o s

F.&.-_ _ _ ~ _ _ _...&S FL-_ _ _ _ ~ _ _...l.O

A B
(a) (b)

FIGURE M12. Extraction diagrams: (a) triangular; (b) replotted triangular.

100%

Xs

0% ~_ _ _ _ _-';:::::"I--_ _~ FIGURE M13. Extraction diagram: rec-

0% 1"'0 100% tangular
 LEACHI~G AND EXTRACTION

s 100%

0-.1--_-""""----100%
0\ IXo/La
\ / /
\ I /
\ //
(<1
\
\
\
FIGURE M14a. Construction of stages: FIGURE M14b. Construction of stages: rec-
triangular diagram. tangular diagram.

100 (;~) F). The line from 100/0 X s to 100 7~ X 0 is the line of 0 %F. If the curved
line Lo-Ln is the locus of raffinate, we know that, if F is not soluble in S, the locus
of extracts will be on the S-O straight line on the triangular diagram (or on the
100~~ Xs-I00%Xo straight line on the rectangular diagram).

c. Countercurrent Construction of Stages

The construction of stages is the same whether you use triangular coordinates,
as in Figure M 14u, or the rectangular coordinates of Figure M 14b.

d. Minimum Solvent Ratio

(For processes with a component that is insoluble in the solvent.) Given a coun-
tercurrent process such as that shown in Figure M 15, a total balance around the
system gives

v2

.. ~
v,
l~
~ .1
~
~ Vn+l
1 \ • Vo n
La Ln
L1 Ln-l
FIGURE MIS. Staged countercurrent extraction system.
98 MASS TRANSFER

and if one of the components of the L stream is not soluble in the solvent, we
may write a balance on that component as (call it a)
o fraction of A
Xa = in stream 0
We write a solvent balance

Now if there is no solvent in the feed, x? = 0, and if the solvent is pure solvent
x: + 1 = 1, so the solvent balance becomes

Define the solvent to feed ratio as P, so P = v..+dLo. Then the total material
balance becomes
1 + P = (vdLo) + (Ln/Lo) = VI + L.
and the solvent balance

Solving the total balance for VI' plugging it into the solvent balance, and re-
arrangmg, gIves

We see that if Ln and x: are fixed by the process, the minimum P occurs for
x; = o.

IV. GAS ABSORPTION

We have a soluble vapor mixed with an inert gas. In gas absorption, a liquid in
which the solute gas is soluble, is passed countercurrent to the gas stream. This
unit operation is most often carried out in a packed tower, but plate columns are
sometimes used.

A. Packed Towers

1. Principles

a. Material Balances

In a packed tower there are no discrete sudden composition changes as there are
in a plate column. The composition changes are continuous. Nevertheless, a
,..

GAS ABSORPTION

I'----l-
I 1
I 1
I
I L,x 1
1
- -- --I
y,V

t FIGURE \116. Packed absorption column.

total material balance is the same as for a plate column. A total balance around
part of the packed column of Figure M 16 gives
La +V = L + V.
Likewise, a soluble component A balance is
Laxa + Vy = Lx + VaYa
with V the molal flow rale of the gas phase and L the liquid phase rate. V, L, y,
and x are all at the same unspecified location.
We can also write the overall balances

Based upon these balances, the operating line may be written as

L VaYa - Laxa
y=-x+ - --
V V
or

LIC-~aXa - 1 ~ x) = VIC ~aYa - 1 ~ J

on a solute-free basis, where
L' = L(I - x), V' = V(1 - y)

h. Limiting Gas-Liquid Ratio

If we reduce the liquid flow at constant gas rate, the operating line slope L/V
decreases as shown in Figure M 17. The operating line top then moves in the
100 MASS TRANSFER

FIGURE '\117. Determining limiting gas-liquid

ratio.

direction of the equilibrium line. Thus the maximum liquid outlet composition
and the minimum liquid rate occurs when the operating line just touches the
equilibrium line at x~, Yb'

c. Differential Material Balance

If we consider a thin slice of the packed column and do a differential balance on

the solute, component A:
d(Lx) = d(Vy)
Note that each of these terms is the rate at which A is transferred from one phase
to the other through the interface between the phases. So we may write
d(Lx) = d(Vy) = dNa
where Na is the transfer rate of component A in moles per hour.

2. Absorption Rate

a. Theory of Double Resistance

Solute A must go in series through a gas resistance, the interface, and a liquid
resistance. If the interface offers no resistance, the gas and liquid phases are at
equilibrium there. This is usually true.
Look at a slice of a packed absorption column dz at a distance z from the top.
The rate of absorption is dN a and the area of the interface is dA. Then, assuming
equimolal diffusion of one component,
dN = ~il'_=- Yi) dA
a (1 - y)L
where ky = gas-phase mass transfer coefficient,
mo1!(area)(hr)(mol fraction difference)
y. Yi = mole fraction A in bulk and interface gas phase, respecti\el~
 GAS ABSORPTION 101

Considering the liquid phase, we can also write

dNa = kx(xj - x) dA
with
kx = liquid-phase mass transfer coefficient

( mol )
area hr mol fraction difference
Thus we may write

dN a = .. - - y)
kv(Y - d A= k x(xj-x) d A= d()
Vy = d( Lx)
(1 - Y)L
and we may choose sets of two at will. For instance, the gas-phase equation is

and the liquid-phase equation is

We define dA = as dz where a is the interface area per unit volume of packing, S

is the cross-sectional area of the tower, and dz is the length increment.
In addition, since V = V'/(l - y),

d(Vv) = V'd(-
. 1- Y
= V' L_)dy
(1 - y)2
= V~
1- y
we have

kya(y - yJ dz = GM _dy _
(1 - y)L y I- Y
with GMy = V IS = molal mass velocity of gas, and, for the liquid-phase equation

dx
kxa(xj - x) dz = G\lx - ~
I-x
with GMx = molal mass velocity of liquid

b. General Case-In Lly Triangle

Since the transfer rate is the same for both phases,

dy dx
GMy -- = GMx -l~-
I-y -x
it follows that

!' =-.l:'! = (1-= Y)Lkx~

Xj - x kya
102 MASS TRANSFER

FIGURE MIS. Packed column· ~ t.xt.y triangle

x method.

or

This equation has a slope of

k x a(1 - Y)L
kyu

and goes through the point (Yi' xJ on the equilibrium curve (point c of Figure
MlS). So ab is Y - Yi and bc is Xi - X. By drawing many of these triangles we
know Y - Yi and Xi - X as a function of X or y. We use these data to integrate
our equations graphically:

f Yb

Ya
~)L dy
(1 - y)(y - yJ
= kyu
GMy
f
0
ZT
dz = (kya)ZT
GMy

and

I Xb

Xa (1 -
dx
X)(Xi - x) = G~x
(k a)
ZT

Both equations give the same result.

c. Overall Coefficients

The general method works for either curved or straight equilibrium lines. But it
uses the individual coefficients that are difficult to determine experimentally. If
the equilibrium line is straight, easily obtained overall coefficients may be used.
The overall coefficients are defined by

dNa/dA
K =---
y y - y*
 GAS ABSORPTION 103

];or the overall gas-resistance coefficient, and

~0r the overall liquid-resistance coefficient. The overall coefficients are related to
the individual coefficients by

.:md
1 1 1 1
--= - + - - = - - -
Kxa kxa mkya mKya
where m = (Yi - y*)/(x i - x) is the slope of the equilibrium line. In terms of the
overall coefficients,

f Yb

Ya
dy
(1 - y)(y - y*) =
(K
C:a)
y ZT

and

f Xh

Xb
dx
(1 - x)(x* - x) =
(K a)
G: x ZT

d. Height of Transfer Unit (HTU) Method

With this method, the total height of packing ZT is

ZT = NtH
with Nt the number of transfer units and H = HTU. The number of transfer
units may be defined by

N ty =
jYa
Yh

(1 - y)(y -
dy
yJ
or

N
toy
= jYa
Yb
--
dy
(1 - y)(y - y*)
---

or

or

N
tox
= fXa
Xb
- -
dx
---
(1 - x)(x* - x)
104 MASS TRANSFER

Likewise we have the following equations for the HTU:

Hy = GMy/kya
Hx= GMx/kxa
Hoy = GMy/Kya
Hox = GM,,/Kxa
The overall HTU's are related to the individual HTU's by

e. Controlling Resistance

With a slightly soluble gas, the slope of the equilibrium line m is large, and Hox is
close to H" if GMJG My is not large. Conversely, for soluble gases, the m is small
and Hoy is nearly equal to Hy if GMy/G Mx is not large.

f. Lean Gases

For systems with a small y, we have a lean gas. For these systems, the operating
line is essentially straight. If, in addition, we have a straight equilibrium line, all
of our previous equations (and integrations) are simpler. For this case, we have
GMY(Yb - yJ = Kya(y - Y*)LZT
and

where
( _ *) _ (Yb - Y6') - (Ya - Y:)
Y Y L - In [(Yb - Y6')/(Ya - Y:)]
and

(X* _ X)L = (x6' - x b ) - (x: - xa)

In [(x6' - xb)/(x: - xa)]
In addition, the HTU equations become
N tOY = (Yb - Ya)/(Y - Y*)L
and

3. Coefficients and BTU's

Little general data exist.

 HUMIDIFICATION PROCESSES 105

B. Plate Columns

Gas absorption in a staged system is straightforward. For a straight operating

iliIne and a straight equilibrium line, the absorption factor (Kremser) equation is
.lipplicable. If either, or both, lines are curved, the McCabe-Thiele graphical
method is appropriate.

v. HUMIDIFICA nON PROCESSES

A. Definitions

The humidity .Yf is defined to be the mass of vapor carried by each unit of mass
of vaporjree gas, or

where M a, Mb are component molecular weights (A being the vapor), Pa is the

partial pressure of A, and PI is the total pressure. We may use this equation with
Dalton's law to find the mole fraction y of the vapor in the gas phase

When the vapor is in equilibrium with the liquid at the gas temperature we have
a saturated gas. According to Dalton's law the partial pressure of the vapor in
the saturated gas is equal to the vapor pressure of the liquid at the gas tempera-
ture. Thus

where Pa = vapor pressure of the liquid.

The relative humidity .Yf r is defined to be the ratio of the partial pressure of
the vapor to the vapor pressure of the liquid, expressed as a percentage, or

.Yf r = G:)100
On the other hand, the percentage humidity (sometimes called absolute hu-
midity) .Yf a is defined to be the ratio of the actual humidity to the saturation
humidity, eApressed as a percentage,
106 MASS TRANSFER

The humid heat Cs is the heat necessary to raise the temperature of one pound
of dry gas and the accompanying vapor by 1°F, or
Cs = C pb + Cpa X'
where Cpa, Cpb are the specific heats of vapor and gas.
The dew point is the temperature to which a vapor-gas mixture must be
cooled (at constant humidity) to become saturated. Note that the dew point of a
saturated gas phase is the same as the gas temperature.
The humid enthalpy or total enthalpy H y is the enthalpy of one pound of gas
plus the vapor it contains, or
Hy = C,(T - To) + x)'o
where is the latent heat of the liquid at To, a datum temperature.
,{O
The adiabatic-saturation temperature 1'. is the final steady-state liquid temper-
ature for a process where a gas, initially at a humidity :Yf and temperature T.
flows through a liquid spray chamber where the gas is cooled and humidified
adiabatically. An energy balance for the process gives the equation
:Yf - Xs
---= --=-
Cs Cpb + CpaX'
1'.
T -
with Cs and A" evaluated at 1'..

B. The Humidity Chart

For standard pressure, humidity charts may be used. Perry has charts for air-
-water as well as for several other systems.

C. Theory of Wet-Bulb Temperature

At the wet-bulb temperature, the rate of heat transfer from the gas to the liquid is
set equal to the rate of vaporization times the sum of the vapor sensible heat and
the latent heat of vaporization
q = MaNaVw + Cpa(T - Tw}
where q is the sensible heat transfer rate to the liquid, the molal rate of vaporiza-
tion is N a' and Aw is the liquid's latent heat at the wet-bulb temperature Tw'
We can also write this heat transfer rate in the usual form
q = hyA(T - TJ
where hy is the gas-to-liquid surface heat transfer coefficient, the interface tem-
perature is T;, and A is the surface area of the liquid.
The mass transfer rate, as for gas absorption, can be written

N = k A (Yi -12
a y (1 - Y)L
 --
HCMIDIFlCATION PROCESSES 107

with the )--'s as mole fraction of vapor. Expressing the mole fractions in terms of
humidities while assuming that the interface temperature is the same as the wet-
bulb temperature.

hiT - Tw) = k..- ( Yf-"----

w
- (1 - yh \(1 M b ) + (Yfw'M.)

- ~Mb) :CYf/Ma»)UW + Cp.CT - T.JJ

With the reasonable simplifying assumptions that (1 - Y)L ~ 1, CpaCT - Tw) ~
i.,.., and that Yf w .•\.1. and Yf/Jl. ~ 11M b ,

or

and for air-water systems, the Lewis relation

hy/J1 b ky ~ Cs
is valid.

D. Humidification Processes

1. Introduction

The various types of humidification processes are shown on the humidity chart
of Figure M 19. Heat fiow and diffusion of vapor through the gas interface char-
acterize the wet-bulb temperature process. If the liquid is a constant tempera-
ture, the adiabatic humidifier is a wet-bulb temperature process. But with
dehumidifiers and liquid coolers, if the temperature of the liquid is changing,
there will also be a heat fiow in the liquid phase. With heat fiow a temperature
gradient results, and thus a liquid phase resistance to heat flow.

Humidification
Evaporative
Cooling
Sensible
SE"nsiblE" HE"ating
H
eDolin,; -----~k_---~

Cooling and
Dehurnidif - :U tion Dehydration

Dehumidification FIGURE M19. Humidification pro·

Dry Bulb Temp cesses.

-
108 MASS TRANSFER

2. Gas-Liquid Interaction

The more important humidification processes can be characterized by

Adiabatic humidifier Yy> T;; 1'. < T;; J'f j > J'f
Dehumidifier 1'y > T;; 1'. < T;; J'f > J'f j
Countercurrent cooling tower
Upper part 1'y < T;; 1'. > T;; J'f j > J'f
Lower part 1'y > T;; 1'. > T;; :lt j > :It
where 1'. = bulk stream liquid temperature,
1'y = bulk stream gas temperature,
T; = interface temperature.
J'f j = interface humidity, and
J'f = bulk gas stream humidity.

3. Packed Gas-Liquid Contactors

Consider the packed gas liquid contactor of Figure M20. The mass velocity of
the gas is G~, mass of vapor-free gas per cross-sectional area per hour. Let the
cross section of the tower be S, so the volume of the test section is S dz. Assume
the liquid is warmer than the gas. We write the following balances over the
volume S dz.

dz

G '
TY
if'
Hy

GY '
~b
Hyb FIGURE M20. Packed gas-liquid contactor.
 HUMIDIFICA nON PROCESSES 109

Enthalpy balance
G~ dHy = d(G;xH;x) (a)

Heat transfer rate: liquid to interface

(b)
QH = heat transfer area per unit volume
Heat transfer rate: interface to gas
G~Cs dT" = hiT; - T,,)QH dz (c)
Vapor mass transfer rate: interface to gas
G~ dy'f = kyM b( y'f i - y'f)Q Mdz (d)
QM = mass transfer area per unit volume
With these equations and the simplifying assumptions

• G;x changes little with z

• H;x = CL(T,. - To); To = base temperature
• QH = aM =a
• Cs changes little with y'f

then
(e)

which is used with Eq. (b) to give

G;xC L dT,. = h;x(T,. - T;)Q H dz
or

(f)

Rearrange Eq. (c)

(g)

and Eq. (d)

dy'f = kyMbQM dz (h)
y'fj -y'f G~

Using Eq. ta) with Eq. (e)

IlO MASS TRANSFER

Now we multiply Eq. (d) by Ao and then add Eq. (c) to get

G~(Cs dI'y + AO d.Yf') = (AokyMbCYfi - Yf)a M + hiI; - I'y)a H ) dz (i)

Substituting
dH y = Cs dI'y + ).0 dYf
and using the simplifying assumptions, Eq. (e) gives

(j)

If we now direct our attention to the usual air-water system, using the Lewis
relation to eliminate hy , Eq. (j) becomes

Using

where Hi gives the air enthalpy at the interface, and the prior equation for H y ,
Eq. (k) becomes

or finally

(I)

Now, from Eqs. (a) and (b)

Solving this for dz and substituting into Eq. (1) gives

(m)

and we ha vc used the Lewis relation again.

Finally, dividing Eq. (1) by Eq. (g)

But. the Lewis relation defines the right-hand side to be unity. thus

dI'y 1; =-
~ (nl
dH) Hi - H,
Our working equations for the air--water system are Eqs. (1), (m), and (n).
 HUMIDIFICATION PROCESSES 111

4. Adiabatic Humidification

Because the air leaving an adiabatic humidifier may not be saturated (if it were,
it would be an equilibrium problem), we must use rate equations to determine
the size of a packed tower. To develop the working equations for this system, we
assume:

• inlet and outlet water temperatures are equal,

• makeup water enters at the adiabatic-saturation temperature,
• aM = a H = a,
• wet-bulb and adiabatic-saturation temperatures are constant and equal and
• Txa = TXb = T; = Tx = T, = constant

with T, = inlet air adiabatic-saturation temperature.

With these assumptions, Eq. (g) becomes

f Tya

Tyb
--y-=dT
T, - Ty
f
ZT

0
h a
-Y-dz
CsG~

which integrates to

In (Tyb - T,) = hyaZT = hyaSZT = hyaVr

Ty. - T, CsGy CsG~S Csm'
with VT = SZT = contact volume and
rh' = G~S = total dry air flow.
Had we used the mass transfer equation (h),

dYf = kyM b a dz
Yfs - Yf G~

which can be integrated since Yf., the saturation humidity at T" is constant:
In Yfs - Yf b _ kyMb a _ kyMbaVT
Yf - Yf - G' ZT - rh'
5 • Y

Notice that we may cast these equations into the HTU form, since, by definition
Yts-Yfb
Nt = In
.K s - Yfa
= number of humidity transfer units.

Then, by the definition of HI>

ZT G~
Ht =-=---
Nt kyMb a
= height of one humidity transfer unit
tt2 MASS TRANSFER

Of course, we get the same answer if we base the number of transfer units on heallt
transfer

VI. DRYING OF SOLIDS

A. Introduction

Many types of equipment are used for drying a wide variety of solids. Thus there
is not a single theory of drying. Most dryers are designed using general princI-
ples with experimental data.

B. Equilibria

The moisture content of a solid is expressed as the mass of water per unit mass of
bone-dry solid. So the driving force for mass transfer in a wet solid is the free-
moisture content X, which is the difference between the total moisture content
X T and the equilibrium moisture content X*.

C. Rate of Drying

The capacity of a dryer depends upon both the rate of mass transfer and the rate
of heat transfer. The heat of vaporization must be supplied to the vaporization
zone to vaporize the water. The water may be either at or near the solid surface
or within the solid, depending upon the solid and the conditions.

D. Rate-of-Drying Curves

With constant dryer conditions. the moisture content X T of a solid usually de-
creases, as shown in Figure M21. The graph is linear at first then curves up, and
finally levels off. Figure M22 shows the drying rate. It is horizontal at first, then
it curves downward and eventually reaches zero. Combining the two figures, we
obtain, for different materials. the drying-design curves of Figure M23.

XT
moisture dt
content drying
rate

time t time t
FIGURE M21. Moisture-content curve. FIGURE M22. Drying-rate curve.
 DRYING OF SOLIDS Il"

B A

R
Drying Rate
Ib/ft 2hr

X, Free moisture, lb/lb dry solid

FIGURE M23. Drying-design curve.

The differences in the shapes are the result of differences in the mechanism of
moisture flow in the different materials.

1. Constant-Rate Period

Every drying-rate curve has at least two distinct parts. After a warm-up period
I notshown in the figure), each curve has the horizontal line AB. This period is
the constant-rate period. The rate of drying is independent of moisture content.
The solid is so wet that a continuous film of water covers the entire drying
surface and this water behaves as if the solid were not there. If there is no heat
transfer by radiation or by direct contact of the solid with hot surfaces, the
temperature of the solid during the constant-rate period is the wet-bulb temper-
ature of the drying air.

2. Critical iWoisture Content and Falling-Rate Period

As the moisture content decreases, the constant-rate period ends at a definite

moisture content. With further drying, the rate decreases. The point at the end of
the constant-rate period (point B) is called the critical point. At this point, there
is not enough liquid water on the surface to maintain a continuous film over the
solid. Of course, if the initial moisture content of the solid is below the critical,
there will be no constant-rate period.
The period after the critical point is called the falling-rate period. The shape
of this curve varies from one type of solid to another. If also depends upon
processing conditions and on the solid thickness.

E. Calculation of Drying Time

The size of the equipment needed for a given capacity is fixed by the drying time.
With constant drying conditions, the drying time can be determined from the
rate-of-drying curve. If we consider the drying of a slab of solid of half- thickness
114 MASS TRANSFER

s and face area S dried from both sides, the total area for drying is 2S. Now the
ordinate R of the rate-of-drying curve is

dm
R= - - -
2S dt
where m = mass of total moisture in the solid and
t = drying time.

The abscissa of the curve is X, the mass of free moisture per unit mass of bone-
dry solid. If Ps is the density of the solid, in mass per unit volume of bone-dry
material, assumed constant, then

dm = 2sSps dX
Thus
dX
R = -sps dt
Integrating this equation between X 1 and X 2, the initial and final free-moisture
contents,

tT =spS I
XI

X2
dX
-R

where tT is the total drying time. The integration may be performed numerically
using the rate-of-drying curve. It can be done analytically if R is known as a
function of X.

1. Equations for Constant-Drying Conditions

The equation may be formally integrated for two conditions: during the con-
stant-rate period and during the falling-rate period, if the rate-of-drying curve is
linear.
During the constant-rate period, R is constant at its value Re. Therefore, the
time of drying in the constant-rate period te is given by

When the drying rate is linear in X during the falling-rate period

R = aX + b; dR = a dX
The integral becomes
 PROBLEMS 115

Of course, the constant a is the slope of the rate-of-drying curve or

Rc - R'
a= -~-- --~

Xc -X'
where Rc = rate at critical point,
R' = rate at final processing condition,
Xc = free-moisture content at critical point, and
X' = free-moisture content at final processing condition.
So

tf
_ sPs(X c
- R _ R'
- X) I
n
(]{l)
R
c 2

When the drying process covers both a constant-rate period and a linear failing-
rate period, the total drying time is the sum of tc and t f •

PROBLEMS

A. How many cubic feet per minute of entering air is needed to evaporate 10 Ib
of water per hour from a rayon, if the air enters at 80°F and 25 % humidity
and leaves at 170°F and 55 % relative humidity. The operating pressure is
14.3 psia.
B. A nitrogen-hydrogen chloride mixture (to vol % hydrogen chloride) is to
be scrubbed with water to remove the hydrogen chloride. To satisfy envi-
ronmental concerns, 99 7~ of the inlet HCl must be removed. Assuming that
the gas leaving the scrubber will be at 125°F and 1 atm saturated with
water, what will be the volume and composition of the gas leaving if we
must process 200 lb mOl/hr of dry entering gas.
C. We wish to extract nicotine from water using kerosene. If we have 100 Ib of
a 2 ~/o nicotine solution extracted once with 200 lb of kerosene, what per-
centage of the nicotine will be extracted?

Ib nicotine Ib nicotine
Equilibrium: y - - - =0.90X- - -
lb kerosene lb water

D. We must cool 2500 gal/min (gpm) of water from 120 to 80°F. A cooling
tower, at 70"F and 1 atm. is to be designed to operate with entering air of
40 (>~ relative humidity_
(a) How many cubic feet per minute (cfrn) of entering air must be sup-
plied?
(b) How many gallons per minute of makeup water must be supplied if
windage loss is 95 % of the amount lost by evaporation?
116 MASS TRANSFER

E. A solvent extraction will be performed on a solid B that contains a soluble

component A (mass fractions Xa = 0.25, Xb = 0.75). The solvent to be
used, C, is mutually insoluble with solid B. If we neglect the entrainment of
B in the overflow solution and screw press the extracted soLd to 1 lb of
solution per pound of B, calculate the pounds of solute-A-free solvent C
which we must feed the extractor, per pound of A-B feed, in order to obtain
95 % of the solute A in the extract overflow solution. What is the concentra-
tion of A in the resulting extract overflow solution?
F. We wish to scrub acetone from an air stream containing 0.020 mol fraction
acetone. If we use a countercurrent packed scrubber designed so that the
exit gas does not exceed 0.0002 mol fraction acetone,

(a) what is the number of overall gas-phase transfer units;

(b) what is the height of one transfer unit;
(e) what is the total height of packing used?

Base your calculation upon the following data:

Gas rate is 700 ft 3 /hr per square foot cross section at standard condi-
tions;
Water rate is 1000 lb/hr ftz;
Kga is 1.75 lb mol/ft 3 hr mol fraction difference;
Henry's law is valid and Ye = 1.75x e (Ye is the acetone vapor in equili-
brium with the mole fraction of acetone Xc in the liquid).
G. We wish to make our office building more comfortable by conditioning the
air. We feel that the scheme shown in Figure M24 is appropriate.

Air to building 10 5 ft 3 /hr at 70°F, 50 /~ relative humidity

Recycle air 5 x )0 4 ft 3 /hr at 70°F, 50 % relative humidity
Fresh air 40°F, 10/~ relative humidity
Steam 30 psig

The humidifier will be designed so that the air leaving it will be at 85 %

humidity.
Design coefficients: humidifier hl/a = 80 Btu/ft 3 hr of
preheater U = 20 Btu/ft 2 hr of

Fresh Adiabatic Heater Office

Preheater r----+ Humidifier f----- ~ Building

I
Air
I
Recy cle +
Exhaust
FIGURE M24. Schematic for problem G.
/
 PROBLEMS 11-

Based upon this scheme, find:

(a) Cubic feet per minute of fresh air required, at entrance conditions.
(b) Temperature and absolute humidities at (i) preheater inlet, (ii) hu-
midifier inlet, and (iii) humidifier outlet.
(c) Temperature of water required in the humidifier.
(d) Humidifier volume.

H. We need to cool 11,000 gal/min of process water from 100 to 80°F. Our
engineer suggests a mechanical (induced) draft cooling tower with 2 cells,
each 45 x 45 ft x 55-ft high. She wants to use air at a rate of 750,000
ft 3 jmin for each cell (measured at 80 c F dry bulb and 70°F wet bulb). The
design wet-bulb temperature was 70"F. Is this design approximately cor-
rect? If hya = 10 Btujft 3 hr of, calculate the tower height required.
I. How many stages and how much water is needed for the counter-current
extraction ofNaOH from a-feed consisting of 80 Ib NaOH, 400 Ib H 2 0, and
100 Ib CaC0 3 . The final extract solution will contain 10% NaOH with the
recovery of95 ~.~ of the NaOH. We shall set the underflow at a constant 3.0-
Ib solution/lb CaC0 3 .
J. We have a binary mixture, 50 % (mol) A, to feed to a distillation column.
We need 97% A overhead and a minimum of97 % B in the bottoms. For an
atmospheric system with a total condenser and a reboiler and a tempera-
ture such that the feed is half vaporized as it enters the column, determine
(a) the minimum reflux ratio; (b) the theoretical stages needed at total ref-
lux; (c) the theoretical stages needed using a reflux ratio of two times the
minimum; and (d) the feed plate number.
The equilibrium data are

XA 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

YA 0 0.242 0.418 0.551 0.657 0.741 0.811 0.870
x A 0.8 0.9 1.0
-0.920 0.963 1.0
YA

K. An ideal mixture of A and B is to be distilled continuously. If the relative

volatility 'l. is constant at 1.75 and

feed = saturated liquid with 50 mol % A,

feed rate = 200 lb mol/hr,
distillation composition = 90 % A,
t- lttoms composition = 10% A,
find analytically the minimum number of theoretical stages and the mini-
mum reflux ratio.
118 MASS TRANSFER

L. When a porous solid was dried under constant-drying conditions, 5

hours were required to reduce the moisture from 30 to 12 Ib HeO/lb dry
solid. Critical moisture content is 18 lb H 20/ib dr) solid and the
equilibrium moisture is 5 lb H 20/ib dry solid. If the dr} mg rate during
the falling-rate period is a straight line through the origin, determine the
time needed to dry the solid from 30 to 8 lb HeO/lb dry solid.

PROBLEM-SOLVING STRATEGIES

A. Although this seems to be a drying problem, it is actually just a material

balance using humidities. But be careful-both humidity and relative humi-
dity are used. Check the definitions. Remember that most humidity charts
are based on atmospheric pressure, not 14.3 psia. So you must perform
calculations instead of using a chart.
B. This scrubber problem is just a material balance problem. The leaving gas,
saturated with water vapor, will have 99 % of the inlet HCI. A humidity
chart will give you the outlet lb H 2 0 vapor/lb dry air. Since this is mostly
nitrogen, the humidity must be calculated. At standard conditions, the vol-
ume occupied by lIb mol of any gas is 359 ft 3 /lb mol.
C. This is an extraction problem, with one stage. We must assume that it is an
ideal stage. The given equilibrium equation is a straight line. Since, on a
nicotine-free basis, the material balance is also a straight line, the problem
could be solved analytically using the Kremser equation (sometimes called
the Souders-Brown equation). But with one stage, writing a nicotine bal-
ance and using the equilibrium equation is quicker. Since there is only one
stage, this extraction problem is just a simple material balance problem.
D. This cooling tower must be designed and there is not enough information
given to assume that it is simply a material balance problem. We need a
design strategy. The outlet air should be saturated with water vapor. But to
calculate the saturated humidity of the outlet air, we need to know the
outlet air temperature in order to find the correct vapor pressure. Heat
transfer ideas suggest that it is unlikely that the outlet air temperature
would equal the inlet water temperature. The inlet air temperature is lOOF
lower than the outlet water temperature. It is logical to assume that the top
temperature approach is the same as the bottom temperature approach or
lOoF. So assuming an outlet air temperature of 120°F - 1OF = 110°F
fixes the humidity. A heat balance will then set the inlet flow rate of air.
E. Since this solvent extraction problem has but one stage, a material balance
provides the answer. A total balance and balances on A and C make this
pro blem a gift.
F. This is a packed scrubber, so we rule out the ideal stage concept. The equili-
brium equation is a straight line, based upon mole fraction of acetone. In
 PROBLEM-SOLVING STRATEGIES

general, a material balance (the operating line) based upon the total m0;~'
of a stream is not a straight line. One way to get a straight-line maten2-;
balance is to use acetone-free air and water streams. Should we choose thl~
approach, the equilibrium relationship would no longer be straight. The
other possibility occurs if the acetone is present in such small amounts that
the material balance is approximately straight. This is the case for this
problem. Because both operating line and equilibrium line are approxi-
mately straight, an analytical equation is available. This is always prefera-
ble to a numerical integration. Thus we may use the lean gas absorption
equations. Materi'al balances supply the unknown data for the equation.
With the equation, the solution is simple.
G. This is a humidification problem at atmospheric pressure. A humidity chart
can be used. With the data given, we know the lb HzO/lb dry air leaving the
preheater is the same as the preheater inlet since the preheater simply in-
creases the temperature. To determine the preheater inlet humidity, we do a
H 2 0 balance around the fresh-air-recycle-air mixing point. An energy bal-
ance sets the inlet temperature. The humidifier outlet humidity is the same
as the office humidity since the heater adds no moisture. The humidifier
outlet temperature is found on the chart where the humidity line intersects
the 85/;' relative humidity line. To determine the humidifier volume, we
need an adiabatic humidifier design equation. Fortunately, McCabe and
Smith has one.
H. We do not have enough information given to solve this cooling tower de-
sign problem. Since this is a packed tower, we cannot use ideal stage con-
cepts. We know that the operating line is an energy balance. We know that
the equilibrium line is the enthalpy at saturation. We know that theory is
available to solve for packing height, given kya. But we also know that
theory is not available to determine bed cross-sectional area. We must
know, or be able to find, common practice concepts for this. We must check
Perry's Chemical Engineers' Handbook. If we find what we need, we can
judge whether to attempt the problem. If not, we should move to another
problem.
I. This is a staged solvent extraction problem. When looking for easy extrac-
tion problems, we want a straight material balance (operating) line and a
straight equilibrium line. We remember that a constant underflow is a
straight equilibrium line, and that a liquid-solid system usually gives a
straight material balance line. It looks as if the Kremser equation will give
us an easy solution. But we remember that with constant underflow, the
inlet stream is rarely set at the same rate as the underflow. We may have to
use just a material balance on the feed stage (first stage) and the Kremser
equation .,here the underflow is indeed constant.
J. This binary distillation problem is probably best solved graphically. We
could analytically find solutions for the minimum reflux and theoretical
120 MASS TRANSFER

stages at total reflux if the equilibrium data were based upon a constanm
relative volatility r:t.. But we would still need graphical solutions for the
other answer. So we would use graphical solutions for the complete prob-
lem.
K. For the minimum number of stages in a binary distillation with constant
relative volatility, we use the Fenske equation. I do not know of an analytI-
cal equation for minimum reflux ratio, but I know graphically how to ob-
tain it. So I can develop an equation upon which the graphical solution IS
based. I know that the minimum reflux line is straight and that its two
intersection points are (a) the distillate composition X D on the 45° line, and
(b) the point where the feed q line intersects the equilibrium line. The equilI-
brium line is defined by a = y(1 - x)/x(1 - y).
L. This is a drying problem. Problems like this are solved analytically or by
numerical integration. Since the falling rate is a straIght line, an analytical
solution is possible.

SOLUTIONS

A. See Figure M25.

Note: Humidity charts are based on l-atm pressure, so we must use equa-
tions, not charts
Vapor pressure (VP) of H 2 0 [McCabe and Smith, (McC & S) 3rd ed., App.
8]; at 80°F, 0.5069 psia; at 170°F, 5.992 psia. D = dryair.

, (0.5069)( 18 ) lb H 2 0
·Yfi = 0.25 14.3 _ 0.5069
28.9 = 0.0057 lb D air

:Yt = ~5(5.992) (~) = 01865 lb H 2 0

o 14.3 - 0.55(5.992) 28.9 . lb D air
So we remove
lb H 2 0
or 0.1808 lb D air

lb .. 10 lb H 2 0 lb D air
-- D aIr reqUIred = - . = 55.3 - - -
hr hrO.1808 IbHzO/lbD aIr . hr

. . . (540)( 1 0.0057)
humId vol of Inlet au = 359 492 28.9 + -18
= 13.76ft 3 /lbDair
.. lb D air ft 3 1 hr
cfm Inlet aIr = 55.3 h 13.76 1b . --.
r D aIr 60 mIn
= 12.68 cfm
 SOLUTIONS 121

Air
170°F
55% reI. hum.
10 Ib/hr H20
Ho

Air
80°
25% humidity Hi
no. of cfm? FIGURE M25. Schematic for solution A.

B. See Figure M26.

Basis: 200 lb moljhr dry inlet gas

Gas lb mol

HCl 20
Nz 180

Outlet gas: 180 lb moljhr N z ; 0.01 (20) = O.2lb moljhr HCl and H 2 0.

ffs = - - .MW - - VP
- -H20· - -H20
---
MWN2 (1 atm - VPH20)
Steam tables:
VP H20 at 125 = 1.942 psia; MWH20 = 18, MW N2 = 28 .

.Yf =}~ (- 0
42 __ ) = 0098 ~H20
s 28 14.7 - 1.942 . lb D gas
or
1.942 mol H 2 0
---~- = 0.152 - - ---
14.7 - 1.942 mol D gas

125°F
1 atm
sat. with H20
1% of inlet HCI

200 lb mol/hr
dry
10 vol % HCl
90 vol % N2 FIGURE M26. Schematic for solution B.
122 MASS TRANSFER

H2O

:
H2O 98 lb nic
X phase
nic 2 lb
nic
kero 200 lb :1 kero
Y phase
FIGURE M27. Schematic for solution C.

Outlet gas:

Gas Moles Percent

N2 180 86.7
Hel 0.2 0.1
H 2O 27.43 13.2
--
207.63 100.0

H 20 =
mol
0.152 ( - -H20) (180 + 0.2) mol D gas
mol D gas
= 27.43
Volume of gas leaving:

20 . l~ol x 359 ft 3 (460 + 125) = 88 630 f~

7.63 hr lb mol 492 'hr
C. Nicotine balance (see Figure M27) (nic = nicotine, kero = kerosene)
input = output in X phase + output in Y phase
lb nic lb nic
Y lb ker~ = 0.90X lb H 2 0

lb nico Ib nic
2 lb = 98 Ib H 2 0X + 200 lb kero Y ~lb
k
Ib H 2 0 ero
Want: pounds nicotine out.
Now X = Y/0.9

2 = 98(~)
0.9
+ 200Y = (980.9 + 200)Y;
Y = 2 -7- (2! +
0.9
200) = ~
310
= 6.45 X 10- 3

lb niC)
lb nico out = 200 lb kero ( 6.45 x 10 - 3 lb kero = 1.29 lb
so
lb nicotine extracted = 1.29.
%extracted = (1.29/2) 100 = 64.5 %
 SOLUTIONS 123

2500 gpm H~
120 0 P Air sat.
Hs = .0589

1
atm

Air
70°F
40% rGl. hum.
Ib HzO
H = .0063 Ib 0 air FIGURE M28. Schematic for solution D.

D. See Figure M28.

Basis:
gal Ib Ib
2500 - x 8.25 - = 20 623 ~-
min gal ' min
ft3 lb Ib
0.13368 -1 x 61.71 [3 = 8.25 -
ga t gal
(a) Entering air. Assume same outlet approach as inlet approach 120 -
Yao = 80 - 70. So Yao = 110°F; Yfs = 0.05891b H 20/lb D air; and

H 20 picked up/lb D air = 0.0589 - 0.0063 = 0.0526 lb H2<?

lb D air
lb 1 Btu
heat removed from H 20 = 20,623 - - x -.- (120-80°F)
min IboF
5 Btu
= 8.25 x 10 --.-
mm

. I . 0 Btu . Btu
Cs (m et air) = .242 0 lb .; C s (out Iet air) = 0.267 - - - -
F Dalr OF lb D air
H = Cs(T - To) + Yf)'o
If To = 32°F; )'0 = 1075.8 Btu/lb,
Houl = 0.267(110 - 32) + 0.0589(1075.8) = 84.2 Btu/lb D air
H in = 0242(70 - 32) + 0.0063(1075.8) = 16.0 Btu/lb D air
Heat Balance: heat removed from H 20 = heat gained by air

8.25 x 10 5 Btu
-.- = (84.2 - 16.0)( -lb'-
mm
Btu
Dalr
)(D D Ib . air)
mm
124 MASS TRANSFER

Specific vol of inlet air = 13.3 fe /lb D air at 70°F, so

lb D air 13.3 ft3 ft 3
cfm entering = 1.21 x 104 . lb D . = 1.61 X 10 5 - .
mIn aIr min
(b) Makeup water
lb H 2 0 lb D air lb H 2 0
H 2 0 lost by evap = 0.0526 1b D . 1.21 X 104 • = 636 .
aIr mIll mm

= 636 lb H 2 0/min = 77.2 g~l = evap

8.25 lb/gal mIll

makeup required = 95 % evap + evap = gal)

1.95 ( 77.2-.
mIll
= 150 gpm
E. See Figure M29.
Basis: lIb of A-B feed
E is solvent and A, with no solid; V is solution (A + '::') and solid; So-
lution is same composition in both E and U. With our basis, V is 0.75 lb
solid and also 0.75 lb solution

A balance: 0.251b = e.E + u. 0.75

C balance: Sib = esE + Us 0.75
total balance: S + 0.25 = E + 0.75 or S = E + 0.5
Now

e.E = 0.95(0.25); u. V = 0.05(0.25); U = 0.05(0.25) = 00167

• 0.75 .
and
0.05(0.25) 0.05 2.95
eo = u., es = Us = 1- U. = 1- 0.75 = 1 - -3- = -3- = 0.9833

and then
E = 0.95(0.25) = 14.25
0.0167
Using total balance, S = 14.25 + 0.5 = 14.75.

8 Ib E Ib sol'n
801 vent C 95% of A

Feed 1 Ib U Ib sol'n
0.2 5 Ib A 1 Ib sol'n
0.75 Ib B Ib B FIGURE M29. Schematic for solution E.
 SOLUTIONS

.0002 mf 1000~
hr ft

Xa=O mf acetone

.020 mf
700 ft3 STl'
hr ft2 FIGURE M30. Schematic for solution F.

F. See Figure M30. This is a lean gas; we use the lean gas absorption equa-
tions
N Yb - Ya GMy
toy - ( *)' Hoy = - K '
Y - Y LM ya
( *) _ (Yb - Y6) - (Ya - Y:) .
LM = log mean
Y - Y LM - In [(Yb - Y6)/(Ya - Y:)] ,
Y6 = 1.75x b , Y: = 1.75x a

700 ft3 (lb mOl) Ib mol

G = hr fe 359 ft3 = 1.950 ~ftT;

_ 1000 Ib H 2 0 (lb mOl) _ 6 Ib mol

L - hr ft 2 181b - 55.5 hr ft2

To get xb-acetone balance

G(0.02U - 0.0002) = Lx b , Xb = (1.98 x 1O- 2 G)/L = 6.94910- 4 mf
Yb - Y6 = 0.020 - 1.75(6.949 x 10- 4 ) = 0.01878
Ya - Y: = 0.0002 - 1.75(0) = 0.0002
0.01878 - 0.0002 0.01858
(y - Y*)LM = In (0.01878/0.0002) 4.542
(a)

N = (0.020 - 0.0002) = 4.84

toy (0.01858/4.542)

(b)

H = GMy = 1.950 Ib mol ft3 hr = 1.114 ft

oy Kya hr ft2 1.75lb mol
(c)
ZT = 1.114 ft (4.84) = 5.39 ft
126 MASS TRANSFER

F Fresh air Air to Bldg

105 cfh
70 0 F,50%RH

Recycle
5 X 104 cfh
70°F ,50%R.H FIGURE M31. Schematic fm solution G.

G. (a) Cubic feet per minute of fresh air required-based on dry air balance
(see Figure M31).

humidity of air to bldg = 0.00794 Ib H 2 0

lb D air

- + -:if) - -
( T)( .28.9 3
humid volume = 359 - 1 ft -
492 18 Ib D air

So

ft3
humid volume of air to bldg = 13.55 - ---
Ib D air

and

huml'd vo I ume of fresh air

. = 12.63 -Ibft3- .
D air

air to bldg = 10 5 cfh/13.55 -I ~ .. = 7.379 X 10 3 ~~D air

b D aIr hr

ft3 Ib D air
recycle = 0.5 x 10 5 cfh/13.55 - - = 3.69 X 10 3 ----
Ib D air hr

3 Ib D air 3 Ib D air
F + 3.690 x 10 --- - = 7.379 x 10 - - - _.,
hr hr

Ib 0 air
F=3689 x 10 3 ---
. hr

3 Ib D air ft3 1 hr
F = 3.689 x 10 --hr - x 12.63 --- --:- - --
Ib D aIr 60 min

= 776.5 ft 3 /min inlet fresh air

(b) Temperatures and humidities.

(i) Preheater inlet -composed of fresh air and recycle.
 SOLUTIONS

We do a mass balance for humidity and an energy balanc~ f<x

temperature

CPH20vap = 0.45, C pair = 0.25 Btu/lb o f Mc&S

3.689 X 10 3 lb Dh air (0.000516 lb H 2

r lb D air
0)
fresh feed
+ 3.69 x 10 3 (0.00794)
recycle

= 7.379 x 103 lb D a~ Yf lb !'2 0

hr lb D air
preheater entrance
so

:ff = 0.00423 lb H 2 0
lb D air

3 lb D air " 3
3.689 x 10 - h-r- Cs (40 F - To) + 3.690 x 10 Cs (70 - To)

fresh air recycle

to preheater

If all C s are approximately equal,

3.689 x 103(40°F) + 3.690 x 103(70°F) = 7.379 x 10 3 (T)

or T = 55°F

(ii) Humidifier inlet and (iii) Humidifier outlet.

At humidifier exit, Yf is same as office bldg Yf, that is, 0.00794 lb
H 2 0/lb D air and this must be at 85 % relative humidity (RH) to
satisfy the design. Intersection of 85 % RH line and Yf = 0.00794
give T = 55°F. At humidifier inlet, Yf is same as that for air to
preheater or 0.00423. Follow adiabatic saturation line to this
humidity gives T = 70°F.
(c) Temperature of H 2 0 required in humidifier is at 100 ~~ saturation on
the adiabatic saturation or 52°F.
(d) Humidifier volume.
The adiabatic humidification design equation is given by McC & S as
128 MASS TRANSFER

where lit' = total flow of dry air and

IS = adiabatic saturation temp. of inlet air
I'yb = 70°F, IS = 52°F
Btu o 0 Btu
es(at 70°) = 0.243 ib~' es(at 55 ) = .245 lb of'

Btu
es(avh) = 0.244 1boF

H. According to Figure 15-6, p. 852 of the first edition of McC & S (also found
in Perry's Handbook), for a hot water temperature of 100°F and a cold
water temperature of 80°F, with wet-bulb temperature of 70°F, the water
concentration should be 2.7 gal/min ft2.
For 11,000 gpm, the cross-section area should be

11,000 gp~ = 4074 ft2

2.7 gpm/ft2
Our designer suggested 45 x 45 ft = 2025/cell or 4050 ftl total McC & S
(same edition) suggests a superficial air velocity of about 6 ft/sec. Our de-
signer suggested

(2min50,000 ft3 .)( min )

2025 ft 2 60 sec
= 6 17 J't
. sec
Now, we know that we are in the ballpark with these rules of thumb. We
want to find the height. McCabe and Smith gives several different (but
equal) equations for doing this gas-liquid contact

dH y = kyaA-!b dZ
Hi - Hy G~

but for air-water, the Lewis relation gives

~=e
Mbkra s

So

_ d~ = }!y CJ.- dZ
Hi - Hy esG;
where Hi = saturation enthalpy (equilibrium line) and
H y = actual air enthalpy (operating line)
 SOLUTIONS I~

or

If both equilibrium and operating lines were straight, we could use the
enthalpy equation analogous to the lean gas absorption equation or

f HYB

Hyb
dHy
Hi - Hy
= Hy. - Hyb
(Hi. - H y.) - (H ib - H yb)
/In (Hia - H y.)
(H ib - H yb )

But we see that the Hi line, shown in Figure M32a, is not linear with tem-
perature so we use graphical integration. The operating line is given by

or

We see from the H - T diagram that T..b = 80°F and H yb = 34 Btuflb D air

. .. (1 0.0205) 540 14.1 ft3

humld vol. of mlet alr = 359 28.9 + 18 492 = Ib D air

80 , -________- , , -________- , , -________- ,

.....~
ol
Cl 60
.0
.....
---..,
;:j

:Il

>:
0-
......
ol
..,
..c:
40
.:
Cil

I terIl>
wb I terIl>
db

20
70 80 90 100

FIGURE M32a. Enthalpy-temperature for solution H.

130 MASS TRANSFER

So
750,000 ft 3/min 1 Ib D air
G = = 53 9 1 - - -
y 14.1 ft3 lIb D air ' min
and

L 5500 -gal)( 61.8 Ib)/ 7.48 -gal

= ( --. 3 = 45,441
Ib H. 2 0 ,
3
mm ft ft mm
So

LC pl = 45,441 x 1 = 0.854 Bt~

Gy 53,191 Ib D mroF
Hy = 34 + 0.854(T - 80)
and we have plotted this, along with the saturation enthalpy in Figure
M32a. We read data from this graph

T 80 85 90 95 100
Hi 43 49 55 63 71
Hy 34 38.3 42.5 46.8 51.1
1/(Hi - Hy) 0.1111 0.0935 0.08 0.0617 0.0503

0.11

0.10 1---_ _ +_+-____--1

0.08 i - - - - - - + - . . . . - - - - - - - l

0.06 1 - - - - - - + - - - - 1 1 \ - - 1

30 40 50
Hy
FIGURE M32b. Graphical integration for solution H.
 SOLUTIONS Dl

Wash H2O Final 10% NaOH

95% of orig. NaOH
Extract
Underflow
80 lb NaOH
3 lb sol'n Feed 400 lb H20
lb CaC0:3 100 lb CaC0:3
(a)

Wash Ya Extract 10% NaOH

Yb
H2O N- 1 First 95% of orig. NaOH
derflow Stages a Stage Feed 80 lb NaOH
400 lb H20
b xa 100 lb CaC0:3
(b)

FIGURE M33. (a) Schematic for solution I; (b) modified schematic.

and plot l/(Hi - Hy) versus Hy and graphically integrate Figure M32b to
J
obtain = 1.377. But

CsG~ = (0.255 BtU)(~l x 106lb D air)( ft 3 hrOF) = 40.2 ft

hya IboF, 2025 ft2 hr 10 Btu
so
Z = (40.2 ft) x (1.377) = 55.4 ft
I. See Figure M33a . •:'or a straight equilibrium line and a straight operating
line, we use the Kremser equation. A constant underflow gives a straight
operating line. But note that this equation cannot be used for the whole
system if the solution entering with the unextracted solids is different from
the underflow within the system, as in this case. Thus we must break the
problem into two parts, as shown in Figure M33b.
Basis: Feed mixture.
lb NaOH in extract = 0.95 (80) = 76 Ib
but NaOH is 10% of the extract solution;
lb extract = 76 IbjO.l = 760lb
lb final underflow = (3 lb solnjlb CaC0 3 )100 lb CaC0 3 + 100 lb
CaC0 3 = 400 lb
Now the solution leaving a stage with the extract is the same concentration
as the solution leaving a stage with the underflow. So point a has a solution
containing 10~~ NaOH: Xa = 0.1 with 300-lb soln and 100-lb CaC0 3 . To
detl [mine the wash water, we do a fluid balance around the total unit:
in out
wash water + 400 + 80 = 760 + 300 or wash water = 580 lb
132 MASS TRANSFER

Kremser equation

Final underflow will contain the unrecovered NaOH: 80 lb - 76lb = 41b.

It is contained in 300 lb of soln, so
Xb = 4/300 = 0.0133, Yb = 0
since wash water contains no NaOH.
To determine Ya' we do a NaOH balance around the first stage
In out
80 + 580Ya = 76 + 0.1(300) or Ya = 0.0448
and
v
• b = 0' Xa = 0.1, Xb = 0.0133
So

N _ 1 -In (0 - O.Q1~)/'ln ( 0 - 0.0448) -In ( -0.133.)/

0.0448 - 0.1 0.0133 - 0.1 - 0.0552

In (-0.0448) = 2.16
-0.0867
Thus N = 3.16.
J. See Figure M34a. Slope of feed line = q/(q - 1), where q = moles of liquid
flow in stripping section resulting from 1 mol of feed or! feed vaporized or!
feed liquid, so
q=! and slope = !I(! - 1) = - 1
Minimum reflux ratio is determined from slope of rectifying operating line
through (Yd' x d ) and the intersection of feed with equilibrium curve. Slope
of this line is RD/(R D + 1), where RD = LID. Operating line goes through
the point x = 0.37, Y = 0.63 or

0.63 = ~D(0.37) + 0.97

RD +1

Xd = 0.97
Xf = 0.5
1/2 vapor

'-----1._..- Xg = 0.03 FIGURE M34a. Schematic for solution J.

 SOLUTIONS 133

I
I
.tl I
feedline
I
I
.6 I
I
I
I
.4
I
I
I
.2 I
I
I
xf XD
O~~-------------------L~------------------~
o .2 .4 .6 .8 1

FIGURE M34h. McCahc Thiele diagram for solution J.

(a)
RD = 1.308 minimum reflux

actual RD = 2 x 1.308 = 2.616, actual slope = 2.616/3.616 = 0.723

(b) Theoretical stages using total reflux: At total reflux, the y = x di-
agonal line is the operating line. From the construction shown in
Figure M34b, we need 6+, or 7, theoretical stages. This construc-
tion is indicated by the intersections with the equilibrium curve
(Tl, T2, etc.).
(c, d) We see from the construction using the actual reflux ratio of
2.616 that we need 9 theoretical stages, feeding onto the 5th theo-
retical plate from the top (A 1, A2, etc.)
K. The minimum number of stages may be determined analytically with the
Fenske equation

So with XD = 0.9, X B = 0.1, ex = 1.75, IV min = 6.85.

134 MASS TRANSFER

I
XF FIGURE M35. Equilibrium curve for solution K.

We have a saturated liquid feed, so q = I. Operating line for minimum

reflux ratio goes through the points (y*, x F ) and (Yo = x o , xo) as indicated
in Figure M35. So

Ro Yo - y* Xo - y*
slope = --- = :....=.-~
Ro +I Xo - XF

or

or
Xo - y*, or

We need y*, the equilibrium value at x = X F . Now

y(1 - x)
a=----
x(1 - y)

where for this case y = y*, x = XF or

5 18 30
Total moisture lb/lb FIGURE M36a. Drying data for solution L.
 SOLUTIONS I ..~

So

Ro = (XO - 1 + ~:F_ 1) XF) I(. + (:X~ l)xF - XF)

Xo + (ex - l)xFxo - XF
exxF - x F - (a - l)xF

= Xo - XF[XD + ex(1- xo)] = 9 - 0.5[0.9 + 1.75(0.1)] = 1.93

(a - l)x F(1 - XF) 0.75(0.5)(0.5)
L. Figure 36a uses total moisture; drying equations use free-moisture values,
so

. ( Ib HzO ) .
Free MOIsture ( I lb H z0r )
Total mOIsture Ib D solid b D so Id

30 =30-5=25
18 = 18 - 5 = 13
12 = 12 - 5 = 7
8 = 8-5=3

We must use the given data to determine the rate-of-drying curve of Figure
M36b.
Now
dX dX
R=- or dt=-
dt R
Constant rate:

t = f Z5
dX = 25 - 13
c 13 Rc Rc

o 13 25 FIGURE M36b. Rate-of-drying curve: solution

Free m:>isture. X L.
136 MASS TRANSFER

Falling rate:
R = aX,
so

or

J13 , ~
R In(~)
tr = [f 13

7
dX ~ XIIRXc =
X
Rc
£ In X
7
t =
r 7
! c

So

13 (13) _ 12 + 8.05 _ 20.05

5 -_ t + tr -_ 12- + -In - - - --
c Rc Rc 7 Rc Rc
or
Rc = 4.01
Our equations for this system are

t
c
=
X-X
Rc
c =
X-13
.
4.01'
Xc (Xc)
tr=--In - - = -13I n (13)
--
Rc X end 4.01 X end
So for X = 25 and X end = 3,
25 - 13
tc = - 4.m = 2.99 hr; tr = 13 In (13)
4.01 3 = 4.75

and
total time = 2.99 hr + 4.75 hr = 7.74 hr
 HEAT TRANSFER

I. Conduction in Solids 139

A. Introduction 139
1. Fourier's Law 139
2. Thermal Conductivity 139
B. Steady-State Conduction 139
1. Introduction 139
2. Compound Resistances in Series 140
3. Heat Flow through a Cylinder 140
C. U nstead y-State Heat Conduction 141
II. Heat Flow in Fluids 141
A. Introduction 141
B. Energy Balances 142
1. In Heat Exchangers 142
2. In Total Condensers 143
C. Rate of Heat Transfer 143
1. Heat Flux 143
2. Average Temperature of Stream 143
3. Overall Heat Transfer Coefficient 143
4. Log Mean Temperature Difference 144
5. Variable Overall Heat Transfer Coefficient 144
6. Overall Coefficients from Individual Coefficients 144
7. Resistance Form of Overall Coefficient 144
8. Fouling Factors 145
III. Heat Transfer to Fluids-No Phase Changes 145
A. Forced Convection- Laminar Flow 145
1. Introduction 145
2. Heat Transfer to a Flat Plate 145
3. Graetz and Peclet Numbers 146
B. Forced Convection- Turbulent Flow 146
1. Introduction 146
2. Dimensional Analysis Results 146

W. L. McCabe and J. C. Smith, Unit Operations in Chemical Engineering, 3rd ed., McGraw-Hili, New
York, 1976.
F. Kreith, Principles of Heat Transfer, 3rd ed., Harper & Row, New York, 1973.

137
138 HEAT TRANSFER

3. Theoretical Results 147

C. Forced Convection outside Tubes 147
1. Introduction 147
2. Flow Normal to a Single Tube 147
3. Flow past a Single Sphere 148
D. Natural Convection 148
IV. Heat Transfer to Fluids-Phase Changes 148
A. Condensing Vapors 148
1. Introduction 148
2. Coefficients for Filmwise Condensation 149
3. Practical Use of Equations 150
4. Superheated Vapors 150
B. Boiling Liquids 150
1. Introduction 150
2. Boiling of Saturated Liquid 151
3. Subcooled Boiling 151
V. Radiation Heat Transfer 151
A. Introduction 151
B. Emission of Radiation 151
1. Introduction 151
2. Radiation Wavelength 152
3. Blackbody Radiation and Emissivity 152
4. Practical Source of Blackbody Radiation 152
C. Absorption of Radiation by Opaque Solids 152
1. Introduction 152
2. Reflectivity and Absorptivity of Opaque Solids 153
3. Kirchhoff's Law 153
D. Radiation between Surfaces 153
1. Introduction 153
2. Angle of Vision 154
3. Square of the Distance Effect 154
4. Quantitative Calculation of Radiation between Black Surfaces 154
5. NonblackSurfaces 155
E. Radiation to Semitransparent Materials 155
1. Introduction 155
2. Radiation to Absorbing Gases 155
VI. Heat Exchangers 156
A. Heat Transfer Coefficients in Shell and Tube Exchangers 156
B. LMTD Correction for Crossflow 156
VII. Evaporators 157
A. Introduction 157
B. Evaporator Capacity 157
1. Boiling-Point Elevation 157
C. Evaporator Economy 158
D. Single-Effect Evaporation 158
E. Multiple-Effect Evaporation 159
Problems 159
Problems-Solving Strategies 161
Solutions 162
 CONDUCTION IN SOLIDS

1. CONDUCTION IN SOLIDS

A. Introduction

1. Fourier's Law

The basic relation for heat flow by conduction is the proportionality between
the rate of heat flow across a surface and the temperature gradient at the surface,
or

dq = -k aT
dA an
where q = rate of heat flow normal to surface,
A = area of surface,
T = temperature,
n = normal distance to surface, and
k = proportionality constant.

2. Thermal Conductivity

The proportionality constant k is called the thermal conductivity. It is a physical

property of the substance. It is independent of the temperature gradient, but it
may be a function of the temperature itself.

B. Steady-State Conduction

1. Introduction

Consider a flat slab with a constant thermal conductivity. We look at the steady-
state heat flow perpendicular to the slab. For this case, q will be constant. The
direction normal to the slab will be called x. Fourier's law becomes

CL=_k dT
A dx
or

f T

T,
2
dT = - -
q
kA
I X

Xl
2
dx

or

1 = k Tl - T2 = k i1T
A x 2 - Xl B

where B = thickness of slab, X 2 - Xl' and

L1 T = temperature drop across slab, T1 - T2
140 HEAT TRANSFER

RA RB He

BA Bs Be

......- - - - - 6T - - - - _ _ FIGURE HI. Compound resistances in series.

or, in resistance form

q = ATjR
with
R = BjkA = thermal resistance

2. Compound Resistances in Series

Consider a flat wall made of a series of layers as shown in Figure HI. For the
first material (orlayer), RA is the resistance, BA is the thickness, and ATA is the
temperature drop across BA- We can write

AT. + AT. + AT. = AT = qABA + qBBB + qcBc

ABC AkA AkB Akc

and, since qA = qB = qc = q,

3. Heat Flow through a Cylinder

Consider the concentric tube shown in Figure H2. Heat is flowing from the
inside to the outside. Note that the area perpendicular to this flow changes with
r, that is,
dT
q = -k-21trL
dr
L = length of tube
Integrating
 HEA T FLOW IN FLUIDS 141

FIGURE H2. Concentric tubes.

or
k2nL(7; - 7;,) kA L (7; - 7;,)
q =--:-----
In (ro/rJ ro-rj
where the last form is the way we wish to write the equation. So, in order to use
this form, we must have

We have just defined the logarithmic mean radius.

C. Unsteady-State Heat Conduction

Consider the earlier flat slab. We now look at it, not in the steady state, but as
time varies. The conduction equation becomes
aT k a2 T iJ 2 T
-=--=(X--
at pCp iJx 2 iJx 2
Because the solution to this equation is a series of terms, charts (such as those
~t0und
in Kreith) are convenient.

II. HEAT FLOW IN FLUIDS

A. Introduction

Consider Figure H3 in which a liquid is being heated by a condensing vapor.

The inlet and outlet fluid temperatures are Toa and Tob' The constant vapor
temperature is Th . At some length from the entrance, the liquid temperature is To
142 HEAT TRANSFER

Outlet

~ar; ".ir~

L,.~
Inlet Liquid
~t,J
Condensate
Th = Temy of Cond. Vapor
____- - - - - - - - - - - - - 6T 2

Length
FIGURE H3. Heat exchanger with condensing vapor.

and so the local difference is Th - T.,. This is called a point temperature differ-
ence. A T. At the inlet. the point temperature difference is Th - I::a = AT 1 ; at the
exit Th - Tch = ATz . The terminal point temperature differences A Tl and AT:
are called approaches. The change in the liquid temperature T.,b - T.,a is called
the temperature range.
Single pass systems can be either parallel or countercurrent. Countercurrent
almost always is the preferred method.

B. Energy Balances

1. In Heat Exchangers

The usual energy balance is

where q = Q/t = rate of heat transfer into the stream,

tit = flow rate of the stream. and
Ha, Hb = enthalpy per unit mass, in and out.

If we assume two fluid streams exchanging heat. with no loss to the surround-
mgs,

h = hot stream, c = cold stream.

Note the sign of qc is positive, while that of qh is negative, that is, q" = -qh'
So, the overall enthalpy balance is
 HEAT FLOW IN FLUIDS 143

If we assume constant specific heats,

mh C Ph( Tha - Thb ) = me C pe( T.:b - tea) = q

2. In Total Condensers

With a total condenser, the hot stream gives up its heat of vaporization only,
that is,
mh ) = mcCpe(T.:b - Tea) =q
If the hot stream also is sub cooled
mh[A + Cph(Th - Thb )] = meCpe(T.:b - T.:a) = q

C. Rate of HeafTransfer

1. Heat Flux

Our heat transfer calculations will be based upon the area of the heating surface.
It is usually expressed as British thermal units (Btu) per hour per square foot or
watts per square meter. This rate of heat transfer per unit area is called a heat
flux. For tube systems, it can be based either upon the outside tube area or upon
the inside tube area.

2. Average Temperature of Stream

Bear in mind that there is usually a temperature difference across the cross sec-
tion of the fluid. But we use the mixing cup or average temperature of the cross
section.

3. Overall Heat Transfer Coefficient

Analogous to the Fourier law of continuous mediums, we can write for the flux
dq/dA,
dq
- = VIlT = V(Th - T.:)
dA
where the proportionality factor V is called the local overall heat transfer coeffi-
cient. But note that if A is taken to be the outside tube area Ao, then V becomes
V o , the coefficient based upon that area; likewise for Ai and Vi' Now both IlT
and dq arc' 'ldependent of the area chosen, so

Vo = dAi = Di D = tube diameters

Vi dAo Do'
144 HEAT TRANSFER

4. Log Mean Temperature Difference

If we assume that the overall heat transfer coefficient is constant, the flux equa-
tion can be integrated over the total heat transfer surface to get

qT = UAT'~TL

~ = I1T2 -I1Tl = LMTD

TL In (I1T2/I1T1 )

5. Variable Overall Heat Transfer Coefficient

If we assume that U varies linearly with temperature drop, a more accurate

equation is

6. Overall Coefficients from Individual Coefficients

The overall heat transferred depends upon heat transfer in the hot stream, in the
tube wall, and in the cold stream. Thus the overall coefficient is a function of
individual coefficients:

if Ao is the area, and

if Ai is the area used, where

hi' ho = inside and outside heat transfer coefficient,

Di , Do = inside and outside tube diameters,
Xw = tube wall thickness,
DL = log mean diameter of tube, and
k = thermal conductivity of wall.

7. Resistance Form of Overall Coefficient

The overall coefficient can be written in resistance form

 HEAT TRANSFER TO FLUIDS-NO PHASE CHANGES 145

8. Fouling Factors

While in service, heat transfer surfaces tend to get dirty or fouled. This is taken
into account by using fouling factors hdi or h do ' depending upon whether dirt is
on the inside or outside of the tube. They may be incorporated into the overali
coefficients in the following way:

III. HEAT TRANSFER TO FLUIDS-NO PHASE CHANGES

A. Forced Convection-Laminar Flow

1. Introduction

With laminar flow, heat transfer occurs only by conduction. This type of prob-
lem may be solved using mathematical analysis. An advanced text on fluid flow
and heat transfer will give the details; only the results will be presented here.

2. Heat Transfer to a Flat Plate

The equation for heat transfer to a flat plate (shown in Figure H4) is
O.332NW Ni~:x
NNu.x = {I _ (xo/X)3/4p/3

where NNu,x = hxx/k = Nusselt number,

CpjJ
N pr = k = Prandtl number, and

= Reynolds number.

Xo = Unheated portion
of plate FIGURE H4. Heat transfer to a flat plate.
146 HEAT TRANSFER

This equation gives hx, the local value of h. Analytically it can be shown that the
average h, over the whole length x, is given by
h = 2hL
where hL is hx evaluated at x = L.

3. GrfUtz and Peclet Numbers

Dimensionless groups are commonly used in heat transfer to fluids. The Graetz
number is defined as

The Peelet number is defined to be the product of the Reynolds number and the
Prandtl number
DVp CpJi VCpD
N Pe = NReNpr = 7T = p - k -

B. Forced Convection-Turbulent Flow

1. Introduction

The most important heat transfer situation is heat flow to a turbulent fluid flow-
ing in a tube. Turbulent flow has Reynolds numbers greater than 2100. The first
results obtained were empirical correlations using dimensional analysis; later
results have had a more theoretical basis.

2. Dimensional Analysis Results

Most turbulent-flow forced-convection heat transfer data can be correlated by

equations of the form

or with G the mass velocity

NSI = ¢(NRe , Npr), NSI = Stanton number
For sharp edged entrances, the equation is
_h_ (CpJi)2/3(Jiw)O.14 = 0.023[1 + (D/L)o.7]
CpG k Ji (DG/Ji)O.2
All physical property data are evaluated at the bulk temperature T, except Jiw
which is evaluated at the wall temperature. To estimate the wall temperature, an
iterative calculation is performed using the resistance equations.
 HEAT TRANSFER TO FLUIDS-NO PHASE CHANGES 147

3. Theoretical Results

Some theoretical results may be obtained using the Reynolds analogy-the ratio
of the momentum diffusivity to the thermal diffusivity is equal to the ratio of the
momentum eddy diffusivity to the thermal eddy diffusivity. It is theoretically
expressed as

The following experimentally determined equation is more accurate:

h N 2 / 3 ("~
__
CG Pr
)0.14 -2f--
p Jl
f = 0.046(DG/Jl)-0.2

More accurate still is the form

but you must know ¢(N Pr) for various cases.

C. Forced Convection outside Tubes

1. Introduction

Heat flow in forced convection outside tubes differs from that inside tubes due to
differences in mechanisms. No form drag exists inside the tubes and all friction is
wall friction.

2. Flow Normal to a Single Tube

The same variables affecting heat transfer within tubes affect heat transfer out-
side, but the constants are different. The form is

For heating and cooling fluids flowing normal to single cylinders, use

hD° (CpJlr )-0.3 = 0.35 + 0.56(D G)0.S2

0. 0

kf kr Jlf

where.f mt·dns evaluated at the film temperature,

T;- = film temperature = (Tw + 1')/2, with Tw = wall temperature, and
l' = bulk temperature.
148 HEAT TRANSFER

3. Flow past a Single Sphere

The equation for this case is

hoDo = 2.0 + 0.60(D p G)o.50(C p J1r)I!3

kr J1r kf
with Dp = sphere diameter.

D. Natural Convection

A fluid, at rest, in contact with a hot vertical plate will be warmer near the plate
Since the density of a hot fluid is less than that of a cold fluid, this buoyanC}
effect will cause the hot fluid to rise and be replaced with a cooler fluid. A vek,('-
ity gradient will be formed near the plate. This is natural convection.
For single horizontal cylinders, the equation is of the form

hDo = 4>(C PJ1f, D~p1~q!n;))

kf kr J1r
where the last term is the Grashof number N Gr and
[J = coefficient of thermal expansion of the fluid
II To = average difference in temperature between the bulk fluid
and the cylinder wall.
Mathematically, the coefficient of thermal expansion is

{J= 1 -
v aT
(av) p
, v = specific volume

For liquids, you can use

{J = PI - 1'2 , PI
Pa = -2--
+ P2
Pa(T2 - TI )
F or ideal gases
{J = liT

IV. HEAT TRANSFER TO FLUIDS-PHASE CHANGES

A. Condensing Vapors

1. Introduction

The condensation of vapors upon the surfaces of tubes which are cooler than the
vapor condensing temperature is important commercially. The important areas
will be discussed.
 HEAT TRA:\fSFER TO FLUIDS-PHASE CHANGES ]49

There are two ways in which a vapor may condense upon a cold surface:
dropwise and filmwise. The average coefficient for dropwise condensation may
be five to eight times that for film wise condensation. Some generalizations are

• Filmwise condensation occurs if both the fluid and the tube are clean.
• Dropwise condensation occurs only when the surface is contaminated.
• A monomolecular film is sufficient to cause dropwise condensation.
• Substances that merely prevent surface wetting are ineffective for causing
dropwise condensation.

To be conservative, use filmwise condensation for design.

2. Coefficients for Filmwise Condensation

These equations have a strong theoretical basis. For vertical tubes, the local heat
transfer coefficient at a distance L from the surface top is
_ (pfg )113
hx - kr 31J..ir

or equivalently by

for an average coefficient, where

1 = rh divided by wetted perimeter,
;. = heat of vaporization,
LT = total tube length, and
1b = condensate loading at bottom of tube.
To remove the temperature difference, some manipulations are required, obtain-
mg
J..il
h ( -3-2
)1;3
= 1.47 - -
(41b) -
113

kr Prg J..ir
The physical property data are evaluated at Tr where

with Th = temperature of condensing vapor and

Tw = wall temperature.
This equation ;s equivalent to
150 HEAT TRANSFER

For horizontal tubes, use either

h(
~)1/3
3 2
_ (4r')-1/3
- 1.51
kr Prg J.lr
or

h = 0.725 (
k3p2g2
r f
)1/4
i17;, DoJ.lr
where r' is the condensate loading per unit length of tube.

3. Practical Use of Equations

These theoretical equations work well for single tubes. Theoretically for a bank
of horizontal tubes, the average loading per tube could be used

r' = mT
S (LN)

where m=T total condensate flow rate for stack,

L = length of one tube, and
N = number of tubes in the stack.
Because of splashing, better results are obtained by using
r~ = mT/(LN2/3)

4. Superheated Vapors

If the vapor is superheated, both the sensible heat (of the superheat) and the heat
of condensation must be transferred through the cooling surface. When the va-
por is highly superheated and the outlet temperature of the cooling fluid is close
to that of condensation, the tube wall temperature may be greater than the
saturation temperature and condensation will not occur. In this case, it may be
necessary to consider the equipment as two separate parts: a desuperheater (just
a gas cooler) and then a condenser.

B. Boiling Liquids

1. Introduction

When boiling is caused by a hot immersed surface, the temperature of the liquid
is the same as the boiling point of the liquid under the pressure in the equipment.
Vapor bubbles are generated at the heating surface, rise through the liquid, and
leave the liq uid surface. This is boiling of saturated liquid.
In some cases, the liquid temperature is below the boiling point but the heat-
ing surface temperature is above the boiling point. Vapor bubbles formed on the
heating surface are absorbed by the liquid. This is subcooled boiling.
 RADIATION HEAT TRANSFER 151

2. Boiling of Saturated Liquid

The general equation is given by

q/A = atJ.Tb

but a and b change depending upon the liquid and the temperature range.
Experimental data are required. McCabe and Smith (pp. 363 -366) report the
available equations, such as they are.

3. Suhcooled Boiling

Very little is known about this topic.

V. RADIATION HEAT TRANSFER

A. Introduction

Radiation is energy streaming through empty space at the speed of light. All
materials at temperature above absolute zero emit radiation. Thermal radiation
is radiation that is the result of temperature only. Since radiation moves through
space in straight lines, only substances in sight of a radiating body can intercept
radiation from the body. The fraction, of radiation hitting a body that is reflected
is called the reflectivity p. The fract.ion absorbed is called the absorptivity :x,
while the fraction transmitted,~1ed the transmissivity r. The sum of these
three fractions must be 1, or
:x+p+r=1

Of course, the maximum possible absorptivity is 1, and this is obtained only if

the body absorbs all incident radiation and transmits or reflects none. Such a
body is called a blackbody.

B. Emission of Radiation

1. Introduction

The radiation emitted by a material is independent of radiation being emitted by

other bodies in sight of, or in contact with it. Of course the net energy is the
difference between the energy emitted and that absorbed by it. With different
temperatur bodies in sight of each other within an enclosure, the hotter bodies
t'

lose energy by emission faster than they receive energy by absorption from
cooler bodies. Eventually the process reaches equilibrium when all bodies reach
the same temperature.
152 HEAT TRANSFER

2. Radiation Wavelength

Radiation of a single wavelength is called monochromatic, but actual radiation

beams consist of many monochromatic beams. The higher the temperature of
the body radiating, the shorter is the predominant wavelength emitted by it.
The monochromatic energy emitted by a surface depends upon the surface
temperature and wavelength of the radiation. The monochromatic radiating
power W is the monochromatic radiation emitted from unit area in unit time.
For the entire spectrum, the total radiating power W is So~ W'" dA.

3. Blackbody Radiation and Emissivity

A blackbody has the maximum attainable emissive power at any given tempera-
ture; it is the standard. Thus the ratio of the total emissive power W ofa body to
that of a blackbody, Wb • is defined to be the emissivity e of the bodies. So
G = W/Wb
Likewise, the monochromatic emissivity G", is defined by
G", = W;/Wb,A

A gray body is defined to be a body whose eA is the same for all wavelengths L

4. Practical Source of Blackbody Radiation

There are no actual blackbodies, but carbon black approaches it. There is a
basic blackbody relationship, the Stefan-Boltzmann law,
Wb =(JT 4 , (J=0.1713 x 10- 8 Btu/ft2hr OR4
with (J a universal constant. Planck's law gives the energy distribution in the
spectrum of a blackbody
C 1 = 3.742 X 10- 16 Wm 2
C 2 = 1.439 cmK
At any temperature T, the maximum monochromatic radiating power is ob-
tained at a wavelength Amax given by Wien's displacement law:
TAmax = 5200; T= OR, I. max = micrometers

C. Absorption of Radiation by Opaque Solids

1. Introduction

A definite fraction p of radiation falling on a solid body may be reflected and the
remaining fraction 1 - P enters the solid and is either transmitted or absorbed.
In most solids the transmissivity T is negligible. In fact, most of the absorbed
 RADIATION HEAT TRANSFER 153

radiation in an opaque solid is absorbed in a thin layer near the surface. Then
the heat generated by the absorption can flow into the remaining solid by con-
duction.

2. Reflectivity and Absorptivity oJOpaque Solids

An opaque solid's reflectivity depends mainly upon temperature and the type of
surface. Reflection is usually either specular or diffuse. Specular reflection makes
a definite angle with the surface, usually occurring on polished metals. Diffuse
reflection is reflected in all directions, usually the result of a rough surface.

3. Kirchhoff's Law

At temperature equilibrium, the ratio of the total radiating power of a body to

the absorptivity of that body depends only upon the body temperature. So for
two bodies 1 and 2,

with W = total radiating power and

CI. = absorptivity.
If one of the bodies is a blackbody
6 = W/Wb = CI.

that is, emissivity and absorptivity are equal when a body is at temperature
equilibrium with the surroundings.

D. Radiation between Surfaces

1. Introduction

The total radiation from a unit area of an opaque body of area AJ , emissivity 6 1 ,
and absolute temperature T is given by
q/A! = en! Ti
But, in the usual situation, the energy emitted by a body is intercepted by other
bodies (or substances) in sight of the body. And these substances are also radia-
tors and their radiation falls on the original body. So the net loss by radiation is
less than our equation indicates.
Consider the simplest case of two parallel surfaces 1 and 2 with both surfaces
black. The energy emitted by surface 1 is oT;. But all radiation from each of the
surfaces fa~h on the other surface and is completely absorbed. Thus the net loss
of energy by the first surface and the net gain of energy by the second is given by
154 UEA T TRANSFER

Usual real world situations are more complex, since one or both surfaces usually
see other surfaces and actual surfaces are rarely black.

2. Angle of Vision

Interception of radiation from an area element of a surface by another surface of

finite size can be seen in terms of the angle of vision, the solid angle subtended by
the finite surface at the radiating element.

3. Square of the Distance Effect

The energy from a small surface dA 1 that is intercepted by a large one depends
only upon the angle of vision and is independent of the distance between the
surfaces. But the energy received per unit area of the receiving surface is in-
versely proportional to the square of the distance between the surfaces, or
dI = (Wdnr2) dA 1 cos ¢
where I = rate of energy received per unit area of the receiving surface, and the
angle ¢ is determined as shown in Figure H5.

4. Quantitative Calculation of Radiation between Black Surfaces

The net rate of energy transfer dq 12 between two area elements dA 1 and dA 2 is
given by

where ¢1 = angle between dA 1 normal and line of sight,

¢2 = angle between dA 2 normal and line of sight, and
r= line of sight distance between areas.
The problem arises when we attempt to do the double integration for different
geometries. In any case, the resulting equation would be
q12 = aAF(Ti - Ti)
where q12 = net radiation between surfaces 1 and 2,
A = area of either of the surfaces, arbitrary choice, and
F = dimensionless geometric factor.

FIGURE US. Angle of vision.

 RADIATION HEAT TRANSFER

The F factor is called the angle factor or the view factor. If surface Al is chosen.

If surface A2 is chosen.

Thus

We can consider F12 as the fraction of radiation leaving area A1 that is inter-
cepted by area A 2 . If surface Al sees only A 2 , then F 12 = 1. If surface Al sees
several surfaces,

where F 11 occurs in case area 1 can see area 1. Regardless, the net radiation
associated with an F 11 factor is zero.
Many view factors have been calculated. See McCabe and Smith, pp.
384-386.

5. Nonblack Surfaces

If we have gray surfaces. the preceding equation becomes

where 3"";.2 and ff21 are overall interchange factors and depend upon 8 1 and 8 2 ,
There are several other cases that may be developed analytically; see McCabe
and Smith, pp. 387-389.

E. Radiation to Semitransparent Materials

1. Introduction

Many industrially important substances are transparent to the passage of

radiant energy. Glasses and many gases are in this category.

2. Radiation to Absorbing Gases

The fracti,)n of incident radialion absorbed by a given amount of gas depends

upon the radiation path length and on the number of molecules met during the
passage. Designing gas fired heaters. because of this, is a difficult task. For
methods, consult a heat transfer text.
156 HEAT TRANSFER

VI. HEAT EXCHANGERS

A. Heat Transfer Coefficients in Shell and Tube Exchangers

Tube side heat transfer coefficients cause no difficulty. But because of the shell
side flow patterns, shell side, or outside, coefficients must be approached differ-
ently.
The mass velocity Gb parallel to the tubes is the mass flow rate divided by the
free area for flow in the baffle window Sb (the portion of the shell cross section
not occupied by the baffle). This area is calculated by
_ r nD; nD~
Sb -.!b 4 - Nb 4

where j~ = fraction of shell occupied by baffle window (usually 0.1955),

Ds = shell inside diameter,
Do = tube outside diameter, and
N b = number of tubes in baffle window.
For shell side crossflow, the mass velocity Gc is based upon the area Sc for
transverse flow between the tubes in the row nearest the shell centerline. It can
be estimated from

Sc = PD{l _~o)
where p = center-to-center distance between tubes and
P = baffle pitch.
With the area determined. the mass velocity is calculated from

for parallel shell flow

and

for crossflow in the shell

The mass velocities are used with the Donohue equation

h~:o = 0.2(D~Qe)O.fi(Ct)O.33(1Jo.14

where Gc = jG~G:

B. LMTD Correction for Crossflow

When a fluid flows perpendicular to the tube bank, the usual log mean tempera-
ture difference (LMTD) only applies if either the inside or the outside tempera-
 EVAPORATORS

ture is constant. otherwise corrections must be used. These are obtained from
figures such as those in McCabe and Smith, pp. 405-407, where

are used to give FG to multiply the usual LMTD by, with

T ha , Thb = hot fluid, inlet and outlet temperatures
~a' ~b = cold fluid, inlet and outlet temperatures

VII. EVAPORATORS

A. Introduction

To concentrate a solution of a nonvolatile solute in a volatile solvent, evapora-

tion is usually preferred. For most evaporations the solvent is water. In an actual
evaporator, the liquid head over the tubes affects performance. Entrainment of
solids and the presence of noncondensables in the vapor stream can be a practi-
cal problem. In preliminary designs these effects are ignored. Heat losses are
always assumed negligible as are the superheat in the steam and the subcooling
in the condensate.

B. Evaporator Capacity

Capacity is defined as the amount of water vaporized per hour. To determine

capacity, the rate of heat transfer through the heating surface of the evaporator
must be known. The evaporator functions as a heat exchanger so the heat
transfer rate is given by
q = UA AT
The capacity is proportional to q if the evaporator feed is at the boiling tempera-
ture corresponding to the pressure in the vapor space. But, if the feed is colder,
some of the q must be used to heat the feed to the boiling point. This reduces the
heat available for evaporation and the capacity decreases. Increased capacity
occurs if the feed enters above the boiling point.

1. Boiling-Point Elevation

At a given pressure, the boiling point of aqueous solutions is higher than that of
pure water. This temperature difference is called the boiling-point elevation of
the solu'tion It must be subtracted from the temperature drop that is predicted
from the steam tables. Duhring's rule~the boiling point of a given solution is a
linear function of the boiling point of pure water at the same pressure~is used
for concentrated aqueous solutions.
158 HEAT TRANSFER

C. Evaporator Economy

Economy is the amount of solution water vaporized per pound of steam fed to
the evaporator. If the water vapor from one evaporator becomes the heating
medium (steam) for a second evaporator, the economy increases. How much it
improves can be determined by enthalpy balances.

D. Single-Effect Evaporation

Consider the evaporator shown in Figure H6. If we neglect steam superheat and
condensate subcooling the enthalpy balance for the steam chestside is
qs = lits(He - Hs) = -litsAs
where He = enthalpy of condensate,
Hs = enthalpy of steam,
lits = steam flow rate,
qs = heat transfer rate through heating surface to steam, and
As = latent heat of condensation of steam.
For the liquor side, the enthalpy balance is
q = (litf - Iit)Hv - litfHf + litH
where H = enthalpy of thick liquor,
H f = enthalpy of feed liquor,
Hv = enthalpy of vapor, and
q = heat transfer rate from heating surface to liquid.

Vapor
(ffit-ffi) Ib/hr
HV Btu/Ib

Feed Tf
!TIt Ib/hr
Hf Btu/Ib

Steam Condensate
IDs Ib/hr Ts
Hs Btu/Ib ins
He

Thick Liquor
in Ib/hr
H Btu/Ib FIGURE H6. Evaporator.
 PROBLEMS 159

If we assume no heat loss, q = - qs or

q = m),s = (mr - m)H, - mrH r + mH
Bear in mind that Hand Hr depend upon both temperature and concentration.
Perry's Chemical Engineers' Handbook has enthalpy concentration diagrams for
many aqueous systems. Once the enthalpy balance (and material balances) has
been completed, we note that
capacity = mr - m
economy = (mr - m)/ms

E. Multiple-Effect Evaporation

In multiple-effect evaporation, the first effect is heated with steam, but the subse-
quent effects are heated by the prior effect's vapor stream. But each effect uses
the single-effect material and enthalpy balances. Common practice dictates that
all of the effects have the same heating area. Trial-and-error calculations are
usually required. McCabe and Smith recommend the following calculation pro-
cedure:

• Assume unknown boiling temperatures.

• Using enthalpy balances, find steam and liquor flows between effects.
• Determine the needed heating surface A.
• If the areas are not equal, estimate new boiling temperatures.

To help in this procedure, it can be shown that in a multiple-effect evaporator

train, the temperature drops are inversely proportional to the heat transfer coef-
ficients.

PROBLEMS

A. We wish to design a kiln to withstand a temperature of 2000°F while limit-

ing the heat loss to 250 Btu/hr ft2 with an outside temperature of 100°F. We
have available the following types of brick:

k(hrBtuft2 ftOF) Thickness (in.)

Max Allowable
Temp (OF)

(a) Fin'day 0.90 4.5

(b) Insulating 0.12 3.0 1800°
(c) Building 0.40 4.0 300°
160 HEAT TRANSFER

What will be the minimum wall thickness and what is the actual heat loss
through the wall?
B. We wish to build an economical coldroom 20 ft square and 10ft high. The
outside temperature is 80°F and the room is to be maintained at OaF. The
boss has suggested parallel wooden outer and inner walls separated by a 6-
in. air space. What will be the heat loss through each wall? Our bright new
engineer suggests that the parallel outer and inner wall be divided in halfby
a thin (O.OOl-in.) sheet of aluminum foil. Should we fire her?
C. To cool a large slab of glass 4 in. thick from 400 to lOO°F (center tempera-
ture), we put it into a constant temperature bath at 50°F. How long must
the glass remain in the bath if:

D. We wish to heat 5000 lb/hr ft2 (based on minimum free area) of 50°F air
by passing it across an available tube bank that contains i in. OD tubes on
1t in. centers with square in-line arrangements of 12 rows in each direction.
If the tube walls are at 200°F, what wil~be-\he outlet temperature ofthe air?
E. Forty-five pounds per minute of water is heated from 60 to 180°F in a i in.,
16 BWG (Birmingham Wire Gage) heat exchanger tube. If the steam
condensing, at 260°F, on the outside of the tube gives an ho of
1000 Btu/hr ft2 OF, calculate the inside coefficient, the overall coefficient,
and the tube length required. After a half a year of operation, the overall
heat transfer coefficient decreased by half. What is the inside scale coeffi-
cient?
F. A tHn. Schedule 40 steel pipe is insulated with a 2-in. layer of asbestos
covered by a 3-in. layer of 85 %magnesia. If the temperature inside the pipe
wall is 300°F and the atmospheric temperature is 80°F, what is the heat loss
per foot of pipe and what is the temperature between the two insulations?

ksteel = 26, kasb = 0.087, kS5% = 0.034 Btu/hr ft OF

G. Our old 1-2 heat exchanger heats 55 gpm of water from 60 to 130 c F using
210 gpm of shell side hot water entering at 240°F. We estimate a dirt factor
of 0.002 and a dirty U 0 of 300. What were the two outlet temperatures when
the exchanger was first installed?
H. We wish to concentrate a colloidal solution from 15 to 65 % solids. The
specific heat of the feed, entering at 50°F is 0.93. The pressure of the satur-
ated steam to the evaporator is 10 psia; the condenser pressure is 5-in. Hg
abs. The steam chest has an overall heat transfer coefficient of
200 Btu/hr ft 2 OF. Water must be evaporated at the rate of 35,000 lb/hr. Ne-
glecting boiling-point elevation, what must be the heating surface required?
What is the steam consumption?
 PROBLEM-SOLVING STRATEGIES 161

PROBLEM-SOLVING STRATEGIES

A. This is a conduction problem with resistances in series. It is also an optimi-

zation problem, that is, the minimum wall thickness is required. But it
would be difficult to set up as a classical minimization problem. We shall
use

with R j = xj/k j A. Start with q = 250 and!! T across firebrick a minimum of

2000-1800° (to insure that insulating brick temperature is less than 1800°
everywhere). This determines a minimum Xl' The actual Xl will probably
be larger since Xl must be a multiple of 4.5 in. Continue in this way for each
type of brick, in order. Finally, knowing ail of the actual Xj, recalculate the
q.
B. This looks like a natural convection problem. But, with most natural con-
vection situations, radiation should also be considered. Natural convection
problems use the Grashof number. Radiation problems require a shape-
dependent equation.
C. This is an unsteady-state conduction problem. For simple shapes, analyti-
cal solutions arc available, but they are series solutions that are time con-
suming. The analytical solutions have been plotted in most textbooks. The
fact that

means that we can ignore convection from the surface.

D. This is a forced convection problem and we need a heat transfer coefficient.
Since heat transfer varies with tube arrangement, we must be able to find a
graph or correlation for this type of tube arrangement. If we can find it, we
have an easy problem.
E. This is a forced convection problem. We have enough information given to
calculate the q needed to heat the water. We must find a correlation for the
heat transfer coefficient for flow inside the tube-a Colburn equation. As-
suming no fouling, we can calculate U. Knowing U and q, with all tempera-
tures known, we can calculate the heat transfer area (and thus the tube
length) needed. If the overall heat transfer coefficien t decreases wi th time on
stream, it is the result of a fouling factor term in U. An easy problem.
F. This is a steady-state conduction problem with resistances in series. We can
assume that the outside temperature of the 85 ~< magnesia is at atmospheric
temperature and the inside pipe wall temperature is at the inside tempera-
ture. Because of the circular geometry, we must use log mean areas. This is
a straightforward, simple problem.
162 HEAT TRANSFER

G. This is a forced convection heat exchanger problem. We are given enough

information to calculate the q required to heat the cold water. Then we can
calculate the outlet hot-water temperature. Since we are given the U, we
can calculate the heat exchanger area. All of the foregoing wa., for presen:
operation. For the initial design conditions, the area is the same and the [.
is reduced by the dirt factor. So we know everything but the two outlet
temperatures. We have enough equations to calculate these temperatures.
but because of the log mean temperature, this would be a trial-and-error
solution.
H. This is a standard evaporator problem requiring only heat and materiai
balances. All the data needed are given and the solution should be straight-
forward.

SOLUTIONS

A. See Figure H7. Xl' x 2 , and X3 must be multiples of 4.5,3, and 4 in., respec-
tively.
Basis: 1 ft2 of surface area
We use the series formula

q = !1 T/L R, R = x/kA, A = 1 ft 2
q= 250(Btu/hr ft2)(1 ft2) = 250 Btu/hr

For firebrick (a), !1 T1 ~ (2000-1800°F) = 200°F

Btu ft 2 200°F .
Xl = 0.9 -hf ~ (1 ft ) 2 0 'h = 0.72 ft = 8.64 m.
rtF 5 Btu/ r

So we need two layers of firebrick (Xl = 2 x 4.5 in. = 9 in. = 0.75 ft).
Then
X /0.9 Btu ft
!1TI = ~kI q = 0.75 ft 250 Btu/hr ~h
f 2 0 - (1 ft2) = 208°F
IA , rtF

FIGt:RE H7. Schematic for solution A.

 SOLUTIONS 163

Then

Btu ft 2 1492 c F .
X2 = 0.12 - . - (1 ft ) 2-· /h = 0.716 ft = 8.59m.
hr ft~F 50 Btu r

So we need three layers of insulating (b), [X2 = 3(3 in.) = 9 in. = 0.75 ft].
Then

t-,.T = ~ = 0.75(250) = 1563 F 0

2 k2A q 0.12(1)

Now

t-,.T3 Z (2000°F - 208°F - 1563°F - lOO°F) = 129°F

X3 = 0.4(1)(129/250) = 0.206 ft = 2.48 in.
So we need one layer of building (c), [X3 = 1(4 in.) = 0.33 ft].
Now minimum wall thickness = 9 in. + 9 in. + 4 in. = 22 in. and actual
heat loss is

q = t-,.T/I R = (2000 _ 100t F/(0.75

0.9
+ ~.7~ + 0.333)
0.12 0.4
1900 Btu
= -=240--
7.916 hr ft2

B. See Figure H8. This is a natural or free convection problem-and possibly

a radiation problem. Gr is the Grashof number.
F or convection

qc = hA t-,.T A = 10 X 20ft 2 = 200ft 2

t-,. T = 80-0°F = 80°F

Vb = 0.0317 Grg.37 (Kreith, Eq. 7-36)

kr

_ --- b ~ 6 in.

20 ft
FIGURE H8. Schematic for solution B.
164 HEA T TRANSFER

For radiation
aA(Ti - T~)
q = -- - (McCabe and Smith, Eq. 14-39)
r (1/10 1 + l/e z - 1)
qppZ 6 1 .
at ~ 40°F, -Z = 3.16 x 10 ~f:3 (KreIth),
f-I. of t
k = 0.0140 Btu/hr ft°F
Grb = (3.16 x 106;oF ft 3 )(80°F)(! ft)3 = 3.16 X 10 7
V = h = 0.0317(3.16 X 10 7)°.37 (0.0140 Bt~) = 0.5286 _~!y__
(t ft) hr ft°F hr ft2 of
qc = 0.5286 Btu/hr ft2 of (200 ftl)(80°F) = 8457 Btu/hr
10 1 = £2 = 0.93, (Kreith Appendix), T\ = 460 + 80 = 540 o R,
T2 = 460 +0= 460 R 0

-8 Btu
a = 0.1714 x 10 hrft20R4

qr = 0.1714 x 10- 8 (200)/'(_1- + _1 - - 1)(5404 _ 460 4) = 11,994 !!.~u

0.93 0.93 hr
q = qc + qr = 20,451 Btu/hr from one wall
For Alfoil £1 = 0.93, £2 = 0.04, (Kreith), b = t(6 in.) = 3 in. = i ft. Assume
physical properties to be about the same; AT = !(80°F) = 40°F.
Gr h = (3.16 x 10 6 ;oF ft 3 )40°F(i ft)3 = 1.975 X 10 6
V = h = 4[0.0317(1.975 X 10 6 )°.37 x 0.0140] = 0.3790
qc = 0.3790(200)(40) = 3032 Btu/hr, Tl = 40 + 460 = 500 0
R
T2 = 0 + 460 = 460 0
R

qr = 0.1714 x 1O-8(200)/(0.~3 + 0.~4 - 1)[(500)4 - (460)4] = 242~t;

q = qc + qr = 3032 + 242 = 3274 Btu/hr for one wall
Use the new engineer's idea and give her a raise.
e. This is an unsteady-state conduction problem in an infinitely large slab.
Kreith, p. 166, has charts for this type of problem.
We need the following parameters: thickness = 2L = 4 in., so L = 2 in.;
x/L = 0/2 in. = 0 centerline; Biot modulus hr m/k, ~ 00; Fourier modulus
aO/L 2 , a = thermal diffusivity = k/Cpp.
TX/L =
To - Teo
Tcyc = .!.OO°F - 50~F
400°F - 50°F
= _5~ = .!. = 0.143
350 7
a8
[) = 0.9, 8 = 0.9(~.i~' __ )2 hr ft°F 0.]8 B~u~6Q ~
12 Ill./ft 0.36 Btu IboF ft 3
8 = 2 hr
 SOJ,UTIONS 165

D. q = hA L\ I;m = mCi r;, - T.), so to get To we must first find the heat transfer
coefficient. Kreith has a section on tube banks, pp. 415-425. Pr is the
Prandtl number.
From his equations,
min free area) _ _
Amin
( unit length - Sr Do

where Sr = distance between centers.

Basis: 1-ft length of bundle

A . = 1.5 in. - 0.75 in. = 00625 [ 2/f Do = i/12 = 0.0625 f1

min 12 in./ft . t t, Ao = 0.1963 ft 2 /ft
and

hDo = 0.33(G max Do)0.6 Prr 3 1'f GmaxDo

----;:;:. 6000 , (Kreith, Eq. 9-10)
kr /If J1r

where f says use mean film temperature, that is,

t[1(r;, + 7;) + TwJ

Gmax -_ 5000 lb hr _ 8~-
2 - 1.3 9 2
hr ft 3600 sec ft sec
Assume To = 100°F, then
If = ![!(100 + 50) + 200J = 137SF

From Kreith, Appendix A-3,

Pr = 0.72, /lr = 1.34 x 10 - 5 Ib/ft sec,

k = 0.0162 Btu/hr feF, C p (75°F) = 0.24 Btu/lboF

GmaxDo = 1.389 ~ (0.0625 ft) ft sec = 6479

/If ft2 sec 1.34 x 10 - 5 lb

0.6 0.3(0.0162 BtU) 1 Btu

h = 0.33(6479) (0.72) hr ft0F 0.0625 ft = 15.0 hr ft 2 of

ft2) 5000lb lb
As = 144 ( 0.1963'ft = 28.67 ft2, m = - hr ft (0.0625 ft)12 = 3750 hr

= _~OO- 50) - (200 - 100) = 150 - 100 = 123.3 F 0

L\1im In [(200 - 50)/(200 - 10)] In (150/100)

So

q = 15.0 hr ~:~OF (28.67 rtl)(123.3°F) = 3750 l~~F (0.241!~~}r;, - 50tF

166 HEAT TRANSFER

or
7;, = 58.9 + 50 = 108.9°F
Second approximation:

'4 = 1[1(108 + 50) + 200J = 140°F

So physical properties remain the same

(200 - 50) - (200 - 108) 150 - 92

To = -- = - = 118.7°F
1m In [(200 - 50)/(200 - 108)] In (150192)
q = 15(28.67)(118.9) = 3750(0.24)(7;, - 50)

or
7;, = 56.8 + 50 = 106.8,

E.

Initially, hdi = 0; 16 BWG, (Kreith, Appendix Table A5), Nu is the Nusselt

number,
Do=iin., Di = 0.495 in., Xw = 0.065 in.
A c• = 0.0013 ft2,
A,o = 0.1636 ft2 1ft,
Asi = 0.1296 ft2 1ft
Btu
Dim = log mean D, kw = 26 ~h f
rtF
assumed steel
0

- i- 0.495 .
D\-m = In (ij0.495) = G.55? m.,

Evaluate H 2 0 properties at T = (60 + 180°F)/2 = 120°F (Kreith, Appen-

dix Table A3),
Btu Btu
Pr = 3.81, k = 0.372 hr ft0F' C p = 1 1boF '

p = 0.392 X 10- 3 Ib/ft sec

GD
Re=- G= 451b ~=577~-
p , min(O.OOt 3ft2) 60 min ft 2 sec
Re = 577 Ib ft sec 0.495 in.
ft2 sec 0.392 x 10 3 Ib 12 in./ft = 60,717

h·D·
Nu = ~ = 0.023(60,717)°·8(3.81)1/3 = 241
 SOLUTIONS 167

h. = 241 _~ = 241 ().3_72_Btu 12 in·/ft = 2173 _B--:tu_

I Di hr ft°F 0.495 in. hr ft3 OF

~i =
1
21 73 + C~~~~)(f2)(~:::~) + C;00}!)(0.495)
= 1.437 x 10- 3
U i = 696 Btu/hr [t 2 of
q = U i AjA7;m = mCp(I;, - 7;);
A7;m = [(260 - 60) - (260 - 180 /ln (.(260 - 60l ) n = 131°F
(260 - 180)
45lb
m= mm
-.-,

q = (45 ~~n )C~:~ )(180 - 60tF

Btu ( ft2) L ft 131°F

= 696- - 0.1296 - .
hr ft 2 OF ft 60 mm/hr

So L = 27.4 ft. With scale

1
U i = 1(696) = 348, -1
+ 1.4373
x 10- = - ,
hdi 348
1
. . = 1.437 X 10- 3 or hdj = 696 Btu/hr fe OF
hdi

F. Kreith, Appendix Table A6,

Do = 1.660 in., Di = 1.380 in., Xw = 0.140 in.

For resistances in series,
q = AT/I.R
For hollow cylinders,

Basis: 1 ft of length

then A = IID(1 ft) = lID

. Xw - Do - Di 2 Di = 1.380 in.
pIpe wall Rw = -, Aw = II - - - . = 0.3968 ft wall .
ksAw In (DoIDJ Do = 1.660 m.
Rw = 1.1308 x 10 - 3
168 HEAT TRANSFER

asbestos

D j = 1.660 in .
.ita = rr po - Dj)a = 0.8537 ft2 asbestos Do = 5.660 Ill.
.
In (Do/Dj)
Ra = 2.244

132 ft
magnesia R =--
m 0.034A m '

Am = rr Po - Dj)m
. D
= 2.1734ft 2 magnesJa j = 5.660 in.
.
In (Do/DJ Do = 11.660 Jll.
Rm = 3.3832
300 - 80
q = 1.1308---::: TO--3.}-2.2440+i3832

220
5.6283
Btu
= 39.1 - h·· per ft of length
per r
300 - T 300 - T
39.1 =---------.
1.1308 x 10- 3 + 2.244 2.2451'

G. See Figure H9. We need the F G factor from McCabe-Smith, Fig. 15.6a

Tha - I;,b
Z = .. - - - - ,
Tcb - T.:a

Z = ri1cCPE
mHCpH
q = U oA oFG,1.I;m = mHCph(Tha - I;,b) = mcCpc(T.:b - T.:a)

Tcb -----'--~
Tca~l)
tr: Thb FIGURE H9. Schematic for solution G.
 SOLUTIONS 169

To get C p values, we need an estimate of Thb :

mH(240 - Thb ) = mc(130 - 60)

Thb = 222°F, Th = 230°F, T.: = 95°F

Kreith, Appendix Table A3

C ph = 1.005, Cpe = 0.9975 Btu/lboF,

P240 = 58.9 Ib/ft 3 , P60 = 62.3 Ib/ft 3
me = ?5 ga!_JJ:!=_ 62.31~ = 458 1 .~
min 7.48 gal ft3 . min
= 2.749 x 104 lb/hr
. 210 gal 1 ft3 58.9 Ib lb
mH = _._- - - - - - = 1654--
min 7.48 gal ft3 min
= 9.924 x 104 lb/hr
qe = 2.749 x 104 (0.9975)(70) = 1.919 x 10 6 Btu/hr;
qh = 9.924 x 104 (1.005)(240 - Thb )
240 - Thb = 19.24, Thb = 220.8°F

11 7;m = [(220.8 - 60) - (240 - 130)]/ln [(220.8 - 60)/(240 - 130)] = 133.8

Z = 0.274, FG = 0.99, 11 = 0.389

So our exchanger has an Ao of

q 1.919 X 106 2

VJ;l1 7;m - 300(0.99)(133.8) = 48.3 ft

Now from Kreith, p. 505

So

We need to find Thb and T.:b with this new Vo

(1) q ~ UoAoFGl1l;m = 750(48.3)(O.99)117;m = 35863117;m

(2) qe = mCCpc(T.:b - 60) = 2.749 X 104 (0.9975)(T.:b - 60)
= 27421(Tcb - 60)
170 HEAT TRANSFER

(A)
(2) = (3) Thb = (256.5 - 0.2749~b)
(B)
a b
_ [(196.5 - 0.2749~b) - (240 - ~b)J
(1) = (2) Tcb - 1.3079 I
n (a /b ) + 60

(C) (A) a b (B)

'I;,b
~bCF) (256.5-0.2749~b) (196.5-0.2749~b) (240 - ~b) ~b (B) - (C)

140 218.0 158.0 100 225.8 85.8

160 212.5 152.5 80 207.0 47.0
180 207.0 147.0 60 187.0 +7.0
184 205.9 145.9 50 182.8 -1.2
183 206.2 146.2 57 183.9 +0.9
183.5 206.1 146.1 56.5 183.3 +0.2

H. See Figure H 10.

Q + mrCp(If - T) = (mr - m)Hv + mCp(T - T)
T = 32, mr - m= 35,000 lb/hr
C p = 0.93 Btu/lboF
If = 50°F, T = 133.8°F
Enthalpy balance:

Q + 0.93(50 - 32)mr = 35,000 x 1119.4 + 0.93(133.8 - 32)m

Solids balance:
0.15mr = 0.65m, mr = (0.65jO.15)m
Total balance:
0.65
mr = 35,000 + m or 0.15 m = 35,000 + m

0.65 - 0.15 35,000

--0-
.15
m= 35,000, m= 3.333 = 10,500
 SOLUTIONS 171

Hy = 1119.4 Btu/hr

Feed 5" Hg abs

15% solids 133.8°F
50°F

35,000 Ib/hr
Steam
P = 10 psia H:P
T = 193.2°F OJndensate
lIHy = 982.1
Btu/lb

65% solids
FIGURE HI0. Schematic for solution H.

Then
me = 45,500
Q = 35,000 x 1119.4 + 0.93(133.8 - 32)10,500 - 0.93(50 - 32)45,500
= 39,411,407 Btu/hr
tlt = 193.2°F - 133.8°F = 59.4°F

A = ~ = 39,411,407 = 3317 ftl

VAt 200 x 59.4
. 39,411,407 Btu/hr
steam reqUired = = 40,130 lb/hr
982.1 Btuj1b
 CHEMICAL KINETICS

1. I n trod uction 174

II. The Basics of Kinetics 174
A. Definition of Reaction Rate 174
B. Single and Multiple Reactions 175
C. Molecularity and Order 175
D. Rate Constant k 176
E. Representation of a Reaction Rate 176
F. Material Balance in Complex Reactions 178
G. Temperature Dependency from the Arrhenius Law 179
III. Interpretation of Batch Reactor Data 179
A. Introduction 179
B. Constant-Volume Batch Reactor 180
1. Irreversible Unimolecular-Type First-Order Reaction 180
2. Other Reaction Types 181
C. Variable-Volume Batch Reactor 181
1. Integral Method of Analysis 183
IV. Single Ideal Reactors 183
A. Single Ideal Batch Reactors 183
B. Steady-State Backmix Flow Reactor (CFSTR) 184
1. Space-Time and Space-Velocity 186
C. Steady-State Plug Flow Reactor 186
Problems 188
Problem-Solving Strategies 189
Solutions 191

o. Levenspiel, Chemical Reaction Engineering, 2nd ed., Wiley, New York, 1972.

173
174 CHEMICAL KINETICS

I. INTRODUCTION

We shall discuss the basic principles of kinetics and reactors. Only topics likely
to be of use on the test will be covered. The reference on p. 173 ~ Carberr: or
Smith texts in McGraw-Hili's Chemical Engineering Series are also good ill
should be consulted for points not discussed here. Remember that there will
probably be more kinetics problems than any other type.

II. THE BASICS OF KINETICS

A. Definition of Reaction Rate

In homogeneous reactions all reacting materials are found within a single phase
gas, liquid, or solid. In addition, if the reaction is catalytic, the catalyst must alsc
be present within the same phase as the reactants and products. Although there
are a number of ways of defining the rate of reaction, only one such measure is
used in homogeneous systems. This is the intensive measure of rate based on
unit volume of reacting fluid or system.
Consider a reaction in which substances A and B react to form C. (It is irrele-
vant and confusing to put down the chemical equation at this point.) Let N a , N b •
and N c be the number of moles of these substances present at any moment in a
system of fixed mass and of volume V. If the system is uniform in composition
- as we shall suppose - the concentrations of the various substances are

[A] = NjV; [C] = NjV

The reaction rate r is the rate, per unit volume of the reacting fluid, at which an:
one of the substances, chosen as a key substance, is being produced or destroyed
by reaction. Let it be the product C which is so chosen. Then if-and only
if - there is no process involving C other than the reaction, the rate r may be
written

r = -1 dN
- c = 1 d(V[C])
-~- ---~

V dt V dt

This equation would not be applicable if substance C were being continually

introduced into or removed from the given volume of a physical process. The
equation as it stands applies only to closed systems and it does not apply, for
example, to a CFSTR (continuous flow stirred tank reactor). In the latter the
substance C is being continually removed through the outflow and dN edt (but
not r) is zero at the steady state of the system.
It would be expected that the variables that affect the progress of homoge-
neous reactions are the composition of the materials within the phase as well as
the temperature and pressure of the system. The container shape, surface prop-
erties of solid materials in contact with the phase in question, and the diffusional
 THE BASICS OF KINETICS 175

characteristics of the fluid should not affect the rate of homogeneous reaction for
component C:
r = !(temperature, pressure, composition)

The variables temperature, pressure, and composition are interdependent, since

the pressure is fixed given the temperature and composition of the phase. So we
can write without loss of generality
r = !(temperature, composition)

B. Single and Multiple Reactions

Before finding the form of the concentration term in a rate expression, we shall
distinguish between a number of types of reactions. This distinction is based on
the form and number of kinetic equations used to describe the progress of a
chemical reaction. Since we are concerned with the concentration-dependent
term of rate equations for the time being, we assume in the following discussion
that the temperature of the system is kept constant.
When materials react to form products it is usually easy to decide after exam-
ining the stoichiometry, preferably at more than one temperature, whether we
should consider a single reaction or a number of reactions to be occurring.
When a single stoichiometric equation and single rate equation are chosen to
represent the progress of the reaction, we have a single reaction. When more
than one kinetic expression is needed to follow the changes in composition of all
components of the reaction we have multiple reactions.
Multiple reactions are classified as consecutive or series reactions,
A---+R--->S
parallel, competing, or side reactions,
A--->R
or
A ---+ S
and mixed reactions

The mixed reactions illustrated are parallel with respect to B and consecutive
with respect to A, R, and S.

C. MoJecuJarity and Order

The moleclllcHity of an elementary reaction is the number of molecules involved

in the rate-determining step of a reaction. Molecularity of reactions has been
found to be one, two, and occasionally three. Needless to say, the molecularity
refers only to an elementary reaction and can only be in whole numbers.
176 CHEMICAL KINETICS

Often we find that the rate of progress of a reaction, involving materials A,

B, ... , D, can be approximated by an expression of the following type:
r. = kC:Ct .. · C~, a+b+ .. ·+d=n
where the a, b, . .. are not necessarily related to the stoichiometric coefficients.
We call the powers to which the concentrations are raised the order of the reac-
tion. Thus the reaction is
ath order with respect to A
bth order with respect to B
nth order overall
Since the order refers to the empirically found rate expression, they need not be
whole numbers, but the molecularity of a reaction must be in terms of whole
numbers since it refers to the actual mechanism of the elementary reaction.

D. Rate Constant k

When a rate expression for a homogeneous chemical reaction is written in the

form shown in the previous equation, the units ofthe rate constant k for the nth-
order reaction are
(time) -l(concentration)l-n
which becomes, for a first-order reaction, simply
(time)-I

E. Representation of a Reaction Rate

In writing a rate equation we may use any measures equivalent to concentration,

such as the partial pressures of the components. So
r A = kp:p~ .. · p~

Whatever measure is used leaves the order unchanged; but it will affect the units
of the rate constant k.
For shortness, elementary reactions are often represented by an equation
showing both the molecularity and the rate constant. Thus

2A~2R
r~presents a bimolecular irreversible reaction with second-order rate constant k 1
implying that the rate equation is
-r. = rr = ktC;
It would not be proper to write the stoichiometric equation for this reaction as
 THE BASICS OF KINETICS 177

for this would imply that the rate expression is

We must be careful to distinguish between a stoichiometric equation that can be

multiplied by any constant and the equation that represents an elementary reac-
tion.
Even writing the equation correctly may not be sufficient to avoid ambiguity
for some reactions. For instance, consider the reaction
k2
B + 2D ...... 3T

If the rate were measured in terms of B, the rate equation is

But if it refers to D, the rate equation is

Or if it refers to the product T,

But from the stoichiometry

so

To which of these primed k2 values are we referring in the stoichiometric equa-

tion? We must be told.
There is an easy way around this problem, but the method is not widely used.
First we set the stoichiometric equation up with all quantities on one side, the
product stoichiometric coefficients being positive, those for the reactants nega-
tive. The previous stoichiometric equation would be written
3T - IB - 2D = 0
The rate would be defined as
1 dN i
r=
:Xi V dt

where the :Xi is the stoichiometric coefficient of the ith component. Thus the rate
would always be a positive quantity. The rate expression corresponding to this
would becomt.
2 1 dN t 1 dNb 1 dNd
r=k CC = --- ----
2 b d 3 V dt V dt 2V dt
178 CHEMICAL KINETICS

F. Material Balance in Complex Reactions

In simple cases, the relations between the conversions of the several participan ts
can be found easily by inspection, but in complex ones a systematic procedure is
helpful. The reactions of a group can be numbered, and the initial and final
amounts of the participants identified for each reaction. For those substances
that participate in more than one reaction, the initial amount for a particular
reaction is the same as the final amount from the last preceding reaction in
which this substance participated. It is convenient to subscript the final amount
from any reaction with the number of that reaction. Then all the possible mater-
ial balances are written down, between initial, intermediate, and final amounts.
The balances may then be combined and rearranged to elimipate the interme-
diate amounts, leaving only the initial and final amounts for the complete reac-
tion.

For example,

Reaction 1: A+B-+C+E
Reaction 2: A+C-+D+E
Reaction 1, initial amounts: Ao, Bo, Co, Eo (no D involved)
Reaction 1, final amounts: AI' B, C b EI
Reaction 2, initial amounts: AI, C I, Do, E1 (no B involved)
Reaction 2, final amounts: A,C,D,E
Reaction 1, balance: Ao - A1 = Bo - B = C 1 - Co = E1 - Eo
Reaction 2, balance: A1 - A = C 1 - C = D - Do = E - E1

Using algebraic elimination of the quantities with subscript 1, the material

balance is expressed in these three relations:
Ao - 2Bo - Co = A - 2B - C
Bo + Co + Do = B + C + D
Ao + Eo = A+ E
As another example, consider

Reaction 1: 2A -+ B
Reaction 2: 2B -+C
Reaction 1, initial: Ao, Bo
Reaction 1, final: A,B 1
Reaction 2, initial: B1 , Co
Reaction 2, final: B,C
Reaction 1, balance: ~(Ao - A) = B1 - Bo
Reaction 2, balance: !(B1 - B) = C - Co
 INTERPRETATION OF BATCH REACTOR DATA 179

Upon eliminating B), the final balance is

Ao - A + 2(Bo - B) = 4(C - Co)

G. Temperature Dependency from the Arrhenius Law

For many reactions, and especially elementary reactions, the rate expres-
sion may be written as a product of a temperature-dependent term and a
composition-dependent term, or
r i = !)(temperature)'!2(composition)
= k· !2(composition)

For these reactions the temperature-dependent term k, the reaction rate con-
stant (velocity constant), has been found in most cases to be well represented by
Arrhenius' law:

k = ko ex p ( - RET)

where ko is called the freq uency factor and E is called the activation energy of the
reaction. This equation fits experimental data quite well over wide temperature
ranges and is a first approximation to the temperature dependency derived
according to several different theories.

III. INTERPRETATION OF BATCH REACTOR DATA

A. Intrpduction
..-----./
A rate equation characterizes the rate of reaction. The value of the constants of
the equation can only be found by experiment. The determination of this rate
equation is usually a two-step procedure: First the concentration dependency is
found at fixed temperature, and then the temperature dependence of the rate
constants are found.
The equipment used to obtain rate data are of two different types: the batch
and the flow reactor. The batch reactor is simply a container to hold the con-
tents while they react. An experimental batch reactor is usually operated
isothermally and at constant volume because of the ease of interpretation of the
results of such runs.
There are two procedures for analyzing experimental kinetic data, the in-
tegral and the differential methods. Tn the integral method. we select a kinetic
model with it, corresponding rate equation and, after appropriate mathematical
integrations and manipulations, predict that a plot of C versus t data on specific
x versus.\' coordinates should yield a straight line. The data are plotted and if a
good straight line results. we accept the assumed kinetic model.
180 CHEMICAL KINETICS

For the differential method of analysis we select a kinetic model and fit its
corresponding rate expression to the data directly. But since the rate expression
is a differential equation, we must first find (l/V)/(dN/dt) from the data before
the fitting procedure is attempted.
The integral method is easy to use when fitting relatively simple mechanisms
or when data are so scattered that we cannot reliably find the derivatives needed
for the differential method. The differential method is more useful for the more
complicated mechanisms but it requires accurate data, as well as more data.

B. Constant-Volume Batch Reactor

The constant volume refers to the volume of the reactor taken up by the reactiQll
mixture. This is thus a constant-density reactor. Most liquid-phase reactions as
well as all gas-phase reactions occurring in a constant-volume bomb fall into
this category.
With a constant-volume reactor, the measure of the reaction rate becomes
1 dNi d(C i V) 1 Ci dV + V dC i dC i
r·=--=---=-------=-
I V dt dt V dt dt

or for ideal gases

1 dpi
r·=--
I RT dt
These measures of rate can be followed directly in most systems. Because of the
ease of interpretation of such data, the constant volume system is used whenever
possible, even though the plant built for the commercial application of the reac-
tion will probably be a constant-pressure system.

1. Irreversible Unimolecular- Type First-Order Reaction

Suppose for the reaction

A -> products
we wish to test the first-order rate equation

_ dC. = kC
dt a

Separating and integrating we obtain

Ca
-In- = kt
CaO

Now, by definition, the fractional conversion X a of a given reactant is defined as

the fraction of the reactant converted into product or
Na = N ao (1 - Xa)
 INTERPRETATION OF BATCH REACTOR DATA 181

where NaO is the starting moles of A. Fractional conversion (or just conversionl
is often used in place of concentration in engineering work; thus, most of the
following results will be given in terms of both C a and Xa' Now

So in terms of conversion, the first-order reaction equation becomes

-In(! - Xa) = kt

2. Other Reaction Types

Equations for other orders of constant-volume reactions are shown in Table K 1.

c. Variable-Volume Batch Reactor

The general form for the rate of change of component i in either the constant-or
variable-volume reaction system has already been given as

1 dN 1 d(C V) 1 V dC i + C i dV
r· = - - - i = - - -i - = --- - - - --- -
I V dt V dt V dt

or

We see that two terms must be evaluated from experiment if r i is to be deter-

mined. Fortunately, for the constant-volume system the second term dropped
out, leaving only the simple expression

dC i
r· = - -
I dt

In a variable-volume reactor, we can also avoid the use of the unwieldy two term
expression if we use fractional conversion instead of concentration and if we are
willing to assume that the volume of the reacting system varies linearly with
conversion, or

\\here e. is the fractional change in the volume of the system between no conver-
sion and complete conversion. Then

VeX. = 1) - VeX. = 0)
&. = - - - - ---- -
V(Xa = 0)
182 CHEMICAL KINETICS

TABLE Kl. Some Rate Equations

Reaction Order Rate Concentration/Time Equations·

A->B 0 -dCaldt = k Ca = CaD - kt; T = C aO /2k

A->B 1 -dCaldt = kC a In(Ca/C ao ) = -kt; T = (In 2)lk
A+A-->P 2 -dC.ldt = kC; (l/Ca ) - (l/C ao ) = kt; T = l/(k Cae
aA + bB
->P 2 -dC.ldt = kC.C b In(Ca/C b ) - In(CaO/CbO )
= [(beao - aCbo)/aJkt;
T = a/k(bC ao - aCbO) X
In [aC bo /(2aC bO - bCao)]
-dCa/dt = kC. - k'C b In [(CaO - A')/(C a - A')]= kt;
k = k(K + l)!K; K = k/k';
A' = (CbO + Cao)/(K + 1)
A+B<=l:
2 In {~ + 2cF/(b -=-Jf)} = Jft;
1 + 2cF/(b + fi)
a = k[CaOC bO - CeoCdo/K];
C = k (1 - l/K);
b = - k[(C ao + C bO )
+ (CeO + Cdo)/K]
f = b2 - 4ac; F = CaO - C a
a T is the half-life for the reactions.

To illustrate this point, consider the isothermal gas-phase reaction

A --> 4R

if we start with pure A,

4-1
t;a = - = 3
1

If we had started with 50 ~~ inerts present, two volumes of reactant mixture (I

pure A, 1 inerts) yield upon complete conversion five volumes of product mix-
ture (4 R, 1 inerts), so
5-2
13=--=15
a 2 .

We thus note that Ga accounts for both the reaction stoichiometry and the pres-
ence of inerts.
Now since
 SINGLE IDEAL REACTORS 183

We obtain

or
Ca 1 - Xa
CaO 1 + BaXa

This is the conversion-concentration relationship that we shall use if we assume

linear volume change. This assumption is quite good for isothermal constant-
pressure systems (if there are no series reactions). The rate expression with the
linearity assumption becomes

1 dNa
-r = --~= - - - - 1- - - -NaO
- -d(l
-- - Xa)
--
a V dt Vo(1 + BaXa) dt

= __~ dX a
1 + BaXa dt

1. Integral Method of Analysis

The integral method of data analysis of isothermal variable-volume reactors

requires that we integrate the rate expression we wish to test. Then the C versus t
function obtained is compared with our experimental data. So for reactant A

1 dNa C. o dX a
-r = --~=-----
a V dt 1 + BaX. dt
or, upon formal integration

t = C. O IX. __d_X_a=---__
o (1 + Ba X ,)( -r.)
This is the expression to be used for all batch reactors for which we can assume
that the volume varies linearly with conversion.

IV. SINGLE IDEAL REACTORS

A. Single Ideal Batch Reactors

We make a material balance for any component A. (We usually select the limit-
mg compOl,ent.) In a batch reactor, since the composition is assumed uniform,
'ilt-e make the balance around the whole reactor. Thus we have

production of A ~ accumulation of A
184 CHEMICAL KINETICS

since there is no input or output. Now

production of A = ra V(moles A formed/time)

I . dNa d dX a
accumu atlOn of A = dt = dt [N ao (1 ~ Xa)] = ~N.o dt-

so

or, as it is usually written

dX a
( ~r)V=N --
a .0 dt

which integrates to

This is the general equation giving the time required to get a given conversion of
reactant A for either isothermal or nonisothermal operation. We leave both V
and ( ~ra) under the integral since, in general, they both vary as the reaction
proceeds.
Of course, for a number of situations the integral simplifies. If the volume of
the mixture remains constant we have

F or all reactions in which the volume of the mixture changes linearly with con-
version, we have

B. Steady-State Backmix Flow Reactor (CFSTR)

Since the composition is uniform throughout the tank (by definition), we make
the material balance over the whole tank. So, if we consider reactant A,

input-output + production by reaction = 0

 SiNGLE IDEAL REACTORS 185

If Fao= Co CaO is the molar feed rate of component A into the reactor shown in
Figure Kl, we have

input of A = Fao(l - XaO) = Fao mOl/time

output of A = Fa = Fao(l - Xa) mol/time
production of A = ra V mol/time

or

which reduces to

Rearranging,

~=_V_=~
FaO voC aO (-ra)
or
V CaOXa
!=--=-= - - -
S vo (-ra)
Note that Xa and (-ra) are evaluated at exit stream conditions, which are the
same as the conditions within the reactor (by definition). If the feed enters the
reactor partially converted, we could rewrite the equation as

!"_ = ~ = Xa,"--- X ai
Fao roC aO (-ra)f
V VC aO CaO(X af - X ai )
r=-=-=--
S Vo FaO (- ra)r
Thus the design equation for this reactor shows that knowing any three of the
four terms X., (-r.), V, or Fao gives the fourth term directly. So in kinetic
experiments, each run at a given r will give a corresponding value for the rate of

Caf=Ca
xaf=xa
(-ralf=(-ral
Fa FIGURE Kl. Stirred tank reactor.
186 CHEMICAL KINETICS

reaction directly. For this reason, the backmix reactor is frequently used [('II"
kinetic studies.
1. Space-Time and Space-Velocity

A good way of expressing the relation between the feed rate Fao and reactor
volume V in a flow system is to apply the terms space-time and space-velocit)
. 1
space-tIme = r = ~
s
time required to process one reactor volume of feed .
= = time
measured at specified conditions
. 1
space-ve I oClty = s = -
r
vol of entering feed at specified conditions/time . _1
=- - --=(tlme)
void vol of reactor
We usually select the conditions as those of the entering feed, thus

r = ! = ~aO J-" = [(mOl A entering)(VOI of reactor)J/(mOI A entering)

s FaO vol of feed (time)
V reactor vol
vo volumetric feed rate

C. Steady-State Plug Flow Reactor

In a plug flow reactor the composition of the fluid varies from point to point
along the flow path; thus the material balance for a component must be made
for a differential element of volume dVas shown in Figure K2. So for component
A we have
input-output + production by reaction = 0
For a differential volume of dV,
input of A = Fa mol/time
output of A = Fa + dF a mol/time
production of A = ra dV mol/time

dV

FIGURE K2. Plug flow reactor.

 SINGLE IDEAL REACTORS 187

So
Fa - Fa - dF a + ra d V = 0
but
Fa = FaoO - Xa)
then
dF a = -FaO dX a
so
FaO dX a = - ra dV

This is the balance over a differential slice of the total volume. For the whole
reactor, we must integrate

1 IV V Ix_r dX
FaO 0 dV = Fao = 0 (-r:)
or
1 V Ix_r dX
r = s=~= CaO 0 (-r:)

);otice that for the plug flow reactor ( - ra) varies while for the backmix reactor
it is constant. If the feed enters partially converted, the equation would be

The design equations for all reactor types are summarized in table K2.

TABLE K2 Ideal Reactor Design

Equations

Reactor Type General Equation

Batch a t=N
a
0 f---
dXa
V( -raj

Plug flow T=C

aO
f
dX a
-
-ra

Stirred tank

a For changing volume, use

v= Vo(1 + £,Xol
188 CHEMICAL KINETICS

PROBLEMS

A. For the reaction stoichiometry

A + B -> R,

a batch reactor gave the following test data:

(min)

Time 0 10 20 30 40 50 60 70

C a (g mol/L) lAO 0.89 0.64 OA7 0.37 0.29 0.23 0.19

C b (g mOl/L) 3.00 2A1 2.07 1.80 1.61 1A5 1.32 1.22
-'-
~

If you can show that the rate equations

are consistent with these data. what arc the values for k 1 and k2?
B. A gas decomposes at 900°C according to

2A(g) -> 2R(g) + S(g)

with a rate constant of 1000 cm 3 /g mol sec using pure A.
(a) How long will it take to react 80 °0 of A in a batch reactor at 900°C
and 1 atm?
(b) What plug flow reactor volume, at 900 c C and 3 atm. will be required
to react 80 % of A with a feed rate of 500 lb/hr? Molecular weight of
A = 45.
C. We wish to produce D by reacting A over a solid catalyst:

A(g) <=2 D(g) + E(g)

Running at 0.2 atm and 1000 F, how many pounds of catalyst are needed
to get a 30 % conversion of A using a feed of 10 lb mol of pure A per hour'?

PdPe) , 2 lb mol A converted

ra = Ki ( Pa - K - ,( 1 + KaPa + KdePdcl = (ib~atalyst)(hr)

P = partial pressure atm: Pde = (Pel + Pc) 2:

Ki = 0.070: Ka = OAO; Kde = 1.50; K = 0.25.
 PROBLEM-SOLVING STRATEGIES 189

D. The isothermal, irreversible, aqueous phase reaction

A+B=E
at 100°F obeys
ft3
k = 15 lb mol hr

Using a 1000-ft 3 stirred tank reactor with an aqueous feed of 2000 ft3/hr,
what will be the outlet concentration of E if the inlet concentration of A and
B are both 0.251b moljft3?
E. A constant-volume batch reactor is used to get the following data on the
decomposition of A:

Ca (g moljL)

15.0 12.0 10.5 7.5 6.0 4.5 3.0 1.5

At 100°C, time of run (sec) 0 56 89 173 229 300 402 573

At 120°C, time of run (sec) 0 28 86 150 201 287
At 150°C, time of run (sec) 0 11 35 61 82 116

Using this data, determine the reaction order and the Arrhenius constant.
F. The liquid-phase hydrolysis of dilute aqueous acetic anhydride solution
(CH3COhO + H 20 ------> 2CH 3 COOH
is second-order irreversible (according to kinetic books). Our data on rates
as a function of temperature were

tCC) 10° 15° 25° 40°

Rate (g mol/cm 3 min) 0.0567C 0.0806C 0.158C 0.380C

where C = acetic anhydride concentration (g mol/em 3).

Why do these data indicate that the reaction is first order? What is the
Arrhenius constant?

PROBLEM-SOLVING STRATEGIES

A. My first reaction to this reaction problem would be to try to solve the two
differential equations for Ca and Cb as a function of time. The k's could then
be calculated at each experimental point. But because these are second-
order reactions, the differential equations may not have easy, analytical
190 CHEMICAL KINETICS

solutions. Because of time limitations it is probably better to do numerical

differentiation directly. To do this we first need to do a material balance on
the stoichiometry equations to relate Cs and C r to C a and C b , the experi-
mental data. Then for each time period, we would use

to calculate k 1 . If the calculated kl values do not vary much, the rate expres-
sions are consistent with the experimental data.
B. This gaseous reaction problem is variable volume due to the stoichiometry.
So use the E concept to relate volume to conversion. Inserting the reaction
rate equation into the variable-volume batch reactor equation and per-
forming the integration analytically should provide the solution for part
(a). To get the mitial concentration, use PV = nRTor C = P/RT. For part _
(b), the rate equation is inserted into the variable-volume plug flo~
equation and integrated for the solution. Because we do not know upon
which component the rate constant is based, we must specify one for the
sake of the test grader.
C. This catalytic reaction problem is just a plug flow reactor; in place of reac-
tor volume we have pounds of catalyst. We must relate the partial pressures
to conversion using the reaction stoichiometry. If the integration can be
done analytically, it would be better than a numerical or graphical integra-
tion.
D. Aqueous phase reaction means that we can assume that the reaction is
constant volume. The solution results from inserting the rate expression
into the stirred tank equation.
E. Assume that the rate expression is

ra = -kC:

where n is unknown. Then, since for a constant volume batch reactor,

dC n
ra = ata = -kC a

we integrate

I e a2 dC

Cal
_a
cn
a
= -k
ft2
e.J t1
dt

Bear in mind that n = 1 must be integrated differently from other n's. The
reason for integrating from t 1 to t 2 is that we notice that some of the con-
centrations are just doubled. So we could choose several different sets of
Cal and C a2 at each temperature. If these sets give the same k. we have
 SOLUTIONS 191

assumed the correct n. Once we have the correct n, we know the k at each
temperature. then from

In k = In ko - (EIRT)

we can calculate ko and EIR.

F. For this second-order reaction,

If the acetic acid is diluted, the concentration of water is large and small
changes in water concentration will have negligible effect on C H10 - it is
constant. This problem is a gift.

SOLUTIONS

A. A+B-4R R+B-4S
CaO C bO 0 Crt C bl 0
C a C b1 Crl Cr C b Cs
CaO - C a = C bO - C bl = C rl Crt - C r = C b1 - C b = C s
Crt = CaO - Ca, C aO - C a - C r
C b1 = C bO - C aO + C a; = C hO - C aO + C a - C b = C s

So

C s = (C bO - C b) - (C aO - C a), C r = 2(Cao - C a) - eChO - C b)

t (min) 0 10 20 30 40 50 60 70

Cb 3.00 2.41 2.07 1.80 1.61 1.45 1.32 1.22

Ca 1.40 0.89 0.64 0.47 0.37 0.29 0.23 0.19
CbO - Cb 0 0.59 0.93 1.20 1.39 1.55 1.68 1.78
CaO - Ca 0 0.51 0.76 0.93 1.03 1.11 1.17 1.21
Cs 0 0.08 0.17 0.27 0.36 0.44 0.51 0.57
Cr 0 0.43 0.59 0.66 0.67 0.67 0.66 0.64
CaCb 4.20 2.14 1.32 0.85 0.60 0.42 0.30 0.23
CbC r 0 1.04 1.22 1.19 1.08 0.97 0.87 0.78
192 CHEMICAL KINETICS

t average
(min) 5 15 25 35 45 55 65

(Ca Cb)av 3.17 1.73 1.09 0.73 0.51 0.36 0.27

(CbCr)av 0.52 1.13 1.21 1.14 1.03 0.92 0.83
-I'1Ca /M 0.0510 0.0250 0.0170 0.Q100 0.0080 0.0060 0.0040
I'1C s/l'1t 0.0080 0.0090 0.0100 0.0090 0.0080 0.0070 0.0060
kl 0.0161 0.0145 0.0156 0.0137 0.0157 0.0167 0.0148 (0.0153).,
k2 0.0154a 0.0080 0.0083 0.0079 0.0078 0.0076 0.0072 (0.0078).,

kl = _(~~a)/CaCb' k2= (~CS) I

M ICbCr

kl = 0.0153 L/g mol min, k2 = 0.0078 L/g mol min

a Ignored since it is most likely in error.

B. (a) 2A --+ 2R + S, k = 10 3 cm 3 /g mol sec ra = -kC;

Batch reactor with V = Vo(l + ex).'

Assume k based upon A, as written

f t k dt = fX C ao (1 + 2ex)2 dx 2 = fX 1 + ex 2 dx
o 0 (1 + ex)C (1 - x) ao 0 Cao(l - x)

Caokt = f: (1 :XX)2 + t; f: (l X-"~2

= [_1I-xo
_Jx + (1 _ e[ln x) + ~.t.-...Jx
l-xo

= _1_ _ 1 + e[ln (l - x) + __1_ - 1]

I-x 1-x
I--l+x
= - ~- (1 + e) + e In (l - x)
I-x

= ---
(l+ e)x + f. In (1 - x)
I-x
 SOLUTIONS

n P 1 atm
PV = nRT or -=C=-·
V RT' C ao = 82.06(117YK)

t = 82.06(1173) [~0.8) + lin (0.2)J = 500.08 sec = 8.33 min

1000 0.2 2

(b) Plug flow

V 1 IX (l + eX )2
Vo = kC ao ° (1 - X)2 dx
= k~ao [f: (1 ~XX)2 + 26 J: Cf=-d:)2 + 62 f: ({~ d:)2 ]

= _1 ~ {_1_ + 26[lll (1 - x) + _1_J

kC. o 1 - x 1- x

- [1 - x - 21n
62 (1 - x) _ _ 1_J}0.8°
1- x

-k1- [(1. + 26 + 6
2
) J O•8
= + 2e(1 + 6) In (1 - x) - 6 2 (1 - x)
CaO 1- x °
C _ 3 atm
.0 - 82.06(1173)

V = 82.06(1173) [1 + 2(t) + t
Vo 3(1000) 0.2

+ 2(2)(:2)
1 3
In 0.2 - 1
4{0.2) - + 11 +
- (1 t) + *(1)
1 J
V _ 82.06( 1173) ( £ J I 02 _ 0.2 _ ~ ~)
Vo - 3(1000) 0.2 +2 n. 4 4 +4
= 217.73 sec = 3.63 min = 0.0605 hr
ft 3 feed 500 Ib lib mol 454 g 82.06(1173)cm 3
Vo = - - = - - ---- ----
hr hr 451b Ib 3 g mol 10 3 cm 3 28.3 L
= 5719.2 ft3/hr
V = v o(0.0605 hr) = 5719.2 ft 3 /hr (0.0605) = 346 ft 3
194 CHEMICAL KINETICS

c. A ~ D + E
t = 0 Ao 0 0
t = t Ao(1 - x) Aox Aox
1-x x x
mf
1+ x 1+x 1 + x
I-x xP xP
-~P
P
l+x I +x I + x

L= Ao(l + X); P = total pressure; atm = 0.2

xP
Pde = 1+ x

fo x dx fW dw Ib mol A reacted
-
--;:,:= 0 Fao'
r=-~~-

a (lb catalyst) (hr)

So

-ra = ra as given
Ib mol A feed
Fao=~~-­
hr

-r =
a
K.(I
'1 +
- x P __ x 2 P 2
X K(1 +
_)/(1 + 1-
X)2 !
K P
a 1+
x
X
+K ~
de I+ x
_)2
= K{(l - x)(1 + X) - x;;Jp I
[I + X + K a P(1 - X) + Kde XP ]2

1 [(1 + KaP) + (1 - KaP + KdeP)x]2

- - - - - - - -
(1.08 + l.22X)2
ra K i P[1 - (l + P/K)x 2 ] 0.014(1 - 1.8x 2 )

or

0.014 ~ = fX ~8 + 1.2~X)2 dx = fX 1.1664 + 2.6352x ~ 1.4884x 2 dx

F aO 0 1 - 1.8x 0 1 - 1.8x

=
1 +- -
1.1664[ - -1 I n - .,j1.8XJx
2j1:8 1- j1:8 x 0

+ 2.6352[- -~. In (I
20.8
- 1.8X 2)Jx 0

+ 1.4884[- ~+ 1 In 1 + ~XJx
1.8 2(1.8)J1.8 1 - Ji.8x 0
 SOLUTIONS 195

= 1.1664 In 1 + fo~ _ 2.6352 In (1 _ 1 8x 2 ) + 1.4884

2fo 1 - Jl~x 2(1.8) . 1.8

x
(
-x + 2~1.8ln
1 1+
1- fox
foX)
0.014 ~= 0.5151
FaO

So
0.5151 0.5151
W = - - F = ---10 = 368lb
0.014 aO 0.014
15 CaoXa (mOl A reacted)
D. A + B-E, r = ra = f 3· at exit conditions
-ra t mm
V 1000 ft3
r = - = - - - - = 0.5 hr
Vo 2000 ft 3 /hr

so

or
2 xa
(1 - Xa) = - 0 --2- = 0.5333xa
.5·15· O. 5
1 - 2xa + x; = 0.5333xa or x; - 2.5333xa + 1= 0

2.5333 ± ,/2.5333 2 - 4
xa =--- 2 -= 20441
. or 0.4892

Now
CE = CaO xa = 0.1223 lb mol/ft 3
E. For an nth-order reaction, C~ -n - C~o-n = ken - l)t, n i= 1; for a first-or-
der reaction, In (Ca/C aO ) = -kt.

Ca/CaO O.~ 0.5 0.3 0.2 0.1 T

t 0 56 173 300 402 573 at lOO°C

[ ·In (Ca/C,o)]/t 0.00398 0.00401 0.00401 0.00400 0.00402 ~0.00401

t 0 28 86 150 201 287 at 120C

[ -In ((' alC .o)];t 0.00797 0.00806 0.00803 0.00801 0.00802 -0.00802
t 0 11 35 61 82 116 at 150°C
[ -In (CalC.o)]/t 0 0.02029 0.01980 0.01974 0.01963 0.01985 ~0.01986
196 CHEMICAL KINETICS

In k = In ko - (E/RT)
or

k 150 = e- E'[(1/423)-(1/373)] or E' = 5049

k 100
5065
k 120 = e- E'[(1/393)-(1/373)] or E' = 5080
k 100

ko = 3166,3172,3149, or 3162; so
r. = _3162e-5065/TC.
F. The stoichiometry indicates that the rate should probably be R =
kC Hzo CA' When the reaction is run with large amounts of water CH20 is
approximately constant and the rate looks like R = fC., where f = kCH2o ,
R El
In - = In ko - - -
C. R T

R/C. lIT In (R/C.)

0.0567 0.003534 -2.8700

0.0806 0.003472 -2.5183
0.158 0.003356 -1.8452
0.380 0.003195 -0.9676

- 2.8700 = In ko - 0.003534(E/R)
-0.9676 = In ko - 0.003195(E/R)

E/R = 5612, In ko = 16.96210, ko = 2.326 X 107

 FLUID FLOW

I. Basic Principles 199

A. Hydrostatic Equilibrium 199
B. Total Energy Equation of Steady Flow 199
C. Manometers 200
D. Mass Balance 201
E. Average Velocity 201
F. Mass Velocity 202
G. Macroscopic Momentum Balance 202
H. Momentum of Total Stream 202
I. Bernoulli Equation without Friction 203
J. Mechanical Energy Equation 203
K. The Kinetic Energy Term 203
L. Fluid Friction Correction 203
M. Bernoulli and Pump Work 204
II. Incompressible Fluid Flow in Pipes 204
A. Skin Friction and Wall Shear 204
B. Friction Factor 204
C. Laminar Flow in Pipes 205
1. Newtonian Fluids 205
2. Non-Newtonian Fluids 205
D. Turbulent Flow in Pipes 205
1. Velocity Distribution 206
2. Universal Velocity-Distribution Equations 206
3. Flow Quantities 206
4. Friction Factors, Noncircular Cross Section 207
5. Friction with Direction or Velocity Changes 207
HI. Flow of Compressible Fluids 209
A. General Information 209
1. Continuity Equation 209
2. Total Energy Balance 209
3. Mechanical Energy Balance 209
4. Velocity of Sound 209
5. Ideal-Gas Equations 210
I,. Stagnation Temperature 210

", L. McCabe and J. C. Smith, Unit Operations of Chemical Engineering, McGraw-Hill, New York,
1976.

197
198 FLUID FLOW

B. Compressible Flow Processes 210

1. Introduction 210
a. Isentropic Expansion 211
b. Adiabatic Frictional Flow 211
c. Isothermal Frictional Flow 211
2. Variable Area Conduits 211
a. Introduction 211
b. Isentropic Flow Equations 211
c. Cross-Sectional Area Effects 212
3. Adiabatic Frictional Flow 212
a. Introduction 212
b. Adiabatic Frictional Flow Equations 212
c. Maximum Length of Conduit 213
d. Mass Velocity 213
4. Isothermal Frictional Flow 213
a. Introduction ]1Y/
b. Heat Transfer in Isothermal Flow 213
IV. Flow Past Immersed Bodies 213
A. Drag 213
B. Drag Coefficients 214
C. Particle Motion through Fluids 214
1. Mechanics 214
2. Equations for Particle Motion through a Fluid 215
3. Terminal Velocity 215
4. Drag Coefficients 215
a. Spherical Particle Motion 216
b. Drag Coefficients for Spheres-Approximate 216
V. Metering and Transporting Fluids 217
A. Fluid-Moving Machinery 217
1. Pumps 217
a. Developed Head 217
b. Positive-Displacement Pumps 218
c. Centrifugal Pumps 218
2. Compressors and Blowers 218
a. Adiabatic Compression 218
b. Isothermal Compression 218
c. Polytropic Compression 218
d. Power 219
B. Measurement of Flowing Fluids 219
1. Venturi Meter 219
2. Orifice Meter 220
3. Compressible Fluids with Venturi or Orifice Meters 220
4. Pitot Tube 220
5. Rotameters 220
Problems 221
Problem-Solving Strategies 222
Solutions 224
 BASIC PRINCIPLES 199

I. BASIC PRINCIPLES

A. Hydrostatic Equilibrium

In any cross section parallel to the surface of the earth, pressure is constant,
varying only from height to height. The equation for a vertical column of fluid is
g
dp + -pdZ =0
gc
where Z = height above base of column,
p = pressure at this point,
p = density at this point,
g = acceleration offree fall in earth's gravitational field,
~ 32.174 ft/sec 2 , and
gc = Newton's law proportionality factor for the gravitational
force unit, ~ 32.174 lb ft/sec 2 lb force.

B. Total Energy Equation of Steady Flow

Consider Figure Fl for a single stream of material. An energy balance gives

m[~~ -u; +
2gcl
g(Zb - Za)
gel
+ Hb _ HaJ = Q _ W.
1
where 1 is the mechanical equivalent of heat = 778.17 ft lbr/Btu
Some limitations of this equation are

• surface, magnetic, and so forth, energy changes are neglected;

• heat effects Q are + when heat flows into the equipment; and
• shaft work W. is + when work is done by the equipment.

Input stream
m lb Heat added thru wall
Q Btu
Velocity
Oltput stream
Va ft/sec

~ UilXnent
Enthalpy
Ha Btu/lb

Shaft work
,~

done by equipment
T m lb
Velocity
Vb
Enthalpy
H
b
Za ft
W, a-l,,< 'b 1ft
Datum
j
FIGURE Fl. Generalized process equipment.
200 FLUID FLOW

C. Manometers

Consider the manometer of Figure F2. Using the hydrostatic equilibrium equa-
tion between points 1 and 2 gives

f pa

P2
dp = -
g
-Pb
gc
fza dZ
Z2

between 5 and 4

f Pb

P4
dp = - -g
gc
Pb fZ.
Zb
dZ

between 4 and 3

f P4

P3
dp = - -g
gc
Pa f
Z3
Z4
dZ

Thus

Now

Adding the Pb-P4 equation to the P4-P3 equation gives

1--
Pa

--r
Zm
Pb
5

Pb

t 4

FIGURE F2. Manometer.

 BASIC PRINCIPLES 201

Now P3 = P2' So subtract this Pb--P3 equation from the Pa-P2 (which is now
Pa-P3) equation to obtain

9
Pa - Pb = - Rm(Pa - Pb)
9c
the equation for a manometer, if Pa and Pb are measured in the same horizontal
plane.
D. Mass Balance

Consider the tube shown in Figure F3. At steady state, the mass rate of flow into
the tube must equal the rate of mass out, considered over the period of time.
Thus, if P and u are constant across area S,
m= PauaSa = PbUbSb
where m= rate of flow of mass (mass/time).
This is the equation of continuity, since
m= puS = constant

E. Average Velocity

If the velocity
U is not constant across the area S, we need an average velocity.
N ow if we look at a small part of the cross-sectional area S, we see that
dril = pu dS
and then the total mass flow rate through S would be

m=p LUdS

assuming that density is constant across S. Then the average velocity V would be
defined by
- pS
m = S1
V ==
f sU dS

Of course, we could also write the average velocity in terms of the total volumet-
ric flow rate V,
_ V
V=-
S

Veloci ty pliaa
Oensi ty
Area Sa
t::i:
----rl --} Velocity
Density
Area
lib
Po
Sb

FIGVRE F3. Generalized flow tube.

202 FLUID FLOW

F. Mass Velocity

We can rewrite the average velocity in terms of a mass velocity G,

- m
Vp = S= G

G. Macroscopic Momentum Balance

If we assume steady flow in the x direction, the sum of all forces acting upon the
fluid in the x direction must eq ual the increase in the time rate of momentum of
the flowing fluid, or

where M = momentum flow rate.

H. Momentum of Total Stream

The momentum flow rate M of a fluid with a mass flow rate mand a velocity u is
M=mu
if u is constant. If u is not constant across S, we must use
dM
- = pu·u = pu
2

dS
or, for a constant density stream,

M=
S
f
eS s
u2 dS

Using a momentum correction factor {J,

or

{J = ~
SV
f s
u2 dS = .~S f (!)2
s V
dS

If we know {J, then

M = {3Vm
and our momentum balance becomes
 BASIC PRINCIPLES 203

I. Bernoulli Equation without Friction

Using a momentum balance for a steady potential flow, we obtain the point
form of the Bernoulli equation without friction
.~ dp + !L dZ + d(u 2 /2) = 0
p dL gc dL gcdL
The differential form is
dp g 1 u2
-+-dZ+-d-=O
p gc gc 2
Of course, between two points a and b, this can be integrated to
P. gZ.
-+--+-=-+-+-
u; Pb gZb u~
p gc 2gc p gc 2gc

J. Mechanical Energy Equation

The Bernoulli equation (Section I) is just an energy balance; all terms have the
dimensions of energy per unit mass. Mechanical potential is gZ/gc; mechanical
kinetic energy is u 2 /2gc; pip is the mechanical work done to get the fluid into a
tube or the work recovered as it leaves the tube.

K. The Kinetic Energy Term

The Bernoulli equation was developed for a system with u constant across a
cross section. Instead, we could use a kinetic energy correction factor ex in the
kinetic energy term;

where

and V is the average velocity.

L. Fluid Friction Correction

For incompressible fluids, the friction form of the Bernoulli equation is

-2 -2
P. gZ. 0(. V. Pb gZb O(b Vb h
-+-+--=-+-+--+ r
p gc 2gc p gc 2gc
where hr is the friction generated in the fluid between a and b per unit mass of
fluid.
204 FLUID FLOW

M. Bernoulli and Pump Work

We consider a pump between points a and b. The work supplied to the pump is
shaft work w..
Pump work is negative, due to shaft work sign convention. So
pump work Wp is defined by

W= W.
p
m
Friction occurring within the pump must be accounted for. The mechanical
energy supplied to the pump as negative shaft work must be discounted by the
frictional losses to get the net mechanical energy available to the flowing fluid. If
hrp is the total friction in the pump per unit mass of fluid, then net work to the
fluid is Wp - hrp . But usually a pump efficiency 17 is used. This efficiency is de-
fined by
IJ = (Wp - hrp)/Wp
Then the final Bernoulli equation for incompressible fluid flow is
P gz IX V2 P gZ IX V2
---" + , _ a +"----" + IJ W = -b + '-------" + ---"-- 11 + hr
p gc 2gc p p gc 2gc

II. INCOMPRESSIBLE FLUID FLOW IN PIPES

A. Skin Friction and Wall Shear

If hrs denotes the skin friction between the fluid stream and the wall, it can be
shown that
2 Ow 4 Ow
hr = - -- I':..L = - - I':..L
s prw pD
where Tw = the shear stress at the wall (lbrlft2),

hrs = skin friction (ft lbrllb), and

D = pipe diameter (ft).

B. Friction Factor

The friction factor is defined to be the ratio of the shear stress at the wall to the
product of density and velocity head, that is,

f = Ow _ 2Yc T w
- pV 2 /2g c - pV 2

Measurement of skin friction in pipes can be done with hrS' I':..Ps' Tw , or f. They
are related by
2 Ow I':..ps I':..L V 2
hr = - -I':..L = - - - = 4f - -
s p rw p D 2gc
 INCOMPRESSIBLE FLUID FLOW IN PIPES

so
f = _Ap,-- ge D
2AL pV 2
Note that the friction factor f used here is the Fanning friction factor.

C. Laminar Flow in Pipes

1. Newtonian Fluids

VeloCity as a Function of Radius. Using the definition of viscosity /1,

----
rgc
/1=
dujdr
we can find the equation for velocity

from which we can find

and

Average Velocity
V _1
and - 2
U max

Kinetic Energy Correction

a=2
M omentum Correction

2. Non-Newtonian Fluids

The velocity profile shape for non-Newtonians differs from that of Newtonians.
Many must be determined experimentally. A few simpler ones have analytical
expressions. Advanced fluid flow texts should be consulted for non-Newtonian
systems.

D. Turbulent Flow in Pipes

Turbulent flow is not nearly so easy to analyze as laminar. Much experimental

and theoretical work has been expended. Some of the more useful ideas are
discussed as follows.
206 FLUID FLOW

1. Velocity Distribution

Turbulent flow velocity distribution has been expressed in terms of dimension-

less parameters:

u* = f"
fIctIOn veloClty
' == -vJ72" = JrP
w gc

+ . . U
U = velocity quotient == -.
u
+. yu*p y ~
y = distance == - - = - Y'wgcp
J.L J.L
Y = distance from tube wall with r w = r +y
2. Universal Velocity-Distribution Equations

Laminar Sublayer-Near the Wall

u+ = y+
Buffer Layer-between Sublayer and Turbulent Core
u+ = 51n y+ - 3.05 (empirical equation)
Turbulent Core
1
u+ = - In y + + c1
k
with k = 0.407, C1 = 5.67.

Sub layer y+ < 5; buffer 5 < y+ <30; turbulent core 30 < y+

3. Flow Quantities

Since the sublayer and the buffer layer are only a small part of the flow regime,
the turbulent core equation may be used, to a first approximation, to determine
the various flow quantities:
Average Velocity
V 1
Umax 1 + (3/2k)Jfi2
Friction F actor- Reynolds Number Relationship

Jm = ~ln (NReJ~) - ;k + C1

DV DVp
N Re = - = - -
V J.L
 INCOMPRESSIBLE FLUID FLOW IN PIPES

Corrections/or Kinetic Energy and Momentum

a = 1 + 8{2 (15 - ~ J7 )
5
{3 = 1 + 8k 2 /

but a, {3 are both almost one.

The Constants k and c 1
k = 0.407, C1 = 5.67 == Nikuradse equation, that is,
1
-;= = 4.0 log (N Re 1]) - 0.60
viI y}
4. Friction Factors, Noncircular Cross Section
Long straight channels with constant noncircular cross section can be handled
with the previous friction equations, if we use the equivalent diameter as the
diameter in the Reynolds number and in the friction factor definition. The equi-
valent diameter for a circular tube is defined to be four times the hydraulic
radius rHo
S
rH = -
LP
where S = cross-sectional area of channel and
LP = perimeter of wetted channel.
Note that hydraulic radius should not be used with laminar flow.
5. Friction with Direction or Velocity Changes
Friction Loss/rom Sudden Expansion. (See Figure F4 for nomenclature.)

hre = K{~~)
= friction loss

FIGURE F4. Sudden expansion.

208 FLUID FLOW

FIGURE F5. Sudden contraction.

Friction Lossfrom Sudden Contraction. (See Figure F5 for nomenclature.)

hfc = Kc(~~)
(empirical)
Kc = 0.4(1 - ~:)
Effect of Valves and Fittings

v,. = average velocity prior to fitting

where

Kf

Globe valve, wide open 10.0

Angle valve, wide open 5.0
Gate valve, wide open 0.2
half-open 5.6
Return bend 2.2
Tee 1.8
Elbow, 90° 0.9
45° 0.4

Bernoulli Equation and Form-Friction Losses

To include all form-friction terms in Bernoulli's equation, we use

where the first term accounts for the skin friction loss.
 FLOW OF COMPRESSIBLE FI"UIDS 209

III. FLOW OF COMPRESSIBLE FLUIDS

A. General Information

We now enter the realm of density variations. In compressible flow at regular

densities and high velocities, a basic parameter is the Mach number, the ratio of
the speed of the fluid to the speed of sound in the fluid under conditions of flow,
N Ma = u/a. The following simplifying assumptions are used: the flow is steady
and unidirectional; velocity is constant across a cross-sectional area; shaft work
is zero; the only friction considered is wall shear; gravitational effects and me-
chanical energy are neglected; and the fluid is considered an ideal gas with a
constant specific heat.

1. Continuity Equation

Previously
In p + In S + In u = constant
and so
dp dS du
~+~+~=O
p S u

2. Total Energy Balance

dQ
--=dH+d ( -u 2 )
nz 2gc1

3. Mechanical Energy Balance

4. Velocity of Sound

where the subscript S means at constant entropy.

210 FLUID FLOW

5. Ideal-Gas Equations

R
p=-pT
M
dp dp dT
=-+-
p p T
H = Ho + CiT - To)
dH = CpdT
pp-Y = constant}. .
IsentropIc
Tp - (1 -1/y) = constant

Cp Cp
Y = Cv = Cp (RIM])

dp dp dP)
-=Y-,
P p
( dp s = YpP

6. Stagnation Temperature

If a fluid is brought to rest adiabatically without shaft work, the temperature

attained is the stagnation temperature 1'.

and the stagnation enthalpy Hs is

u2
H =H+-
s 2gJ

B. Compressible Flow Process

I. Introduction

Assume a large supply of gas at specified temperature and pressure at zero velo-
city. This gas is in a reservoir at reservoir conditions (stagnation temperature).
We assume the gas flows from the reservoir, without entrance friction, into and
 FLOW OF COMPRESSIBLE FLUIDS 211

through some conduit. Our gas leaves this conduit at definite temperature, pres-
sure, and velocity into the exhaust reservoir. Here the pressure may be con-
trolled at a constant value less than the reservoir pressure. Several processes
may occur within the conduit.

a. Isentropic Expansion

Cross-sectional area of conduit must change. Stagnation temperature is con-

stant in the conduit.

b. Adiabatic Frictional Flow

Cross-sectional area constant, stagnation temperature constant; irreversible.

c. Isothermal Frictional Flow

Cross-sectional area constant, heat flow through the wall to keep the tempera-
ture constant.

2. Variable Area Conduits

a. Introduction

An isentropic flow device is a nozzle, illustrated in Figure F6. It usually consists

of a convergent section, a throat section, and a divergent section. These devices
are frequently used in the measurement of fluid flow.

b. I sentropic Flow Equations

Gas Properties Change during Flow

P - Po
- -
_T_ _ To
P6' pl-(1!YJ P6 -(lfy)

Subscript 0 indicates reservoir conditions.

Convergent Throat Divergent

FIGURE F6. Nozzle.

212 FLUID FLOW

Nozzle Velocity

u 2 = -,!-!'?JcPo
c' -
1)po
[1 _(p )1-(I IY)]
Po
or

N2
Ma
= _2_
I' - 1
[( Po )1-(1 11 )-1
P
1
from which

P = 1/1
Po
'{ + [Y---
2
1] 2 }Y!(Y-l)
N Ma

When u = a and N Ma = 1, the critical pressure ratio rc is

rc = p(u = a, N = 1) = ( 2 )Y!()'-I)

Po Y+ 1

c. Cross-Sectional Area Effects

The changes in velocity and Mach number caused by area changes can be deter-
mined from

du1
- ( ~.) + dS = 0
u a2 S
or

du 2 dS
-(NMa - 1) =
u S

3. Adiabatic Frictional Flow

a. Introduction

If heat transfer through the pipe wall is negligible, flow through a straight pipe of
constant cross section is adiabatic.

b. Adiabatic Frictional Flow Equations

Property equations

Pa = N Ma.h
Pb
N
Ma.a
J( 1+"
~ 1 N2 )
2 Ma.b
( 1 +'','
,
-2 1N2, )
Ma.a

Ta _ (
Tb - 1+
r' - 1
2 NMa,b
2)(1 + -2-
~' - 1 ,2 )
1\\1a,a
 FLOW PAST IMMERSED BODIES 213

and
Pa Pa Tb
Pb Pb T"
where a is the entrance and b is the exit.

c. Maximum Length of Conduit

The maximum value oflength L, consistent with a given entrance Mach number,
is given by

JLmax =! [ 1 __ 1 _I'. +--.! In~.!....± (y - DL2N~a,a)J

rH Y N~a,a 2 N~a,.(Y + 1)
where J = t(fa + j~).
d. Mass Velocity

G = pN Ma J9 yTR2
~- .. =
c-
N MaV pgcYp

4. Isothermal Frictional Flow

a. Introduction

The fluid temperature during compressible flow through a pipe of constant cross
section can be kept constant by heat transfer through the pipe wall. The basic
equation is

where f..L = Lb - La'

b. Heat Transfer in Isothermal Flow

IV. FLOW PAST IMMERSED BODIES

A. Drag

Drag is the force in the direction of flow exerted by a fluid on a solid. When the
solid's wall is parallel to the flow direction, the only drag force is the wall shear
Tw' The fluid pressure may possess a component in the direction of flow, and this
214 FLUID FLOW

also contributes to drag. The total drag is the sum of the two. Wall drag is the
total integrated drag from wall shear; form drag is the total integrated drag from
pressure. The phenomena in actual fluids are complicated; experimental or ap-
proximate methods are usually employed.

B. Drag Coefficients

A drag coefficient is used for immersed solids, analogous to a friction factor. Just
as the friction factor is defined as the ratio of Tw to the product of fluid density
and the velocity head, the drag coefficient Co is defined as the ratio of F ojAp to
this same product,

where F 0 = total drag,

Ap = projected area; area perpendicular to flow, and
Uo = velocity of approaching stream.

Using dimensional analysis, you can determine that the drag coefficient of a
smooth solid in an incompressible fluid depends upon a Reynolds number and
shape factors. For a given shape

with NRe,p = GoDp!/.l,

Dp = characteristic length and
Go = uop·

C. Particle Motion through Fluids

1. Mechanics

Particle movement through a fluid requires a density difference between the

fluid and the particles. In addition, an external force must provide motion to the
particle relative to the fluid. This force is usually gravity. The forces acting on a
particle that is moving through a fluid include

• an external force: gravity or centrifugal,

• a buoyant force acting parallel to the external force but in opposite direction.
and
• a drag force acting to oppose the motion.
 FLOW PAST IMMERSED BODIES 215

2. Equations for Particle Motion through a Fluid

A force balance gives

.h mae (external force)

WIt Fe =-
gc a e = acceleration of the particle
(buoyant force)
Pp = particle density

(drag force)

Then
du Pp - P C o u 2 pAp
- = a --- - --=-'-----"
dt e Pp 2m

If the external force is gravity

ae = g (the acceleration of gravity)

If the external force is centrifugal

where r = particle path radius and

ill = angular velocity (radians/second).

3. Terminal Velocity

The particle quickly reaches a constant velocity, the maximum attainable if gra-
vity is the force, and this is called the terminal velocity. We obtain it from du/
dt = 0, or

2g(pp - p)m
AppppCo

4. Drag Coefficients

The use of the above equations requires that Co be known. When a particle is at
sufficient distance from the container boundaries and other particles, we have
free settling. If particle motion is impeded by other particles, we have hindered
settling.
216 FLUID FLOW

a. Spherical Particle Motion

Assume a sphere of diameter D p ' then

nD2
A =-~p
p 4

and

-du-=a Pp - P--~--~~~
3C D U 2 P
dt e Pp 4ppDp
also

at steady state.

b. Drag Coefficients for Spheres-Approximate

Stokes Law Range: NRe,p <2

24
CD = y;;-'
Re,p

Intermediate Range: 2 < NRe,p < 500

18.5
CD = N o.6
Re,p
'

Newton's Law Range: 500 < NRe,p < 200,000

CD = 0.44,

To use, calculate K from

K = D [aeP(pp -
P 1-12
p)Jl /3
If K < 3.3 use Stokes law
3.3< K < 43.6 Intermediate law
43.6 < K < 2360 Newton's law
 METERING AND TRANSPORTING FLUIDS 217

v. METERING AND TRANSPORTING FLUIDS

A. Fluid-Moving Machinery

1. Pumps

a. Developed Head

The Bernoulli equation can be used. The friction term hr can be neglected since
the only friction is that occurring in the pump (and this is taken into account
with the mechanical efficiency '1),

The parentheses are the total heads H. Since a is the pump inlet and b the pump
outlet, usually Za ~ Zb' In any case Ha is the total suction head and Hb the total
discharge head, or
W = Hb - Ha _ I1H
p '1 '1

The power supplied to the pump driver by some external source is PH'

rhl1H
PH = rhWp = rh = mass flow rate
'1

The power delivered to the fluid P r is defined by PrlPB = '1. These equations
could also be used for fans by letting P = (Pa + Pb)/2.
The power is independent of pressure level, depending only on pressure differ-
ence. But the lower limit of the suction pressure is fixed by the vapor pressure at
the temperature of the suction. If the suction pressure is lowered, some liquid
flashes to vapor (called cavitation) and no liquid can be drawn into the pump.
Of course, this will not occur if the su'm of the velocity and pressure heads at the
suction is greater than the liquid vapor pressure. The excess is called the net
positive suction head (NPSH) or Hsv.

Pv = vapor pressure

If the reservoir is located at a' below the pump entrance at a with friction hrs in
the suction line,

H
sv
= fa: - Pv _ h _ gZ~
fs
P gc

where Za is the height of the pump above the reservoir.

218 FLUID FLOW

b. Positive-Displacement Pumps

A definite volume of liquid is trapped in a chamber, alternately filled from ~ *

inlet and emptied through the discharge at a higher pressure. The two maJ.Jil
types are reciprocating and rotary.

c. Centrifugal Pumps

The basic equations for power, developed head, and capacity have been derived
for an ideal pump in McCabe and Smith, pp. 187-193. Their results are
U2[U 2 - (qr/Ap tan (2)]
I1H r = Hb - Ha = -
gc
where U 2 = peripheral velocity of tips of blades,
qr = volumetric flow rate, and
P2 = angle that tangent to impeller tip makes with tangent of circle
traced out by impeller tips.

2. Compressors and Blowers

The general equation is

~= rPbdp
P JPa
a. Adiabatic Compression

If the unit is uncooled, the process is isentropic, so

W = PaY [(fJb)(Y-l)/Y - 1J
p (y - 1)Pa Pa

b. Isothermal Compression

If the unit is cooled, the process is isothermal, so

W = Pa In Ph
p P. Pa

c. Polytropic Compression

Rather than being isothermal or isentropic, the machine is likely somewhere in

between. It is useful to use
 METERING AND TRANSPORTING FLUIDS 219

with n somewhere between 1 and y. The constant n is determined experimentally

from

Use this value n in place of the y for the adiabatic compression.

d. Power

The power for adiabatic compression or polytropic compression can be calcu-

lated from

P B = 0.0643T.,yqo [(Pb)(Y-l)/Y - 1J
520(y - 1)17 Pa

where P B = brake horsepower and

qo = volume of gas compressed (std ft3 fmin).
\

(Use n in place of y if polytropic.)

The power for isothermal compression is

B. Measurement of Flowing Fluids

1. Venturi Meter

Using the Bernoulli equation for incompressible fluids and the continuity equa-
tion written as

we obtain

for the Venturi meter shown in Figure F7. This is ideal. To account for the
friction loss, the Venturi coefficient C v is used:

v, - -- ~v_._ J2gcip~ - P~
b - -/1 _-f34 P

Cv ~ 0.98 for Da 2-8 in.

~ 0.99 for Da> 8 in.
220 FLUID FLOW

FIGURE F7. Venturi meter.

2. Orifice Meter

The same formula is appropriate for an orifice meter. If the Reynolds number,
based on orifice diameter, is greater than 20,000, C v = 0.61. In fact, if 13 < 0.25,
the equation could be written

Uo = 0.61 J2gc(P~ - Pb)

3. Compressible Fluids with Venturi or Orifice Meters

For compressible fluids, multiply both formulas by Y. For Venturi meters,

Y = (Pb)l /Y
Pa
J y(l - 13 4)[1 - (Pb/pJey l)/Y] .
(y - 1)[1 - (Pb/Pa)][1 - f34(PbipJ2!Y]

For orifice meters

_
Y-l- 0.41 + 0.3513 4 (1 - Pb)
y Pa

4. Pitot Tube-------~\

.j
F or incompressible fluids

and for compressible fluids

using the notation of Figure F8.

5. Rotameters

These must be calibrated for a given tluid.

 PROBLEMS 221

- a
--
b
Ps

FIGURE F8. Pitot tube.

PROBLEMS

A. What is the settling time for the following spherical particles to settle under
free settling conditions through,5ft of water at 70°F. (From McCabe and
Smith.) '" )

Particle Material Specific Gravity Diameter (in.)

Galena 7.5 0.01

2 Galena 7.5 0.001
3 Quartz 2.65 0.Q1
4 Quartz 2.65 0.001

B. A lO-ft diameter tank with a quick-opening value to a sump has a 10-ft

depth of liquid. How long does it take to drain the tank if the valve is
assumed to be equivalent to a 4-in. diameter orifice? The liquid density is
60 Ib/ft 3 .
C. What size sharp-edged orifice was in use if a flow of 120 gpm produced a
4-in. Hg differential in a 4-in. Schedule 40 pipe. The fluid, measured at 60°F,
had a viscosity of 1 centistoke and a specific gravity of 1.
D. Seven hundred gallons per minute of water are being discharged by a cen-
trifugal pump 5 ft above the pump inlet. The inlet pressure is 4 in. Hg above
atmospheric and the discharge pressure is 29 psia. If the pump has an 8-in.
diameter inlet and a 4-in. diameter discharge, find (a) the pump efficiency,
(b) the new flow rate, net head, and brake horsepower if the pump speed is
increased from 1600 to 3200 revolutions per minute (rpm). Assume the
pump input is 8 hp.
E. In case of a temperature runaway in our reactor (shown in Figure F9),
cooling water (60 F) flows from a cooling tank (vented to the atmosphere,
C

14.7 psia) through the reactor to a waste system kept at 7 psia. The piping
 222 FLUID FLOW

I
125 ft

1
Reactor

- ____a..-_ Waste FIGURE F9. Reactor for problem E.

is 6-in. Schedule 40 steel containing two gate valves, three standard 90°
elbows, and three couplings and it has a length of 400 ft. Assume the pres-
sure loss through the reactor is equivalent to another 500 ft of this pipe.
What will be the rate of water flow in an emergency (gpm)?

F. A centrifuge with a 20-in. diameter bowl 14 in. deep operates at 1200 rpm
giving a 2-in. thick liquid layer when used with a sludge whose liquid has a
specific gravity of 1.2 and a viscosity of 2.5 CPo The solid component of the
sludge has a specific gravity of 1.7. If the cut size of the particles is 35
microns, what is the capacity of the centrifuge in gallons per minute?

G. Determine the net positive suction head available to a centrifugal pump

when the level of water is just at the suction point. The pumped fluid is
water at 140'F and it is being pumped from an atmospheric tank at
250 gpm. The tank suction point is 7 ft above the pump. The piping is 5-in.
Schedule 40, lO-ft long with two short radius 90° elbows.

H. Water at 60°F flows through a 3-in. Schedule 40 pipe. A pitot tube in the
~"" pipe shows a 3-in. Hg differential. If the pitot tube is located at the center of
the pipe, what is the mass flow rate of the water?
~

PROBLEM-SOLVING STRATEGIES

A. This is a free settling problem. How does the drag on a spherical particle
affect settling time? Look up free settling or drag in the index of McCabe
and Smith. Ahah! The equations are given. Just plug the data into the
equations. A simple problem if the equations are available.
B. Draining a tank is an unsteady-state process. Possibly Bird, Stewart, and
Lightfoot! have the equation. If you cannot find the equation, develop one

tR. B. Bird. W. E. Stewart, and E. N. Lightfoot, Transport Phenomena, 1st. ed., Wiley, New-York,
1960.
 PROBLEM-SOLVING STRATEGIES 223

using a force balance. An orifice equation relates force to pressure drop

across the orifice. And the pressure drop is proportional to the height of
fluid in the tank. We should be able to set up and solve the differential
equation.

c. This orifice problem has all the data needed for a solution. Usually the size
is given and the flow requested. Our problem will be trial and error since we
will need to start by assuming that the orifice coefficient CD is 0.61. With
this assumption we use the orifice equation to determine the diameter. With
the diameter, we calculate the Reynolds number, which allows us to check
the value of CD'

D. We can determine the shaft work with the Bernoulli equation since all the
needed quantities are given. Since the pumP inp,ut is given, we find the
efficiency by dividing the shaft work by the pu}hp input-using correct
units of course. To solve the remainder of the problem, we need to know
how pump speed affects the variables. Perry's Chemical Engineers' Hand-
book is the most likely source. If you cannot find it discussed, you should
not attempt the problem unless you are willing to settle for half credit.

E. This is a Bernoulli equation problem with pipe friction. We know all quan-
tities needed except the outlet velocity and the pipe friction. But the velocity
and the pipe friction are related. So this must be a trial-and-error solution.
One way to solve it is to assume a zero velocity and calculate the pipe
friction hr from the Bernoulli equation. Knowing hr, we use the friction
factor chart to find the Reynolds number and then the outlet velocity. Use
this velocity in the Bernoulli equation to recalculate hr. Then recalculate
outlet velocity, and so forth. Except for the trial and error, the problem is
simple.

F. This is a centrifuge problem and I know nothing about centrifuge design.

To even start working this problem, J need an equation. Checking the
McCabe and Smith index leads me in a short time to a centrifugal sedimen-
tation design equation. All of the data needed for the equation is given in
the problem. I have discovered that a seemingly difficult problem is a snap.

G. This problem can be solved with the pump Bernoulli equation. It looks like
all needed data are given. Because the volumetric rate is given, no trial and
error will be required. If I can find the definition of net positive suction
head, I should have no difficulty with the solution.

H. This is a pitot tube problem and they are usually simple. I may need a
correction since the velocity is a function of radius. A quick review of pi tot
tube equations should allow me to decide if I can handle the problem.
224 FLUID FLOW

SOLUTIONS

A. From McCabe and Smith (McC&S), pp. 153-156 on free settling:

K -- D p Q were
h Q _ [aep(pp - p)]l J3 ae (Eq. (7-54), McC&S)
- 2 ' = 9
I-l
If K < 3.3 Stokes region,
3.3 < K < 43.6 Intermediate,
43.6 < K < 2360 Newton's law,
If Stokes,

[Eq. (7-50), McC&S]

If intermediate,

Particle Material Pp - P gp/)12 Q K Region

1 Galena 405 4.61 x 109 12.33 X 10 3 0.000833 10.28 Intermediate

2 Galena 405 4.61 x 109 12.33 X 10 3 0.0000833 1.03 Stokes
3 Quartz 102.9 4.61 x 109 7.79 X 10 3 0.000833 6.49 Intermediate
4 Quartz 102.9 4.61 x 10 9 0.0000833 0.65 Stokes

where

(0.153)32.20' 71( 10 4)°.43(0.0008331.14)405°' 71 ft

1. Ut = 29 = 0.273 -
(62.3°' )6.60°·43 sec
5ft
t = 0.273 ft/sec = 18.3 sec

(32.2)8.33 2 x 10- 1°(104) ft

2. Ut = (18)6.60 (405) = 0.0076 sec

t = 5/0.0076 = 658 sec

3. Ut = 0.1034 ft/sec, t = 5/0.1034 = 48.3 sec
4. ut = 0.001933 ft/sec, t = 5/0.001933 = 2580 sec
 SOLUTIONS 225

B. Consider a tank of cross-sectional area A and liquid height h as shown in

Figure FlO and do a momentum balance with
input = 0, output = F, accumulation = A(dh/dt)
Then input-output = accumulation or
dh
A-= -F
dt
Now we know that the equation for an equivalent orifice is

F= CoAo J-p-
2g~P

and Co = 0.61 and ~p/p = h. So

A(dh/dt) = - coAo./29h
or

or

2hl/2 = a -(CoAo)2gt/A)
At t = 0, h = hi = 10 ft or a = 2hf/2. So
t = 2(hf/2 - hl/2)A
C o Ao)2g
IT.D2 1T. 1T. ( 4 in.)2 1T. 1 2
A =4 = 4" (10 ft)2; Ao = 4" 12 in./ft = 4" 9 ft
2g = 32.2 x 2 ft/sec 2
When empty, h = 0

t = GoAo
2A Jh;
2g =
21T.100 fe(4)(9) J-l~
4(0.61)1T. ft 2 64.4 ft/sec 2
= 1163 sec or 19.4 min

FIGURE FlO. Schematic for solution B.

226 FLUID FLOW

C. We use McC&S Eqs. 8-39 to 8-43, following Example 8-5.

3 gibe 4 in. lb lb e
p = 62.3 lb/ft , tlp = . - . - ---:- . (13.6 - 1)62.3 - = 261.7 -
gc lb 12 m. ft 3 ftl

m= 120 gal_l~_ 62.31b 1 min = 16 66~-

min 7.48 gal ft3 60 sec . sec
4 in. Schedule 40 = 4.026 in. ID
4.026 lb ft
Da = -12- = 03355
. ft , gc = 32.2 ~2-1
sec be
p2 = . 4(16.66Ib/sec)
_ . __
0.61n(0.3355 ft)\/2(32.2 lb ft/sec Ib r)[261.7 Ibr/ftl)(62:3 Ib/ft 3 )
2

= 0.3015
P= 0.5491
So Do = 0.5491(0.3355 ft) = 0.184 ft = 2.21-in. orifice diameter. We must
now check to see that the Reynolds number is > 20,000 in order to have
used CD = 0.61
/1 m 2 /sec m2
- = 1 cS x 10 - 6 - - - - = 10 - 6 .
P cS sec

= 10- 6 m 2 (3.2808 ft)2 = 10.76 X 10- 6 fe

sec m sec
/1 = 10.76 x 10- 6 (62.3) Ib/ft sec
4rh 4(16.66Ib/sec)
N Re 0 = -- = ----- -- ~.- - - -
. nDo/1 n(0.184ft)(62.3)(10.76 x 1O- 6 )lb/ftsec
= 177,976
So Co = 0.61 and our calculation is correct.

D. Flow rate:

700.gal _ 1 ft3 1 min = 1.560 ft 3

mm 7.48 gal 60 sec sec
or
ft3 Ib lb
1.5560 - x 62.3 - = 97.2 -
sec ft 3 sec
Use McC&S, Eq. 8-2.
Pb-P. g • C(bV~-C(.V; a = inlet
(a) '1Wp = - - -- +-- (Zb - Z.) + - - - - -
P gc 2gc b = outlet
 SOLUTIONS 227

Wp = 8 h x 550 ft Ib r ~ = 45.3 ft Ib r
p sec 1 hp 97.21b lb

p. = 4 in. Hg + 29.92 in. Hg = 33.92 in. Hg x 14.7 psia

29.92 in. Hg
= 16.7 psi a
_ ft3 4 144 in. 2 ft
V. = 1.560 - ----- = 4.47 - -
sec n 8 2 in. 2 ftl sec
Now

so

Vb = 4(4.47 ftjsec) = 17.88 ft/sec

If we assume kinetic energy correction factors, cx. and CX b = 1

(
ft lbr) _ (29 - 16.7) lbr ft 3 !44 in. 2 + 5lbr ft
'1 45.3 lb - in. 2 62.3 Ib ftl lb
(17.88 2 - 4.47 2) ft2
+ --~------~~--~
2(32.2 Ib ft/sec 2 lbr) sec 2
or
3 ft Ibr _ 284 !! lbr 5 ft lbr 47 [t lbr _ [t Ibr
45. '1 Ib - . lb + lb +. Ib - 38.1 lb

So '1 = 0.84 or efficiency = 84 %.

(b) Capacity is proportional to speed. Head is proportional to (speed)2.
Brake horsepower proportional to (speed? So if speed is doubled, at
the new conditions
gal/min = 2(700) = 1400 galjmin
head = 4(37) = 148 ft lbrllb
bhp = 8(8 hp) = 64 hp
E. We use the Bernoulli equation corrected for fluid friction, McC&S, Eq.
(4-30).

-P. +gZ.
-+ ;{A = -
Pb ~+ -
+./6, CXb V~
- + hr
a = inlet (tank)
p gc gc P gc 2gc b = outlet (waste)
o (at rest in tank) 0 (baseline)
We know neither Vb nor hr; to determine hr, we need Vb' If we know hr, we
can cakJlate Vb from

[Eq. (5-63), or Fig. 5-10, McC&S]

228 FLUID FLOW

As a first approximation, assume Vb = 0; then

or
(14.7 - 7) Ib r ft 3 144 in. 2 ft lb f _ )iR h
in. 2 62.31b fe + 125 Ib - ge. + f

0, 1st approximation
or
hr = 17.8 + 125 = 142.8 ft lbrllb
Then
N ri _ 6.065 in. _ 62.4 Ib ft sec
ReV f - 12 in./ft ft 3 7.59 X lO- 4 1b

X )142.8 ft Ib r 6.065 ft 32.2 ft l~

Ib (2)900 ft 12 sec 2 Ibr

= 4.155 X lO4Jl29l = 4.72 X 10 4

Using Fig. 5-10 with K = 0.00015,

K 0.00015 12 = 0000
6.065
X
. 3
or 0.004 = f
D
So

N _ 4.72! 10 4 4.72 X 104

Re - .jj - - - = 74.63 X 104
jO.004
DVp
N Re = - - or v =L
Dp
X 74.63 X 104
J1
v 7.59 x lO- 4 lb 12ft 3 4 ft
vb = -- x 74.63 x 10 = 17.96-
ft sec 6.065 ft 62.4 lb sec

We now check this in the Bernoulli equation, using

( 4f ~
D
+ V~
Kr) 2ge
in place of hr (see McC&S, Eq. 5-74).
So

142.8 ft lb = r C(b V~ + (4f~ + Kr) V~

Ib 2ge D 2ge
 SOLUTIONS 229

From Table 5-1, McC&S

Kr
2 gate valves = 2(0.2)}
3 90° elbows = 3(0.9) 3.1 if Q(b =1
3 couplings = 3(0)

142 8 = [1 4(0.004)900(12) 31J V~ = 0 06 -2

. + 6.065 +. 64.4 .5 1 Vb

Vb = 16.80 ft/sec
This value of Vb will not change the f significantly.
inside area = 0.2006 ftl
ft gal sec
gpm = 16.8 - (0.2006 ftl)7.48 -r 3 60 - . = 1512 gpm
sec t mm
F. See Figure Fl1. We use McC&S, Eq. 30-55.

_ nbm 2(pp - p )D~c ( 2 2)/1 (~)

qc - 18 r2 - r1 n
J1. r1 + r2
. 3.973 x 10- 5 in. ft
Dpc = 35 mIcrons x .
mIcron
x 12I.n.

= 1.148 X 10- 4 ft
lb Ib
pp - p = (1.7 - 1.2)62.4 ft 3 = 31.2 ft3

2n min rad
m = 1200 rpm - - = 125.7-
60 sec sec

J1. = 2.5 c 6.7197 x 10- 4 ~ = 16.80 X 10- 4 ~

P cp ft sec ft sec
10 in. 8 in.
r2 = 12 in/ft = 0.833 ft; r 1 = 12 in./ft = 0.666 ft

14 in.
b = 12 in./ft = 1.167 ft

~lUJi
~
.D rl
I
I
1--8 in.-!
\-10 in.-j
r2 FIGURE Fll. Schematic for solution F.
230 FLUID FLOW

n(1.167 ft)125.7 rad 2 /sec 31.2Ib/ft 3 (1.148 X 10- 4 )2

qc = - 18(16.80 x 1O- 4 Ib/ft sec)

2 2 2 [ 2(0.833) ]
. (0.833 - 666 ft )/In (0.833 + 0.666)

ft 3 7.48 gal 60 sec gal

= 1.879 - --.- = 843 - .
sec ft3 mID mID

G. See Figure F12. From McC&S, Eq. 8-8, VP at 150°F = PV'

oc. V; P. - p,
H sv =-2- +--
gc p
250 gal ft3 1 min ft3
vol rate = -. - - - - = 0.557 -
mID 7.48 gal 60 sec sec

Ai = cross-sectional area of 5 in. Schedule 40 pipe =

0.139ft2
_ 0.557 ft 3 ft 5.047 in.
V = - = 401- Dj = 12' f = 0.4206 ft
• 0.139 ft 2 sec . sec' ID. t

So we know everything but P•. We use Bernoulli with p = 61.381b/ft 3 , Pv =

2.88861br/in. 2 , and Jl. = 0.470 cPo

Kr = 2(0.9) = 1.8 McC&S, Table 5-1

0.4206 ft 4.01 ft 61.38 Ib cp ft sec 5

NR = -- - = 3.28 x 10
e sec ft 3 0.470 cp 6.7197 x 1O- 4 1b
k 0.00015
- = - 2-06 = 0.00036
D 0.4

FIGURE F12. Schematic for solution G.

 SOLUTIONS 231

Figure 5-9 gives f = 0.0043, so

h = [4(0.0043 _7_ 1.8J (4.01 ft/sec)2 sec 2 1br = 1 ft Ibr

r ) 0.4206 + 2(32.2) ft Ib 0.52 Ib
Now for the Bernoulli equation

So
, Z'
Hsv = Pa - Pv + ~ - hr
p gc
Ib Ib 144 1·n.2
14.7:--% - 2.89 :--%
m. m. ft2 Ibr 7 f _ 052 ft lbr
61.38Ib/ft3 + Ib t . 1 lb

= (27.7 + 7 - 0.5) ft Ibrllb = 34.2 ft lbrllb

H. Using McC&S, Eq. 8-49

Uo = J29~!J.P, Uo = centerline velocity, max = umax

Then use Figure 5-7,

to obtain fI, with J.l = 1.129 cP

!J.p = -gcg 12'3~./f (13.6 -

m. t
lli
t
~
1)62.3 f3 = 196.2 -f2
t

Uo = J2 (32.2 ft lb) !96.2Ibr ft 3 = 14.24 ~

sec 2 lbr ft2 62.3 Ib sec
3.068 in. ft 62.3 lb ft sec 5
NRe = 12 in./ft 14.24 sec 1.129 ft3 (6.7197 x 1O-41b) = 2.99 x 10
So
iJumax = 0.815
So
fI = 0.815(14.24 ft/sec) = 11.61 ft/sec

or
(11.61 ft/sec)0.0513 ft2(62.31b/ft 3 ) = 37.1lb/sec
 ENGINEERING ECONOMICS

1. Interest and Interest Formulas 235

A. The Time Value of Money 235
B. Types of Interest 235
1. Simple Interest 235
2. Compound Interest 235
C. The Cash Flow Diagram 235
D. Interest Formulas 236
1. Single-Payment, Compound-Amount Factor 236
2. Single-Payment, Present-Worth Factor 237
3. Equal-Payment-Series, Compound-Amount Factor 237
4. Equal-Payment-Series, Sinking-Fund Factor 237
5. Equal-Payment-Series, Capital-Recovery Factor 237
6. Equal-Payment-Series, Present-Worth Factor 238
7. Uniform-Gradient-Series Factor 238
8. Format for the Use ofInterest Factors 239
9. Interest Factor Relationships 239
E. Nominal and Effective Interest Rates 239
II. Bases for Comparison of Alternatives 240
A. Net Cash Flows for Investment Opportunities 241
B. Present-Worth Amount 241
C. Capitalized Equivalent 242
D. Annual Equivalent Amount 242
E. Capital Recovery with Return 243
F. Future-Worth Amount 243
G. Rate of Return 244
H. Some Relationships between Methods 244

E. L. Grant, W. G. Ireson, and R. S. Leavenworth, Principles of Engineering Economy, 7th ed., Wiley,
New York, 1982.
H. G. Thuesen, W. 1. Fabrycky, and G. 1. Thuesen, Engineering Economy, 5th ed., Prentice-Hall,
Englewooc l'liffs, New Jersey, 1977.

233
234 ENGINEERING ECONOMICS

III.Decision Making Among Alternatives 244

A. Types of Investment Proposals 244
1. Independent Proposals 245
2. Dependent Proposals 245
a. Mutually Exclusive Proposals 245
b. Contingent Proposals 245
B. Mutally Exclusive Alternatives and Decision Making 245
1. Comparing Mutually Exclusive Alternatives 245
2. Forming Mutually Exclusive Alternatives 246
C. Decision Criteria for Mutually Exclusive Alternatives 246
1. Selecting an Interest Rate 246
2. The Do Nothing Alternative 246
3. Criteria 246
a. Present-Worth on Total Investment 246
b. Present-Worth on Incremental Investment 247
c. Rate of Return on Incremental Investment 248
D. Applying Decision Criteria When Money Is Limited 248
E. Comparison of Alternatives with Unequal Service Lives 248
1. Study Period Approach 248
2. Estimating Future Alternatives 248
a. Expanding Short Life to Long Life 248
b. Common Multiple Life 249
IV. Benefit-Cost Analysis 249
V. Depreciation and Depreciation Accounting 249
A. The Value-Time Function 249
1. Straight-Line Method of Depreciation 250
2. Declining-Balance Method of Depreciation 250
3. Sum-of-the- Years-Digits Method of Depreciation 251
4. Sinking-Fund Method of Depreciation 251
B. Capital Recovery with Return - All Methods of Depreciation 251
VI. Expected Value Decision Making 252
Problems 255
Problem-Solving Strategies 257
Solutions 258
 INTEREST AND INTEREST FORMULAS 235

I. INTEREST AND INTEREST FORMULAS

A. The Time Value of Money

Money can be invested so as to produce more money. If it is invested at some

specific rate of interest it will increase in value with time. So money has a time
value. With this in mind, the comparison of sums of money separated by time
must be made at a common time.

B. Types of Interest

The rental rate for a sum of money is expressed as the percentage of the sum that
is to be paid for the use of the sum for a period of one year. Interest rates are also
given for periods other than one year. This will be discussed later.

1. Simple Interest

In simple interest, the interest to be paid on repayment of a loan is proportional

to the length of time the principal sum has been borrowed. Let P represent the
principal, n the interest period, and i the interest rate. Then the simple interest I
is given by
1= Pni

2. Compound Interest

When a loan is made for a length of time equal to several interest periods, inter-
est is calculated at the end of each interest period. A number ofloan repayment
plans are available. These range from paying the interest when it is due to accu-
mulating all the interest until the loan is due. If the borrower does not pay the
interest earned at the end of each period and is charged interest on the total
amount owed (interest plus principal), the interest is said to be compounded.

C. The Cash Flow Diagram

To help in recording and identifying the economic effects of alternative invest-

ments, a graphical description of each alternative's cash transactions can be
used. This pictorial description, called a cash flow diagram, will provide all the
necessary information for analyzing an investment proposal.
The cash flow diagram represents any receipts received over a period of time
as an upward arrow (a cash increase) located at the end ofthe period. The height
of the arrow is usually made proportional to .he magnitude of the receipts re-
ceived during that period. Likewise, the disbursements are represented by a
downward arrow (a cash decrease). These arrows are placed on a time scale that
represents the duration of the proposal.
236 ENGINEERING ECONOMICS

D. Interest Formulas

The interest factors presented in this section apply to the common situation of
annual compounding of interest and annual payments. We use t11t'! following
symbols:

i = the annual interest rate,

n = the number of annual interest periods,
P = a present principal sum,
A = a single payment (annuity), in a series of n equal payments, made at the end
of each annual interest period, and
F = a future sum, n annual interest periods hence, equal to the compound
amount of a present principal sum P, or equal to the sum of the compound
amounts of payments A, in a series.

The symbols are chosen so that each is an initial letter of a key word. Thus i
applies to interest, n applies to number of periods, P applies to present worth (or
sum), F applies to future worth, and A applies to annual payment or annuity.

1. Single-Payment, Compound-Amount Factor

If an amount P is invested now with the amount earning at the rate i per year,
how much principal and interest are accumulated after n years?
This is determined from the single-payment compound amount factor

F/P(i, n) = (1 + on
or
F = P{F/P(i, n)]

as shown in Figure E 1.

o
1 2 3 n-l n

p FIGLRE El. Cash flow diagram-single-

payment compound-amount.
 INTEREST AND INTEREST FORMULAS 237

2. Single-Payment, Present-Worth Factor

The single-payment, compound-amount relationship can be solved for P

P=F 1 =F'{P/F(i,n)j
(1 + on
The resulting factor is known as the single-payment, present-worth factor.

3. Equal-Payment-Series, Compound-Amount Factor

In engineering economics studies it is often necessary to find the single future

value that would accumulate from a series of equal payments occurring at the
end of succeeding annual interest periods. The factor needed is

its cash flow diagram is illustrated in Figure E2.

o 1 2 3 n-1 n

FIGURE E2. Cash flow diagram-series-

A A A A A payments compound-amount.

4. Equal-Payment-Series, Sinking-Fund Factor

The equal-payment-series, compound-amount relationship can be solved for A

A = F[ i ] = F· (A/F(i. f1)]
(1 + on - 1

This is called the equal-payment-series, sinking-fund factor.

5. Equal-Payment-Series, Capital-Recovery Factor

A deposit of amount P is made now at an annual interest rate i. The depositor

wants to withdra w the principal plus earned interest in a series of eq ual year-end
amounts ov~r the next n years. When the last withdrawal is made there should
be no funds left on deposit, as shown in Figure E3.
238 ENGINEERING ECONOMICS

A A A A A

o
1 2 3 n-I n

p
FIGURE E3. Cash flow diagram-series-
payments capital-recovery.

The relationship needed is

A
i(1
= P [ (1 + i)n
+ i)n
_
J
1 = P . {AIP(i, n)j

Note that for perpetual or infinite time, the capital-recovery factor is i.

6. Equal-Payment-Series, Present-Worth Factor

To determine what single amount must be deposited now so that equal end-of- •
year payments can be made, we solve for P in terms of A or

P = A [(1 + i)n+ - 1J
i(1 i)" = A . (PIA(i, n)}

Note that for perpetual or infinite time, the present-worth factor is 11i.

7. Unifotm-Gradient-Series Factor

Frequently, annual payments do not occur in equal-payment series. A linearly

increasing or decreasing series, as illustrated in Figure E4, can be reduced to an
equivalent equal-payment series. If we let
Al = payment at the end of the first year,
G = annual change or gradient, and
A = Al + A2,
then

I
n -A
A2 = G[ ~1 - ~. .-
I F(I, n)
J = G . (A/G(i, n)}

AI+(n-l)G
AI+(n-2)G

AI+2G
AI+G
Al

0
t 2 3 n-I n
FIGURE E4. Cash flow diagram-
un iform-gradien t -series.
 INTEREST A"ID INTEREST FORMULAS 239

8. Format for the Use of Interest Factors

In engineering economics studies, disbursements made to start an alternative are

considered to take place at the beginning of the period embraced by the alterna-
tive. Payments occurring during the period of the alternative are usually as-
sumed to occur at the end of the year or interest period in which they occur.
In the use of interest factors for annual payments:

• The end of one year is the beginning of the next year.

• P is at the beginning of a year at a time regarded as being the present.
• F is at the end of the nth year, from a time regarded as being the present.
• A occurs at the end of each year. When P and A are involved, the first A of
the series occurs one year after P. When F and A are involved, the last A of
the series occurs at the same time as F.

9. Interest Factor Relationships

P/F(i, n) = l/[F/P(i, n)]

F/A(i, n) = l/[A/F(i, n)]
P/A(i, n) = I/[A/P(i, n)]
F/A(i, n) = 1 + F/P(i, 1)+ FIP(i, 2) + ... + FIP(i, n - 1)
PIA(i, n) = P/F(i, 1) + PIF(i, 2) + ... + PIF(i, n)
AIPO, n) = AIFCi, n) + i
FIG(i, n) = [A/GO, n)][F/A(i, n)]
FIP(i, n) = [F/P(i, nl)][FIP(i. n2)]'" [F/P(i, nk)]
where n = nl + 11 2 '" + 11k

E. Nominal and Effective Interest Rates

Up to this point, we have looked at interest periods of only one year. But agree-
ments can specify that interest will be paid more frequently, such as each half-
year, each quarter, or each month. These agreements result in interest periods of
one-half, one-quarter, or one-twelfth year, and the compounding of interest 2, 4,
or 12 times a year, respectively.
The interest rates associated with this more frequent compounding are nor-
mally quoted on an annual basis according to the following convention. When
the actual or effective rate of interest is 3 % interest compounded each six-month
period, the annual or nominal interest is quoted as 6 % per year compounded
semiannually For an effective rate of interest of 1.5 % compounded at the end of
each three-month period the nominal interest is quoted as 6 % per year com-
pounded quarterly. So the nominal rate of interest is expressed on an annual
basis and it is determined by multiplying the actual or effective interest rate per
240 ENGINEERING ECONOMICS

interest period times the number of interest periods per year. Thus the actual or
effective interest rate per year is higher than the nominal interest rate. If we let

r =
nominal interest rate (per year),
i = the effective interest rate (per period),
c = the number of interest periods per year,
then
i = effective annual interest rate = (1 + r/cy - 1

The previous interest formulas were derived on the basis of an effective inter-
est rate for an interest period, specifically, for an annual interest rate com-
pounded annually. But they can be used when compounding occurs more
frequently than once a year. This can be done in either of two ways:

1. find the effective annual interest rate from the previous formula and use
this rate when time periods are years; or
2. match the interest rate to the interest period and use the factors directly
or its corresponding tabulated value.

F or example, we wish to find the compound amount of $1000 four years from
now at a nominal annual interest rate of 6 % compounded semiannually.
Since
c=2 and r = 0.06

i = (1 + 0.~6y - 1 = 0.0609

F = $1000(1 + 0.0609)4 = $1267

or
F = $1000·F/P(6.09~o, 4) = $1267

Likewise, since the nominal annual interest rate is 6 % compounded semian-

nually, the interest rate is 3 % for an interest period of one-half year. So

F = $1000 (1 + 0.030)8 = $1267

or
F = $1000-F/P(3%, 8) = $1267

II. BASES FOR COMPARISON OF ALTERNATIVES

All decision criteria must incorporate some index, measure of equivalence, or

bases for comparison that summarizes the significant differences between invest-
 BASES FOR COMPARISON OF ALTERNATIVES 241

ment alternatives. A basis for comparison is an index containing information

about a series of receipts and disbursements representing an investment oppor-
tunity. This reduction of alternatives to a common base is needed so that appar-
ent differences become real differences with the time value of money considered.
These real differences become directly comparable and can be used in decision
making. The most common bases are

• Present-worth amount.
• Annual equivalent amount.
• Capitalized equivalent amount.
• Future-worth amount.
• Rate of return.

A. Net Cash Flows for Investment Opportunities

An investment is described by the cash receipts and disbursements that are anti-
cipated. The representation of the amounts and timing of these cash receipts and
disbursements is called the cash flow of the investment.
When an investment has both cash receipts and disbursements occurring
simultaneously, a net cash flow is calculated for the investment. The net cash
flow is the arithmetic sum of the receipts ( + ) and the disbursements ( - ) that
occur at the same point in time. This utilization of net cash flows in decision
making implies that the net dollars received or disbursed have the same effect on
an investment decision as does the separate consideration of the total receipts
and disbursements of the investment.
To describe investment cash flows we shall use the notation

Fjt = net cash flow for proposalj at time t

Then if Fjt < 0, Fjt represents a net cash cost or disbursement. If Fjt > 0, Fjt is a
net cash income or receipt.

B. Present-Worth Amount

An index that reflects the differences between alternatives by considering the

time value of money is the present-worth amount. For a known cash flow and
given interest rate, an equivalent amount for the cash flow can be calculated at
any point in '_ime. So it is possible to calculate a single equivalent amount at any
point in time that is equivalent in value to a particular cash flow.
The present-worth amount is an amount at the present (t = 0) that is equiva-
lent to the cash flow of an investment for a particular interest rate i. So the
242 ENGINEERING ECONOMICS

present worth of investment proposalj at interest rate i with a life of n years is

PW(i\ = F jo ' P/F(i, 0) + Fjl . P/F(i, 1) + ... + F jn · P/F(i, n)

n
= 'IFjl·P/F(i,t)
1=0

A number of features makes the present-worth amount suitable as a basis for

comparison:

• It considers the time value of money according to the value of i selected for
calculations.
• It concentrates the equivalent value of any cash flow into a single index at a
particular time (t = 0).
• The present-worth amount is unique regardless of the investment's cash flow
pattern.
• The present-worth amount is the equivalent amount by which the equivalent
receipts of a cash flow exceed (or fail to equal) the equivalent disbursements
of that cash flow.

C. Capitalized Equivalent

A special case of the present-worth amount PW(i)j is the capitalized equivalent

CE(i). This method finds a single amount at the present which at a given rate of
interest will be equivalent to the net difference of receipts and disbursements if a
given cash flow pattern is repeated in perpetuity.
To determine the capitalized equivalent of an investment that is to produce
cash flows from the present to infinity, we:

• Convert the actual cash flow into an equivalent cash flow of equal annual
payments A that extends to infinity.
• Discount the equal annual payments A to the present by using the equal-
payment-series, present-worth factor.

D. Annual Equivalent Amount

The annual equivalent amount has characteristics similar to the present-worth

amount. Any cash flow can be converted into a series of equal annual payments
by first calculating the present-worth amount for the original series and then
multiplying the present-worth amount by the interest factor A/P(i, n). So the
annual equivalent amount AE(i) for interest rate i and n years can be calculated
from
AE(i) = PW(i) . A/P(i, n)
 BASES FOR COMPARISON OF ALTERNATIVES 243

E. Capital Recovery with Return

An asset is a unit of capital. Some assets, such as machines, lose value over a
period of time. This loss of value of an asset represents a consumption or expen-
diture of capital.
Capital assets are bought with the idea that they will earn more than they
cost. So a part of the prospective earnings is considered to be capital recovery.
The capital is recovered in the form of income derived from the services rendered
by the asset and from its sale at the end of its useful life.
Another part of the prospective earnings will be considered to be return.
Since the capital invested in an asset is usually recovered bit by bit, it is neces-
sary to consider the interest on the unrecovered balance as a cost of ownership.
Two transactions are associated with the buying and eventual retirement of a
capital asset; its first cost and salvage value. A simple formula for the equivalent
annual cost of an asset can be determined from these amounts. We let

P = first cost of an asset,

F = estimated salvage value,
n = estimated service life in years, and
CR(i) = capital recovery with return.

So the annual equivalent cost of an asset can be expressed as the annual equiva-
lent first cost less the annual equivalent salvage value, or
CR(i) = p. A/P(i, n) - F· A/F(i, n)
or
CR(i) = (P - F)· AI P(i, n) + Fi

F. Future-Worth Amount

The future-worth basis for comparison is an equivalent amount of a cash flow

calculated at a future time for some interest rate. The future-worth amount for a
proposal) at some future time n years from the present is given by
FW(i)j = P jO -F/P(i, n) + P j ! • F/P(i, n - 1) + ... + PjnF/P(i, 0)
or

FW(i)j= IPjt·F/P(i,n-t)
1=0

A simple method for calculating the future-worth amount is to first determine

the present-~ vrth amount of the cash flow and then to convert to its future
equivalent n year hence, or
244 ENGINEERING ECONOMICS

G. Rate of Return

The rate of return is a widely accepted index of profitability. It is the interest rate
that reduces the present-worth amount of a series of receipts and disbursements
to zero. So, the rate of return for investment proposal j is the interest rate ijthat
satisfies the equation

0= PW(ij)j = I Fjt·P/F(ij, t)
1=0

Because of the relationship between the annual equivalent amount, the future-
worth amount, and the present-worth amount, this equation is also equivalent
to

or
0= FW(ij)j
In all the other methods, the interest rate is given (or assumed given). With this
method, we must calculate the interest rate or rate of return. The calculation
becomes a trial-and-error problem. But in return for the extra work, we have a
reasonable way to compare the economic desirability of alternative investments
without assuming that we know highly uncertain future interest rates.

H. Some Relationships between Methods

Given one basis for comparison, we can calculate the others from the following
relationships:
PW(i) = AE(i)·P/A(i, n)
CE(i) = AE(i)/i
PW(i) = CE(i)· i· P/ A(i, n)
FW(i) = PW(i)·F/P(i,n)

Ill. DECISION MAKING AMONG ALTERNATIVES

We now look at various decision criteria that describe how investment decisions
should be made. These criteria will be rules or procedures that describe how to
select investment opportunities so that particular objectives can be achieved.

A. Types of Investment Proposals

An investment proposal is considered to be a single project that is being con-

sidered as an investment possibility. An investment alternative is a decision op-
tion, so every investment proposal could be considered to be an investment
 DECISION MAKING AMONG ALTERNATIVES 245

alternative. However, an investment alternative could consist of a group or set of

investment proposals. It could also represent the option of doing nothing. So if
there are two proposals, A and B, it is possible to have four decision alternatives:
do nothing, accept only A, accept only B, or accept both A and B.

1. Independent Proposals

When the acceptance of a proposal from a set of proposals has no effect on the
acceptance of any of the other proposals contained in the set, the proposal is
said to be independent. Usually, if proposals are functionally different and there
are no other obvious dependencies between them, it is reasonable to consider
the proposals as independent.

2. Dependent Proposals

For many decision problems, a group of investment proposals will be related in

such a way that the acceptance of one of these proposals will influence the accep-
tance of the others. There are several types.

a. Mutually Exclusive Proposals

If the proposals in the set are related so that acceptance of any proposal from the
set precludes the acceptance of any of the other proposals in the set, the propo-
sals are said to be mutually exclusive. These usually occur when there is a need
to be fulfilled and there are a variety of proposals each of which will satisfy that
need.
b. Contingent Proposals

Another type occurs once some initial project is undertaken. Then a number of
other auxiliary investments become feasible. Such auxiliary proposals are called
contingent proposals because their acceptance is conditional on the acceptance
of another proposal.

B. Mutually Exclusive Alternatives and Decision Making

From now on the selection of investment proposals will be viewed as a problem

of selecting a single economic alternative from a set of alternatives. Thus the
proposals are set up so that the alternatives are considered to be mutually
exclusive.
1. Comparing Mutually Exclusive Alternatives

When comparing mutually exclusive alternatives, it is the future difference be-

tween the alternatives that is relevant for determining the economic desirability
of one compared to the other.
246 ENGINEERING ECONOMICS

2. Forming Mutually Exclusive Alternatives

Investment proposals can be independent, mutually exclusive, or contingent.

Rather than develop special rules for each, the approach will be to alrange all
proposals so that the selection decision involves only the consideration of the
cash flows of mutually exclusive alternatives.
All that is required is the enumeration of all the feasible combinations of the
proposals under consideration. Each combination represents a mutually exclu-
sive alternative since each combination precludes the acceptance of any of the
other combinations. The cash flow of each alternative is found simply by adding,
period by period, the cash flows of each proposal contained in the alternative
being considered.

C. Decision Criteria for Mutually Exclusive Alternatives

1. Selecting an Interest Rate

The objective is the maximization of equivalent profit given that all investment
alternatives must yield a return that exceeds some minimum attractive rate of
return (MARR). This cut-off rate is usually the result of a policy decision made
by management. The MARR could be viewed as a rate at which the business can
always invest since it has a large number of opportunities that yield such a
return. Whenever any money is committed to an investment proposal, an op-
portunity to invest that money at the MARR has been foregone.

2. The Do Nothing Alternative

If the funds available are not invested in the projects being considered they will
be invested in the do nothing alternative. This means that the investor will do
nothing about the projects being considered but the funds will be placed in
investments that yield a rate of return equal to the MARR.

3. Criteria

a. Present- Worth on Total Investment

This criterion is one of the most used criteria for selecting an investment alterna-
tive from a set of mutually exclusive alternatives. All that is required is to calcu-
late the present-worth amount for the cash flow representing each alternative.
Then select the alternative that has the maximum present-worth amount (pro-
vided this amount is positive).
We have seen that the present-worth amount, the annual equivalent amount,
and the future-worth amount are consistent bases for comparing alternatives.
Thus, if
PW(i)A > PW(i)B
 DECISION MAKING AMONG ALTERNATIVES 247

then
AE(i)A > AE(i)B
and
FW(i)A > FW(i)B
Note that the future-worth amounts indicate that the receipts from the invest-
ment are actually invested at the MARR from the time they are received to the
end of the life of the alternative. So future-worth calculations explicitly consider
the investment or reinvestment of the future receipts generated by investment
alternatives. So the use of the present-worth amount implicitly assumes the in-
vestment or reinvestment of receipts at the MARR.

b. Present-Worth on Incremental Investment

The present-worth on incremental investment requires that the incremental dif-

ferences between alternative cash flows actually be calculated. We first deter-
mine the cash flow representing the difference between the two cash flows. Then
the decision whether to select a particular alternative rests on the determination
of the economic desirability of the additional increment of investment required
by one alternative over the other. If the present-worth amount for the incremen-
tal investment is greater than zero, the increment is considered desirable and the
alternative requiring this additional investment is deemed best.
To apply this decision criteria to a set of mutually exclusive alternatives, the
following steps must be used:

• List the alternatives in ascending order of their equivalent first cost or initial
dis bursemen ts.
• Select as the initial current best alternative the one that requires the smallest
first cost. In most cases the initial current best would be the do nothing
alternative. Compare the initial current best alternative and the first chal-
lenging alternative. Examine the difference between the two cash flows. If the
present worth of the incremental cash flow evaluated at the MARR is greater
than zero, the challenger becomes the new current best alternative. If the
present worth is less than or equal to zero, the current best alternative re-
mains unchanged and the challenger in the comparison is eliminated from
consideration. The new challenger is the next alternative.
• Repeat the comparisons of the challengers to the current best alternative
until no alternatives remain.

This method is more time consuming than the present worth on total invest-
ment criterion. But both criteria lead to the same solution. In fact,
PW(i)B - PW(i)A = PW(i)B-A
AE(i)B - AE(i)A = AE(i)B-A
FW(i)B - FW(i)A = FW(i)B-A
248 ENGINEERING ECONOMICS

c. Rate of Return on Incremental Investment

This decision criteria uses the same type of incremental analysis as the prior
criterion. The only difference is the decision rule in the third step. For this
method, the rule should be

• The increment of investment is considered desirable if the rate of return re-

sulting from the increment is greater than the MARR. that is, i=_A > MARR.

D. Applying Decision Criteria When Money Is Limited

Up to now the decision criteria have been applied to sets of mutually exclusive
alternatives and it has been assumed that sufficient money is available to under-
take all of the proposals. We now assume that the total money available for
investment is a fixed amount that restricts the number of feasible alternatives.
No changes at all are required in order to include the budget constraint in the
decision process. All that is required is that the proposals be rearranged into
mutually exclusive alternatives. Next, those alternatives that have a first cost
that exceeds the budget amount must be dropped from consideration. Every-
thing else remains the same.

E. Comparison of Alternatives with Unequal Service Lives

To calculate the present worth for cash flows of unequal duration is incorrect.
When comparing alternatives with unequal lives the principle that all alterna-
tives under consideration must be compared on the same time span is basic to
sound decision making. There are two basic approaches that can be used so that
alternatives with different lives can be compared over an equal time span.

1. Study Period Approach

This approach confines the consideration of the effects of the alternatives being
evaluated to some study period that is usually the life of the shortest-lived
alternative. The costs occurring after the study period are disregarded, since the
equivalent costs are being compared only for the study period indicated.

2. Estimating Future Alternatives

The second approach is to estimate the future seq uence of events that are antici-
pated for each alternative being considered so that the time span is the same for
each alternative. Two methods are used to accomplish this.

a. Expanding Short Life to Long Life

Here we have explicit consideration of future alternatives over the same time
span.
 DEPRECIATION AND DEPRECIATION ACCOUNTING 249

h. Common Multiple Life

We assume that an investment opportunity will be replaced by an identical

alternative until a common multiple of lives is reached. The annual equivalent
comparison should be used since it is computationally the most efficient
approach. Since the cash flows for each alternative consist of identical repeated
cash flows, it is only necessary to calculate the annual equivalent for the original
alternative.

IV. BENEFIT-COST ANALYSIS

With the same data used, the decision reached by comparing benefits with costs
is the same decision reached by other methods. But the same proposed project
may have several different values of the benefit-cost ratio depending upon
whether certain adverse items are subtracted from benefits or added to costs. In
general, relevant consequences to the general public should be classified as bene-
fits or disbenefits. Consequences involving disbursements by governmental units
should be classified as costs.
In comparing projects, the larger benefit-cost ratio prevails. But you can see
that what I call a disbenefit, you could call a cost. Therefore your ratio calcula-
tion could differ from mine.

V. DEPRECIATION AND DEPRECIATION ACCOUNTING

A. The Value-Time Function

The pattern of the future value of an asset should be predicted. It is usual to

assume that the value of an asset decreases yearly in accordance with one of
several mathematical functions. Once a value-time function has been chosen, it
is used to represent the value of the asset at any point during its life.
Accountants use the term book value to represent the original value of an
asset less its accumulated depreciation at any point in time. The book value at
the end of any year is equal to the book value at the beginning of the year less the
depreciation expense charged during the year.
In determining book value the following notation is used:

P first cost of the asset,

=
F = estimated salvage value,
B, = book value at the end of year t,
D, = depreciation charge during year t, and
n = estimated life of the asset.

The following sections are concerned with methods for determining D, for t = 1,
2, ... , n.
250 ENGINEERING ECONOMICS

1. Straight-Line Method of Depreciation

This method assumes that the loss in value of the asset is directly proportional
to its age. So
P-F
D,=--
n
and

P-F)
B,=P-t( - n-

2. Declining-Balance Method of Depreciation

This method assumes that the annual cost of depreciation is a fixed percentage
of the book value at the beginning of the year. The ratio of the depreciation in
anyone year to the book value at the beginning of the year is a constant k. This
method gives
book value at end of t years = B f = P(l - k)'
book value at end of n years = Bn = P(l - kt
There are two ways of using this method. Knowing the original cost P, the
salvage value F, and the useful life n, we use

to calculate k and then use

B, = P(l - k)'

to calculate the book value at the end of t years. We shall call this the general
Matheson method.
The usual case, though, is the double-declining balance. For this case, the k
used is double the depreciation rate of the straight-line method. For the straight-
line method the depreciation rate is

~-!
P-F-n
So, for the double-declining-balance method

2
k=-
n
and
 DEPRECIATION AND DEPRECIATION ACCOUNTING 251

3. Sum-of-the- Years-Digits Method of Depreciation

For this method, the digits corresponding to the number of each year of life are
listed in reverse order, from n to 1. The depreciation factor for any year is the
reverse digit for that year divided by the sum of the digits. Since the sum-of-the-
years digits is n(n + 1)/2, the depreciation factor for any year tis

(n - t + 1)
D = 2(P - F) - -~ -
t n(n + 1)

and the book value (after considerable manipulation) becomes

t(2n - t + 1)
B =P -- ~ (P - F)
t n(n + 1)

4. Sinking-Fund Method of Depreciation

This method assumes that a sinking fund is established to accumulate funds for
replacement. The total depreciation that has taken place up to any given time is
assumed to be the uniform annual sinking-fund deposit plus the interest on the
imaginary accumulated fund. This method (unlike the others) preserves the in-
vested capital. So, with this method, the book value at the end of t years will be

B t = P - (P - F)·A/F(i, n)·F/A(i, t)

B. Capital Recovery with Return-All Methods of Depreciation

We have seen that the straight-line, declining-balance, sum-of-the-years-digits,

and sinking-fund methods of depreciation all lead to different value-time func-
tions for book value. But if the retirement of any asset takes place at the age
predicted and the book value equals the estimated salvage value, the deprecia-
tion amount and interest on the undepreciated balance can be shown to be
equivalent to the capital recovery with a return. Thus, for any method of depre-
ciation, the capital recovery with return is

(P - F)· A/P(i, n) + Fi
Depreciation is a cost of production. In economic analysis dealing with
physical assets, it is necessary to compute the equivalent annual cost of capital
recovered plus return, so that alternatives involving competing assets may be
compared on an equivalent basis. Regardless of the depreciation model chosen
to represent the value of the asset over time, the equivalent annual cost of capital
recovered and return will be

CR(i) = (P - F)· A/P(i, n) + Fi

252 ENGINEERING ECONOMICS

VI. EXPECTED VALUE DECISION MAKINGt

If probability distributions are used to describe the economic elements that

make up an investment alternative, the expected value of the cost or tJrofit can
provide a reasonable basis for comparing alternatives. The expected profit or
cost of a proposal reflects the long-term profit or cost that would be realized if
the investment were repeated a large number of times and if its probability dis-
tribution remained unchanged. Thus, when large numbers of investments are
made it may be reasonable to make decisions based upon the average or long-
term effects of each proposal. Of course, it is necessary to recognize the limita-
tions of using the expected value as a basis for comparison on unique or unusual
projects when the long-term effects are less meaningful.
To see how these ideas can be useful for decision making, consider the follow-
ing game. A coin will be tossed twice. If no heads occur, $100 will be lost. If one
head occurs, $40 will be won. If two heads occur, $80 will be won. What is the
expected value of the random variable, profit from game?
For a discrete random variable x, the expected value is defined as

E(x) = L xP(x)
x

where P(x) is the probability of x occurring. Now if we toss a coin twice, simple
reasoning leads to the following probabilities:

no heads P(x) = 1 out of 4 = 0.25,

two heads P(x) = 1 out of 4 = 0.25, and
one head P(x) = 2 out of 4 = 0.50.
Therefore

E(G) = -$100(0.25) + $40(0.50) + $80(0.25) = $15

If one were able to participate in a large number of such games, the expected or
average winnings per bet over the long run would be approximately $15.
Because most industries and governments are generally long lived, the
expected value as a basis for comparison seems to be a sensible method for
evaluating investment alternatives under risk. The long-term objectives of such
organizations may include the maximation of expected profits or minimization
of expected costs. If it is desired to include the effect of the time value of money
where risk is involved, all that is required is to state expected profits or costs as
expected present worth, expected annual equivalents, or expected future worths.

tThis stction is based upon H. G. Thuesen, W. J. Fabrycky, and G. 1. Thuesen, Engineering Ec'ono-
5th cd., Prentice-Hall, Englewood-Cliffs, New Jersey, 1977, pp. 430-432. Used by permission of
n1l',
Pren tice- Hall.
 EXPECTED VALUE DECISION MAKING 253

As an example of decision making using expected values, suppose that a com-

munity has a water treatment facility located on the flood plain of a river. The
construction of a levee to protect the facility during periods of flooding is under
consideration. Data on the costs of construction and expected flood damages
are shown in Table E1.

TABLE El. Probability and Cost Information for Determining Optimum Levee Size

A B C D E
Number of
Years River Probability of Initial Cost
Maximum Level River Being Loss If River Level Building
Was x ft x ft Above Is x ft Above Levee x ft
x(feet) Above Normal Normal Levee ($) High ($)

0 24 0.48 0 0
5 12 0.24 100,000 100,000
10 8 0.16 150.000 210,000
15 3 0.06 200,000 330,000
20 2 0.04 300,000 450,000
25 1 0.02 400,000 550,000
50 1.00

Using historical records that describe the maximum height reached by the
river during each of the last 50 years, the frequencies shown in column B were
ascertained. From these frequencies are calculated the probabilities that the
river will reach a particular level in anyone year. The probability for each height
is determined by dividing the number of years for which each particular height
was the maximum by 50, the total number of years.
The damages that are expected if the river exceeds the height of the levee are
related to the amount by which the river height exceeds the levee. These costs are
shown in column D. It is observed that they increase in relation to the amount
the flood crest exceeds the levee height. If the flood crest is 15 ft and the levee is
10 ft the anticipated damages will be $100,000, whereas a flood crest of 20 ft for a
levee 10-ft high would create damages of $150,000.
The estimated costs of constructing levees of various heights are shown in
column E. The company considers 12 % to be their minimum attractive rate of
return and it is felt that after 15 yr the treatment plant will be relocated away
from the floodplain. The company wants to select the alternative that minimizes
its total expected costs. Since the probabilities are defined as the likelihood of a
particular flood level in anyone year, the expected equivalent annual cost is an
appropriate choice for the basis for comparison.
An example of the calculations required for each levee height is demonstrated
for two of the alternatives.
254 ENGINEERING ECONOMICS

Five-foot levee:
A/P(l2,15)
annual investment cost = $100,000(0.1468) = $14,682
expected annual damage = (0.16)($100,000)
+ (0.06)($150,000)
+ (0.04)($200,000)
+ (0.02)($300,000) = 39,000
total expected annual cost $53,682

Ten-foot levee:
A/P(12, 15)
annual investment cost = $210,000(0.1468) = $30,828
expected annual damage = (0.06)($100,000)
+ (0.04)($150,000)
+ (0.02)($200,000) = 16,000
total expected annual cost $46,828

The costs associated with the alternative levee heights are summarized in Table
E2. The levee height that minimizes the total expected annual cost is the levee
that is 10 ft in height. The selection of a smaller levee would not provide enough
protection to offset the reduced construction costs, while a levee higher than
10 ft requires more investment without proportionate savings from expected
flood damage. The use of expected value in determining the costo()fflood damage
is reasonable in this case, since the 15-yr period under consideration allows time
for long-term effects to appear.

TABLE E2. Summary of Annual Construction and Flood Damage Costs

Levee Height Annual Investment Expected Annual Total Expected

(ft) Cost ($) Damage ($) Annual Costs
($)

0 0 80,000 80,000
5 14,682 39,000 53,682
10 30,828 16,000 46,828
15 48,450 7,000 55,450
20 66,069 2,000 68,069
25 80,751 0 80,751
 PROBLEMS 255

PROBLEMS

A. A manufacturer has purchased a machine for $80,000. Its useful life is 5 yr,
at which time it can be sold for $5000. The hourly costs are $7.50/hr; the
annual costs are $1000/yr. This machine is used to make castings. Each
casting requires about 100 min to produce. The material for each casting is
$8.00. If the cost of money is 8 %, calculate the cost per unit to make each
casting if (a) 120, (b) 220, and (c) 550 castings are made each year.

B. We have at least a million dollars for investments if we can receive a return

of 15 %/yr. Three different proposals for modifying our assembly line have
been made. Which proposal should be accepted?

Proposal A B C

Investment ($) 140,000 165,000 190,000

Annual savings ($) 21,750 26,750 29,750

C. We have two alternatives for our proposed new grease canning line. Plan A
uses an electrically powered belt that costs $1500 with a 5-yr salvage value
of $300. The power cost is $1.00/hr of use and maintenance should run
$150/yr.
Plan B uses a reconditioned gas engine costing $500 with no salvage
value. Gas and oil costs are estimated at $0.60/hr and the operator (usually
the janitor) is paid $3/hr. Because of its age, we assume maintenance costs
to be $0.25/operating hr.
We assume that the canning line will be needed for 5 yr. Management
requires an overall rate of return of 12 % before taxes.
Determine the hours ofline use per year for the two plans to break-even.
Solve by equivalent annual cost methods before taxes.

D. After our refinery was built on an island in the river, we were told that the
island was occasionally under water. Corps of Engineers data indicate that
the chances of a flood are about one in seven each year. The likely flood
damage would cost $25,000.
A levee could be built at a cost of $60,000 with a useful life of 28 yr with
no salvage. If money is available at 10%, should we build the levee?

E. The chairman will be visiting the plant. We must paint. Paint A costs $7.00/
gal and covers 350 sq ft/gal. The manufacturer claims that it will last 4 yr
and can be applied at a rate of 100 sq ft/hr.
Paint B costs $lO/gal and covers 500 sq ft/ga1.1t will last for 5 yr and can
be appbed at a rate of 125 sq ft/hr. If we pay the painter $4/hr, which paint
should we use. Money is available at 10 %.
256 ENGINEERING ECONOMICS

F. We wish to invest some of our excess profits in a new branch office. We

have money enough for one. It would cost $150,000 to set up a Cleveland
office. Annual revenues would be $68,000 and annual costs $58,000. It
would cost $180,000 to open a Pittsburgh office. Annual costs \\ ould be
$50,000 and annual revenues $75,000. If we assume no salvage value after
20 yr for either office, which branch would be the most profitable (i.e., yields
greatest return on investment)?
G. We have the choice of building a new plant in one or two stages. We have
cost estimates for both choices, based upon 1980 costs. Which method,
based upon the following data, should be recommended?
We can build the complete plant or we can build it in two stages. The
increased capacity is not needed until 20 yr from now. We can borrow
money at 10 %. Construction costs will rise at 6 % per year. Maintenance
costs after 20 yr are assumed equal for the two projects.

Annual Construction
Capacity Maintenance Costs (1980)
(lbjyr) ($) ($)

First stage 100,000 80,000 14,000,000

Second stage 175,000 12,000,000
All at once 175,000 85,000 21,000,000

H. We wish to build an R&D office with a first cost of $400,000, annual

operating and maintenance costs of $40,OOOjyr, and a salvage value of
$60,000 at the end of 50 yr. Assume an interest rate of 10 %.
(a) What is the present value of this investment if the planning horizon is
50 yr?
(b) If the building replacement will have the same first cost, life, salvage
value, and operating costs as the original, what is the capitalized cost
of perpetual service?
I. The gas company data shows the following annual frequency of leaks and
cost characteristics for their processing plant:

Number ofleaks 0 1 2 3 4
Probability of this 0.3 0.2 0.2 0.2 0.1
number during
year
Cost of this number of $0 $1500 $2500 $5000 $11,000
leaks to gas company
 PROBLEM-SOLVING STRATEGIES 257

With the installation of new piping, the above probabilities would be

changed to

P(O) = 0.75, P(1) = 0.15, P(2) = 0.10, P(3) = P(4) = 0

If the new piping will have a life of 20 yr, no salvage, and an annual main-
tenance of $500, determine the amount that the gas company could afford
to spend on the project. Assume a 12 % interest rate.

PROBLEM-SOLVING STRATEGIES

A. Find the yearly cost of the equipment and then add to it the annual costs.
These costs will be the same regardless of the number of castings produced.
Then for each different number of castings, determine the hours required to
produce them. Calculate the total hourly costs for each number of castings.
The total annual cost to produce a number of castings is this cost added to
the other yearly costs. Dividing by the number of units provides the answer.
This is a simple, straightforward problem.
B. This is a mutually exclusive alternatives problem. The straightforward
method is to list investments by ascending first costs, comparing percentage
return on both first cost and on incremental investments.
C. For this problem, add up the total annual costs for both alternatives, letting
H be the hours of line use per year. The break-even point occurs when the
two annual costs are equal.
D. The secret to the solution of this problem is to decide how to handle the
probability of a flood. A simple way to do it is to assume that the yearly
cost of a flood that occurs once every 7 yr is ~ of the total cost of a flood.
This value is then compared to the annual cost of building a levee. This
concept fits with the life of the levee being 28 yr-a constant multiple of
seven.
E. This is a break-even problem. We must determine the annual cost for each
paint. Data looks as if the initial cost will be in dollars per square ft year.
Note that one paint will last 4 yr, the other 5 yr.
F. We can determine the annual profit for each office. Since we know the
investment needed to set up each office, we can calculate a rate of return for
each, based on
A/P( %, 20) = profit/investment

The la rger rate of return is chosen.

G. Although there are several approaches to this problem, calculating present
worth is probably the best. We must remember to update the construction
258 ENGINEERING ECONO\1ICS

cost of the second stage by 6 %/yr. Use the plan that has the lower present
worth.

H. From the data given, calculate the present worth, assuming a lif.: of 50 yr.
Use the relationship between present worth and capitalized annual cost to
get the capitalized cost from the present worth.

I. We must know how to handle probabilities in economic analysis to do

this problem. In this case, it is simple. The total cost is just the sum of all
individual costs times occurrence probabilities. With this in hand, we must
calculate the annual cost of new piping, assuming C is the initial cost. A
break-even analysis completes the problem.

SOLUTIONS

A. (a) Cost to make 120 castings

time = (120 x 100 min)/60 min/hr = 200 hr

yearly cost of equipment = ($80,000 - 5000)A/P(8 %,5) = $18,784.50
[A/P = 0.25046J
fixed cost = $5000 x i[i = 0.08J 400.00
annual costs 1,000.00
hourly costs = $7.50/hr x 200 hr 1,500.00
materials = $8.00 x 120 960.00
total cost = $22,644.50

cost/unit = $22,644.50/120 = $188.70

each

(b) Cost to make 220 castings

time = (220 x 100)/60 = 366.67 hr

yearly cost = $18,784.50
fixed cost 400.00
annual costs 1,000.00
hourly costs = $7.50/hr x 366.67 hr 2,750.03
materials = $8.00 x 220 1,760.00
total cost = $24,694.53

cost/unit = $24,694.53/220 = $ 112.25

 SOLUTIONS 259

(C) Cost to make 550 castings

time = (550 x 100)/60 = 916.67 hr

yearly cost = $18,784.50
fixed cost 400.00
annual cost 1,000.00
hourly costs = $7.50/hr x 916.67 hr 6,875.03
materials = $8.00 x 500 4,400.00
total cost = $31,459.53

cost/unit = $31,459.53/550 =$ 57.20

B. Need a return of at least 15 %

A
Invest- Invest- A
ment Savings % ment Savings A%
Proposal ($) ($) Return ($) ($) Return

A 140,000 21,750 15.5

B 165,000 26,750 16.2 25,000 5000 20.0
C 190,000 29,750 15.7 25,000 3000 12.0

Proposal A wins against a simple 15 % investment; B wins against A

because both % return and A % return over A are above J 5 %, and Closes
against B because the incremental return is less than 15 %.
So we accept proposal B.

C. Plan A-annual cost

capital recovery ($1500 - 300)A/P(12 %,5)

=
[A/P = 0.27741] = $332.89
interest on salvage = $300i[i = 0.12] 36.00
power cost = $1.00H = H
lIlaintenance = $150/yr = 150.00
total cost = $518.89 +H
260 ENGINEERING ECONOMICS

Plan B-annual cost

capital recovery = $500 A/P(12 %, 5)[A/P = 0.27741J = $138.71

gas and oil = $0.60 H 0.60 H
operator = $3.00 H 3.00H
maintenance = $0.25 H 0.25H
total cost = $138.71 + 3.85 H
break-even = $518.89 +H = $138.71 + 3.85 H
380.18 = 2.85 H; H = 133.4 hr
Above 133.4 hr favors plan A.

D. Likely cost of a flood = ~ ($25,000) = $3571.43 each year

Annual cost of building levee = $60,000 A/P(1 0 ~~, 28)
= $60,000 x 0.10745 = $6447
So we do not build the levee.

E. Annual cost-paint A [A/P(10%, 4) = 0.31547J

[ $7.0~ 1 gal + $4 ~J
gal 350ft 2 hrl00ft 2
x [0.31547
J
= $0.060/ftz x 0.31547 = $0.01893/ft 2 yr
Annual cost-paint B [A/P(10%, 5) = 0.26380]
$10 1 gal $4 1 hr )
( gai 500 ft2 + hr 125 ft2 x [0.26380J

= $0.052/ft 2 x 0.26380 = $0.01372/ft 2 yr

So we use paint B.

F. Cleveland:

annual revenue less annual cost = $68,000 - $58,000 = $10,000

capital recovery = $150,000 A/P (i, 20)
A/PO, 20) = $10,000/$150,000 = 0.06667

A/P(2!%, 20) = 0.06415; A/P(3 %, 20) = 0.06722

 SOLUTIONS 261

Pittsburgh:

annual revenue less annual cost = $75,000 - $50,000 = $25,000

capital recovery = $180,000 A/ P(i, 20)
A/PO, 20) = $25,000/$180,000 =0.13889

A/P(12 %,20) = 0.13389; A/P(13%,20) = 0.14235

then
0.13889 - 0.13388
i ~ 12% + 0.14235 _ 0.13388
~ 12% + 0.59 or i = 12.59%

So we open a branch in Pittsburgh.

G. First plan-two stages 1st today, 2nd 20 yr hence

present worth (1st stage)[P/A(10%, 20) = 8.514]
$14,000,000 + $80,000P/A(10%, 20) = $14,681,120
present worth (2nd stage)[F/P(6%, 20) = 3.2071; P/F(10%, 20)
= 0.1486]
$12,000,000 F/P(6%, 20) P/F(lO%, 20) = $5,718,901
total present worth-first plan = $20,400,021
second plan-build all today
present worth [P/A(10%, 20) = 8.514]
$21,000,000 + $85,000 P/A(10%, 20) = $21,723,690
Thus we should use first plan with two stages.
H. (a) Present worth [P/A(lO%, 50) = 9.915; P/F(10%, 50) = 0.0085]
$400,000 + $40,000 P/A(10%, 50) - $60,000 P/F(10%, 50) = $796,090
(b) capitalized annual cost [P/A(10%, 50) = 9.915]

_PW _ = $796,090 = $802915

i·P/A(1O%,50) 0.10 x 9.915 '
I. Expected leaks cost of old piping
= I cost (p). pep)
p

= 0.3($0) + 0.2($1500) + 0.2($2500) + 0.2($5000) + 0.1($11,000)

= $2900
262 ENGINEERING ECONOMICS

expected leaks cost of new piping

= 0.75($0) + 0.15($1500) + 0.10($2500)
= $475
Equate annual equivalent cost of old system to new system [A/P(12 %,20)
= 0.13388]

$2900 = $475 + $500 + cost· [A/P(12 %,20) = 0.13388]

= $975 + 0.13388 cost
or
cost = ($2900 - $975)/0.13388 = $14,378.55
We could profitably afford to pay up to $14,378.55 for the new system.
 INDEX

Absorption factor, 83 Balance:

Absorption rate: energy:
controlling resistance, 104 general, 52
double resistance, 100 mass transfer, 82
general case, 101 enthalpy, 84
height of transfer unit, 103 material:
lean gas, 104 gas absorption, 98, 100
overall coefficien ts, 102 general, 52
Absorptivity, 153 mass transfer, 81
Acceleration, 9 momentum, 202
Acentric factor, 17 Batch reactor:
Activation energy, 179 constant-volume, 180
Activity, 26 variable-volume, 181
Activity coefficients, 26 Benefit-cost analysis, 249
Adiabatic: Bernoulli equation, 203, 204
humidification, III Blackbody radiation, 152
process, 16 Block diagrams, 62
Adiabatic frictional flow, 211, 212- Boiling Jiq uids, 150-15 1
213 Boiling-point elevation, 157
Air, excess. 50 Btu, 10
Alternatives. economic comparison of:
eq ual senice lives: Calorie, 10
annual equi\'alent amount, 242 Capacity, evaporator, 157
capi:aJized equivalent. 242 Ca pital-recovery, 237
capital recovery, 243 Carnot:
future worth. 243 cycle, 31
p~e5en; worth. 241 PV diagram, 33
rate of return. 244 refrigeration cycle, 34
uneqi.:a: 'ervice lives. 248-249 Cash flow, 235, 241
Angle f2~tc,. 155 Chemical potential, 24, 29
Angle oj '·ision. 154 Closed system, 11
Area. :'[ee, 156 Coefficient:
Arrhenw5 ,.1\\ ,-9 activity, 26
drag, 214, 215-216
Bait1e pitch. 156 fugacity, 25
Bame WlTId·~'.\. 156 heat transfer, overall, 143

263
264 INDEX

Coefficient, heat transfer (Continued) natural, 148

variable, 144 Conversion, fractional, 180
and HTU, 104 Cooling tower, 108
overall (absorption), 102 Correction factors:
overall from individ ual, 144 fluid friction, 203
of performance, 35 kinetic energy, 203
resistance form, 144 momentum, 202
Columns, absorption: Countercurrent, extraction, 97
packed, 98-104 Critical point, 13
plate, 105
Combustion: Dalton's law, lOS
equation, 51 Degrees of freedom, 12
heat of, 54 Dehumidification, 107
problems, 50 Density, 8
Compound-amount: Dependent proposals:
equal-payment series, 237 contingent, 245
single-payment, 236 mutually exclusive, 245
Compressibility factor, 14 Depreciation, 249
Compressible flow, 209-213 declining balance, 250
adiabatic frictional, 211, 212-213 sinking-fund, 251
isentropic expansion, 211 straight-line, 250
isothermal frictional, 211, 213 sum-of-years-digits, 251
Compression: Dew point, 106
adiabatic, 31 Diagrams:
isothermal, 19, 31 block,62
ratio, 31 P-V, 14,19,33
Corn pressors: T-S, 34, 35
adiabatic, 218 Dimensional analysis, 146
isothermal, 218 Direction changes, 207
polytropic, 218 Distillation, 85-90
power of, 219 continuous, 87
Condensation: differential, 85
dropwise, 148 flash, 85
filmwise, 149 steam, 86
Condensing vapors, 148-150 Donohue equation, 156
Conduction: 139-141 Drag, 213
stead y -sta te, 13 9 Drag coefficients, 214, 215-216
unsteady-state, 141 Drying, 112-114
Constant: calculations, 113
equilibrium, 30 equilibria, 112
rate, 176 rate of, 112
time, 60 Duhring's rule, 157
Constant-pressure process, 15
Constant-rate, drying, 113 Economy, evaporator, 158
Constant-temperature process, 15 Efficiency:
Constant-volume process, 15 mechanical, 217
Continuity equation, 209 pump, 204
Control,60 Energy:
Controllers: activation, 179
derivative, 65 definition, 9
integral, 65 Gibbs free, 22
proportional, 64 Helmholtz free, 22
Convection: internal, 10
forced, 145-148 kinetic, 10
 INDEX 265

potential, 10 Expansion:
Energy balance: adiabatic, 19, 31
evaporators, 158 constant-enthalpy, 36
f1uid f1ow: isothermal, 31
mechanical, 209 louie-Thomson, 36
total,209 ratio, 31
in heat exchangers, 142 Redlich-Kister, 27
mass transfer, 82 Expected value, 252
in total condensers, 143 Extract, 94
Engines, heat, 31 Extraction:
Enthalpy: countercurrent, 97
definition, II data, 94
humid, 106 Ponchon-Savarit, 95
Entropy, 20 principles, 95-98
Equations: rectangular method, 96
Arrhenius, 179 solvent ratio, 97
Bernoulli, 203, 204
Donohue, 156 factor:
kinetic, 182 absorption, 83
Kremser, 84,91 acentric, 17
Margules, 28 angle, 155
rate, 182 compressibility, 14
Rayleigh, 86 Fanning friction, 205
reactor design, 187 fouling, 144
Tait, 18 frequency, 179
van Laar, 28 friction, 204
van't Hoff, 30 interchange, 155
virial, 14 view, 155
Equations of state: F alling-ra te, drying, 113
acentric factor, 17 Fanning friction factor, 205
cubic, 16 Feed plate, distillation, 88
generalized, 17 Filmwise heat transfer:
Redlich-Kwong, 17 horizontal tubes, 150
van der Waals, 16 vertical tubes, 149
Equilibrium: First law, 1l
and activity coefficients, 27 Flow in pipes:
composition, 30 laminar, 205
constant, 30 turbulent, 205-209
criteria, 27 Fluid flow:
definition, 12 compressible, 209-213
drying, 112 incompressible, 204-209
hydrostatic, 119 Fluid now corrections:
leaching, 91 Newtonian fluids, 205
reaction, 28 turbulent flow, 206-207
vapor-liquid, 27 Fluid friction, 203
Evaporative cooling, 107 Force, 7
Evaporators: Forced convection:
boiling-point elevation, laminar flow, 145-146
157 outside tubes, 147-148
capacity, 157 turbulent flow, 146-148
economy, 158 Formation, heat of, 54
multiple-effect, 159 Fouling factors, 144
single-effect, 158 Fourier's law, 139
Excess properties, 26 Fractional conversion, 180
266 INDEX

Free area, 156 Heat flux, 143

Frequency factor, 179 Heat transfer:
Friction: evaporators, 157-159
fluid, 203 in fluids, 141-157
skin, 204 radiation, 151-156
Friction factor, 204 in solids, 13 9-141
Fugacity, 24 Height of transfer unit, 103
Fugacity coefficient, 25 Helmholtz free energy, 22
Functions: HTU, 103,104
point, 10 Humid enthalpy, 106
state, 10 Humid heat, 106
thermodynamics, 22-24 Humidification, 105-111
transfer, 61 adiabatic, 111
gas-liquid interaction, 108
Gas: packed contactors, 108
diatomic, 16 Humidity:
flue, 50 absolute, 105
ideal, 15 chart, 106
lean, 104 percentage, 105
monatomic, 16 relative, 105
polyatomic, 16
Gas absorption, 98-105 Ideal gas:
Graetz number, 146 definition, 15
Graphical method: flow, 210
McCabe-Thiele, 88 law, 13
mass transfer, 82 Ideal stages (leaching):
Grashof number, 148 constant underflow, 91
variable underflow, 92
Head, fluid, 217 Immersed bodies, flow past, 213-216
Heat: Incompressible fluid flow, 204-209
of combustion, 54 Interchange factors, 155
definition, 10 Interest:
of formation, 54 compound, 235
of fusion, 22 formulas, 236-240
humid, 106 simple, 235
of reaction, 55 Interest formulas:
of vaporization, 22 equa1-payment-series:
Heat capacity, 13 capital-recovery, 237
Heat engine: compound-amount, 237
Carnot cycle, 31 present-worth, 238
Diesel cycle, 31 sinking-fund, 237
gas turbine, 31 relationships, 239
Otto cycle, 31 single-payment:
P-V diagrams, 32 compound-amount, 236
performance, 32 present-worth, 237
and second law, 18 uniform-gradient-series, 238
T-S diagrams, 32 use of, 238
Heat exchangers: Interest rates:
crossflow-LMTD, 156 effective, 239
shell and tube coefficients, nominal, 239
156 Internal energy, 10
Heat flow: Investment:
in cylinder, 140 incremental, 247
in fluids, 141-145 total, 246
 INDEX 267

Investment proposals: MARR,246

dependent, 245 Mass, 7
independent, 245 Mass balance, fluid flow, 201
Irreversibility, 21 Material balance:
Irreversible reactions, 180 gas absorption, 98,100
Isentropic expansion, 211 mass transfer, 81
Isobaric process, 15 Matheson depreciation, 250
Isothermal frictional now, 211, 213 Maxwell relationships, 24
Isothermal process, 15 Mechanical effIciency, 217
Mechanical energy equation, 203
Joule, 9 Metering fluids, 217-220
Meters:
Kinetic data: orifIce, 220
differential, 180 pitot, 220
integral, 179 rotameter, 220
Kinetic energy, fluid flow, 203 venturi, 219
Kirchhoffs law, 153 Method, graphical:
Kremser equation: extraction:
distillation, 84 Ponchon-Savarit,95
leaching, 91 rectangular, 96
mass transfer:
Lagrangian, 59 general,82
Laminar flow, 205 McCabe-Thiele, 88
Laplace transform, 60 Ponchon-Savarit, 90
Law: Minimum attractive rate of return, 246
Arrhenius, 179 Molecularity, 175
Dalton, 105 Momentum balance, 202
Duhring, 157 Money, time value of, 235
fIrst, 11
Fourier, 139 Natural convection, 148
ideal gas, 13 Net positive suction head, 217
Kirchhoff, 153 Newton, 7
Newton, 7,216 Newtonian fluids, 205
Planck, 152 Newton's law, 7, 216
second, 18 Nonblack surface, 155
Stefan-Boltzmann, 152 Non-Newtonian fluids, 205
Stokes, 216 NPSH,217
third, 22 Number:
Wien, 152 Graetz, 146
Leaching, 90-93 Grashof, 148
Lean gas, 104 Mach, 209
Length,7 Nusselt, 145
Lever principle, 84, 92 Peelet, 146
Lewis relation, 107 Prandtl, 145
Line, opera ting, 82 Reynolds, 145
Liquifaction, 32, 36 Stanton, 146
Lives, unequal, common multiple, Nusselt number, 145
249
Log mean temperature difference, Opera ting line:
144 leaching, 91, 93
mass transfer, 82
Mach number, 209 Optimization:
Manometer, 200 analytical methods,S 7
Margules equation, 28 constrained,59
268 INDEX

Optimization (Continued) Radiation:

graphical method, 57 to absorbing gases, 155
unconstrained, 57 absorption of, 152-153
Order, reaction, 175 blackbody, 152
Orifice meter, 220 emission of, 151-152
Orsat analysis, 50 between surfaces, 153-155
Overall heat transfer coefficient: wavelength,152
with fouling factors, 145 Radiation heat transfer, 151-156
from individual coefficients, 144 Raffinate, 94
inside, 143 Rate:
outside, 143 absorption, 100-104
resistance form, 144 constant, 176
variable, 144 drying, 112-113
heat transfer, 143-145
Partial pressure, 25 of reaction, 174
Particle motion, 214-216 Rate equations, 182
Pascal,8 Rate of return, minimum attractive, 246
Peclet number, 146 Ratio:
Phase rule, 12 gas-liquid, limiting, 99
Pitot tube, 220 solvent extraction, 97
Planck's law, 152 Rayleigh equation, 86
Point: Reaction:
critical, 13 bimolecular, 176
functions, 11 competing, 175
Polytropic process, 16 complex, material balance, 178
Ponchon-Savarit: consecutive, 175
distillation, 90 coordinate, 28
extraction, 95 elementary, 175
Potential, chemical, 24, 29 endothermic, 54
Power cycles, 31 entropy change of, 30
Power (pump), 217, 219 equilibrium, 29
Prandtl number, 145 exothermic, 54
Present-worth: extent, 28
equal-payment series, 238 first-order, 175
single-payment, 237 Gibbs free energy, 29
Pressure: heat of, 30
definition, 8 mixed,175
partial,25 multiple, 175
reduced, 17 parallel, 175
Probability, in economics, 252- second-order, 175
254 series, 175
Process: side, 175
constant pressure, 13 single, 175
constant volume, 13 unimolecular, 180
real, and second law, 21 Reaction rate, 174
Pump efficiency, 204 representation of, 176
Pumps, 217-218 Reactor:
centrifugal, 218 backmix, 184
head, 217 batch, 180, 183
positive-displacement, 218 CFSTR,184
Pump work, 204 constant-density, 180
PVT behavior: interpretation of, 179
gas, 13 plug-flow, 186
liquid, 18 Redlich-Kister expansion, 27
 INDEX 269

Reflectivity, 153 Stefan-Boltzmann law, 152

Reflux ratio: Stirred tank, 60
definition, 88 Stoichiometry:
minimum, 89 balances, 50
Refrigeration: reaction, 29
air cycle, 34 Stokes law, 216
Carnot cycle, 34 Sudden contraction, 208
vapor-compression, 35 Sudden expansion, 207
Relative volatility, 86 Superheated vapors, heat transfer, 150
Resistance: Symbols, mass transfer, 81
absorption, 100
compound,140 Tait equation, 18
controlling, absorption, 104 Temperature:
overall coefficient, 144 adiabatic-saturation, 106
series, 140 average, 143
Reversible process, 13 definition, 7
Reynolds number, 145 dew point, 106
Rotameter, 220 difference, 142
range, 142
Second law, 18 reduced, 17
Shear, wall, 204 stagnation, 21 0
Sinking-fund, 237 wet-bulb, 106, 113
Skin friction, 204 Terminal velocity, 215
Space-time, 186 Thermal conductivity, 139
Space-velocity, 186 Thermodynamic:
Specific heat: definitions, 22
definition, 13 relationships, 22
mean, 53 Third law, 22
Stability: Time, 7
characteristic roots, 66 Time constant, 60
partial fraction expansion, 65 Transfer functions, 61
Routh's method, 67 Transform, Laplace, 60
Stages: Transmissivity, 151
feed (distillation), 88 Turbulent flow, 205-209
ideal:
general,83 Underflow:
leaching, 91-93 constant, 91
infinite (distillation), 89 variable, 91
minimum (distillation), 89 Uniform-gradient series, 238
principles, 81
Stagnation temperature, 210 Valves and fittings, 208
Standard conditions, 52 van Laar equations, 28
Stanton number, 146 van't Hoff equation, 30
State functions, 11 Variable area conduits, 211-212
Steady-state flow process, 11 Velocity, average, 201