Chemical Reaction Engineering: A First Course

ENGINEERING
ENGINEE149 G
ENGINEERING
ENGINEERING
ENGINEERING
___________ NFFRIty
Chemical Reaction G
Engineering
A First Course
Ian S. Metcalfe
ENGINEERING
ENGINEERING
ENGINEERING
ENGINEERING
ENGINEERING
ENGINEERING
ENGINEERING
ENGINEERING
P. ENGINEERING
ENGINEERING
ENGINEERING
O X F O R D SCIENCE P U B L I C A T I O N S
ENGINEERING
Chemical Reaction
Engineering
A First Course
Ian S. Metcalfe
D epartm ent o f Chemical Engineering, University o f Edinburgh.
Formerly D epartm ent o f Chemical Engineering and Chemical Technology
Imperial College o f Science,Technology; and Medicine
London
OXPORD
U N IV E R S IT Y P R E S S
O X F O R D
U N IV E R S IT Y PR E SS
Great Clarendon Street, Oxford 0 X 2 6DP

Oxford University Press is a department o f the University o f Oxford.
It furthers the U niversity’s objective o f excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid M elbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan South Korea Poland Portugal
Singapore Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade m ark o f Oxford University Press

in th e UK and in certain other countries
Published in th e United States
by Oxford University Press Inc., New York
© Ian S. Metcalfe, 1997, reprinted 2000, 2003, 2005, 2006
The m oral rights o f th e author have been asserted

Database rig h t Oxford University Press (maker)
Reprinted 2011
All rights reserved. No p art o f this publication may be reproduced,

stored in a retrieval system, or transm itted, in any form or by any m eans,
w ith o u t th e prior perm ission in w riting o f Oxford University Press,
or as expressly perm itted by law, or under term s agreed w ith the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside th e scope of th e above should be sent to th e Rights D epartm ent,
Oxford University Press, at the address above
/ ф Л /> m ust n o t circulate this book in any o th er binding or cover
.л. . J And you m ust impose this same condition on any acquirer
C.TopaftfbipoB
si аты н д агы ПМУ-,.\ ч
-к а д е м и к Бейс-. ISBN 978 -0-19-856538-3
■i ятыкг ы fbi;
Printed and bound in Great Britain by CPI Antony Rowe,
•' I X A.b C hippenham and Eastbourne
Foreword
Lynn F. Gladden
University o f Cambridge, Department o f Chemical Engineering
Understanding how chemical reactors work lies at the heart of almost every
chemical processing operation. This particular topic more than any other
represents a true integration of the skills of the chemist and the chemical engineer.
Oxford Chemistry Primers are designed to give a concise introduction to a
wide range o f topics that may be encountered by chemistry and chemical
engineering students. Each primer typically contains material that would be
covered in an 8-10 lecture course. In this prim er Ian Metcalfe presents a
systematic, and very student-friendly introduction to the basic principles of
chemical reaction engineering for students of both chemistry and chemical
engineering. Each topic is illustrated with a wealth of examples, m aking this
prim er not only an excellent self-teaching aid but also extremely useful to
those teaching this subject.
Preface
The book is primarily intended for use by chemical engineering under
graduates to complement their first chemical reaction engineering course at
university. It is also intended to introduce students of other disciplines (such
as chemistry) to the principles o f chemical reaction engineering. It should be
possible for an undergraduate to learn all of the fundamental aspects of
reaction engineering from the book, using it as a self-study guide. F or this
reason the focus of the book is very much on underlying principles and not
on detail. Consequently, a large num ber o f example problems have been
used. These examples come in two forms.
Firstly there are small problems used throughout the book to illustrate
im portant points. These problems are immediately followed by the solution.
It is recommended that the reader tries to solve these problems before viewing
the solution. However, these examples are mainly intended to put the
concepts in context and it is not necessary, or advisable, that the reader feels
he or she m ust be able to solve each one independently before proceeding. The
second type o f problem is found at the end of each main chapter and is based
upon final exams and problem sheets. Again, it is intended that the reader
attem pts to solve these problems before viewing the solutions. However, more
importantly, the solution to the problem should be understood; if the problem
is found to be difficult, viewing the solution is recommended.
I would like to thank Professor R. Kandiyoti for his help, Dr. M. Sahibzada
for reading the manuscript and for making such useful comments, and my
wife, Alison, for her support and patience. I would also like to thank all of the
students I have taught who have helped me refine the ideas presented here.
London I.S.M.
March 1997
Contents
1 Introduction
1.1 Scope
1.2 Starting knowledge required
1.3 The rate expression
1.4 Objectives
2 M aterial balances for chemical reactors
2.1 Batch reactors
2.2 Plug flow reactors (PFRs)
2.3 C ontinuous stirred tank reactors (CSTRs)
3 Calculation o f reactor volume and residence time
3.1 Residence time o f batch reactors
3.2 PFR s
3.3 CSTRs
3.4 Com parison o f PFR s and CSTRs
3.5 Reactors in series
3.6 The recycle reactor
3.7 Problems
4 Multiple reactions
4.1 Parallel reactions
4.2 Parallel reactions of different order
4.3 Series reactions
4.4 Problems
5 The energy balance and tem perature effects
5.1 Tem perature dependence o f reaction rate
5.2 The energy balance
5.3 Steady-state multiplicity in CSTRs
5.4 M ultistage adiabatic PFR
5.5 Problems
6 Non-ideal reactors
6.1 Residence time distributions
6.2 Calculation of mean residence time
6.3 Calculation of conversion from RTD
Solutions
Nomenclature
Further reading
Index
Introduction
1.1 Scope
In Chapter 2 we sow how material balances should be performed for the three
fundamental reactor types used in reaction engineering, namely the plug flow
reactor (PFR), the continuous stirred tank reactor (CSTR), and the perfectly
mixed batch reactor. In Chapter 3 we will see how these material balances can
be used to design reactors (by this we primarily mean how to calculate their
volume or residence time) when one reaction is taking place. We will also
compare the behaviour of the different reactors. In Chapter 4 we will proceed
to look at how the design process must be modified when more than one
reaction is occurring. In Chapter 5 we will see that reactors need not be
isothermal. Therefore, we need to look at how reaction rate depends upon
temperature for different classes of reaction. Then we will formulate the
energy balance for given reactors and use this to investigate the variation of
temperature and therefore reaction rate with time or position in the reactor.
This in turn will be used to allow us to calculate reactor volumes and residence
times for a given duty. Finally, in Chapter 6, there is a brief discussion of non
ideal reactors; this is intended to illustrate the limitations of always assuming
that reactors behave in an ideal manner.
Semi-batch or semi-continuous processes will not be considered. Likewise,
the book will not discuss catalysis, mass transfer, or any related phenomena,
but will lead into a second course covering such material. All reactions are
presented as being homogeneous reactions and reaction rates are always
presented as being volume specific (as opposed to being specific to the mass of
catalyst used). The gas phase will always be treated as ideal.
1.2 Starting knowledge required

It is expected that all readers will have a British А-level or equivalent
knowledge of mathematics such that they are comfortable with the use of
differential calculus in applied situations and that they can perform simple
analytical integration and numerical integration. They should be familiar with
the concepts of material and energy balances and should be able to apply these
in unfamiliar circumstances. This book is primarily intended for readers who
have already had a basic course in chemical kinetics. As some readers may not
have come across reaction kinetics before, there is a short description below,
which, although far from complete, will allow the student to proceed with this
text until they do take a course on chemical kinetics. The reader should also
have some knowledge of thermodynamics for an appreciation of reaction
equilibria, heats of reaction, and energy balances.
1.3 The rate expression

When material and energy balances for batch, CSTR, and PFR reactors are
performed, reaction rate expressions will be required in an algebraic form.
2 Introduction
Reaction rate is usually expressed in terms of the concentrations or partial

pressures of the reactants (and sometimes products) and may be determined
empirically or may, in part, be based upon an understanding of the reaction
mechanism.
Consider the irreversible reaction,
va A ->• vbB
where va and vb are stoichiometric coefficients. The rate of disappearance of

A, rA, can be given by a reaction rate expression depending upon the
concentration of A,
rA = k c C д
where n is the reaction order (not necessarily equal to vA), kc is a rate constant,
1 Throughout this text rates o f reaction and С a is the concentration of A1.
w ill be given in term s of one of the Alternatively, the reaction could be expressed in terms of partial pressures
species participating in the reaction,
(this is very common in the case of gas-phase reactions),
hence the rates fo r oth er species may be
calculated using stoichiom etry, for га = kPP nA
example:
where кp is a rate constant and PA is the partial pressure of A.
Vb
гв = — Га For a gas-phase reaction, k? and kc can be related,
Va
w here rB is the rate of appearance of B.
kc C \ = kPP l
As regards the sign convention for the For an ideal gas,
rate o f reaction, in general, the rate of PAV = N a RT
disappearance of a reactant w ill be
defined as a positive rate and the rate of
where V is the volume of the batch reactor, N a is the total number of moles of
appearance of a product w ill be defined A in the reactor, R is the gas constant, and T is the reactor temperature (which
as positive. There are of course other must of course be in Kelvin).
ways o f handling the sign convention; it
is m ost im p o rta n t tha t w hatever PA = ^ y R T = CART
convention is used is applied
consistently.
where С a is the concentration of A.
k c C nA = kpC nA(RT)n
* p W = kc
This relationship between kP and kc is true for all reactor types.
Consider a general reversible reaction,
va A + vBB vcC + vdD
rAi = k xC A
nAC nBB
where rAi is the rate of forward reaction in terms of moles of A disappearing,
nA and «в are reaction orders in those species, and,
rA_! = fc_ 1C £ C g )
where rA- i is the rate of the back reaction in terms of moles of A being
formed. Therefore the net rate of disappearance of A is,
га = rAi - га- i
га = к \ С пА Съ — k - i C g C p
At equilibrium the net rate of production or removal of A is equal to zero,
0 = JfciC£C£B - k - i C ’g C 'g
Chemical reaction engineering 3
ъ. г "Aс "в — t , f ncr Dn°

— л- 1 ^ с
k i = C £C £
к- i C n£ C %
where К is the equilibrium constant.
1.4 Objectives
This book has a number of different kinds of objectives. First of all there is a
knowledge of the definitions and principles within reaction engineering that
the reader is expected to gain while reading the book. Secondly, there are a
number of skills that the reader should acquire. All of these skills combined
will enable the reader to design basic chemical reactors (each one of these
skills is graded to illustrate how it relates to the overall understanding of the
reader). Thirdly, the reader’s attitude towards reaction engineering (and in a
larger sense engineering in general) should change by reading the book.
1.4.1 Knowledge
At the end of the book the reader should know
(a) the difference between batch, semi-batch, and continuous modes of

operation;
(b) the principles of perfect mixing and plug flow and how they relate to the
three basic reactor models used in the book—the perfectly mixed batch
reactor, the continuous stirred tank reactor (CSTR), and the plug flow
reactor (PFR);
(c) the definitions of space time and space velocity;
(d) that an infinite number of CSTRs in series behave like a PFR and that a
PFR with infinite recycle rate behaves like a CSTR;
(e) that, for a first-order reaction a PFR will give a greater conversion than a
CSTR because of the higher effective concentration of reactants;
(f) the definitions of selectivity and yield when multiple reactions occur;
(g) qualitatively, how the concentrations of reaction intermediates in a series
reaction may vary with reactor residence time;
(h) in what maimer reaction rates are influenced by changing temperature
and composition;
(i) that there exists an optimum temperature for a reversible exothermic
reaction where rate is maximized;
(j) what ‘interstage cooling’ and ‘cold shot’ are and when they are used.
1.4.2 Skills
At the end of the book the reader should be able to
(a) perform mass balances for the basic reactor models so as to derive the
appropriate design equation from first principles *;
(b) apply the basic reactor models for the design of isothermal reactors given
any set of reaction kinetics*;
(c) perform energy balances for the basic reactor models as necessary*;
(d) apply energy balances for the design of reactors given any set of reaction
kinetics*;
(e) apply basic reactor models for the design of isothermal reactors when
multiple reactions are taking place**;
(f) in the case of multiple reactions, find the optimum reactor type and
residence time for given kinetic behaviour and economic constraints**;
(g) for non-isothermal reactors, maximize reaction rates by an appropriate
choice of operating temperature, as necessary**;
(h) design more complex reactor networks, e.g. reactors in series and
parallel**;
(i) generate more complex reactor models as necessary***;
(j) critically evaluate the applicability and limitations of any reactor model
for a given purpose***.
‘Constitutes a basic understanding of the material covered.

**A more advanced understanding.
***A thorough understanding and appreciation of the material covered with the
ability to apply this in unfamiliar circumstances.
1.4.3 Attitudes
At the end of the book the reader’s attitude should move away from:
‘All reactors are too complex to model’ or ‘All reactors are either perfectly
mixed or plug flow’,
and towards:
‘It is possible to describe chemical reactors using models. However, all models
have limitations. It is an engineer’s responsibility to be aware of these
limitations and to choose a model which is of sufficient complexity to give an
answer of the required accuracy.’
2 Material balances for
chemical reactors
The material (or mass balance) for a reactant can be written in a general form
applicable to any type of reactor. Consider a small volume element of a reactor
and what happens over a small interval of time. The material balance for any
individual reactant or product is,
accumulation of = moles entering - moles leaving - moles disappearing (2 . 1)
moles in element element per element per due to reaction per
per unit time unit time unit time unit time
(1) (2) (3) (4)
In general, some of these terms will be equal to zero, e.g. for batch reactors
terms 2 and 3 are zero (if the element is the whole reactor), for a steady-state
process term 1 is zero. We will apply this equation to the three types of basic
reactor illustrated in Fig. 2.1.
In a batch reactor, see Fig. 2.1(a), all of the reactants are supplied to the
reactor at the outset. The reactor is then sealed and the reaction is performed.
There is no addition of reactants or removal of products during the reaction.
The vessel is kept perfectly mixed so that there are no concentration or
temperature gradients.
A plug flow reactor (PFR), see Fig. 2.1(b), is a special type of tubular
reactor. Feed is continuously supplied to the reactor and products are
continually removed. There is no attempt to induce mixing in the reactor. The
velocity profile is flat, i.e. uniform over any cross-section normal to the
direction of fluid motion.
The continuous stirred tank reactor (CSTR), see Fig. 2.1(c), like the PFR,
has a continuous supply of feed while products are continually removed.
However, in this case perfect mixing is achieved, i.e. there are no
concentration or temperature gradients within the reactor (in a similar manner
to the batch reactor).
2.1 Batch reactors

As previously mentioned, the total feed is introduced at the outset and no Uniformly mixed Product
withdrawal is made until the reaction has reached the degree of completion
desired. Fig. 2.1 S chem atic representation of:
(a) a batch reactor; (b) a plug flow
reactor; (c) a continuous stirred tank
1. This is a fundamentally unsteady process. We expect all variables to reactor.
change with time.
2. The mixture in the volume is assumed to be perfectly mixed. This means
that there will be uniform concentrations.
6 Material balances fo r chemical reactors
3. The temperature will also be uniform throughout the reactor—however,

it may change with time.
4. The volume may be kept fixed (most common) or may be varied to keep
the pressure constant (i.e. there are two modes of operation, constant
volume or constant pressure).
1 In fact it would be v ery difficult to use First of all we must identify the element over which we will perform the
any oth er elem ent because w e do not material balance, then we can simplify our fundamental material balance (eqn
know enough about the flow -field w ithin
2.1). Because the concentrations are uniform within the reactor it is most
the reactor. A fuller explanation is
com plicated, however; the fluxes of any
convenient to treat the entire reactor as this element1. With this element, terms
spe cie s into and ou t o f a differential 2 and 3 in eqn 2.1 go to zero as there is no addition or removal of material. For
elem ent in a p e rfe c tly m ixed reactor will reactant A in the volume V between times t and t + A t where,
be ve ry large but this do es not mean that
they are equal. In fact, this inequality of Na is number of moles of A in volume V (the system)
fluxes is needed to m aintain . , £ , moles of A disappearing
concentration gradients at negligible rA is reaction rate defined as (unit time) (unit volume)
levels— negligible in so fa r as one
average concentration can be the material accumulated (term 1) in the time interval is given by,
accurately used for the determ ination of
the reaction rate.
moles accumulated = N^lt+M —N \ |,
and the material reacted in the same time interval in the reactor (term 4) is
given by,
moles reacted = rAV A t
Therefore, using the material balance (eqn 2.1) we can substitute mathematical
expressions for all of the terms,
ЛГА|,+д, —N&\t = 0 + 0 —r^V A t
WaI,+a, Nk\, „ T/
--------A , ------- ГаУ
In the limit as A t —> 0,
cW a
— = - r AV or
,A = - I ^ (2 .2 )
A V dt K '
Equation 2.2 is often known as the design equation for batch reactors. We will
come back to this equation and apply it when we look at batch reactor design
in more detail.
Note that we have defined the rate in terms of the disappearance of A,
hence, a minus sign appears in the equation as d/VA/df is negative (the amount
of A in the reactor is decreasing with time). Furthermore, the greater the
reaction rate, i.e. the larger rA, the quicker A will disappear.
If the rate is defined in terms of the rate of production of a product В we can
say,
, _ , moles of В formed
rn is reaction rate defined as 7— r —:— г-7— r-— ,----- г
(unit time) (umt volume)
(2.3)
2 Once again, fo r sign conventions it is V dt
im po rtan t to think about the physical
m eaning of the equation and to remain
This time there is no negative sign in the equation as the number of moles of В
consistent. in the reactor will increase with time2.
The time necessary for achieving a given degree of reaction can be found
by integrating the design equation with the initial condition, t = 0 ; iVA = NAo
Г fNk 1 cLVa
Jo d t = ~ J nao vV ~ rA
T (24)
3 If w e have m ore than one reactant, e.g.
Remember: the reaction will proceed (unless externally aiTested) until: (a)
equilibrium is reached (reversible reaction); (b) limiting reactant is exhausted A + В - » Products
(irreversible reaction). then the conversion of A and В w ill not, in
The degree of reaction that is required in a reactor is often specified in general, be the same. Mostly, but not
terms of conversion, i.e. xA being the conversion of A3. At any time, t, always, w e w ill w o rk w ith the conversion
of the lim iting reactant.
_ N ao - N a We w ill not use the concept o f extent of
1 4 - i t (2 5 ) reaction in this text.The extent of reaction
is defined as,
In the limit of 0% conversion,
( = N k, - N ,
NA = A'ao Vi
while in the limit of 100 % conversion, w here v/ is the stoichiom etric coefficient
of the ilh species. No m a tter w hich
na = o species is considered, the extent of
reaction w ill then be the sam e (unlike
Differentiating eqn 2.5 we get,
conversion w hich is species dependent).
However, w hen w orking w ith extent of
dxA = - ^ (2 .6 ) reaction, care m ust be taken to w rite the
ÂO reaction in one consistent form so that
and this can be used to change the variable of integration in eqn 2 .4 , the stoichiom etric coefficients rem ain
fixed.
Г N ao dÂ
fA
Jo'-f Jo V rA
(2.7)
A first-order irreversible reaction,

Example 2.1
A -> B; rA = kCA
is to be carried out in a batch reactor. If k = 0.01 s-1, calculate the time to
reach 30% conversion.
From eqn 2.4,

Solution
f_ Л * 1 <WA
J na0 V rA
A —> В; rA = kCA
The reaction rate is expressed in terms of concentration but we must put
everything on the right-hand side of the equation in terms of one variable to
integrate. This can be done by recalling that the concentration is simply the
number of moles divided by the total volume.
rA = kCA = kNA/ V
{=_ f N- dNA = _ l ]n Na_

JNtfi kNA к N ao
8 Material balances fo r chemical reactors
N ao - N a _N a
A N ao
t = —- ln(l —дед) = 35.7 s
2.2 Plug flow reactors (PFRs)

Now that we have ‘designed’ a simple batch reactor we will take a look at
plug-flow reactors.
Tubular reactors are used for many large-scale gas reactions, e.g.
• Homogeneous reactions (tube contains only the reactant and product

gases and any inert gases), e.g.
2NO + 0 2 2NOz (HN0 3 from NH3)
• Heterogeneous reactions (tube is packed with catalyst), e.g.
CO + 2H 2 CH3OH
N 2 + 3H 2 & 2NH 3
In tubular reactors, there is a steady movement of reagents in a chosen

direction. No attempt is made to induce mixing of fluid between different
points along the overall direction flow.
A formal material balance over a differential volume element requires that
we know (or assume) patterns of fluid behaviour within the reactor, e.g. the
velocity profile. The simplest set of assumptions about the fluid behaviour in a
tubular reactor is known as the plug flow (or piston flow) assumption.
Reactors approximately satisfying this assumption are called plug flow
reactors (PFRs).
The plug flow assumptions are as follows,
(a) flow rate and fluid properties are uniform over any cross-section normal
to fluid motion;
(b) there is negligible axial mixing—due to either diffusion or convection.
The plug flow assumptions tend to hold when there is good radial mixing
(achieved at high flow rates Re > 104) and when axial mixing may be
neglected (when the length divided by the diameter of the reactor > 50
(approximately)).
Feed Product
This means that if we consider a differential element within the reactor
(with its boundaries normal to the fluid motion), it can be taken to be perfectly
Differential
batch reactors mixed and, as it travels along the reactor, it will not exchange any fluid with
Fig. 2.2 S chem atic of a PFR showing
the element in front of or behind it. In this way, it may be considered to behave
how differential elem ents may be as a differential batch reactor (see Fig. 2.2). We will see later that there are
considered to be w ell-m ixed batch consequently many similarities between the behaviour of a PFR and the
reactors. behaviour of a batch reactor.
Furthermore, we will make an additional assumption that the reactor is at
steady state. In a batch reactor, composition changes from moment to moment.
In continuous operation at steady state in a PFR, the composition changes with Vr

position, but at a given position there is no change with time. «Т VTc
«А
nTe
We are now able to perform a material balance (we now know enough *A «At
about the behaviour of the fluid within the reactor). *A=
We must choose an appropriate element over which to perform the material
balance. It might seem natural to follow one of these ‘differential batch 0 I l+dl
reactors’, i.e. to perform the material balance over such an element and Fig. 2.3 S chem atic o f a PFR with
integrate over the time taken for the element to travel along the tube. However, nom enclature and show ing a differential
this would lead to the same equation as the batch reactor and, because of its elem ent used in th e form ulation of the
time dependency, would imply that the concentrations in the reactor were m aterial balance.
changing with time (this approach is not incorrect but we must remember that
the time refers to the residence time of the fluid in the reactor). Alternatively,
we choose to observe what happens in a differential element that is in a fixed
position at any point along the length of the reactor (we do not follow
elements—we watch them go by). See Fig. 2.3.
The following symbols are used4, 4 A low er case n is now used to denote
th a t w e are considering m o la r flow rates
I length (m) (and not an ab solute num ber of m oles as
L total reactor length (m) in the case of th e batch re actor w hen an
V reactor volume (m3) upper case N w as used).
A cross-sectional area (m2)

ид molar flow rate of component A (mol s_1)
пт total molar flow rate (mol s-1)
vT volumetric flow rate (m3 s-1)
0 denotes inlet conditions (subscript)
e denotes exit conditions (subscript)
The material balance over the differential element has the usual form,
Accumulation = input — output — loss through reaction
We now substitute in mathematical expressions for each of these terms.
There is no accumulation in the element as the reactor is at steady state.
Therefore, the element always contains fluid of the same composition. The
input and the output from the element are just the molar flows of whatever
component we are considering and the loss through reaction will just be the
reaction rate (which is volume specific) multiplied by the volume of the
element.
0 = « a I; —n A h+dl - rAdV
Dividing throughout by dI the differential is obtained,
dnA dV
0 = - щ - - ' * d7
Dividing by the cross-sectional area and remembering that Adl = dV,

d»A
га = —
dV
(2 .8)
Equation 2.8 is known as the design equation for PFRs. As can be seen, the
higher the reaction rate the faster the molar flow of A will decrease as the
Material balances fo r chemical reactors
reactor length or volume is increased. Whereas in the case of a batch reactor

compositional changes take place with time, in a tubular reactor (or our special
case of a PFR) the compositional changes take place spatially.
The design equation could also be written for a product of the reaction,
d»B (2.9)
dV
For greater reaction rates, the flow of В will increase even faster as we pass
along the length of the reactor.
The dependence of the volume of a PFR on conversion may be found by
performing the integration of the design equation (eqn 2 .8 ),
f dV = V = - ( "AdnA (2.10)
J0 Jn,
ЛА0 r A
The definition for conversion (eqn 2.5) can be rewritten in terms of molar flow
rates,
и д о ~ «А
*A = ■ (2 . 11)
«А0
Differentiating gives,
dwA (2 . 12)
d*A — ~
nAo
Expressing eqn 2.10 in terms of conversion of reactant we get,
rXA nM dxA
= Jo f r\
(2.13)
Example 2.2 A first-order irreversible reaction,

A -> B; rA = kCA
is to be carried out in a plug flow reactor. If к = 0.01 s_1 and the volumetric
flow rate is 10-3 m3 s ~ \ calculate the reactor volume and residence time
required for 30% conversion.
Solution From eqn 2.10,

.= _ Г dnA
JtiM) rA.
B\ ra = kCA = knA/ v j
= _ V t f '”A dnA
кL
Vt ,
nA
«А
= I n ----
к nA0
= - ^ l n ( l - xA) = 3.57 x 10~2 m 3
к
Residence time, г - V/vr = 35.7 s
If we compare the result from Example 2.2 to that from Example 2.1 we can
see that the residence time required in both cases is the same. This is because we
can consider the elements in the PFR to be differential batch reactors. Their
«АО
composition-time history is just the same as if they were in a batch reactor. The
difference between batch and PFR reactors is that composition changes take
place temporally in the first and spatially in the second.
2.3 Continuous stirred tank reactors (CSTRs)

The key feature of a CSTR is that perfect mixing occurs in the reactor. Perfect
mixing means that the properties of the reaction mixture are uniform in all ^"Ae- Сд
parts of the vessel and identical to the properties of the reaction mixture in the *Ae=*A
exit stream. Furthermore, the inlet stream instantaneously mixes with the bulk
of the reactor volume (see Fig. 2.4). Fig. 2.4 S chem atic of a CSTR with
We will also assume that the CSTR has reached steady state. Therefore nomenclature.
reaction rate is the same at every point, and time independent.

Again we must choose an element for our material balance. Like the batch
reactor it would be very difficult to consider a differential element because we
do not know enough about the flow field in the reactor. Furthermore, because
the properties of the reaction mixture are uniform throughout the reactor
volume we can perform the material balance for component A over the whole
volume, V.
Again,
Accumulation = input — output — loss through reaction
We now substitute in mathematical expressions for each of these terms.
There is no accumulation in the element as the reactor is at steady state. The
input and the output from the element are just the molar flows to and from the
reactor and the loss through reaction will just be the reaction rate (which is
volume specific) multiplied by the volume of the reactor.
0 = «АО - «Ае - rAV
«A0-«Ae (2.14)
ГА~ V
This equation is known as the design equation for a CSTR.
The design equation can be rewritten because the conditions within the
reactor are the same as those in the outlet stream and therefore there is no need
to distinguish between them. More commonly, the design equation is simply
written as,
«АО —ИA
га= -~ г ^ (2.15)
The design equation for a CSTR is not a differential equation as it is in the

case of a PFR or batch reactor. The PFR needs a differential equation to
describe it because compositional changes take place spatially and the batch
reactor obeys a differential equation because compositional changes take place
in time. The equation for a CSTR, in contrast, is not a differential equation
because composition does not change spatially or temporally. So we could
ask, if there are no changes of composition spatially or temporally how can the
CSTR behave as a reactor? The answer lies in the fact that there is a
discontinuity where fresh feed is introduced into the reactor. Because of
perfect mixing there is an instantaneous change in the composition and it is
here that the compositional change takes place.
Material balances for chemical reactors
Example 2.3 A first-order irreversible reaction,

A -> В; rA = £Сд
is to be carried out in a CSTR. If к = 0.01 s " 1 and the volumetric flow rate is
10-3 m 3 s-1, calculate the reactor volume and residence time required for 30%
conversion. (Remember that the concentration in the reactor is the same as the
outlet concentration because of perfect mixing.)
Solution From eqn 2.15,

пар — па
V=
га
A В; га = кСА = knA/vj
ИАО — «А
V—
_ / «АО _ Л
к \ па )
П\
But хА = 1 -------
ПАО
1
- 1 : — - *А - = 4.29 х 10~2 т 3
* -т .С1 - *а) к (1 - х А)
Mean residence time, г = V/vj = 42.9 s
As Example 2.3 shows, the volume or residence time required for a given
conversion in a CSTR (for a first-order reaction) is greater than that required in a
PFR or batch reactor (compare with Examples 2.1 and 2.2). This is because when
we introduce fresh feed into the CSTR we immediately dilute it to the exit
concentration and therefore we see lower rates and hence longer residence times
are required. In a PFR the concentration of reactants gradually reduces as we
travel along the length of the reactor so, although the rate at the exit of the PFR is
similar to that in the CSTR, the upstream rates are higher. In a batch reactor at
short times we have higher rates and the rate gradually decreases as time passes.
Imagine having a reaction taking place in the liquid phase where the reactant
is red and the product is colourless. The feed to the reactor will be red. In the case
of a CSTR, the exit stream will be a light shade of red. This will be the same
colour as the liquid within the reactor. In the case of a PFR the exit stream will
also be a light shade of red; however, the colour of the stream will change
gradually from inlet to outlet. For reactions of positive order in concentration,
deeper shades of red would of course correspond to higher reaction rates.
However, not all rates are higher at higher reactant concentrations. We
could be designing a reactor for a reaction which obeyed kinetics in which the
rate depended, for instance, on the inverse of a reactant concentration. In this
case lower concentrations would give higher rates and we would find therefore
that a CSTR would require a smaller volume or residence time than a PFR or
batch reactor for the same conversion. This comparison between CSTR and
PFRs is a subject that we will return to and investigate in more detail in the
next chapter.
3 Calculation of reactor volume
and residence time
We will now apply the material balances or design equations derived in

Chapter 2 to the design of isothermal reactors with one reaction
occurring.
3.1 Residence time of batch reactors

Recalling the design equation (eqn 2.2)
ldAfc
ГА V dt
The initial condition is that at time t — 0, = N \q
Integration gives eqn 2.4,
1 dN a
/ * - ГJ гV rA
Jo Jn
nma0 *
In general, for PFR and batch reactors, where we need to perform an

integration, we will need to put everything on the right-hand side of the
1 There are, however, cases w here we
equation in terms of one common variable. This common variable will m ay use a different varia ble to make the
usually be total conversion1. Recalling the definition of conversion integration easier.
(eqn 2.5),
N ao - N a
Xa = — Tt-------
Differentiating eqn 2.5 gives eqn 2.6,

. dNA
N ao
and eqn 2.6 can be used to change the variable of integration in eqn 2.4
yielding eqn 2.7,
f ‘dt=z [ Xa N a o ^ a
Jo Jo V rA
In general, the irreversible reaction:
A -> Products
has a rate expression of the form
rA = k C \
Expressing this in terms of conversion,
This can be substituted into eqn 2.7,
t - —1 - Г У" - 1— ^ — (3.1)
k N ^ 'J o (1 - x AT
Na0 is a constant and к is a constant (we assume that the reactor temperature
does not change as the conversion changes).
There will be a change in the total number of moles present if the number
of moles of products is different from the number of moles of reactants. For a
gas-phase reaction, this means that, if the reactor is at constant pressure, the
2 When reactions take place in the liquid
phase betw een dissolved species there
reactor volume may change (like a piston) or if the reactor is of constant
is no volum e change o r pressure volume the reactor pressure may change2. To proceed we need to know which
change. type of reactor to consider.
3.1.1 Constant volume batch reactor

The reactor volume is constant and so can be taken out of the integration
(eqn 3.1):
t-z^f
yn- 1
Wn
AO
o' Jo
JO
[XA
*A)
= —f
kC"Ao' Jo
'AO JO
*A
(1
dxA
*A )
and this expression can be easily integrated.

If n ф 1,
-1
kCn~'
коА0 ( 1 - n)
-1
t л [(1 - x A)'~ n - и
kCn~'
AO ( 1 - n)
For n = 1,
r ckA i
= ~
к Jo( (1 -*a ) к
t= - J ln(l - xA) = \ In — - — - and (3.2)

к к ( I - xA)
xA = \ - e ~ k' (3.3)
This means that, for a first-order irreversible reaction, as the residence time of
a batch reactor approaches infinity the conversion of the reaction will approach
unity in an exponential manner.
Alternatively, expressing eqns 3.2 and 3.3 in terms of the moles of A
present,
t = -j- In and (3.4)
к Na
NA = N Aoe' k' (3.5)
3.1.2 Constant pressure batch reactor

Recalling eqn 3.1,
1 ГА
я- l
t = ------г / vn~
ЩVAo'0 Jo
^0 0 -* л )"
For a constant pressure batch reactor, the reactor volume is no longer

constant and must remain within the integration. To perform the integration
a relationship between the reactor volume and the conversion must be
found.
Consider a general reaction,
va A —> v b B + v c C
where all of the reactants and products are gas-phase species. Let us consider
the number of moles of each component present. First of all, from the
definition of conversion (eqn 2.5), the number of moles of A at any time is
given by,
N a - N ao — N \ oXa
If we know how many moles of A have disappeared then we can calculate
from the stoichiometry of the reaction how much В and С must have been
produced,
Vg
Л'в = Л^во -I----- N ao * a
va
VC
Nc — N qо H-----N aqXa
va
The number of moles of inert present will remain constant,
Ni — Mo
.And if we sum over all of the components present we obtain an expression for
'-he total number of moles present in the reactor.
% = л »го+ ( ',в + ^ ~ ‘|> ) ^ а
vb+ vc —va\N ao ,,
£ - 1+( VA
h r - *a = 1 + £axa
J N TO
where eA is a constant that depends upon the stoichiometry and the feed
conditions,
'Ч’в + vc — va \ Nao
«А
-c - va JN ito
If we are dealing with an ideal gas,
PV = NrRT; PV0 = N tqRT
The volume is simply proportional to the number of moles present (at constant
temperature and pressure),
V Nj
V0 = n ^ = 1 + E aXa
V = Vo(l + ёаХа ) (3.6)
Equation 3.6 can now be substituted into the integration (eqn 3.1),
^ S f d + ^ A )"-1 ^
M lo J o ' ' (1 - x A)"
1 Г Ч1+ВАХАГ1 .
kC tfJo (1 -x a )” A
Calculation o f reactor volume and residence time
Example 3.1 Consider the gas-phase reaction,

2A -» В + С + 3D with rA — kC \
Initially 50% A and 50% inert are present. If the pressure and
temperature remain constant, derive an expression relating the volume
of the system to its initial volume and the conversion of A. (Assume ideal
gas behaviour.)
Solution Na = Nao —NaoXa, = Nm + -N aoâ
1 3
Nc = N qq + - N \ oxa, N o — No о + ~Nao*a
N i = Mo
3
Л^т = Л/то + -Л^ао^ а
Л'т | , 3 Aâo ,,3

Tf — 1 + ~TT~Xa- — 1 + T*A
/Vto 2 Л/jo 4
PV = NTRT\ PVо = Л^то/гг
V Nj 3
Vo A^to 4 A
У = Уо( 1 + ^ А)
3.2 PFRs
Again we take the design equation (eqn 2.8),
dnA
rA = ~ d ^
and integrate it to give eqn 2 . 10 ,
у Г dnA
-f JJ nA,
«АО ^
Putting everything in terms of one common variable, xA, by using

eqn 2 . 12 ,
dnA = —«AodÂ
and substituting for nth-order kinetics,
»аоО ~ хк Т
ГА = kC l = ,k»A
^ _= k/ ; nA0(v n XAj gives,
Vj
v= Г ^ d * A..... (3 7 )
Jo kn\Ql ( \ - x A)n
However, the volumetric flow rate, Vj, need not be a constant (it will be
constant for a liquid-phase reaction or for a gas-phase reaction in which the
total number of moles and the pressure do not change). If it is not constant, we
need an expression relating the volumetric flow rate to the conversion. Again
for a general reaction,
3
У дА —^ V B -f- v c C
where all of the reactants and products are gas-phase species, we can look at
how the individual molar flow rates will depend upon the conversion (in a
similar way to that of the batch reactor).
« a = «до — « ao* a
, ув
«в - п в о + — «ао*а
va
«С - ПСо Ч------ ПА0Х А

Va
п\ - Пю
, / Vb + VC - VA \
«Т — «то + I ----------------- )«ао* а
flj А/ 1'в + vc — vA \ п А0
— = 1+ 1 " ----- }— —хА•- 1 + еАхА
«то V vA )'«то
г
Ideal gas behaviour,
Pvj — njRT', Pvto = njoRT
and we will assume that the pressure is equal everywhere within the reactor.
(This is generally a good approximation but is not always true, particularly if 3 C om pare eqn 3.8 w ith eqn 3.6.
the reactor contains a catalyst.) Therefore, the volumetric flow rate is

proportional to the total molar flow rate3,
4 Question: if the volum etric flow rate
ut
—— — 1+ ea x a changes as w e proceed along the length
v to «TO of a reactor, how d o we know w hat the
v i = i>to (1 + £a*a) (3.8) residence tim e will be?

Answ er: w e need to perform an
And substituting into eqn 3.7, integration. If w e consider a differential
elem ent we know that the residence tim e
v = *40 0+
К о Jo rf £a * a T
(1 - x A)n
dbcA in this elem ent w ill be d ( //v r. Summ ing
the residence tim e s of all such elem ents,
rdV
This equation is true for any nth-order irreversible reaction occurring in a PFR. T o ta l re s id e n c e t i m e :
If we take a look at one special case, a reaction for which the volumetric J Vj
flow rate is constant (so eA = 0 and Vj = i>to) and which obeys first-order For real reactors the residence tim e can,
consequently, be very difficult to
sdnetics (n = 1), then the reactor volume as a function of xA is given by,
calculate. A s a result, w hat is known as
th e ‘space-tim e’ is often used. This
(1 - * a ) space-tim e is equal to w hat the
1 residence tim e would be if the volum etric
V = -^ [ ln ( l xA)]f/ or (3.9) flow rate rem ained unchanged at its
(1 - * A) original inlet value.
1 /kV\ V
(T^ =“pUJ=exp<‘') S p a c e -tim e = —
vto
The term 'space-velocity', Sv, is also
« here г is the residence time of the reactor (this is equal to the volume of the
used. This is sim ply equal to the
reactor divided by the volumetric flow rate)4. Therefore, reciprocal of space-tim e,
v - . сраигыров
xA - 1 -kx (3.10) Vto
Sv :
академик С Бейсе?
атынд^'-ы f ЫЛЫ;
k it a h x a h m ■ I;
1
This means that in a PFR (as in a batch reactor, eqn 3.3) the conversion will
approach unity exponentially as residence time is increased.
Expressing eqns 3.9 and 3.10 in terms of the molar flow of reactant, we get,
for a first-order reaction with no volume change,
„
V — ^—Т , «А0
In— or, Л 111\
(3.11)
к nA
nA = n AOe~kx (3.12)
Example 3.2 For gas-phase reactions, rates are often expressed in terms of partial pressures,
e.g A ->• В + 2C; rA -- kPA, к — 50 mol s _1 m -3 bar -1
In a PFR, the feed is pure A at a flow rate of 10 mol s -1 and the reactor
operates at a pressure of 20 bar. Calculate the reactor volume required to reach
50% conversion.
f AdnA f AdnA f *AnAOdxA

Solution V= - / — = - / ТГ =
JnnA0
АО ' A ■ 'Идо *^" A
■'"AO A J0
JO kPA
«А ■ «А0 — n A0x A
«В = nAOxA
nc = 2nA0xA
m — njo + 2nAoxA
p _ Wa p _ Пао^ ~ p = 1 ~ *A p
A их wao(1 + 2 jca) 1 + 2 xa
P «aq( 1 + 2 xa)
■-г
Jo 1 k ( l - X A) P
= nAQ f*a з
( - 2 + T ------- r)dxA
W o (1 - xA)
= ^ [ - 2 х а - 3 1 п ( 1 - х а )Гоа
= ^ ( - 2ха - 31п(1-*а))
Ю ---- - 3 _ , „„ 1П_2 „3
(1.079) m 3 = 1.08 x 10~z mJ
(50)(20)
Example 3.3 Calculate the reactor volume required to achieve 90% conversion in an
isothermal PFR for the second-order irreversible reaction,
2A -> В + С
The feed consists of pure A at a molar flow rate of 1.2 mol s_1. The reaction
rate constant is 6.7 mol s_1 m " 3 bar-2. The pressure of the reactor is 1.4 bar at
the inlet, falling linearly with reactor length, to 1.0 bar at the outlet. The
reactor is of uniform cross-sectional area.
Solution
rA = ^d v , ГД _- kP1
rA
nA
«А = «А0(1 ~ XA ) , Идо =«т, PA = — P
Их
иАо(1 - *а)
гa = к
«АО
- d пА d*A , [ « ао( 1 ~ * а )
-Т7Г- — «А О “ ТТГ = к \ ------------------------ Р2
dV dv I и А0
ПАО&СА
f iV=fJo
Jo k { \ - X Af P 2
7г.г above equation cannot be integrated as the pressure cannot be expressed
in terms of the conversion. However, the pressure can be expressed in terms of
length which can be related to reactor volume.
fv rx nApdxA
/ P2dV =
Jo Jo k ( l - x A)2
I
P = P0 - - A P
■-here I is the length variable and L is the total reactor length.

I «А0 dxA
A P I dV ■
L ' ) - к J0
Jo ( 1 - * A)2 f
Using the length variable in the integration, dV = Adi,
i f XM dxA
'fAp° L
■AP d I
)
_ «АО
к Jo i - x Ay
(1
for the limits of integration remember that, formally, when I — L, xA = xAe)

1 «АО
.3 L к [(1 - х л Г 'В
AL «АО 1
[(P0 - AP)3 - (P0)3] =
3 AP к [(1 ~ Х Ае)
ЗАР nA0 *Ае 3(0.4) (1.2) (0.9)

AL= V = m = 1.11 Щ3
P0l - P1]е -к V
(1-Х де)
1 лАе/ (1.4) —(1.0) (6.7) (0.1)
The principle demonstrated in this example is an important one. We may
encounter variables that cannot be related to the conversion and we therefore
need to relate them to position, in a PFR, or time, in a batch reactor (another
demonstration of this can be found in Problem 5.5).
3.3 CSTRs
In an analogous manner to Sections 3.1 and 3.2, we start with the design
equation (eqn 2.15) and substitute in for nth-order kinetics,
_ « д о - «А _ «А0*А _ n A p X A U j _ «A 0*A *4
га kC A kn\ k n \ „(1 - xA)n
For a gas-phase reaction we use eqn 3.8,

vT = vto(1 + eAxA)
_ nAo 1 (1 + eAxA)n
к Сд 0 (1 —хА)” Xa
If we take a look at the special case of a reaction for which the volumetric flow
rate is constant and for which first-order kinetics apply, eA = 0 ,
VT=VT0,n= 1 ,
Fig. 3.1 Conversion versus residence
tim e for a PFR and for a CSTR (first-order v \ xA
r = — = - - ------- (3.13)
reaction kinetics). i>r к 1 —xA
5 As w e will see in C hapter 6, not all
m olecules in a CTSR have the sam e
where x is the mean residence time5. Rearranging,
residence tim e (unlike a PFR o r batch
reactor) fo r this reason the volum e kx , ..
xA = -- ,- (3.14)
divided by the volum etric flow rate does 1 + kx
not give us an actual residence tim e but
rather a mean residence time. As can be seen from the above expression, the conversion in a CSTR
approaches the ultimate conversion in a manner that depends upon the
reciprocal of mean residence time. This is illustrated in Fig. 3.1. Furthermore,
for first-order kinetics a PFR will always need a shorter residence time to reach
a given conversion or for a fixed residence time a PFR will always reach a
higher conversion.
Expressing eqns 3.13 and 3.14 in terms of the molar flow of reactant,
1 «АО «А /-о c\
т = - ----------- (3.15)
к пА
—ш? <ЗЛ6)
3.4 Comparison of PFRs and CSTRs
We will now make a more detailed comparison of the behaviour of PFRs and
CSTRs. Figure 3.2(a) shows a plot of rate versus concentration, i.e. it shows a
kinetic dependency, for a reaction with positive-order rate dependence of
reactant concentration. Obviously, the outlet concentration of reactant will be
lower than the inlet concentration. A CSTR will operate at the outlet condition
(as the concentration in the reactor will be the same as the concentration in the
exit stream) and the CSTR can therefore be represented by an operating point.
Conversely, the PFR needs to be represented by an operating line: the
concentration in the reactor continuously changes as we proceed from the inlet
0 Хд X Ae
to the outlet.
We can now take the same kinetic expression and rearrange it to plot the
Fig. 3.2 (a) Rate versus concentration
fo r a po sitive-order reaction.The PFR is
inverse of the reaction rate versus the conversion (see Fig. 3.2(b)). Again
represented by an operating line w hile the PFR is represented by an operating line and the CSTR is represented by
the CSTR is represented by an operating an operating point. The volume of the PFR is given by,
point, (b) Inverse of reaction rate versus
conversion for a p o sitive -order reaction.
f XAe d*A
VpFR = «АО I ------
The shaded area under the curve is Jo rK
proportional to the volum e of a PFR while
the total shaded area is proportional to
and therefore the volume of the PFR is proportional to the area under the
the volum e of a CSTR. operating line (the heavily shaded area) in Fig. 3.2(b).
The volume of the CSTR is given by, (a) outlet inlet

*A e Operating point
V c s t r = Л ао)---- for CSTR i
га
Operating line
and is therefore proportional to the total shaded area (the heavily shaded area for PFR or
plus the lightly shaded area) in Fig. 3.2(b). batch reactor
From Fig. 3.2(b) we can easily see that, for a given conversion, the volume
of a PFR will always be less than that of a CSTR for a positive-order reaction.
The PFR tends to operate at higher reactant concentrations that the CSTR, C\ Ca,
since, in the CSTR, instantaneous dilution with the product takes place. This
inlet outlet
means that for positive-order reactions, PFRs will exhibit higher overall rates
of reaction and therefore will have lower volumes than CSTRs for a given
conversion.
Figure 3.3(a) shows a kinetic relationship for a reaction of order less than
zero. Figure 3.3(b) shows the same kinetics transposed to show the inverse
of the rate versus conversion. It can be seen that for such kinetics the volume
of the CSTR (proportional to the lightly shaded area) will be always less than
that required with a PFR (the total shaded area). This is because of the
'dilution’ of reactants associated with the CSTR which, in this case, results
Fig. 3.3 (a) Rate versus concentration
in higher rates. fo r a n e gative-order reaction, (b) Inverse
of reaction rate versus conversion fo r a
negative-ord er reaction. The total
3.5 Reactors in series shaded area under the curve is
p roportional to the volum e o f a PFR while
the lightly shaded area is proportiona l to
3.5.1 C ST R s o f equal volume in series
the volum e o f a CSTR.
(a) Calculate the volume of a PFR and a CSTR required for 90% conversion Example 3.4
of reactant by a first-order reaction:
A — B; rA = k C \ ‘, v j / k = l m 3
(b) Calculate the total volume of two CSTRs (both of the same volume) in
series required for 90% conversion (see Fig. 3.4).
(a) PFR Solution

Cac V jdCA _ ^ CAq
V—
Сдо к Ca к Сде
Сдо
= 10 -АО СА1
С As.
V = 2.3 m 3
CSTR
«АО — n Ae 1>т C aO — С A t _ ut / Cap _ Л
ГА к Сде к \C Ae /
V = 9 m3 CSTRs in series.
Сдо — CAl Vr Cai — Сде

(b) Vi = v 2=
CAi Сде
2 к { C ai ) к \ С Ае )
Therefore,
C ap C ai _ ( C ap V _ C ap
C Al С де \С д ; ) C az
^ =ш,,г=^
Vi = V2 = (l)(V lO - 1) m 3 = 2.16 m 3
0 v *A i *A e
Total volume = Vi + V2 + 4.32 m3

Fig. 3.S Plot o f inverse of reaction rate
versus conversion showing tha t the
If we look at the two CSTRs of Example 3.4(b) in terms of the plot of inverse
volum e of tw o CSTRs in serie s is less that
required of one CSTR alone but m ore
rate against conversion, we can see that two CSTRs in series would be expected
than th a t of a PFR. to have a lower total volume than one CSTR but a greater total volume than that
required for a PFR (see Fig. 3.5). Remember that, for the second reactor, the
volume depends upon the difference in the outlet conversion and the inlet
conversion divided by the rate of reaction in the reactor,
„ « Ap (* A e - * A i)
V i = -------------------
ГАе
From Fig. 3.6 (which shows the behaviour of four CSTRs in series, all
operating at approximately equal individual conversion), we can see that as the
number of CSTRs in series is increased the total volume will eventually
approach that of a PFR. This can be shown rigorously in the following way.
Consider a first-order irreversible reaction performed with N CSTRs in series
(see Fig. 3.7), all of the same volume, V„ and with a constant volumetric flow
rate and the same reaction rate constant, then, from eqn. 3.13,
Ут C a p — C m vt C ai ~ С дг _ vr C a i-i — C ai t^ r C a j v - i ~ C a n
*a 1 к C ai к Cai к C A; к Can
Fig. 3.6 Plot of inverse of reaction rate

versus conversion for fou r CSTRs in Therefore,
serie s (all w ith approxim ately equal
conversions). C ap _ C ai _ C u -i _ C a jv -i
C ai Ca i Cai Can
If we multiply together all of these N relationships we get,
( C ai-i Y V _ Caq
\ C ai / Can
C a O. «АО C Ai, nAi C a /. i , «а/-1 Ca N-I >n AN-l
A ,R t X A,R2 ■ u _ XA ,RI u X A,RN
CA2, >, nA2

А2
CAf, n Ai C a v , йдлг
Fig. 3.7 S chem atic showing N CSTRs in series.

Substituting into the expression for the volume of any individual reactor,
1/N
V i= 4
V' ~ к № ) -•]
Hence, the total volume for all N reactors is,
NV t
Ш ‘'Ч
As N oo
\/N
C ap \
Уюы = Lim ^
(
N-+oc к \ C a n )
But,
( ^ ) /=exp|^ 4^)1
and as N -> oo, because the conversion in the series is finite, then the
exponent will tend to zero and therefore the exponential can be re-expressed as
a Unear approximation,
exp
Substituting this into the expression for the total volume,

^1
Vtmai = Lim ^
N->oО к N
vt , ( Сдо \
This expression is, of course, the same as the expression for a PFR reactor
compare to eqn 3.11). For the same conversion, as the number of CSTRs
increases, the total volume approaches that of a PFR (this is true for all types
of kinetic behaviour—not just first order). This is consistent with our
understanding of these two reactors. Previously we made an analogy
between a PFR and infinitesimally small batch reactors. Each differential
element in the PFR could be considered a batch reactor as we followed it
-Trough the PFR. However, if we fix our frame of reference to a particular
rlace along the length of the PFR and now, instead of following each
element, we observe what happens in that fixed element, the element will
appear to be an infinitesimally small CSTR—it will be perfectly mixed on
this scale and will have flow in and flow out from and to the adjacent CSTRs.
Therefore, we would expect to be able to model our PFR as an infinite set of
differential CSTR reactors in series.
3.5,2 C ST R s o f different volume in series

Now let us consider a number of CSTRs of differing volume that we intend to
-se in series. How are they best employed to maximize conversion? In the
case of first-order kinetics, we have already shown that the conversion in a
CSTR is independent of its feed concentration. For a constant volumetric flow

rate the conversion in a CSTR (eqn 3.14) is given by
kx
*A = 1 кт
r o r’
1
(1 - x A) =
1 + кт
so conversion is dependent only on the rate constant and the mean residence
time and not on the inlet concentration. If we have N reactors in series (see
6 The analysis does not only ap ply to Fig. 3.7), we can derive an expression relating the overall conversion to the
CSTRs as shown. conversions in the individual reactors.6
«Ai = « ao(1 - * a,ri )

«А2 = « A l(l -*A,R 2) and
nAN — «AiV-l(l -Â.iw)
where x a ,r ; refers to the conversion of reactant A in the rth reactor.
Therefore,
Wan = ”Ao(l —*A,Rl)(l —*A,R2) • • • (1 ~ *A,RJv)
and the overall conversion of the series, xa , is given by,
Xa = 1 ---- — = 1 —(1 —XAiR i ) ( l —*A,R2) • • • (1 ~ *A,Rn) or,

n A0
(1 —д:А) = (1 ~ XA,Rl)(l —*A,R2) . .. (1 —xa.rn) (3.17)
Therefore, for N CSTRs in series with a first-order reaction,
( i " *a) = ( Г + ^ х Г Т Щ Т Г о Т ь л О
and we can see that the order of the individual reactors cannot influence the
overall conversion.
What happens with a first-order reaction is a limiting case. For other
reaction orders the sequence of reactors is important. If the reaction kinetics
are of order greater than one it is now important for us to keep concentrations
as high as possible for as long as possible to benefit from the associated high
rates of reaction. Therefore, we should place the smallest CSTRs first (see Fig.
3.8(a)). Conversely, if the reaction is of order less than one it is advantageous
to lower reactant concentrations as early as possible. Therefore, the largest
CSTRs should be placed first (see Fig. 3.8(b)).
There is also the special case where the reaction is zero order and the size
7 We w ill not discuss the case of negative of individual reactors becomes unimportant (and therefore the order must be
reaction orders. M ultiple solutions are unimportant); the only thing that matters is the total reactor volume
available and analysis is m ore com plex. available7.
3.5.3 C ST R s o f optimum volume in series

Now let us consider how to minimize the total reactor volume (for a given
conversion) if we are allowed to choose the sizes of individual reactors in the
reactor chain (a single optimum exists in all cases for n > 0 ; for the zero-order
9 Again the analysis for negative reaction case only the total volume is important8). For instance, consider using two
ord e rs is beyond the scope of th is text. CSTRs to achieve a particular conversion in a system displaying positive-
Rg. 3.8 (a) For reactions w ith po sitive-order kinetics small CSTRs should be used e a rly to
-.axim ize conversion, (b) For reactions w ith ne gative-order kinetics large CSTRs should be used
ia rty to m axim ize conversion.
rider kinetics. Figure 3.9 shows three choices of reactor sizes. In all three
cases,
Vmax + Vi + V 2 = constant
-here the constant is the volume that would be required for one CSTR
:perating alone. V\ and Vi are the volumes of the individual reactors and,
rserefore, if we are trying to minimize their sum we should maximize Vmax- In Fig. 3.9 Plot of inverse o f reaction rate
rig. 3.9(a) the first reactor is much smaller than the second, in Fig. 3.9(b) the versus conversion show ing the volum e
reactors are of a similar size, and in Fig. 3.9(c) the second reactor is much of tw o CSTRs in s e rie s ; (a) the first
•mailer than the first. It can clearly be seen that the intermediate case gives the reactor is m uch sm aller than the second;
(b) the reactors are of sim ila r siz e ; (c)
smallest total reactor volume. In fact, it can be shown (see Example 3.5) that
the second reactor is m uch sm aller than
гэг first-order kinetics the reactors should be chosen to be of equal size to the first.
minimize the total volume. For kinetics of order n > 1, the first reactor should
re smaller than the second reactor. For n < 1, the first reactor should be larger
nan the second. The saving in total reactor volume by having reactors of
afferent size tends to be very modest and is therefore, in general, not
-orthwhile economically.
A given conversion is to be achieved using two CSTRs in series. The total Example 3.5
reactor volume is to be minimized. Show that, for first-order kinetics, the
reactors should be chosen to be of equal size.
For a CSTR (eqn 3.14), Solution

kr
Xa -
1 + kr
1
(1 - * A) =
1+ kr
For two reactors in series (eqn 3.17),

(1 - XA) = (1 - *A,Rl)(l - *A,R2)
For two CSTRs,
(1 - xA) =
(1 + fc ti)(l + kx2)
= (1 + & ti )(1 + kx2)

(1 - *a)
Hence, for a given conversion, xA, since A: is a constant, x\ is a function of x2
alone. We wish to minimize the total volume or total residence time, (ti + x2),
by choosing an appropriate value of r 2,
d , ч „
j - ( t i + r 2) = 0
ат2
* L + 1 = 0 ;£ L = -1
d t2 a z2
Remembering that,
1 = (1 + f c r i ) ( l + kr2)
(1 - xA)
and differentiating with respect to r2, with *д constant,
0 = k ^ p - ( 1 + k x 2) + fc(l + kxi)
d r2
d ri
Substituting for - —
d t2
&(1 + kx2) = fc(l + kx\) or,

t2 = ti
3.5.4 PFR and C STR in series

When PFRs and CSTRs are both available and to be used in series, the order
of the reactors can affect the total conversion. The conversion in any one
reactor will, in general, be dependent upon the inlet concentration to that
reactor. Therefore, the overall conversion for a CSTR and PFR in series will,
in general, depend upon the order in which the reactors are placed since the
order will affect the inlet concentration to each reactor. There are two
exceptions to this. The overall conversion for a CSTR and a PFR in series is
independent of reactor sequence for first-order reactions as the individual
reactor conversions are independent of inlet concentration. The overall
conversion is also independent of sequence for zero-order reactions as only the
total reactor volume available is important.
Let us consider the case of first-order reactions. We have already seen that
for a constant volumetric flow rate the conversion in a CSTR (eqn 3.14) is
given by,
kxCSTR
A,CSTR J £Tcstr
which is dependent only on the rate constant and the mean residence time. The
conversion in a PFR (eqn 3.10) is given by,
*A,PFR - 1 - e~kX™
The total conversion for the combined CSTR and PFR is then given by (using
eqn 3.17),
g —fctPFR
( 1 — x A ) = (1 — * a , c s t r ) ( 1 — x a , p f r ) = . , ,-----------
1 + K rcSTR
As the conversion in both a PFR and a CSTR is independent of inlet

concentration, the total conversion for the two reactors in series must be the
?ame regardless of the order in which the reactors are placed.
For reaction orders greater than one, high rates (at high concentrations)
siiould be taken advantage of if conversion is to be maximized. As a PFR
lakes advantage of the high reaction rates it should be placed before the CSTR.
If more than one CSTR is available, then the smallest ones should be used first
immediately after the PFR) so as to keep concentrations higher. However, if
the reaction order is less than one, then it is better to dilute the stream as early
^5 possible. Therefore, the conversion would be greater if the CSTR were
г laced before the PFR.
A CSTR and a PFR are to be used in series to decompose a reactant A via a Example 3.6
second-order process with rate constant of 0.002 litre mol-1 s_1. The reaction
s performed in the liquid phase. The inlet concentration of A to the reactor is
5 mol litre-1 at a flow rate of 0.02 litre s_1. The volume of the CSTR is 2 litres
a d the volume of the PFR is 2 litres. What order should the CSTR and PFR
re placed in to maximize conversion? What is this maximum conversion?
The PFR should be placed first to take advantage of the higher concentrations Solution
iad therefore reaction rates.
-here Cai is the concentration between the reactors,

У_ vt Cap - CAj
к Сдо СЛ1
CAi 1
сАО 1 + ^гС д р
As ктСАО — 1> C Ai = 0 . 5 С Ар
ТЪе PFR is then followed by a CSTR,

у __ C . \ e )
— 1 ± у /Т + Ш С ^
A e = ---------- 2 kr----------
с _ (= 1 ± ^ ± 1 )о .
Сде = О.ЗббСдо
63.4% conversion
If the CSTR were placed before the PFR, then the conversion achieved would
be less.
_ *>t(Cao ~ Âi)
-1 ± V I + 4A:t C ao
C A i = ---------- 2 &r----------
Cai = ^ ^ C ao - 0.618Сдо
This is followed by the PFR,

d«A dCA . _2
'•* = - d V = - ^ d V = t C *
f ^ d C A _V Сде
^ L k C i~ l CA Cai
Vj Cai — Сде
v =
^ Сд^Сде
CAe 1 1 1
= 0.618
Cai 1 "Ь ктСд! 1 -Ь 0.618&тСдо 1 -|- 0.618
Сде = (0.618)(0.618)Сд0 = 0.382САО

This gives an overall conversion of 61.8%.
We can see that there is an advantage to be gained if the PFR is placed
before the CSTR for a second-order reaction. However, the increase in
conversion is modest (63.4% versus 61.8%).
3.6 The recycle reactor

Figure 3.10 depicts a recycle reactor. Fresh feed is mixed with a recycle stream
and then fed to the PFR. The outlet stream from the reactor is split; part of the
stream is recycled to the reactor inlet and the rest is the product stream. We
define the recycle ratio in the following way,
moles returning
moles leaving system
If we first look at the reactor alone and ignore the fact that there is a recycle,
we can write down an expression for conversion (known as the per-pass
conversion),
(*+l)vTe
( Л + 1 ) 1 'Те
Fig. 3.10 S chem atic o f a recycle reactor.
nA\ — (R + 1)nAe
xA = (3.18)
«Ai
If we now treat the reactor and its recycle to be one ‘overall’ reactor then,
'.coking at the inlet and outlet streams, we can write down an expression for
die overall conversion,
v __n A 0 n Ae v .
ЛА —------------ . «Ае = Яао(1 —л a ) (3.19)
ПАО
We now need to undertake some process analysis to understand the behaviour
of the system. Performing a material balance at point A in Fig. 3.10,
nAi = Ядо + RnAe (3.20)
Substituting eqn 3.19 into eqn 3.20,
wAi = «ao(1 + ^ —ЛХ^а) (3.21)
Substituting eqn 3.21 into eqn 3.18,
и а о (1 + R - RX a ) - waq(1 + R)(l - XA)
XA nA0( l + R - R X A)
'‘ =(idbw < 3 ' 2 2 )
So when R = 0, xa = X a , i.e. per-pass conversion and overall conversion are 9 The fraction recycled, a, is also
tbe same thing (which is of course to be expected as the reactor is then simply som etim es used,
in ordinary PFR). As R tends to infinity for finite XA then xA must tend to
zero9. Я+ 1
Now we need to be able to calculate the reactor volume. The volume will Rearranging,
be given by the usual integral from the inlet condition to outlet condition,
R= -
<Л+1>"Ае dnA 1
-f Га
(3.23) Substituting into eqn 3.22.
=
XA
In performing the integration we will leave the variable as nA because the use xa
(1 + * - * X A)
of conversion as the variable can introduce confusion (the inlet conversion to (1 - a )X A
die reactor is not zero as part of the product stream is being recycled). (1 - a X A)
For a first-order reaction,
па
rk = kCA = к
Vj
Let us assume that we are dealing with a reaction for which there is no volume
change, i.e. a liquid-phase reaction or a gas-phase reaction with an equal
number of moles on both sides of the reaction. Then,
UTO = Vje
VTi = Vr — (R + O^Te = (R + lX’TP
m , nA , ПА
rA = kCA = к — = k-
ut (R + l)^ o
Substituting the above into eqn 3.23, the integral for volume becomes,
M R+ 1
,(й+1)ЛАе ( f l+ l) i> r o dnA
JnA
/ПД)i=
= n
ЛА0+ЙПАС к ПА
(R + l)t>ro г,_ -|(Я+1)ВА.

fr L А-1лА0+Япас
__ (R + 1)^то ^ пар + RnAe
к (R + l)nAe
Now we will consider what happens in the limits of R tending to zero and
infinity.
As R tends to zero,
l’to . «А0
V = — In —
к nAe
as would be expected because the reactor behaves as a simple PFR (compare
to eqn 3.11).
As R tends to infinity,
v _ (R + l)i>Tp ^ j 1 + «ар - nAe
(R + l)«Ae
As R increases this exponential can be approximated; for small x,
ln(l + x ) & x,
у _ (R + l)l>TP nA0 ~ wAe _ UTP nA0 ~ иАе
к ( R + 1)nAe к nAe
This is of course the expression that would be expected for a CSTR (compare
to eqn 3.15). So a PFR with an infinite recycle ratio behaves as a CSTR. By
using the recycle we are effectively inducing mixing in the reactor. As the
recycle ratio approaches infinity the concentration gradients within the reactor
become negligible (as far as determining reaction rates are concerned) and this
is analogous to perfect mixing. Therefore, we would expect the recycle reactor
to behave like a CSTR. This is consistent with eqn 3.22 which shows that as R
is increased the per-pass conversion must go to zero, i.e. concentration
gradients disappear.
Example 3.7 For a first-order, irreversible reaction,

A -*■ В; rA — kCA\ кт — 2
sketch xa (per-pass conversion) and XA (total conversion) versus fraction
recycled, a, for a PFR with recycle.
PFR: recall eqn 3.10, Solution

*A = l - e - * r
CSTR: recall eqn 3.14
kx
XA=TTkx
a = 0 , overall conversion is the same as per-pass conversion,
xA = XA = 1 - e“ 2
a = 1, there is negligible per-pass conversion and the reactor gives an overall
conversion the same as that for a CSTR, xA — 0; XA — 2/3
See Fig. 3.11.
Fig. 3.11 Plot of conversion versus
3.7 Problems fraction recycled for a PFR w ith recycle.
Both per-pass conversion, xA and overall
3.1 A gas-phase reaction is performed conversion, XA, are shown.
mao _ 2 n во _ 1
m a batch reactor (constant volume),
«то 3 ’ лто 3
A -> 2B гд = kP A
P = 1 bar; &i = 0.05 mol m -3 s~'
Initially, 50% A and 50% inert are present bar- 2 , k_i = 0.025 mol m “ 3s_l
in a volume basis. bar- 1 , i>fo (600 К, 1 bar)
= 1 litre s“ \ T = 600 K , R = 8.314
Po = 1.00 bar, Pg = 1.38 bar,
J m o l-1 К " 1
-here P0 is the initial pressure and Pe is
ihe final pressure (a) W hat are P a , P b at xA = 0.25?
(b) W hat is the reaction rate at
t = 170 s, T = 500 K,
x A — 0.25? (c) How many moles o f A
R = 8.314 J m o r 1 K~' are fed to the reactor, i.e. пдо? (d) What
a) What is the final conversion? reactor volume is required for x \ = 0.25?
>b) Show that r^V = kRTN*. (c) W hat
ts the value o f k l 3.4 A second-order liquid-phase reac
tion is to be carried out in a PFR:
3.2 It is intended to carry out a A+ В Products ra = кСАСв\
gas-phase dimerization reaction in an
к = 10~6 m3 s“ ' mol” 1
isothermal batch reactor,
2 A -+ В The inlet concentration o f both A and В is
103 mol m -3 and the flow rate is 10-5 m3
The reaction is second order: rA = k C \ s~‘. Calculate the reactor volume re
A conversion of 90% must be achieved. quired for 50% conversion.
Calculate the ratio of the residence
lin e of a constant volume reactor to the 3.5 Consider a gas-phase reaction per
residence time of a constant pressure formed in an isothermal PFR,
reactor. Initially, both reactors are o f the
A + B —»• С
same volume and at the same pressure,
and no product, B, is present. which is zero order with respect to A and
Why is the ratio o f residence times first order with respect to B:
greater than, or less than, unity? (Both rA = kP B-, к = 2.3 x 103 mol s-i
reactors operate at the same temperature m~3 bar-1
and the mixture behaves as a perfect gas.)
At a total feed rate of 120 mol s_l
calculate the reactor volume required to
3.3 A gas-phase reaction is performed
obtain 45% conversion of A. The feed is
ш a CSTR,
A and В in a molar ratio o f 2:1. The total
2A О В rA = k \ P \ - k - \ P B operating pressure is 20 bar.
3.6 T h e first-order, reversible reaction, (a) Calculate the percentage conversion of

A. (b) W hat will be the new overall
A <3-B rA = fe ]P A - k - i P B
conversion if half of the reactor outlet is
is conducted in an isothermal PFR of recycled and mixed with the feed? (c)
volume 2 x 10_3m 3. The total pressure is W hat is the limiting conversion reached
5 bar. The rate constants are 1.2 and 0.8 as the fraction recycled is increased?
mol m -3 s-1 bar for the forward and
reverse reactions respectively. The feed is
pure A at a rate o f 0.01 mol s_1.
Up until now we have only considered what happens if just one reaction is
taking place in the reactor. However, the desired reaction is often one of a
number of reactions occurring. Consider for example the oxidation of ethylene
in air. The desired product is ethylene oxide (Reaction R4.1) but total
oxidation products, carbon dioxide and water, are produced through Reaction
R4.2. Furthermore, our desired product can itself further react to total
oxidation products through Reaction R4.3.
C2H 4 + ^ 0 2 ^ C 2H40 (R4.1)

C 2H4 + 3 0 2 -> 2C 0 2 + 2H20 (R4.2)
C 2H4O + 2 1 0 2 -* 2C 0 2 + 2H20 (R4.3)
In a case such as this, the important performance indicator is the amount of

ethylene oxide produced and not the amount of ethylene that is consumed.
Reactions R4.1 and R4.2 are said to be in parallel; they both compete for the same
reactant. Reactions R4.1 and R4.3 are said to be in series; ethylene oxide must be
produced by Reaction R4.1 before it can be consumed by Reaction R4.3.
To proceed it is helpful to introduce some definitions. Consider a general
reaction network which includes both series and parallel reactions,
vaiA -» v>biB + vciC
ЧлгА —y VD2D
VB3B —»• уезЕ
where vAl is the stoichiometric coefficient of A in reaction 1, etc. Using A as
the reference reactant we can define the yield and selectivity of the products
here we will use the notation corresponding to a continuous reactor). The
yield of В (with reference to the amount of reactant, A, fed to the reactor),
J b / a , is given by,
(4.1)
I'm n Ao
: .e. the yield of В is the number of moles of В formed for every mole of A fed
to the reactor, with the stoichiometric coefficients chosen so that if there are no
competing reactions the yield will be equal to the conversion of A1.
The overall selectivity for В (with reference to the amount of reactant, A,
that has disappeared), Sb/ a , is given by,
, A,
VBl «АО - иA
This is the same as saying that the yield of В is the num ber o f m oles of A required (to produce the
B). v^ i (« b —«BoV^Bb divided by the total amount of A fed to the reactor, лдо. If we compare this
-efmition o f yield to the definition o f conversion (the total num ber o f moles o f A reacted divided
by the total A fed to the reactor), we can see that yield and conversion are closely related (as
previously m entioned they are the same if there are no competing reactions). In fact, the yield can
ne viewed as an individual conversion to a given product.
i.e. the overall selectivity for В is the number of moles of В formed for every
mole of A reacted, with the stoichiometric coefficients chosen so that if there
are no competing reactions the overall selectivity will be 100 %2.
Using eqns 4.1 and 4.2 the yield and overall selectivity can be related,
Ув/ а = яа^ в/ а (4-3)
Rather than evaluate selectivity based on the reactant and product flows in and
out of the reactor, we can also evaluate selectivity at any point in the reactor
(continuous reactor) or at any time (batch reactor). This selectivity is known as
the local (continuous) or instantaneous (batch) selectivity, Фв/ а ,
, »ai dnB
Ф в /а = -------- j — (4-4)
vbi dnA
Note that eqn 4.4 could be arrived at by writing eqn 4.2 for a differential
element of reactor length (PFR) or differential element of time (batch reactor).
The yield, overall selectivity, and local (or instantaneous) selectivity are
defined in the same manner for all of the other products3.
4.1 Parallel reactions

Now consider the simple reaction scheme,
A —> R тД1 — k\CA (R4.4)
A ->• 2S fa 2 = k2CA (R4.5)
where rAl is the rate of disappearance of A through Reaction R4.4 or reaction
1, and rA2 is the rate of disappearance of A through Reaction R4.5 or reaction
2. Therefore, the total rate of disappearance of A, rA, is simply the sum of the
two individual rates,
га = rAi + rA2 (4.5)
For simple parallel reactions (i.e. only one product produced per reaction), the
sum of the individual product yields must equal the conversion of reactant,
jca = Yi/A (4.6)
i = products
Equation 4.6 is true because the yield of each product is simply equal to the
fraction of reactant that reacts to produce that particular product. So if we sum
all of these fractions we must recover the total fraction of reactant that
disappears.
The proportion of the total reaction which produces the desired product is
the overall selectivity for that product,
Sg/t = j m = (4 .7)
XA IR/A + Ys/A
S s, a = ^ = y Ys!Ay (4-8)
xA 1 r / A + 1 s /a
2 This is the same as saying that the selectivity for В is the num ber of moles o f A required (to
produce the B), uAi (« b - «во)/«вь divided by the total am ount of A reacted, nAо - «А-
3It is best to always think o f yield and selectivity in terms o f the amount of reactant required to make
the product (rather than the amount of product formed) divided by the reactant supplied (yield) or the
reactant consumed (selectivity). This avoids problems associated with the stoichiometry.
The instantaneous or local selectivity is based upon the instantaneous or local

reaction rates,
0R/A = — =
ГA rA 1^+ ГA2
(4.9)
0 s/ a = — = (4.10)
rA 1
ГA + r A 2
4.1.1 Batch reactors and PFR s
Consider again Reactions R4.4 and R4.5. For a batch reactor (PFR behaviour
is analogous except that residence time is used instead of reaction time),
1 dN a
Га = - y - f a - ; rAi + rA2 = fc]CA + k2CA (4.11)
*•
r ^ = - r (tl+w,
•'Wao Jo
Na = NAo exp{—(/cj + fc2)?} (4.12)

And we can see that A will disappear with a time constant that depends upon
the inverse of the sum of the rate constants (compare to eqn 3.5 for a single
reaction). What we are more interested in is how the number of moles of R and
S will vary with time.
From eqns 4.9 and 4.10 the instantaneous selectivities are given by,
Фк / а ---- ---- — = T ~ t r an d ; (4 Л З )
r A l + r A2 k \ + kn
r A2 k2
0S/A = ----- ;----- = 7— —r- (4.14)
ГA1 + rA2 k\ + k2
The instantaneous selectivities tell us what fraction of reactant is consumed by
each reaction, i.e. what fraction of reactant is responsible for formation of each
product at any given time. Note that, in this case, both of the instantaneous
selectivities are constant and, therefore, the overall selectivities must be equal
to the local selectivities. Let us show this rigorously by considering the whole
reactor. We can form two design equations, one each for the two individual
reactions,
1 cWA,
rA 1 = y ~ d f = k l C A ( 4 Л 5 )
1 dNA2
Г А 2 = - - ^ = к2СА (4.16)
where NAi and NA2 are the number of moles of A that disappear through
reactions 1 and 2 respectively4. If we add eqns 4.15 and 4.16 together,
1 cWAi 1 dNA2
rAi ^Аг = у + у = k\ CA + k2CA (4.17)
4 JVai and N/a are closely related, through the stoichiometry, to the amounts of products formed
and equations 4.15 and 4.16 are therefore similar to design equations for products.
But because, N& = jVAo —Nai —Â 2 (4.18)
dN A _ (W A i cW A2
dt dt dt
If the above equation is substituted into eqn 4.17, then eqn 4.11 is recovered.
However, by treating the two reactions separately we can get the extra
information that is required to know the product distribution.
We can divide the design equations, eqns 4.15 and 4.16, by one
another,
(4 1 9 )
dNA2 k2
Integrating eqn 4.19 with the initial condition, N&\ — N a 2 = 0,
s ar2 = hr
N (4-20>
5 If all o f the reactions o c cu rring are of The above equation applies for any set of irreversible parallel reactions with
equal reaction rate orders, then in a equal kinetic orders. In such a case, the moles of reactant that disappear due to
batch reactor,
a particular reaction are simply proportional to the rate constant. This means
N ai that if we want to change the reaction selectivity we can only do this by
Wa2 ^2 changing temperature and so changing rate constants5.
and in a continuous reactor, We can now proceed to obtain expressions for the number of moles of A
ПА1 that disappear through each individual reaction,
ЛА2 к2
As Na = N до —Na\ —N A 2
and the selectivities can only be
m odified by changing the rate constants using eqn 4.20,
(the choice of reactor cannot influence
selectivity). Rate constants are a function iVA =iVA0 - N A i ( l + ! )
of tem perature alone,
f c i = ^ exp( - ^ l ) N ai = 7 ~ ~ ~ T ( N ao —N a ) (4.21)
h +k2
*a = *'2 e x p ( - ^ ) N A 2 = r ~ r (.NA0-NA) (4.22)

h + k2
w here к \ and к '2 represent the
a p propriate A rrhenius pre-exponential
The overall selectivity for R is given by the number of moles of A that
factors and Л£й and Д E2 are the
activation energies associated with
disappear through reaction 1 divided by the total number of moles of A that
reactions 1 and 2 respectively. disappear through both reactions,
Therefore, the rate o f one reaction
relative to another can be m odified by S R /A = Nm = = -A L _ (4.23)
changing the tem perature, R/A N a i + N a1 N a o -N a h+ k2
NM _ k , _ k \ . . J (A £ 2 - A £ i ) 1
and the overall selectivity for S is given by,
Wa2 k2 k'2 P1 RT j
If > АЕг then an increase in

tem perature will increase reaction rate 1
5 — — ——— (4.24)
S/A iVAi+iVA 2 N ao - N a h + k 2
m ore than reaction rate 2 and the
selectivity for the firs t reaction will The overall selectivity is therefore equal to the instantaneous selectivity
increase. (compare eqns 4.23 and 4.24 to eqns 4.13 and 4.14) as expected. Although
this may seem trivial, the principle demonstrated is important. The
instantaneous or local selectivity depends upon the instantaneous or local
reaction rates while the overall selectivity depends upon integration of the
design equation for the relevant reaction.
It is now straightforward to get expressions for the moles of R and S
present6 as a function of NAi and NA2.
N r — N ro + N A\
Using eqn 4.21,
N r = N ro + ' (N ao ~ N a )
k\ + k2
and we already know how the amount of A depends upon time (eqn 4.12), so
substituting,
N r = N r0 + j —^ T j^ N A °[l - exP (—(*i + *2>}]
Likewise, Ns — NSo + 2 /VA2
Fig. 4.1 Reaction netw ork from

= NSo + £ NM [l - ехр{-(&! + k2)t]j
Example 4.1.
Consider the reaction scheme in Fig. 4.1. All reactions are first order in A with Example 4.1
rate constants,
ti = l x 10-3 s-1 , k2 — 2 x 10-3 s_1, Лз = 3 x 10~3 s-1
What residence time is required to achieve a 10% yield of В in a batch reactor?
Orders are equal so, ^ ^ ± ^ ^ Solution

k2 2 NA2 N q ki 3 Na 3 Afo
N q = 2Nb; Afo = 3jVb

Because the rate of reaction 2 is twice the rate of reaction 1, two moles of С
will be produced for each mole of В produced. Furthermore, because of the
one-to-one stoichiometry, the number of moles of A consumed must equal the
total number of moles of В, C, and D produced.
N Aox a = N b + N q + N& = 6Nft
We require a 10% yield of B,
— = — = Y =0 1
N ao N ao B/A '
N ^ + N c + N& Nn
*A = ------- r:----------= 6 —- = 0.6
N ao N Ao
^Ув/ а = 0 . 1 ; Ус/а = 0.2; FD/ a = 0.3; SB/ a = ^
N a — N ao exp[-(fci + k 2 + k3)t]
6 Notice that we have not had to use the reaction stoichiometries for any of this analysis until we
nnally calculate the amounts o f the products formed.
t = ------------------ In — = 4 — Г ln(0.4) = 153s

*i -I- *2 + h N\o 6 x 10
Example 4.2 The following reactions are performed isothermally in the gas phase in a PFR,
A -> В rAi = k \P \
A — 2C r A2 — ki Pд
A -► 3D газ - k3PA
The reaction rate constants are given by,
*i = 5.4 x 102 exp^ ^m ol s_1 m ^ b a r " 1
k2 = 6.5 x 103 exp^ y 0 )m o l s-1 n T 3bar -1
*3 = 2.1 x 102 exp^ ) m°l s~* т - 3Ьаг-1
where RT is in J mol-1.
(a) At what temperature should the reactor be operated to achieve a 20%
selectivity for D (i.e. one mole of A reacts through reaction 3 for every
five moles of A reacted)?
(b) At this temperature, what reactor volume is required for a 10% yield of D
(i.e. one mole of A reacts through reaction 3 for every ten moles of A fed
to the reactor) if pure A is fed to the reactor at a rate of 2 mol s-1 . The
reactor pressure is 1 bar.
Solution (а) фи/a = — ГАЗ ---- = , i , 3 i = 5 d/a = °-2

rA + ГД2 + Газ k\ + k2 + КЗ
0.8*3 - 0 .2*1 = 0 .2*2
' —40 000 \

60 exp ^ 2^ ° Qj = 1.3 x 103 exp^-
RT )
T = 782.IK
*i = 24.92 mol s-1 n T 3 bar-1
*2 = 13.85 mol s- 1n T 3 bar-1
*3 = 9.69 mol s_1 m -3 bar-1
(Ь ) «А - «АО — «А1 — «А2 - Па з
п в — пА1 , nc = 2 иА2 , «D = Зпдз

Ит = Идо + «А2 + 2«А З
*В /А = — . *С /А = — . *D /A = — . *А = У в/А + *С /А + ^Ъ /А
«АО «АО «АО
Chemical reaction engineering
«А2 _ *2 «A3 _ *3
«А1 *1 ’ «А1 *1
«т = иао+ 7 - и а 1 + 2 — «А1 and
*2 *3
«А — «АО — «А1 — 7~ИА1 - -гП М
ki к\
Пм - и ■ 11 I I ( ” А0 " « а )
*1 + к2 + к3
, *2 + 2*3
ИТ - « А О + - ^ — ^ ( « А О - П а)
Let а — ; = 0.686
М "I" л2 + лз
dnA
■ - / (* 1+ * 2 + * з ) ^ / >
Jo
1
-г
Р (*1 + *2 + *з)) Уо
Л>
Г «АО + я(ЛАО -
«А
«а )
dnA
1
[(1 + а)иАо In «А - олаК*.
Р (*1 + *2 + *з)
| Г(1 + а)иАо In ~ - ап а + алдо

Р(к\ + *2 + *з) L «АО
«АО
/*(*1 + *2 + *з) )[(1+“)1птгЬг‘“А]
We require a 10% yield of D. As the selectivity for D is 20% we must react
50% of the A fed to the reactor,
Jd/a = Sd/a*a
xA - 0.5
V = 0.034 m3
4.1.2 C ST R s
Continuing with the same reaction scheme as previously (Reactions R4.4 and
R4.5), the design equation for A will be,
rA = - Л-° ~ ПЛ = (*i + *2) — (4.25)

V 1>T
Rearranging, nA = ” A0 (4.26)
1 + (*1 + *2) t
To calculate «r and ns again we use the individual design equations for the
two reactions,
«А1 j «А /л \
rKi = — = fci — (4.27)
V
ГА 2= —
” A2 7
= к2 —
" A !Л ООЛ
(4.28)
к L^x
w here nAl and «А2 are the m olar flows o f A that disappear through reactions 1
and 2 respectively. So,
»ai _ k\
(4.29)
«Л2 ^2
Rearranging eqns 4.27 and 4.28,
nAi = k\xnA
«А2 = кгТПА
«R = «R0 + «А1 = «R0 + fcl rnA
«S = «SO + 2nA2 = «SO + 2^2 tn A

Substituting for « a from eqn 4.26,
к\ГПAO
«R = «R0 +
1 + (& i + А :г)г
2^2 та AO
«S - «SO +
1 + (& i + & г ) г
4.2 Parallel reactions of different order

4.2.1 Two reactions
Our treatment must be different when there are reactions of mixed reaction rate
orders proceeding at the same time, e.g.
A -*■ R rA\ — k\ Сa (R4.6)
2A -> S r A2 = k2C \ (R4.7)

and the local or instantaneous selectivity depends upon the concentration of
reactant,
0 r/a = — — — = -----k lC —- T (4.30)
f A1 + rA2 k \C A + k2 C A
Л rA2 k2C2A
S/A ГА1 + Г А2 kiCA + k 2C l
In such a case, the choice of reactor will influence the overall selectivity. For
a CSTR the overall selectivity is of course equal to the local selectivity as the
local selectivity is a constant because of the lack of concentration gradients.
Hence, calculation of the product distribution is straightforward. For a PFR
or batch reactor it is more complex to evaluate the overall product
distribution as the local selectivity or instantaneous selectivity will vary with
position or time.
As we have two reactions we can write down two independent design

equations. For a PFR, let us look at one design equation for both reactions
combined and the other for Reaction R4.6 or reaction 1 alone,
d«A . .
'•» = - - 5 y ( « 2)
d«Ai
га. = Ж (4-33)
Integration of the first design equation is how we get a relationship between

7 It is assum ed that the reaction rates are
the outlet concentration of reactant and the volume of the reactor7. The second
only dependent on reactant
design equation cannot be integrated. There are three variables; the rate concentrations and not product
depends upon reactant concentration and we also have the nM and the reactor concentrations; if not, the differential
volume. We cannot relate nAi to nA because of the variable selectivity. To equations w ill be coupled and they must
proceed to the product distribution we divide one design equation by the other be solved sim ultaneously.
(remember nM increases as nA decreases), dividing eqn 4.33 by eqn 4.32,
Фк/ а = (4.34)
dnA
This eliminates the volume variable and we recover the definition of local
selectivity (compare to eqn 4.4).
The amount of A that disappears through reaction 1 is given by,
/■exit тАе
иА1 = / dnAi = - / 0R/AdnA (4.35)
J inlet J Лдо
Therefore, the overall selectivity is given by,
S r/a = — — — = 7------------ т [ 0 R/AdnA (4.36)

«АО - «Ае («АО - n Ae) J„A0
If the volumetric flow rate is a constant, then eqn 4.36 can be rewritten in
terms of concentration,
Sr/a — тр------ тт- r f Фя/а&Са (4.37)

((-AO — t-Ae) J Сao
and the overall selectivity can be seen to be effectively an average of local
selectivities over the whole reactor8.
8 R em em ber th a t the m inus sign is
We can now compare the performance of CSTRs and PFRs (or batch before the integral because w e are
reactors). In general, if the ‘useful’ reaction is of higher order than the integrating from the inlet to the outlet
‘wasteful’ reaction then reactant concentrations should be kept as high as w hich correspo nds to decreasing
possible. This means that a PFR would perform better than a CSTR. If CSTRs concentration of A and therefore the
integral w ill have a negative value.
must be used, then selectivity can be increased by using a number of CSTRs in
series. If the volumes of the CSTRs are different then the smaller CSTRs
should be placed first (large CSTRs early in the cascade would have the effect
of diluting the reactant concentration early).
Consider what happens when the second-order reaction (Reaction R4.7
or reaction 2) is the desired reaction. The local selectivity for S is given
by,
hC\ _ k2CA
0 S /A =
k\CA + /с2Сд k\ + kiCA
| outlet inlet; Figure 4.2 shows a plot of local selectivity versus concentration. At zero
concentration of A the selectivity for S is equal to zero. As concentration of
reactant is increased the selectivity increases approaching unity for high
concentrations. Because the CSTR operates at the outlet concentration of A,
the yield of S depends only on the selectivity at that concentration and is
therefore proportional to the heavily shaded rectangular area. The yield from
the PFR involves integration of the local selectivities and from eqn 4.37 is
proportional to the total shaded area under the curve. It can be seen that, for
the same conversion, the yield from the PFR will be much better than for the
Fig. 4.2 Local selectivity for S versus CSTR.
concentration.The yield of a CSTR is Similarly, if the ‘useful’ reaction is of lower order than the ‘wasteful’
proportiona l to the heavily shaded area;
reaction then reactant concentrations should be kept as low as possible9. This
the yield o f a PFR is proportional to the
total shaded area.
means that a CSTR would perform better than a PFR. Consider the case where
the first-order reaction (Reaction R4.6 or reaction 1) is desired. The local
selectivity for R is given by,
, _ k\CA _ k\
^ k\ Сд + k2C \ k\ + k2C \
9 Unlike the previous case, this m eans Figure 4.3 shows a plot of local selectivity versus concentration. At zero
that all reaction rates w ill be decreased. concentration of A the selectivity for R is equal to unity. As concentration of
Process econom ics w ill dictate how far
reactant is increased the selectivity decreases, approaching zero for high
reaction rates can be reduced, and
hence reactor volum es increased, in
concentrations. The yield of R from the CSTR is proportional to the total
tryin g to im prove selectivity. shaded rectangular area. The yield from the PFR is proportional to the heavily
shaded area under the curve. It can be seen that, for the same conversion, the
yield from the CSTR will be much better than for the PFR.
Example 4.3 Consider two competing liquid-phase reactions:
a + b ^ r ^ ( D d ) - 04
§RJA
A—
outlet inlet
k\ = l.Omol litre- 1s-1 , k2 — 2.0mol litre- 1s—1;

С — reference concentration = 1 mol litre-1
Cao = CBo = 1 mol litre-1
(The above rate data are expressed in terms of dimensionless concentrations.
This is a common practice for empirical data as, otherwise, the units of the rate
Fig. 4.3 Local selectivity fo r R versus constants would become very inconvenient.)
concentration.The yield o f a PFR is What is the yield of R for: (a) a CSTR; (b) a PFR?
proportiona l to the heavily shaded area;
Both operate at 80% conversion. (If the initial concentrations of A and В
the yield of a CSTR is proportiona l to the
were not in the stoichiometric ratio we would need to say whether the desired
total shaded area.
yield was defined in terms of consumption of A or B.)
, га i _ кЛ%)(%)__________
Solution
,A
l
h&mr'+h&r®?2
- l + 2 ( ^ ) - ° - 6( £ ) L6
But, CA = CB; <Pr/a = ~ —
(a) PFR
From eqn 4.37,
2 (-Cm ] rCAc ]
s^ = - K c ; L f c ' A dC A = 5 c i i c„ T T i ( S ) dCA
where, Д CA — CAe - CAo
С 1.4
Sr/a =
2 A CA ln(1+2t)' In — = 0.476
3
F r / a = 0 .4 7 6 д :Ае = 3 8 %
(b) CSTR
S r/a - <fa./A = = 0.714

1 + 2^
yR/A = 0.714лAe = 57%

The CSTR is much more selective for reaction 1 because it works at the outlet
concentration. Low concentrations relatively favour reaction 1 over reaction 2,
because the overall reaction order is lower. Note that while a CSTR will give a
better yield of reaction 1, it would need to be larger in volume than the PFR
for a given conversion.
For bimolecular reactions, e.g.

A+ B C
2A - D I
where the desired product is C, it is advantageous to keep the concentration of
Fig. 4.4 S chem atic representation of
A low. This can be achieved by having a primary process stream containing В advantageous reactor system s for
which is fed to the reactor inlet and then A is injected into the reactor or increasing selectivity w hen the desired
reactor chain as required (see Fig. 4.4). In this way the concentration of A can reaction is bimolecular.
be kept low and the selectivity improved.
4.2.2 Three or more reactions

When three or more reactions are taking place in parallel and the desired
reaction is of higher order than some of the competing reactions but of lower
order than others, then an interesting problem is encountered. At low
conversions the reactant concentrations are high and therefore the important
competition may be between the desired reaction and reactions of higher
orders. Therefore, concentrations should be reduced by using a CSTR first. At
higher conversions, where reactant concentrations are lower, the important
competition may be between the desired reaction and reactions of lower orders
and now a PFR should be used to keep concentrations high. Therefore, it is
possible that the best approach may involve a combination of CSTRs and
PFRs.
Example 4.4 Consider three competing gas-phase reactions of different orders,

A —►В r \\ = k\
A -* С rA 2 = k2PA
A -*■ D газ = h P 1^
with ki = 1 mol s-1 m~3, k 2 = 2 mol s-1 m ~3 bar-1, k 3 = 1 mol s_1 m -3

bar-2
The feed is pure A at a pressure of 5 bar. 96% conversion is required. What
would the overall selectivity for С be if: (a) a PFR is used; (b) a CSTR is used;
(c) the optimum reactor configuration to maximize the overall selectivity for С
is used?
Solution The local selectivity for С is given by,

i _ ЬРа
Ф с / А - к , + к 2Р А + к 3Р 1
This function is plotted in Fig. 4.5(a). At a reactant partial pressure of zero the
selectivity is zero, as the reactant partial pressure tends to infinity the
selectivity tends to zero. For intermediate values of reactant partial pressure, a
maximum in selectivity is obtained.
(a) From eqn 4.36, for a PFR,
-1 f PAe
S c / a = Tp~~— p \ / Фс / а ^ Р а
(.Г AO — “ Ae) J p A0
-1 /**■ k2PA
-d PA
(PAO — P Ae)>JPM
JPf, k\ + k2PA + къР\
The overall selectivity for С is proportional to the area under the curve in Fig.
4.5(b).
Leaving PA in bar and substituting for the rate constants,
k2PA 2P a 2P a 2 2
h + k 2P A + h P \ 1 + 2 P a + P2
a ( 1 + P a )2 1+â ( 1 + P A)2
Sc,a = (p ,*~ -7 m ) L (т Г ъ ~ (Г
J "]^*Ae
:[
~ { P a o - P az) 1
-1
21n(l + P a ) + -
(1 + ^ a )J
~ (4.8) 2ta¥+iH]=39-3%
(b) For a CSTR the overall selectivity is equal to the selectivity evaluated at
the outlet partial pressure,
- 2 ^°Ae —2 =
S c /a = Фс / а = ---------- --------------- r
ki + k2PAe + k3P2M = (1 + P Ae)2 1.22 = 2 7 -8 %
The overall selectivity for С is proportional to the area of the rectangle in Fig.
4.5(c).
(c) At the inlet conditions, the selectivity is low compared to the maximum
possible selectivity. Therefore, to optimize the overall selectivity, the
partial pressure should be dropped to that corresponding to the (а) outlet inlet
occurrence in maximum local selectivity by using a CSTR.
dPA = 0
The maximum in local selectivity occurs when, dfciA
Differentiating the expression for фс/А,
й ф с /А k 2 (* 1 + k 2P A tmaz + 1
* з Р А П а х ) — * 2 ^ * А ,т а х ( * 2 H" - ^ З ^ Л л а х )
= 0
d?A (*1 + к 2Р А ,тах + W 2A ,mJ 2
к2(к\ + к2Ра ,тах + = ЬРл,тах (к2 + 2 * з Р а ,та х )
Р A,max — — 1 bar
The overall selectivity for the CSTR can then be evaluated,

k2P a ,max
Sc/A,CSTR = = 50%
*1 + k2P A,max + k3p A
\ ,max
A PFR is now used to reduce the reactant partial pressure to its final value.
After eqn 4.36,
-1
Sc/A,PFR = 21n(l + Pa) ■+■
(P a ,max - Р ас) (1 + P a ).
-1 1.2 2 2
= 44.4% (C) outlet inlet
(0 .8) П 2 + 1.2 2
The overall selectivity for the combination of reactors is then an average based
upon the amount of reactant converted in each reactor,
S c /A,CSTr (P a O - P a ,max) + Sc/A,PFr (P a ,max ~ PAc)
Sc/ A = ■
(P a O- P Ae)
0.5(4) + 0.444(0.8)
= 49.1%
(4.8)
The shaded area in Fig. 4.5(d) is proportional to the optimum overall
selectivity for the combination of a CSTR followed by a PFR.
4.3 Series reactions

Consider the two series reactions,
A ^ R -^ S
PA,max P
with both reactions first order.
Fig. 4.5 (a) Local selectivity for С
rA — *iCa (4.38) versus partial pressure (not to scale), (b)
m — k\C A к2Сц (4.39) Overall selectivity of a PFR is
rs = k2CR (4.40) proportional to shaded area, (c) Overall
where A is defined as a reactant (positive rates mean rates of disappearance) selectivity of a CSTR is proportional to
shaded area, (d) Overall selectivity of
and both R and S are defined as products (positive rates mean rates of
optimum reactor configuration (CSTR,
appearance). with outlet at the composition
corresponding to maximum selectivity,
followed by PFR) is proportional to
4.3.1 Batch reactors and PFRs
shaded area.
Again, batch reactors and PFRs have similar behaviour and therefore we will
only look at the behaviour of a batch reactor.
Solving the design equation for the amount of A present in a batch reactor
(see eqn 3.5), we get, as before,
NA = N M e~k'‘ (4-41)
We can write down the differential equation or design equation governing the
number of moles of R present using eqn 4.39,
rR = k \C A — &2C r = (4-42)
= (4.43)
V dt V V v '
and substituting in for the time dependence of the number of moles of A
(eqn 4.41),
—7—+ k2Nn = k\N \o e~k,‘ (4.44)

dt
We now try a solution of the appropriate form and evaluate the
coefficients,
JVR = Ae~kl‘ + Bc~kl‘
and evaluate A and В by substitution into the original differential equation
with the initial condition,
t = 0 ,N R = 0
10 Try substituting this back in the original This gives10, jVr = 1 {e kl‘ — e klt} (4.45)
differential equation to show that it is a 2 — 1
solution. We can find the number of moles of S present because the total number of
moles is conserved,
— Nao —Na —Nr
* L = l - e ~ k lt ------*!— {e~*l( - e^*2'j
Nao k2 - k i x
In many cases R may be a valuable product and hence we may want to
maximize its production. The optimum residence time to maximize the yield
of R can be easily found using eqn 4.45,
Nr _ I -M _ .-hM
N ao k2 - k x [ '
and to maximize Л/r , we need,
d^R _ 0 _ k\N \o - e-*,ropt _|_ e- t 2«opt j

df k2 — k \ l
k\ e_tl,op‘ = k2 e~k2,°'"
In l
( 4 '4 6 )
where k\m is the log-mean rate constant. Substitution of the expression for
optimum residence time (eqn 4.46) into eqn 4.41 will give us an expression for
the maximum yield of R which can be obtained in a batch reactor.
4.3.2 CSTRs
Solving the design equation for A (see eqn 3.16),
«АО
(4.47)
The design equation for R,
(4.48)
For no R initially present,

nR = k\ Т72д —^ « R (4.49)
and rearranging and substituting eqn 4.47 into eqn 4.49,
kirn A ki rnA0
(4.50)
Hr 1 + k2x (1 + k\ r)(l + k2x)
In addition ns = nô —«a —«r
k\T
« ao +fciT
11 + (1 + A^rXl + k2x)
k\k2x2
(1 + fciT)(l + k2x)
For the optimum residence time to maximize the yield of R we differentiate
eqn 4.50 with respect to residence time,
dr
d k xk2x2 Q
d r ( l + k\x){\ + k2x)
Differentiating and manipulating (see Problem 4.5 at the end of the chapter)
gives,
1 1
(4.51)
where km is the geometric-mean rate constant.
4.3.3 Comparison of PFRs and CSTRs

Let us consider what will happen if we have a PFR and a CSTR both operating
at the same residence time (see Fig. 4.6). If we have a positive-order reaction,
the outlet concentration of reactant from the PFR will be lower than that of the
CSTR because of the lower rates exhibited by the CSTR as the reactant is
immediately diluted. This means that, at short residence times, the outlet
concentration of the intermediate R will be higher from the PFR because of its
greater rate of production. This gives the appearance that things happen
quicker in the PFR than in the CSTR. A maximum in outlet concentration of R
is reached in the PFR at a shorter residence time than that required to reach a
r
maximum in the CSTR. Most importantly, the maximum outlet concentration
of R from the PFR is greater than the maximum outlet concentration of R from Fig. 4.6 Concentrations (not to scale)
of species involved in a series reaction,
the CSTR. This is because when we have a series reaction, if we want to
A -*■ R -> S, in a PFR and a CSTR.
maximize the outlet concentration of an intermediate, we want to make sure
that all molecules react for the same length of time (this time being the
optimum residence time). This is what happens in a PFR—all molecules have
exactly the same residence time. However, in a CSTR not all molecules have
the same residence time. The residence time of a CSTR is a mean residence
time, some molecules pass through quickly and some pass through the reactor
more slowly. As a result the maximum yield of an intermediate can never be
as good from a CSTR as it would be from a PFR (or batch reactor).
4.4 Problems
4.1 The following liquid-phase reac 4.4 Two competing reactions, one of
tions were performed in a PFR, which is autocatalytic, are performed
A- 2R r A1 = *1С a
in a CSTR,
A- • 3S ГА2 = k2CA A -> В га\ — k\ СдСв
For a residence time of 40 s the A -*• С га2 = k2CA
conversion of A was found to be 60%
and the ratio of moles of R produced where rAi is the rate of disappearance
to moles of S produced was found to of A through reaction 1, etc. There is
be 4 to 1. The reactor feed was pure A. no В present in the feed to the reactor.
Calculate the values of k\ and k2. (a) Given that reaction 1 is the
desired reaction, indicate how you
4.2 The following reactions are carried would start up the reactor.
out isothermally in the gas phase in a (b) Show that, at steady state,
PFR: C b = C A0- ^ - £
A 2R rAi = k\P a for Cao > 3 - + 1
A 3S га2 = к2Рк
(c) If the reactor is operated at a
P = 1 bar, k\ = 20mols _1 m~3 bar-1;
conversion such as to maximize the
k2 = 40 mol s_1 m -3 bar -1 rate of reaction 1, show that at steady
Pure A is fed to the reactor at 1 mol state the selectivity for В will be,
s-1. What reactor volume is required
for a 30% yield of R?
Sb/a —
4.3 Consider the following reactions, 1+
A ->• В, га\ — Ь\Са
2A -*■ С, га2 = к2С \ 4.5 The intermediate R is to be
produced in a continuous reactor:
В is the desired product and С is a
waste product. A-^>R-^»S k\ = 2s _1 and k2 = 0.5 s“
ifci = 1 s_1, k2 = 10 litre mol-1 s-1,
Calculate the value of г for a CSTR
dt = const = 1 litre s ,
and a PFR for maximum production
СAo = 1 mol litre-1
of R. For this optimum value of x,
(a) What is the volume of a PFR
calculate the conversion of A and the
required for 95% conversion of A?
yields of R and S.
(b) What is the yield of B?
(c) What is the overall selectivity for B?
4.6 Consider the reaction network
(d) What is the conversion of A in a
shown in Fig. 4.7. All reactions are
CSTR of the same volume?
first order with respect to the reactant
(e) What is the yield of В in this case?
kî indicated.
(f) What is the overall selectivity for B?
(g) What would be the yield of В if k3/ki = 0.4; (k2 + fcOAi = 0.2
the volume of the CSTR were in
creased to give 95% conversion? Determine the yield of В when the
Fig. 4.7 Reaction network for Problem (h) What is the overall selectivity for conversion of A is 70% in: (a) a PFR:
4.6. B? (b) a CSTR.
5 The energy balance and
temperature effects
We will now no longer confine ourselves to considering isothermal

reactors. Let us start by looking at the temperature dependence of reaction
kinetics.
5.1 Temperature dependence of reaction rate

5.1.1 Irreversible reaction
For a first-order irreversible reaction,
ra - kCk
Substituting in the temperature dependence for the rate constant and using
conversion,
where k ' is the pre-exponential constant and A E is the activation energy of the
reaction. Figure 5.1 shows how the rate of reaction varies as a function of
temperature and conversion (lines of constant rate are shown). As temperature
is increased reaction rate increases; as conversion is decreased reaction rate
increases as the concentration of reactant is higher. A very low rate will be Fig. 5.1 Lines of constant rate of
reaction shown in conversion-
apparent either at a very low temperature or a very high conversion temperature space for a first-order
(approaching unity). In fact, a line of zero rate would correspond to the y-axis irreversible reaction.
(zero temperature) and the line at a conversion of unity. As we move away
from this envelope, reaction rate increases. Clearly, for any given conversion,
we would wish to operate at the highest possible temperature (this will be
limited by factors such as materials of construction) to maximize the reaction
rate and therefore minimize the reactor size.
5.1.2 Reversible endothermic reaction Reaction coordinate

Figure 5.2 shows an energy profile for an endothermic reaction, А -о- B. The Fig. 5.2 Energy profile for the
reactants are on the left-hand side and the products on the right-hand side. The endothermic reaction, А -о- B.
energy associated with the products is greater than that associated with the
reactants for an endothermic reaction. Therefore, the forward activation 1 The difference between the activation
energy, A E i, must be greater than the reverse activation energy, Л £ _ , The energies is equal to the change in
net reaction rate is given by the difference in the forward and reverse reaction internal energy which, for a reaction with
an equal number of moles on both sides,
rates,
is equal to the heat of reaction,
ra = k\CA — к-\Св ДE, - ДЕ.., = Д U = Д
50 The energy balance and temperature effects
For no В present initially,
rA = k\ e x p f - ^ - ^ C Ao(l - xA) - k'_x e x p ^ - ^ ^ ^ C AOxA
At equilibrium, rA = 0; xA = x*A
Substituting into the reaction rate expression,
4 *i r ^ u J A E ~l ~ A E l \
T ^ A- n - K - i r 1 exp{ rt )
where К is the equilibrium constant.

^ , К
Therefore, *A =
As temperature increases, the value of К increases for an endothermic reaction
and therefore the value of the equilibrium conversion, x*A, increases. Figure 5.3
oi---------------------------------
T shows the form of the equilibrium line for a reversible endothermic reaction.
High equilibrium conversion are favoured at high temperatures. The
Fig. 5.3 Lines of constant rate of
reaction shown in conversion-
equilibrium line is a line of constant reaction rate (this rate of course being
temperature space for a reversible equal to zero). Other lines of constant rate are shown in Fig. 5.3. Low rates are
endothermic reaction. obtained close to the equilibrium line. Moving away from the equilibrium line
serves to increase reaction rate. As in the case of the irreversible reaction, we
would wish to operate at the highest possible temperature to maximize
reaction rate.
(a)
5.1.3 Reversible exothermic reaction
A '
Again, K
The forward activation energy, A E \, is now less than the reverse activation
energy, A E - A s temperature increases, the equilibrium constant will
decrease and therefore the equilibrium conversion will decrease. Figure 5.4(a)
shows the equilibrium line in conversion-temperature space. Once more this
line corresponds to the line of zero reaction rate. However, there is also zero
reaction rate at zero temperature (the y-axis). Again, rate is increased if we
move away from the zero rate envelope.
Let us now consider a line of constant conversion. If we proceed along this
line from low temperature (at a temperature of absolute zero the rate wiU be
T
zero—point A) the rate will increase. However, at some point it will just touch
Fig. 5.4 (a) Lines of constant rate of one of the lines of constant reaction rate (point В— and this will correspond to
reaction shown in conversion- the highest rate achievable at this conversion) and then as temperature
temperature space for a reversible increases further the rate will drop reaching zero at the equilibrium line (point
exothermic reaction. Also shown is a line
C). The rate at constant conversion can be plotted against temperature and this
at constant conversion. On this line the
rate will be zero at point A, will reach a is illustrated in Fig. 5.4(b).
maximum at point B, and will be zero at This means that if we have a reactor operating at a particular conversion
point C. (b) Plot of rate versus there will be an optimum temperature that will give us a maximum in
temperature on line of constant reaction rate. It is important to be able to calculate this optimum
conversion.
temperature.
For the reaction А о В with first-order kinetics,

r A = *iCa —*-iCb
For no В present initially,
Гд — k\ CA0( 1 X \) k—i C\(\X\
га = CA0{k, - (ki + *_i)xA}
The maximum reaction rate for a given conversion will occur when,
^ =° <5Л>
Differentiating the kinetic expression, with Сдо and xA as constants,
drA _ fd*! fd k i dk- Л 1
dT A0| d r d r ) Xa\
dki
.. —---------
... dT 1
xA
d*i d*_i_ d*_i dki
dT dT dT dT
But, in general, к = к! exp Г— ^

Therefore,
d*i , A£i
d f= 4 ^ “ p
d*_!
Therefore,
ХА = А £ _ !* ~ = A£_! Г (5'2)
A.E\k\ A£ , /Г
This means that for any conversion the temperature that results in the
maximum reaction rate can be easily found from eqn 5.2 (or its equivalent if a
2 Conversion will vary with time in a batch
different kinetic relationship is obeyed). Therefore, as a CSTR operates at only
reactor and, therefore, to operate at the
one conversion, it should be operated isothermally at the appropriate minimum necessary residence time to
temperature to maximize rate or minimize its volume. However, conversion reach a given final conversion, the
wifi vary along the length of a PFR and therefore the optimum temperature temperature of the reactor will need to
will also vary. This results in an optimum operating line (in conversion- change with time in the appropriate
manner.
temperature space) for a PFR2.
The exothermic gas phase reaction, Example 5.1

A В + С; rA — k\P a — k -iP ^ P c
is to be carried out in a CSTR operating at 50 bar.
Calculate the maximum rate if the reactor operates at a 22% conversion of A
(feed is pure A).
The energy balance and temperature effects
/ - 2 0 000 \
k\ — 0.435 expl - . " jmol s 1 m 3 bar 1
Л RT )
-6 0 0 0 0 ^
к- 1 = 142.6 e x p ~ j mo^ s 1 m 3 bar z
R T is in J mol-1 ; R = 8.314 J mol-1 KT1
Solution A-&B+C
ПА = «АО — «A0-*A> « В — «АО*А> «С = «А 0*А . п Т ~ «АО + «АО + «АО*А
РА = — Р = р = 31.97 bar
Пт 1 + *А
РВ = Р С = - Р = 9.02 bar
1 +*А
Га = ^ Р д - к -\Р в Рс
drA _ d*, dfc_, _

"Т^Г = “7 ^ * А ----- - T ^ r W c — U
dr dr dГ
/dJfc_! _ PBPc
d ^ /dfc-i
dr / dT
= fc'j exp ( - ^ ) ; *-i = fc,- i exp ( ‘)
dî / dfc_i A E \k i P bPc

d r / ~ d F " Д Е_ 1*_1 ~ P A
ifci , / 4 0000\ A £L,PBPc 60(9.02)2 , .

J - - 3.05 x КГ’ e x p (— ) = = 3 , 1 1 9 7 b" = 7 635 bSI
Г = 614.8 К
k\ = 8.69 x 10~3mol s _1 m ~3 bar -1
jt_i = 1.14 x 10_3mol s_l m -3 bar -2
rA = 0.278 - 0.093 = 0.185 mol s~* m ~3
5.2 The energy balance

Up until now we have only used the material balance in reactor design. Now,
if we want to take account of temperature effects, we must also include an
energy balance.
5.2.1 CSTRs
Consider a process stream at an initial temperature of To with a conversion
defined as zero. This process stream is now passed through a reactor and at
some later time is at a temperature T and conversion x A. Because energy is Final

an intensive property it is not important what path the stream followed in
conversion-temperature space from one point to the other so we will
consider a path that makes the formulation of the energy balance
straightforward (see Fig. 5.5). First, we will heat the mixture up while at
constant composition, and then we will allow the mixture to react at Initial---------------1-------------
constant temperature. T
First, let us consider the energy involved in heating the process stream at
Fig. 5.5 Initial and final states of a
constant initial composition, process stream in conversion-
temperature space.
Qr = nAOcFA(T — To) + nBQCpb(T — T0) + ■■•
where cfi is the mean (mean because it is an average over an appropriate
temperature range) specific heat capacity of the ith component (with units,
for example, J mol-1 K-1). Qr is the heat required to heat up the process
stream or alternatively the heat removed by the process stream (in, for
example, J s-1),
Qt = Y 1 nioCpi(T ~ To) (5-3)

i
Defining an overall mean specific heat capacity as,

__ «;o__
CP0 — / ----- c Pi
“ «то
Q t = nToCpo(T - To) (5.4)
Now let us consider the energy involved with the change in composition due
to reaction,
Qt = nAOx A( - A H R(T)) (5.5)
where Qg is the heat generated by the reaction. It should be remembered that
the heat of reaction will, in general, be a function of temperature. 3 To illustrate this assumption let us
The energy balance is of the form, consider the reaction of hydrogen with
iodine at 300 K,
Qg = Qr + Q (5.6)
H2 + l2 О 2HI
i.e. the heat generated by the reaction goes in heating up the process stream
The specific heat capacities of the
and also supplies any heat, Q, that is removed from the system. Therefore, individual components are as follows,
substituting eqns 5.4 and 5.5 into eqn 5.6,
сРНг = 28.7 J m ol-1 K“ 1
nAOxA( - A H R(T)) = njocpo(T — Tq) + Q (5.7)
Cpi2 = 37.6 J m ol-1 K-1
This is then a general form of the energy balance.
We will now make some simplifying assumptions that will make the cPHi= 30.0 J m ol-1 K-1
energy balance easier to use.
The heat capacity of one mole of
1. We will assume that the heat of reaction is independent of temperature, hydrogen and one mole of iodine is then
i.e. equal to 66.3 J K-1 .This compares to the
heat capacity of two moles of hydrogen
—A H r (T) = —A tfR
iodide of 60.0 J K—1. It can be seen that
2. We will assume that the heat capacity of the stream is a constant, i.e. it is there is only a difference of about 10% in
not function of composition3, the two values, hence the assumption
that total heat capacity remains constant
«ТОСЯО = n-xCp is reasonable.
However, it must be remembered that n j and cj> must be evaluated at the same
composition.
With these two assumptions we can simplify the energy balance
expression,
и а о * а ( — A # r) = n j C p ( T — To) + Q (5.8)
For an adiabatic reactor where no heat is added to, or removed from, the
process stream, i.e. Q — 0,
паоха (—А Н к ) = njCp(T - T0) (5.9)
Example 5.2 A liquid-phase reaction is performed in a CSTR,

A -> Products
(a) What is the operating temperature if the reactor is adiabatic?
(b) What is the operating temperature if the reactor is surrounded
by a cooling jacket with coolant at 300 K? The heat transfer
area, A, is 0.2 m 2 and the overall heat transfer coefficient, U, is
500 W m -2 К ” 1.
Inlet concentration, Cao Ю mol litre-1
Volumetric flow rate, vT 0.1 litre s-1
Conversion, хд
Feed temperature, To 300 К
Heat of reaction, A H r -100 kJ (mol of A )-1
Overall mean heat capacity, Cp 4.2 kJ litre-1 K _1
Solution (a) <2g = л Ао* а ( — A # r ) = i >t C ao * a ( - A # r )
Qt = v jc p (T — T0)
(Note: the units of the mean heat capacity imply that it is the volumetric flow
rate of the stream that should be used and not the molar flow rate—remember,
always check the units!)
Qg = Q t
T _ С ао * а ( — АЯК) _ (10)(0.8)(105) =
cp 4200
T = 490 К
(b) Qg = 1>гСАо*А(—A # r)
Qt - VjCp(T — To)
Q = U A (T - 7 j )
Qg = Qt + Q
vt C a 0x a ( - A H r ) (0.1)(10)(0.8)(105) =
vjc-p + UA (0.1 )(4200) + (500)(0.2)
T = 454 К
5.2.2 PFRs
For PFRs we need to use an energy balance that is in differential form.
Considering a differential change in temperature caused by a differential change
in composition with a differential amount of heat removed from the system d<2 ,
and using a similar approach as for the derivation of eqn 5.8,
па0< 1ха(-А Н к) = njC pdT + d Q (5.10)
If q is the heat flux through the wall of a PFR, eqn 5.10 becomes,
пдосЬса(—A # r ) = n jC p d T + 2 n r d l q (5.11)
as 2nrdl is the differential area of a tubular reactor wall associated with an

element of length dl.
Remembering that the material balance for a PFR (eqns 2.8 and 2.12) can
be written as,
rAdV — —dnA = nAOdxA (5.12)
Substituting eqn 5.12 into the energy balance (eqn 5.11) gives,
rAdV(—A # r ) = njCpdT + 2nrdlq
But, dV = лт 2dl
__d T 2q
rA( - A H R) - njcp — H-----
dV r
dT _ га ( - А Я к) - ^
(5.13)
dV n jc^
which determines the temperature profile along the reactor. The material 4 In the case of a CSTR the concept of an
balance (eqn 5.12), of course, determines the composition profile along the operating line is meaningless; instead,
the operating point of the CSTR is
reactor,
dictated by the simultaneous solution of
d*A _ rA
(5.14) the material and energy balances, i.e.
dV nA0 the intersection of the material balance
line and the energy balance line in
If we divide eqn 5.14 by eqn 5.13,
conversion-temperature space.
dxA njCp
(5.15)
dT ИА0( _ д я к) - ^ ^
Equation 5.15 relates the conversion to the temperature and can therefore be
considered to describe an operating line for the PFR4.
The heat flux can take different forms. In the adiabatic case, eqn 5.15
becomes,
d*A = _ n I c p _
dT иАо(-Д Я к) (ЭЛО)
Fig. 5.6 Operating lines for an
adiabatic reactor in conversion-
and the gradient of the operating line is constant. Figure 5.6 shows such straight temperature space for endothermic and
adiabatic operating lines for different heats of reaction. The gradient of the exothermic reactions.
operating line is small (and positive for an exothermic reaction or negative for an
endothermic reaction) if the reactant is at high concentration or the heat of reaction
is large. This small gradient means that small changes in composition will result in
large changes in temperature.
For electrical heating, the heat flux is constant,
q = qo (5.17)
whereas with a cooling or heating medium the heat flux depends upon the
difference in temperature between the process stream and jacket,
Fig. 5.7 Operating line (from the
q = U (T —7j) (5.18) energy balance) for an exothermic
reaction with adiabatic operation (q = 0)
Figure 5.7 shows the operating lines for an exothermic reaction with heat and operation with heat input (q < 0) and
input and heat removal (heating and cooling with a heating/cooling heat removal {q > 0).
medium) and adiabatic operation. In the adiabatic case the operating line is
a straight line. In the case of heat input, the temperature rises more quickly
than the adiabatic case (this is particularly important at low process stream
temperatures where heating is greatest). In the case of heat removal, the
temperature rises more slowly than in the adiabatic case (this is particularly
important at high process stream temperatures).
Note that the equations dictating the temperature profile and composi
tion profile (eqns 5.13 and 5.14) are not independent but must be solved
Fig. 5.8 Adiabatic operating line for a
reversible endothermic reaction.
together as coupled differential equations (the rate of reaction is a function
of both composition and temperature). The solution, in general, must be
performed numerically. However, let us consider some general forms of the
solution.
R eversible endotherm ic reaction. Figure 5.8 shows an adiabatic operating

line in conversion-tem perature space (lines of constant rate are also
shown). The rate will monotonically decrease from the inlet (Fig. 5.9(a))
as will the temperature (Fig. 5.9(b)). The conversion profile is shown in
Fig. 5.9(c). The gradients of both the temperature and conversion profiles
are directly related to the reaction rate (through eqns 5.13 and 5.14) and
at lower rates (i.e. further along the reactor) the gradients tend to be
lower.
Reversible exothermic reaction. Let us consider the operating line for a

reversible exothermic reaction. Figure 5.10 shows the operating line for the
adiabatic case (q = 0 ) as well as for the case of heat removal (q > 0 ).
Initially, the rate will increase along the reactor due to the increasing
temperature of the process stream and then, as equilibrium is approached,
the rate will begin to drop. Figure 5.11 shows the general forms of the
reaction rate, temperature, and conversion profiles.
For the adiabatic case, the temperature and conversion profiles (Figs
Fig. 5.9 (a) Rate of reaction, (b)
5 .11 (b) and 5.11 (c)) have a similar shape as the temperature must be linearly
temperature, and (c) conversion as a dependent upon the conversion (from integration of eqn 5.15). At the reactor
function of axial position (or reactor inlet the rate of reaction is low and therefore the gradients of the temperature
volume) fo ra reversible endothermic and conversion profiles are low. As the temperature rises the rate increases
reaction.
and finally begins to fall as equilibrium is approached. This results in the
characteristic sigmoidal shape of the temperature and conversion profiles.
For heat addition (q < 0) the solution is not changed significantly from the
adiabatic case (q = 0 ); however, the case of heat removal (q > 0 ) is of
interest.
In the case of heat removal, as the rate of reaction begins to fall, the rate of
heat generation will decrease and will eventually become equal to the rate of
heat removal. At this point the gradient of the operating line will be infinite
(Fig. 5.10). Further along the reactor still, the rate of heat removal will be
Fig. 5.10 Operating line fo ra reversible greater than the rate of heat generation and the temperature of the process
exothermic reaction for adiabatic stream will fall (Fig. 5.11(b)). This results in a ‘hot-spot’ in the reactor. This
operation (qr = 0) and operation with heat heat removal takes the process stream away from the equilibrium condition
removal (q > 0).
and therefore allows higher conversions to be achieved in the reactor. Heat

removal will lower rates of reaction early in the reactor, but, at high
conversions, heat removal will keep the stream further from equilibrium and
therefore result in higher reaction rates (Fig. 5.11(a)).
A diabatic reactor. As previously mentioned, in general, the reactor

volume for a PFR cannot be calculated analytically because of the
coupled material and energy balances. However, in the adiabatic case,
the equation for the operating line can be determined analytically from
integration of eqn 5.16 (this effectively uncouples eqns 5.13 and 5.14).
As we know the relationship between temperature and conversion on the
operating line, the reaction rate can be evaluated at any point and a
numerical integration of eqn 5.14 performed. If лдо/гд can be plotted
against conversion (see Fig. 5.12), the PFR volume is simply equal to the
area under the curve. One simple method of numerical integration, or
estimating the area under the curve, is the trapezoidal rule (see Fig.
5.13). The function to be integrated, nAo/rA, is evaluated at a number of
values of conversion. The approximate area under the curve is then given
by summing the areas of all of the individual elements. For instance, the Fig. 5.11 (a) Rate of reaction,
area enclosed between two function evaluations is given approximately (b) temperature, fo ra reversible
exothermic reaction and (c) conversion
by, as a function of axial position (or reactor
volume) for a reversible exothermic
reaction for adiabatic operation (q = 0)
and operation with heat removal (q > 0).
If all function evaluations are performed at uniform spacing, i.e. A x a is

constant, then the approximate area under the curve will be given by,
f d * A % A xa + fi H--------- fn—i + ^
where, A x a = x Ae/n , as the range of integration is broken down into n equal

subdivisions. As the value of n increases the accuracy of the estimate will
increase. Fig. 5.12 Plot of nM/rk versus
conversion. The area under the curve
corresponds with the volume of the reactor.
The irreversible gas-phase reaction, Example 5.3

A + В -» 2C rA — кхР АРъ
is to be carried out in an adiabatic PFR. Using the data below calculate the
reactor volume required for 80% conversion of the reactants (numerical
integration is required).
Inlet mole fractions 10% A, 10% B, 80% inert

Feed temperature 300 К
Total molar feed rate 8.9 x 10-4 mol s-1
Operating pressure 1 bar
Fig. 5.13 Use of the trapezoidal rule to
estimate reactor volume.
k\ = 2 x 106 exp^ — moi s 1 m 3 bar 2 Г in Kelvin
Heat of reaction -60 kJ (mol of A )-1

Overall mean heat capacity 30 J mol-1 K -1
Solution у = - Г ^
JnA0о rA
JпА
гд — k\P АРв — k\Pn2^ _

— *7, i(l v ^P
(1 —xA) d 2aq
nAockA
- Jfo * ,( 1 ~ x a ) 2P 2
A0
ПА0 ______ d*A __

P A0k i Jo exp( - )(1 - XA)2
Now we must find T in terms of xA through the energy balance,

nA0(-A .H R)dxA — nroCpdT
dT = п ао (-Л Я к) = (1)(60 x 103) K = 2(Ю R

d*A m cp (10)(30)
T — Tq -Ь 2 0 0 jc a
Now the function that we require to integrate can be evaluated as a function of
хд T f(xA,T) x 105/m3 conversion.
0 300 9.80 ft T\ — Пао 1

- p 2 Qjk/ е х р ( _ З Ш ) ( 1 _ ЖА)2
0.1 320 6.48
0.2 340 4.72
0.3 360 3.78 4.45 x 10~9
0.4 380 3.32 ехр(_ЗШ 0 )( 1 - х А)2
0.5 400 3.22
0.6 420 3.52 Trapezoidal rule, V & A xa + fi A-------1-/7
0.7 440 4.52
0.8 460 7.56
« 0.1(38.24 x 10~5) « 3.8 x 10-5 m 3
5.2.3 Batch reactors

Again, for batch reactors we need to use an energy balance that is in differential
form. Considering a differential change in temperature caused by a differential
change in composition occurring in a differential element of time, it,
N AOdxA( - A H R) = NTc fd T + Qdt (5.19)
Remembering that the material balance for a batch reactor (eqns 2.2 and
2 .6 ) can be written as,
гa Vdt = —dNA — N a 0dxA (5.20)
and substituting eqn 5.20 into the energy balance (eqn 5.19) gives a
relationship between temperature and time,
rAV d t( - A H R) = NjCpdT + Qdt
d T _ г д П -А Я к ) - Q
(5.21)
dt NjCp
The material balance (eqn 5.20) relates conversion to time,
ТdtП = N
^ Ao <5-22>
Compare eqns 5.21 and 5.22 to eqns 5.13 and 5.14. Equations 5.21 and 5.22
can be solved simultaneously in a manner analogous to the treatment for a
PFR.
5.3 Steady-state multiplicity in CSTRs

Consider the first-order irreversible reaction,
A -> В
The material balance for a CSTR (eqn 3.10) with substitution for the
temperature dependency of the reaction rate constant gives,
kr 1
xa = — ■;..------------------------------ j ------------- (5.23)
1 + kr
1+
,, ( AE ,
ex p ( - — )
Fig. 5.14 Lines of constant residence
Therefore, for a constant r, the conversion in the reactor can be plotted as time for a first-order reaction in a CSTR in
a function of temperature (see Fig. 5.14). At very low temperatures the conversion-temperature space.
conversion will asymptotically approach zero and at very high tempera
tures the conversion will asymptotically approach unity. Consequently, a
family of sigmoidal curves are obtained, each one corresponding to a
different value of r. Obviously, for a reactor operating at a given
temperature the conversion will be greater if the residence time, r, is
greater.
To find the operating point of the CSTR we need another relationship
between conversion and temperature. This relationship is the energy balance,
eqn 5.8, Fig. 5.15 Simultaneous plot of the
material and energy balances for a
и а о * а (-Л # к) = njCp(T — To) + Q CSTR with a single exothermic
In the case of cooling or heating coils, irreversible reaction.
nAnxA( - A H R) = птЩ Т - Го) + UA(T - 7-) (5.24)
nAOxA( ~ A H R) = (nyCp + UA)T - (nTc^T0 + UATS) (5.25)

Equation 5.25 is a straight-line relationship in conversion-temperature space.
Figure 5.15 shows the material and energy balance plotted together for an
exothermic reaction. Three different energy balances are shown, all with
different gradients. This could be achieved by having, for example, different
heat transfer areas (unless the reactor inlet temperature is adjusted this would ?b.3j
also affect the intercept). Figure 5.16 shows the effect of altering the reactor Fig. 5.16 Effect of altering reactor inlet
inlet temperature or the reactor coolant temperature on the energy balance. In temperature and coolant temperature on
both cases (Figs 5.15 and 5.16) we see that up to three steady-state operating the energy balance.
points (characterized by the material balance and energy balance being

5 What we are doing here is slightly simultaneously satisfied) can be obtained, one at low conversion, one at
different from how we used the energy intermediate conversion, and one at high conversion5. If only one steady-state
balance in Example 5.2. In that example,
the conversion was specified and we
operating point is available it can be at either low conversion or high conversion.
chose an energy balance that would The low-conversion steady-state operating point is favoured by,
allow that conversion to be one of the
possible steady-state operating points. (a) low rate constant
We did not look to see if there were other
(b) low residence time
possible steady states also available.
(c) small heat of reaction
(d) small concentration of reactant
(e) low values of inlet and coolant temperature
(f) large heat transfer areas and heat transfer coefficient (for cooling)
The high-conversion operating point is favoured by the opposite

conditions.
If multiple steady states are available how do we know at which one the
reactor will be operating? We need to think about the stability of the operating
points. Instead of thinking in terms of the material and energy balances let us
think in terms of the heat generated by reaction and the heat removed by the
Q SQ cooling medium and the process stream.
The heat generated, Qg, will be given by (using eqn 5.23),
Qg = nA0( - A H R)xA = иа0( -А Я к) j k?— = ---- » ao ( A HR)---- (5 .26 )
1 + тk' exp( - | f )
and the heat removed by the cooling medium, Q, and the process stream, QT,
т will be given by,
Fig. 5.17 Heat generation, g g, and QT+ Q = nTCp(T - T0) + U A (T - 7-) (5.27)
heat removal, Q, + Q, as a function of
reactor temperature. The three steady- Now if we plot the heat generation and heat removal lines against temperature
state operating points are shown. (see Fig. 5.17), again we see that we can have the same three steady-state
operating points.
Let us now look at the low-conversion operating point (operating point 1)
in more detail. If the reactor is at that operating point and, due to some
perturbation, it moves to a slightly higher temperature, then, at this slightly
higher temperature, the rate of heat removal must be greater than the rate of
heat generation. Hence the reactor temperature falls until the steady-state
operating point 1 is again reached. If the initial perturbation resulted in the
reactor temperature falling slightly, the heat generation would now be greater
than the heat removal and the reactor temperature would increase until it again
reached the steady-state operating point 1. Consequently, operating point 1 is
stable.
Now let us look at the intermediate-conversion operating point in more
detail. If the reactor temperature is perturbed to a higher level, the heat
generation becomes greater than the heat removal and the reactor continues to
heat up (it will eventually reach steady-state operating point 3 where it will
remain). Conversely, if the reactor temperature drops slightly then the rate of
heat removal will become greater than the rate of heat generation and the
reactor will continue to cool until steady-state operating point 1 is reached.
Therefore, the operating point at intermediate conversion is unstable.
By these arguments the operating point 3 will also be stable and we can
write down the stability condition in a general form.
d (6 r + Q )
For stability, (5.28)
dr
i.e. the rate of heat generation must increase slower than the rate of heat
removal with increasing temperature.
So now we know that our reactor at stable steady state will operate at either
point 1 or point 3 (if we do not use external control to force operation at point 2).
How do we ensure that the reactor is at the desired of the two operating points? If
we start the reactor up, its temperature will increase until it reaches operating Fig. 5.18 Schematic showing how
point 1 and it will stay there. However, in general we will want to operate at the preheating the feed stream will lead to
high-conversion operating point—so how do we get there? If we preheat the inlet only one steady-state operating point.
Allowing the temperature of the feed
stream to the reactor we can ensure that the high-conversion operating point is
stream to then decrease ensures that
the only possible operating point (see Fig. 5.18). (It could be that this operating the reactor will operate at the high-
point would be at too high a temperature for the reactor materials, in which case conversion operating point.
the reactant could be initially diluted.) Now once we are at the high-conversion
operating point we can reduce the temperature of the inlet feed and the reactor
will remain at the high-conversion operating point. What we have done is used
preheating to cause ‘ignition’.
Up until now we have only considered a single exothermic reaction. Let us
now look at the heat generation and removal lines (or material and energy
balances) for some other cases of reaction.
Consider a series reaction (with both steps being exothermic), e.g.
A -> В -* С
If the first reaction goes almost to completion before the second reaction starts
appreciably, then the corresponding curves are as shown in Fig. 5.19(a). It can
be seen that up to five steady-state operating points could be obtained. Three
would be stable and two would be unstable.
For an endothermic irreversible reaction the heat generation is negative
so heat must be supplied for steady operation; see Fig. 5.19(b). This means
that only one operating point can ever be obtained and it must always be
stable.
For a reversible exothermic reaction see Fig. 5.19(c). The heat generation
will increase to some temperature where it will go through a maximum and
then start decreasing as equilibrium is approached at higher temperatures.
Remember that heat generation is proportional to rate of reaction and that rate
of reaction will go through a maximum as the temperature is varied. Therefore,
if we design a CSTR to operate at this maximum in reaction rate it must satisfy
the following conditions, Fig. 5.19 Heat generation, Qg, and
heat removal, Q, + Q, as a function of
dGg_ , d(Qr + 0 n
—^ = 0 ; a n d ----- —---- > 0 temperature for: (a) two exothermic
dr dT ' ~ ~ dT reactions in series; (b) an endothermic
and is therefore stable. reaction; (c) a reversible exothermic
reaction.
The exothermic gas-phase reaction, Example 5.4

A ^ В + C; ?"a = k\P A — k -iP s P c
is to be carried out in a CSTR operating at a pressure of 50 bar.
(a) Calculate the minimum reactor volume if the reactor operates at a 22%
conversion of A (feed is pure A fed at 5 mol s-1).
(b) Heat is removed from the reactor via cooling coils. What heat transfer
area is required for steady operation under the above conditions if the
overall heat transfer coefficient is 10 W rrT 2 K_1?
(c) Comment on the stability of the operating point.
k\ = 0.435 exp^ ^ ввв ^ moi s- i m -3 bar -1
/ - 6 0 000 \ , ,
= 142.6expl— — — I mol s m bar
R T is in J mol-1 ; R = 8.314 J mol-1 K -1

Feed temperature 350 К
Coolant temperature 400 К
Heat of reaction -40 000 J (mol of A )-1
Mean heat capacity of А, В, С 30, 20, 10 J mol-1 K -1 respectively
Solution (a) In Example 5.1, we have already shown that the maximum rate of
reaction,
rA = 0.185 mol s~' m -3
As 1.1 mol s -1 of A are converted, the corresponding minimum volume is,
V = 5.94 m3
(b) Heat released

nAOx A( - A H R) = (1.1)(40 000) = 44 kW
To heat up reactants,
Ш па0А Т = (30)(5)(264.8) = 39.7 kW
Must remove 4.28 kW by heat transfer,
4.28 kW = U AAT
A = 1.99 m2
(c) The operating point must be stable as
dr dr
5.4 Multistage adiabatic PFR

For large-scale industrial reactors, heat supply or removal through the reactor
wall can be extremely difficult and operation must effectively be adiabatic.
However, we know that for an exothermic reversible reaction there is, for any
given conversion, a temperature that maximizes the reaction rate (Fig. 5.4 and
the solution of eqn 5.1). This means that for a PFR there exists an optimum
temperature profile or line of maximum reaction rate (if we have the aim of
minimizing the total reactor volume) and it is important to try to approach this
path. Two different methods for achieving this, interstage cooling and cold-
shot cooling, will be discussed here.
(a) XA2 *A 3 XM X A5 X A6
Fig. 5.20 (a) Process diagram for interstage cooling, (b) Operating lines for interstage cooling
in conversion-temperature space.
5.4.1 Interstage cooling

We preheat our process stream so that the initial reaction rate is not too low.
The stream is allowed to react until the temperature becomes high enough that
the rate begins to decrease significantly as equilibrium is approached (this
point where the reactor should be terminated to give a minimum overall
volume can be rigorously determined—however, such a rigorous approach is
beyond the scope of this text). The reactor is then ‘terminated’ (i.e. the first
stage of the PFR is designed to be of the appropriate size) and the exiting
stream is cooled in a heat exchanger to take it away from equilibrium. The
stream is then fed to the second reactor stage in the series and once more
allowed to react, heat up, and approach equilibrium. Once again, when the
reaction rate has fallen sufficiently the process stream is passed on to a heat
exchanger for cooling before the next reaction stage. Figure 5.20(a) shows a
simple process diagram and the behaviour, in terms of a plot of conversion
against temperature, is shown in Fig. 5.20(b). For adiabatic reactors, all of the
reaction stages will have the same gradient (eqn 5.16). The lines representing
the cooling stages have a gradient of zero as the conversion does not change
(xA2 = *аз> *A4 = * a s )- In Fig. 5.20(b) it can be seen that this kind of strategy
has the result of keeping the operating conditions close to the line of maximum
reaction rate.
However, heat exchangers can be expensive or alternatively, if we are
considering a high-pressure process, the volume they occupy in a pressure
vessel can have a very high capital cost associated with it. These factors must
be considered in deciding whether interstage cooling should be used versus the
alternative approach (which does not rely on the use of heat exchangers) of
cold-shot cooling.
5.4.2 Cold-shot cooling

The principle behind cold-shot cooling is that only a fraction of the process
stream is fed to the first reactor stage. When this fraction exits the reactor it is
cooled by using cold fresh feed rather than a heat exchanger (see Fig. 5.21(a)).
To
(a) *A 2 x A3 л:Аз x A5 x A6
Fig. 5.21 (a) Process diagram for cold-shot cooling, (b) Operating lines for cold-shot cooling in
conversion-temperature space.
This serves to eliminate the need for a heat exchanger but could be viewed as
less efficient than interstage cooling because in mixing with fresh feed the
conversion effectively drops. In practice, cold-shot reactors are used for high-
pressure processes such as ammonia synthesis. This reduces the total volume
of the pressure vessel as interstage heat exchangers are no longer needed. Let
us now look at what is happening in terms of temperature versus conversion
(see Fig. 5.21(b)). We preheat a fraction of the process stream from To to T\
and then allow it to react. When it exits the first reactor stage it is at a
temperature T2 and a conversion xA2- This stream is now mixed with the cold
fresh stream at temperature T0 and with zero conversion. The resulting mixed
stream, (хАз, ? з ), must lie on a straight line between the two streams that are
mixed (i.e. between (О, Г0) and (xA2, T2)), in conversion-temperature space.
The position of the mixed stream on the line depends upon the relative
contributions from the two streams. The larger the relative fraction of the
original fresh stream the closer the conditions of the mixed stream will be to it.
This stream is then fed into the second reaction stage. Cooling is then again
achieved in the same manner. Again, the gradients of the reactor operating
lines will all be similar in the case of adiabatic operation.
Example 5.5 An exothermic, reversible, gas-phase reaction is to be performed in a

multistage adiabatic PFR,
A ^B
K - 2 5 x 10~3 exp^
A H r — —20 kJ (mol of A)-1

c? = 140 J mol-1 K -1
R = 8.314 J m o r 1 K '1
The feed is pure A. The temperature of the feed and coolant streams is
290 K.
Calculate the maximum possible conversion of A using: (a) interstage
cooling; (b) cold-shot cooling.
Yl ^
Equilibrium condition, К = = A Solution
n*A 1 - Д*
у* К l
2406\)
1 + 4 0 0 ex p<(-----—
(a) With interstage cooling the maximum conversion achievable is when the
process stream is at equilibrium at 290 К (to obtain this a large reaction
volume and plenty of cooling would be required),
K = 10.0, xA =0 .91
(b) For cold-shot cooling the maximum conversion can be reached using
one reactor stage with inlet temperature of 290 K. This reactor stage, if of a
volume approaching infinity, would give a conversion that would approach
equilibrium. No heat is actually removed from the process stream with
cold-shot cooling; therefore, we can never do better than is possible with
the cold stream fed to one reactor stage as far as the greatest conversion is
concerned (of course, we can do much better using a series of reactor stages
if we are concerned about the total reactor volume). Alternatively, we can
never get past the operating line for a single reactor stage with the cold
fresh feed as its inlet stream as the operating line is coincidental with the
mixing line.
The operating line is,
п а о * а (-Л Я к) = n jC p (T — To)
cp(T - To) ,
*A = -A H r
= 7 x 10 - 290)
T У»
Operating line, x'A = 1 x 10-3(Г - 290) A
1 290 0 0.91
Equilibrium line, x"K =
A 2406 400 0.77 0.51
1 + 400exp( — ) 350 0.42 0.71
375 0.60 0.60
Solve by trial and error, xA = 0.60
5.5 Problem s
5.1 A CSTR is used for carrying out the (a) Calculate the operating tempera
liquid-phase reaction ture required for a conversion of
60%.
A + В —> Products rA = /сСдСв
(b) How much heat must be added to
The feed concentration of both A and В is the reactor to maintain the system at
3 mol litre-1. The volumetric flow rate is steady state if the feed temperature is 290
1 litre s-1 . K?
V = 4 litres temperature of 300 K. A conversion of

к — 1014 e_A£/,Rr litre mol-1 s-1 80% is required.
A H r = 60 kJ per mol of A reacted Calculate (numerical integration required)
Д E = 100 kJ ш оГ1 the reactor volume required.
сp = 4.2 kJ litre-1 K_1 P = 1 bar
к = 28.1 e ~AE/RTmol s_1 m-3 bar
5.2 The reversible gas-phase reaction A H r = —50 kJ mol-1
A E / R = 1000 К
А о В rA = fci P A - fe_i PB
cp = 100 J т о Г 1 K-1
is to be carried out in a CSTR. The feed is
10% A, 90% inert, and a conversion of 5.5 A gas-phase reaction
62% is required.
A -> 2B rA = kPA
Calculate: (a) the minimum necessary
reactor volume; (b) the amount of heat к= (mo1 of A)
to be supplied or removed for steady
m_3s_1 bar-1
operation.
R T in J mol-1
лт = 2.1 mol s-1
Feed temperature, To = 573 К is performed in a batch reactor. Initially,
Pressure, P = 1 bar the reactor and its contents of pure A are
k\ = 103exp(—A E / R T ) mol m~3 s_1 at 300 K. The reaction is then initiated
bar-1 and the reactor temperature is linearly
i = 106exp(—A E - i / R T ) mol m~3 increased with time at a rate of 0.2 К s_1
s_1 bar-1 to a final temperature of 540 К (giving a
A E \ = 41 570 J mol-1 ; total residence time of 1200 s). Ideal gas
A E - \ = 83140 J mol-1 behaviour may be assumed. Calculate the
C p \ = Cp в = Cpinert = 42 J mol-1 final conversion of A (numerical integra
к -1 tion is required).
5.3 It is desired to carry out the same 5.6 The gas-phase reaction
reaction as in Problem 5.2, again with
62% conversion. This time a CSTR with A -> В + С т-д = kCA
a volume of 5 m3 is available and is to be is to be carried out in a CSTR of volume
used in series with another CSTR, which 0.02 m3. Heat is to be removed from the
Ух K = 5 m 3L * -x ,,= 0 .6 2 is to be designed. reactor by cooling coils carrying water at
(a) What is the minimum volume re 373.2 К and discharging saturated steam
Fig. 5.22 Schematic representation of quired of this other CSTR if the feed at 373.2 K. The feed consists of pure A at
Problem 5.3(a). stream is fed to it initially (see Fig. 5.22)? a temperature of 325 К and at a pressure
(b) What are the feed temperatures for of 1 bar and flow rate 10~3 mol s_1.
the two reactors if they are operated Calculate (an iterative procedure is re
adiabatically? quired) the operating temperature and
(c) What would be the minimum vo conversion of the steady-state operating
lume required of this other CSTR if the points.
feed stream were fed to the 5 m3 CSTR UA = 5 x 10~2 W К"1
first? (Iterative procedure required.) / 18 000
0004\
expl
5.4 The gas-phase reaction Т is in Kelvin
A Hr = —70 kJ (mol of A)-
A -*■ В rA = kPA
cpa = 120 J mol-1 K_1
is to be carried out in an adiabatic PFR. Cpi = 80 J mol-1 K_1
The feed consists of 50% A and 50% inert cpc = 40 J mol-1 K_1
at a total flow rate of 6.3 mol s_1 and a R = 8.314 J т о Г 1 K -1
In practice, plug flow and perfect mixing are never achieved. The behaviour of
all reactors is somewhere between these two extremes of no mixing and ideal
mixing. This is because of effects such as stagnant regions and ‘short
circuiting’ (see Fig. 6.1). In a tubular reactor these stagnant regions will have a
longer residence time than the rest of the process stream and mixing will occur
within the stagnant regions. In stirred tank reactors some elements might
‘short circuit’ or bypass the well-mixed bulk of the reactor. This will result in a
larger fraction of the process stream having short residence times and there Short-circuiting
or bypassing
will be incomplete mixing of all the elements in the reactor.
We need an approach to understand the behaviour of real reactors. The
approach we use is built upon the concept of residence time distributions
(RTDs). We can model a continuous reactor as a collection of small elements;
each one of these elements will take a particular path through the reactor and
have an associated residence time (see Fig. 6.2). The RTD is a distribution plot Stagnant regions
of the fraction of these elements exiting the reactor with different residence
times. Fig. 6.1 Short-circuiting, by-passing,
and stagnant regions in reactors.
6.1 Residence time distributions (RTDs)

Residence time distributions, or RTDs, can be obtained by injecting a tracer
into the process stream at the reactor inlet. For instance, if we have a normally
colourless process stream, we could inject an instantaneous pulse of red dye at
the reactor inlet. Effectively, we have labelled a representative group of
elements entering the reactor at the time of the pulse. We then monitor the
concentration of the dye in the outlet. The concentration of dye in the outlet at
any particular time will be proportional to the fraction of labelled elements
leaving the reactor at that particular time.
6.1.1 Ideal CSTR Fig. 6.2 Modelling of a continuous

For an ideal CSTR, after injection of dye, we would instantaneously see some reactor by considering it to behave as a
collection of differential batch reactors
dye in the outlet stream because perfect mixing would mean that all of the dye
with different residence times.
immediately mixed with the liquid in the reactor; the concentration of dye in
the outlet stream from the reactor would then drop exponentially with time as
the dye continuously leaves the reactor while normal feed, with no dye
present, enters the reactor (Fig. 6.3(a)).
Mathematically, let us consider N moles of tracer instantaneously
introduced into the reactor inlet at time t — 0. Instantaneous mixing means
that the initial concentration in the reactor is uniform,
The unsteady material balance says that the rate of accumulation of dye must
equal minus the rate of out-flow of the dye (there is no in-flow and the dye is
not involved in any reaction),
.dC
(6 .1)
Equation 6 .1 is easily integrated for constant volumetric flow rate to give,
(b) C(f) = C0 e x p C 0 e x p *- J (6.2)
If we want to compare different experiments (which could be conducted with

different amounts of injected dye) it is important to normalize this distribution.
The normalized form is known as the residence time distribution (RTD) and is
denoted by the variable, E.
C(t)
E(f) = (6.3)
J ? C ( i)At
Г С014 0
But C(0dr = / dt
Jo 1>T
where n(t) is the molar flow of tracer in the outlet of the stream.
However j T n(t)dt = N
Fig. 6.3 Concentration versus time tor
tracer injected into: (a) an ideal CSTR;
(b) an ideal PFR; (c) a real reactor. Therefore Г C(t)dr = - = — = тС 0 (6.4)
Jo vT vi
and substituting eqns 6.4 and 6.2 into eqn 6.3,
E(t ) (6.5)
The RTD decays exponentially and because of the normalization the area
under the curve is equal to unity.
6.1.2 Ideal PFR

For an ideal PFR, if an instantaneous pulse of tracer were injected, we would
see no dye until the residence time of the reactor had passed and then all of the
dye would come out in a single pulse (see Fig. 6.3(b)). The RTD is simply
given by the delta function,
E{i) = S(t - r) (6 .6)
where &(t —r) = 0 , for t Ф x
S(t - т) ф 0, for t = x
with J 8(t — r)d t - 1
This corresponds to a pulse of tracer (which, because it is introduced

instantaneously and it does not mix with any other fluid elements in the
reactor, is of infinite concentration) and the area under the curve is
unity.
Chemical Reaction Engineering 69
6.1.3 Real reactor

In the case of a real reactor the situation is somewhere between these two
extremes (Fig. 6.3(c)). Because there is a degree of mixing in the reactor we do
not see a sharp peak, as in the case of a PFR; instead, it is more disperse.
Mixing is not instantaneous, as in a CSTR, so we do not see an instantaneous
value for the outlet concentration of the dye at time t = 0 .
6.2 Calculation of the mean residence time

The RTD can be used to calculate the mean residence time of any reactor.
Returning to the discussion about labelling the elements with a pulse of
red dye in the inlet, the fraction of dye that leaves the reactor in any time
interval is simply proportional to the average concentration of dye in the
outlet stream during the interval multiplied by the duration of the interval.
Therefore, the fraction of the total stream that has a residence time
between / and t + dt is simply E{t)dt (see Fig. 6.4). To calculate the mean
Fig. 6.4 Residence time distribution
residence time, we take a weighted average over all residence times, so for an ideal CSTR.The fraction ofthe total
that the mean residence time is simply the individual residence time for stream that has a residence time
each element multiplied by the fraction of the dye in that element, between fa n d f + d fis £ (f)d /.
summed over all elements. Expressing this for infinitely small elements in
integral form,
/*00
t= / tE(t)dt (6.7)
Jo
If we take the case of a CSTR on substitution of eqn 6.5 we get,
?CSTR
=Л ехр('0“'
Integrating by parts,
tcsTR =
гехрВ)Г+Г ехр(-0“*
Table 6.1
—fexp
t/s £/s -1
= [0 —(—г)] = г = —-
Ух
0 0.00
which is exactly the result we would expect for the mean residence time of the 2 0.04
CSTR. 4 0.15
Likewise, we can use the RTD of a PFR, substituting eqn 6.6 into 6 0.15
eqn 6.7, 8 0.10
/»oo 10 0.05
V
fpFR - J tS(t — r)df = r — 12 0.01
Vj 14 0.00
again as expected.
Given the reactor residence time distribution (RTD) shown in Table 6.1, Example 6.1
evaluate the mean residence time.
Solution Equation 6.7, 1= J tE(t)dt
However, we have been supplied with discrete data and therefore we must use
Table 6.2 a summation rather than an integral.
t & At tE
t/s tE
Evaluating tE, as in Table 6.2, gives 7 = 6 s
0 0.00
2 0.08 6.3 Calculation of conversion from RTD
4 0.60
We have already said that we can model a continuous reactor as a collection
6 0.90
of small elements. The concentration of reactant in each of these elements
8 0.80
10 0.50 will be a function of how long the element has been in the reactor and, from
12 0.12 the RTD, we know how long each element will spend on average in the
14 0.00 reactor. Therefore, we can calculate the mean outlet concentration of
reactant,
C~a = [ CA(t)E(t)dt (6 .8)
We may treat each of these elements as batch reactors and, if they are
small enough, we can consider them to be well mixed internally. For a
first-order, irreversible reaction the design equation for a well-mixed batch
reactor is, ,_
dCA
—— = - r A = - khCr K
dt
Therefore, for any element,
С a = СAOexp(-fa)
and substituting into eqn 6 .8 ,
rOO
Ca = CAo J e xp (-kt)E (t)d t (6.9)
Table 6.3 Evaluation of this integral for a PFR is trivial and so we will demonstrate the
case of a CSTR. We substitute eqn 6.5 into eqn 6.9,
t/s E/s
C a . c t s r = C A0 J exp( - k t ) i e x p 0
dt
0 0.00 /•0°
2 0.04 = Cao /
4 0.15 Jo
6 0.15
3
exp( —kt — - )
II
I
I '
8 0.10 V
10 0.05
Cao
12 0.01
1 + kx
14 0.00
----------------- which is the expression that would be expected for a CSTR.
Example 6.2 A first-order, liquid-phase reaction,

A -» В, rA = kCA, к = 0.307 s_l
is performed in a non-ideal reactor with the reactor residence time distribution
(RTD) shown in Table 6.3.
(a) What is the conversion of A?

(b) Compare the conversion to that of a PFR and a CSTR of the same mean
residence time (6 s).
(a) For a first-order irreversible reaction, eqn 6.9 applies, Solution
~ = / exp(—kt)E(t)dt
<"A0 JO
Table 6.4
From Tables 6.3 and 6.4,
t/s e *'£/s -1
Ca ~ a t exp (~ k t)E = 2(0.106) = 0.21
Сдо 0 0.00
cl 2 0.022
Xa — 1 0.19 4 0.045
С AO 6 0.026
(b ) xa ,p f r = 1 - exp(-fcr) = 0 .8 4 8 0.009
kr 10 0.003
•Xa . c s t r = = 0 .6 5 12 0.001
1 + A :t
14 0.000
We can see that the conversion estimated for the real reactor lies between the
two extremes of the CSTR and the PFR.
Solutions
Solutions to Chapter 3
3.1 A -► 2B rA - kPA = k ^ - P
(a) N a = Л/'до —N AQXA
N\i = 2N aoxa
N\ — Nw
N r — N j о + N AoxA
— = ^ О + Л ^А О Х А = = = 0 ? 6
Po NT0 2
(b) rA = k P A = k ^ P ; PV = NTR T
RT
rA = kNA — ; rA V = kRTNA
1 diVA _ Л d/VA
(C) ГЛ V df ’ LWao rAV
1 Г. , , yvA _ ^ 1- -ÂO
f = “ j K r P n iV A ^ « = kRT ln IV a
1
1~kR T Чт^);‘=^ 1п(пУ
к = ----------------------- ln(4.167) = 2.02 x 10~6
(170)(8.314)(500)
it = 0.202 mol bar 1 m 3 s 1
CNa 1 d/VA
3.2
- jf
t= — I - —
Jn„ rA
N2
2A -> B; rA = £C2 -
'" A 1 d JV A y 2
- / J *n до ^ ^ A
C onstant volume, Vo
rv = ,Y±
к
= V j ____ L \
к \N a Nao)
_Vq W I
A: Nao \1 - * a
Vo
kNA0 1 - xA
C onstant pressure, Pq (initially reactor is of volume Vo)
Na = N ao —N ao * a
Nb = Nb + - N AQXA
N t = Nao —2 ^ A0*A
P0Vb = N AoR T ; P0V = NTR T
rN* NTR T dNa

= fn* Po k N \
- JJNM
R T l Г*л
xА (1
П -- i\xx* A)
')
“2 (—NAo)drA
~ ~Po~k J 0 Nao(1 —x A )
RTl r ( _ i _
+ d*A
^ к Л) 11 - *A (1 — xA)2
RT 1 Xfi,
*a) +
' ~2Pok (1 - * a )
V0 XA
- l n ( l - X A)
2kNA0 [ 1 - xA
xA = 0.9; fv = 9V0/k N A0\ tp = 5.65Vo/fcNAo; tv/ t p = 1.59

As the reaction proceeds the volume of the constant pressure reactor
decreases, hence the concentration and therefore the rate are higher.
3.3 2AO B
га = h P \ - к - ХРъ
«АО _ 2 «ВО _ 1
пто 3 ’ пто 3
1 1
па = пао — ПА0ХА , пй = «во + 2 Иао*а. пт = «то —2 Па°Ха
Р = 1 bar, х А = 0.25
(а) РА = 0.545 bar

"A o (f - 5*а )
74 Solutions
пао( \ + { ха ) .
«ао(| - 5 *а)
(Ь) гА = кхр А\ - L_,PB = (0.05)(0.545)2 - (0.025)(0.454) mol m " 3 s '

= 0.0035 mol m ~3 s _1
(c) «то = - ~ r
2 PVT0 2 ( 105)( 10-3) , ,

= ъ - W = 3(8.314X600)mo1 s' = 0 0 1 3 4 шо1 s
(d) v = ^ = (00134X0;2 5 ) m, = 0 9 6 m ,
v ’ rA (0.0035)
3.4 Liquid phase, v j is constant.

^dWA
■ ~ J>rПа
«АО
« А = ЯAO — « А 0 *А , «В = «ВО — п Айх А
«АО = «ВО
_ _ Г А -«AoiCA
Г —/гдрсЬ
/о
4
_
«AO*io Г
Г d-XA
(1 - Ха)2
пАОк
±.(-1--- ---------------
:А0* V1 - -*А/ Сао* V1 “ *а
Л
/
«АО = ^ т С а О
3
(тЛз-1) '001
3.5 А + В -> С
у = _ Г А^ = _ Г dnA
JnA0 ГА Jnдо *РР В
« А = «АО ~ «АО*А
«В = «ВО — «АО*А
«С — « а о * а
ИТ = пто —ПаоХА
«АО = 2/1В0
пто = Зиво = ^ "AO
p R — *}]L . p — nAo(s ~ Xa) p __ \ — xA p

m «то - n\QX\ l~ x A
_ _ F A -» A p (| - *A )d x A
Jo Щ 3 - *a)
=^ f ( 1+Fs;)d,‘4
= ~ ^ XA - Ц - Xa)]
80
= ( 2 .3 x l 0 3 X i 0 ) 104^ ' " (005)1
= 6.0 x 10~3 m 3
3.6
- fJ Wi
У= - I
/«A,
—
'■A
«а = птУа . where уa is the mole fraction of A
y _ _ Г Ле «TdÂ
>A
Val Р (к\У А — & -l(l —Уа))
'>,Де Итйул
- f Jy.
yAi + *-,)Ь а - —
№+*-0
= 0.4
(&i + &_i)
v = — 51— i n W i~ ° - 4
P (fcl+£_l) Â e-0.4
VAi —0.4
Therefore, уде = ------------- 1- 0.4
V
P V{kx + *_Q
where, (p = exp
nj
and we have a relationship between yAe and yAj.

(a) nT = «ao; Mi = 1
P V j b + k . 1) „
nj
1 - 0.4
yAe = ---- ------ 1- 0.4 = 0.481
ez
«лоМе = «тСуа1 - УАе) (amount of A reacted)
76 Solutions
*A e = 1 - jA e = 0 . 5 1 9
(b) nTi - nTe - n j

The flow in the reactor is no longer equal to nM ,
пао = » t( 1 - « )
and rewriting the relationship between yAe and удь

jAi 0.4
УАе = ------------- Ь 0.4
<P
/
where cp = exp I -
«А0 /
The material balance at the mixing point,

«A i = «АО + a «Ae
In terms of mole fractions,

n A0 + a « A e ,, S .
УА1 = --------------------= (1 - « ) 4 OtyAe
nj
Eliminating удь
(1 - a) + ayAe - 0.4
M e = ------------ ------------- 4 0.4
_ 0 A (tp - 1 ) + (1 - a )
tp — a
1 ,
« = - ; <p = e
yAe = 0 .5 3 5
X A = 1 —yAe = 0 .4 6 5
(0.4)0? - 1) 4 (1 - «)
(C ) Me =
(p — a
A s a - * 1 , 0 = e2(1_a) « 1 + 2(1 - a)
(0.4)(2)(1—oQ+ ( ! - « )
УАс 1 + 2(1 - a) - a
!. 8(1 — a ) _ L 8 _
3(1 - a) 3
But total moles of A reacted, nAoXA = «AoCvAi — УАе)
X A = 0.4
4 .1 A —> 2R уА\ = ^\C a
A -> 3S Га2 = *2Â

Solutions
(d)
rA
(CA0 — Ca )l>t = V(k]CA + k 2C 2A )
C ^ ( C ± \$ V b + l\ C ± _ l= 0
vt \C aoJ \ vt J C ao
+2^ - 1=0
Vt-AO / с AO
CA
— = 0.232
Cao
xa = 76.8%
We can see that the conversion of A is much lower in the CSTR.
,ч «в - «во , n
(e) --— = k xС a
Св — Cbo — k\C A —
( 1-00 )
= (l)(0 .2 3 2 )^ y y i = 0.232 mol 1
Kb/ a = - I- ~--Cb° = 23.2%

L ao
However, the yield of В is greater in the CSTR than in the P FR despite
the lower conversion. This can be explained by the more favourable
selectivity obtained at low reactant concentrations (the favoured reaction
is of lower order than the competing reaction).
m г с в - C B0 __ 0.232
® Л ' А - ^ с ; - 0 Л 68 = 30'2 Л
(g) Cao*a^t = У(к\СА + k2C $
(1)(0.95)(1) = V((l)(0.05) + (10X0.05)2)
У = 12.7 litres
— ~ Cb-° = k i ^ — = (0.05X12.7) = 63.5%

Cao С-до Ut
(h) 5b/ A= ^ = 66 .8 %
4.4
(a) Some В must be added to the reactor to start reaction 1. Once the
reaction has started then the В produced will catalyse the reaction and the
reaction will continue.
80 Solutions
(b) Design equation for reaction 1,
v = - wg -
^ i Ca Cb
Ca = F ~
Kit
Design equation for reaction 2,
Пг
V = —-—
к гС \
h
Cc = k2C \T -
k\
CB - CA0 - Ca - Cc = Cao
( 1 kA
(c) M ust maximize C aC b or C a o (1 - * a ) I Cao — ^ — )
Сдо(1 —*a)^Cao — Cao(1 —*a)^Cao*a —^7 ^
C aoâ - ~ r - CAo4 + t^ * a must be maximized,

k\ k\
Differentiating the above expression with respect to xA and setting equal
to zero,
k^ .
Cao —2CAoxA + — = 0
xA
2 \ ^kyC M )
CB Cb Cao \? A ^ ° )
Sb / a =
Cao — С a Cao Cao ~ С a xa
1 ___*2_
a Cao
SB/A = 1 I k2
1 ^ *iCA0
4.5 A -> R -*■ S

PFR: wr is maximum when (eqn 4.46),
l n № = l n4 = 0.924 s
k t-k i 1.5
*A = i _ ^ = i _ е -» ,т = о 8 4 3
«А0
« к / « А о = ^ \ - { е - * 1Г - е - ^ г}
k2 - k i
= —^ ( 0 .1 5 7 - 0.630) = 0.631
Xa = « r /ИАО + ” s / nA0
ns/nA0 = 0.212
k \z
CSTR : nR/n Ao =
(1 + * i t ) ( l + k2x )
dnR
= 0
dr
1 _ d«R _ —* ir[(l + k \ t )*2 + (1 + *2^)*i] + (1 + fcit)(l + k2x)k\
ПА0 dr (1 + * ir)2(l + *2T)2
k2x + k ^ x 2 + k\T + к\к2т^ — 1 —k\ r —k2x — k f a x 2 — 0

k ^x 2= 1
1 = 1.00 s
*Jk\k2
xa = = 0.667
1 + k \X
Z * = _______ kA L _______ = 0.444

идо (1 + *i т) + (1 + k2x)
ns/nA0 = 0.223
4.6 /fc3/*i = 0.4; (k2 + h )/k x = 0.2
N o В present initially, up is constant.
(a) P F R : nA = nA0e~{kl+k,)r
= 1 - xA = 0.3 = е~(*1+*з)г =
«АО
but = k \C a — (k2 + кл)Съ
dwg
— = k XHA - ( k 2 + k 4) n B
^ = (k2 + k 4) n B = k xn A 0e ^ k' +k,)t
Solution of the form:
ив = Ae~(kl+k*)z + B e -(kt+ki)r
82 Solutions
~ = ~ (k 2 + k4)AQ-(h+ki]T - (Jfcj + h)Be-<-kt+h)z

ax
As «в = 0 w henr = 0 : A + В = 0
Coefficients of e_(*1+*3)T:
—(fci + fc3)S + (^2 + k ^ B — k\ идо
n _ fel” A0 A = -B
(к г + fc t) — ( k \ + к з )
n* = {кг
7 Г+ Ткь)Т —
Т (k\
^ 7+Г*3)
^ {е~(к' +кз)Х ~ e~{k2+ki)z}
_ __________ к \П ao __________ ( t 2 + fc,)T _ g - ( i i + *3) r i
(fc i + k 3) — ( к г + к ц )
e - ( h + h ) t = e-i.4*ir _ 0 j
е-(к2+кл)г _ e -0.2*,r _ (0 3)1/7
«В = ' " A0 {(0.3)1/7 - 0.3}

1 . 4 - 0.2
^ - = 0.452
ИД0
(b) CSTR : ---------- = (A:, + *3) —
«АО
nA = •
1 + (&i + кз)х
«в - и в о , иА , , .лв
---- 17---- = ^ ------ (*2 + м ) —
k i r n АО
Ив = —(&2 + &4)гиВ
1 + (*1 +
к\ тпдо
{1 + (&i + £3)т }{1 + (кг + к/С)х}
и t
but ха " А = 11—
= 11 ------- 1
ИАО 1 + (fci + къ)т
(ki + к3)х
= 0.7
1 + (£i + к3) г
(ki + къ)х = 2.333
k ix = 1.667
(&2 + &4)г = 0.333
, ___________1.667 _
ив/идо {1 + 2 .333 } + {1 + 0 .333 }
5.1
. , .. ПДО - ИД « A 0 *A
(a) V =
гд к \С \С в
Сд = Св = —- — CA0(1 —*a)
Vj
nAQxA
V = , ----------- Z2 ' ” A0 = C aoV T
1 AO^ ~ x*0
, xkv-\
M — __ ,* 42
УСА0(1 - * a)
(0 .6)( 1) litres -1 m ner* 1-1 -1
= —- — ------------------- Г = 0.3125 litre mol s
(4)(3)(0.4) litre mol litre ' 1
k x = 1014е_ЛЕ//гг
R / 1014
- 105/8314 - = 360.1 К
In (0.3125 x 1 0 -14)
(b) Energy balance:
vjCp(To — T ) + A # r * a « ao + Q — 0
Q = —(1)(4.2)(70.1) - (60)(0.6)(3) kW = -4 0 2 kW = 402 kW (added)
5.2 A B; r \ — k \P \ — k\P ^
n " A O ,, 4D D «A 0*A D
P A = ----- (1 - XA ) P ; P B = ---------- p
n-\ nj
nj 1 .
r A ------ -- = ki - (ki + k - i)* A
пао Р
nj drA
dra dfci
dk\ (/ dki
dk{ d/c_i \
dfc_i
n AoP dT = dT _ V d r + ~ d F ) XA
and ^ = 0 for maximum rate at a given conversion
d k i/d T
xA =
d k i/d T + d k - i / d T
1
(^L )(-l/r 2)t'1e-i£i/'n'

1
1 _L A£~l *-1
1 ' r ДЯ, кI
1
1 + (2 x 103e-500°/:r)
-5 0 0 0
Solutions
At this tem perature
rA = r^ { k l ~ ( h + k - i ) x A )
nT
k\ = 0.306 mol s_1 m -3 b ar -1
*_i = 0.094 mol s _1 m -3 bar -1
rA = ) „ „ 0 306 - (0.4)(0.62))mol s' 1 m " 3 = 5.8 x 10 3 mol s ' 1
y = w a д (о -и х о -е д т з _ 2 2 4 m ,
rA (5.8 x 10-3)
Energy balance:
Q + птсР(Т - Го) = { - A H R)xAnA0
A H r = A E i - A E - i = -41570 J m ol -1
Q = —(2.1)(42)(45) + (41570)(0.62)(0.21) W = +1443 W
= 1443W (removed)
5.3
(a) Reactor 2 operates at 618 К to maximize rate (see solution for
Problem 5.2),
П \о ( х А 2 — * A l)
v2 =
Г\2
(5)(5.8 x 10-3) nioo

* A 2 - « I = -------- ------------------ = 0 1 3 8
*ai = 0.482
-5 0 0 0
T\ = о гг = 664 к
In 2000^
At this temperature: k\ = 0.537 mol s_l m -3 b ar -1
j = 0.288 mol s -1 m ~3 b ar -1
(0.21)(0.482) з _ з
v . = ---------------------- - ----------------------— m = 7.40 m
1 (0.1){0.537 - (0.537+0.288)(0.482)}
(b) Adiabatic tem perature rises:

n jC p A T — ( —A H r ) ii A o A x a
A T = (4157Q) (°,4- 2-l = 47.7 K, T i (feed) = 664 - 44.7 = 616 К

1 (10) (42)
АТг = = 13.7 K, T2(feed) = 618 - 13.7 = 604 К

2 10 42
(c) W ith 5 m 3 reactor first, to maximize rate in reactor 1:

-5 0 0 0
In 2000
1
« A O Â l
Vi =
, T 4 _ , _ __________ »T*A 1___________

* ’ PVx{ k i - { k x + k ^ ) x ai}
= 0.42 mol s-1 m ~3 b ar -1

PV l
0.42 ДГА1
g(xA, T ) =
h - (k\ + k - i)xAi
We guess xAi and evaluate the temperature, the rate constants, and g,
XA1 A’-i/mol s _1 9 { xa J )
jcai = 0.435 m-3 bar-1
Reactor 2 operates at 618 К again.
0.43 0.663 0.956
_ n A0( * A 2 - * A i ) _ (0.21X0.185) j _ з
о 1 л ^ o ./0 m 0.435 0.649 0.996
ГА2 (5.8 X 10~3) 0.44 0.636 1.037
5.4 A —> В /*д = k \P \

Energy balance (adiabatic PFR):
dr _ гАЛ Я к
dV n\\)Cp
M aterial balance:
dxA __Гд_
dV идо
d T = «до ( ~ A H r ) _ 1 / 5 x 104 \
dxA nT0 c? ~ 2\ 100 /
dr
3 — = 250, Г - T0 = 250 xa
dxA
i d Â j n Â0 ч
t, = t; c x p (^ -^ j
y _ f XA nApdxA______
Jo P B * ( l-X A ^ e -A * /® -
_ «Т FA dxA
PIЦ Jo (1 - XA)e -A£/-R(7’o+250xA)
86 Solutions
0.224
r (1
dxA
. -1000
* a ) e x P jI 300+250*д
300+250* a I
I
0.224
f ( x A) =
(1 *A)exp| 300| ^ Хл
Xa T *1 f(xA)/m3
0 300 1.00 6.285

0.1 325 1.30 5.404
0.2 350 1.61 4.880
0.3 375 1.95 4.610
0.4 400 2.31 4.552
0.5 425 2.67 4.716
0.6 450 3.05 5.172
0.7 475 3.42 6.135
0.8 500 3.80 8.283
Trapezoidal rule:
_ , i f o + f i) A .(Л + Л )*_ , . ( Г 7 + / 8 ) Дт
Integral = — ------ Д*а H------- ~---- Д*А + • • • H------ ^— a *a
1 dNA
5 '5 r* = - v ~ d T
1 dxA
jC *-44 kP a
^ = ^ ’p v= ^ (pinbar)
Л7ао( 1 -ДСа)ЛГ
Pa =
105V
*A
105i *( 1 -* а )Я Г
f kR Tdt = 105 [ * т т ^ Ч
Jo io (1 —*a)
T = 300 + 0.2f
Let [ kR Tdt = f f( t) d t
Jo Jo
J /•1200 /4 no 255 7 \
/(r)d f = 2 0 0 ^— + 11.72 + . .. 165.8+ - ^ - J = 0.982 x 105
t/s 77K f
/л л л л Г 1 /1 \Tl4
u . y s z = | —i n ( i - x a AIo
0 300 4.02
200 340 11.72
In — i — = 0.982 400 380 27.57
1-x A 600 420 55.69
800 460 100.4
xa = 0.62
1000 500 165.8
Note: this question does not actually involve an energy balance and could 1200 540 255.7
have been solved using only the material from Chapter 3 and earlier.
5.6 A - + B + C
Energy balance:
п ю с р а (Т - To) + U A (T - 7j) - - A H RnAOx A
nroCpA . л UA
* a - — — ------- (T - T0) + — — --------- (T - Tj)
— A # r « ao — A H r « ao
120 5 x 10~2 _
(T - 325) + —— — (T - 373.2)
7 x 104 v (7 x 10_4(10~3)
170
= ----- —т ( Г —339.2)
7 x 104 v '
ИА = «АО — Па о Х а , Ив = Па о Х а , ПС = И А0*а ,П Т = «АО + ПАО*А
ПАОХА Па о Х а ,, ,
у = 7---------- = 1— п ------------ч • "а о (1 + * а ) - ^ -
Л п а /V t Ь Ао (1 - х а ) Р
(As v j — n jR T /Р for a perfect gas)
x a (1 + *a ) t «АОR
(1 —JCA) k PV
= 4.16 x 10~6 s' 1 К -1
L e t/(jA ,7 0 = * A^ - _ ^ ) Y oii/i80oo/r(4 -16 x 10' 6)
Use iterative procedure—guess T, calculate xA from energy balance,

substitute in expression Cor f(x A, T) which should equal unity. Г/К Xa f(xA,T)
Г = 592 К , x a = 0.61
550 0.512 5.93
This is the unstable operating point because there are two other solutions 592 0.614 1.013
on either side, 595 0.621 0.904
600 0.633 0.751
x a ^ 0; T = 339.2 K, Therefore x \ — 6.4 x 10 -10
xa « 1; T = 751 K, Therefore 1 - xa « 1.6 x 10 3, xa = 0.9984 T - 592 K, xA = 0.61

Nomenclature
Below is a list o f common symbols used. Typical units are shown in

parentheses; however, kJ, kmol, litres, and kW are used interchangeably
with J, mol, m 3, and W.
A H eat transfer area or cross-sectional area (m2)

Ca C oncentration of species A (mol m -3)
Ca Mean concentration of species A (mol m -3)
cp Specific heat capacity (J m ol -1 K -1 or J m ~3 K -1)
cp M ean specific heat capacity (J m ol -1 K ” 1 or J m -3 K -1)
E (t ) Residence time distribution (s~!)
k Reaction rate constant (units vary)
kf Pre-exponential constant in expression for reaction rate
constant (units vary)
К Equilibrium constant (units vary)
I Length variable (m)
L Total length of reactor (m)
n Reaction order (dimensionless)
па M olar flow rate of species A (mol s-1)
nj Total m olar flow rate (mol s_1)
Na Moles of A in a closed system (mol)
Nj Total moles in a closed system (mol)
иА1 M olar flow of A reacted or produced through reaction
1 (mol s_1)
N ai Moles of A reacted or produced through reaction 1 (mol)
Pa Partial pressure o f A (bar)
q H eat flux (W m -2)
Q Heat transferred from a system (W)
Qs H eat generated by reaction (W)
QT H eat removed by cooling or heating o f the process stream (W)
га Rate of reaction or form ation o f A (mol m -3 s-1)
rAi Rate o f reaction or form ation o f A through reaction
z'(mol m -3 s-1)
R Gas constant ( = 8.314 J m ol -1 K -1)
Re Reynold’s num ber
S b/ a Overall selectivity for В with reference to A consumed
(dimensionless)
Sv Space velocity (s-1)
t Time variable (s)
t mean residence time (s)
T Tem perature (K)
7], Го Tem perature of cooling/heating jacket and feed respectively (Ki
U Overall heat transfer coefficient (W m ~2 K -1)
V Reactor volume variable for PFR , total volume for a batch

reactor or CSTR (m3)
vt Total volumetric flow rate (m -3 s-1)
x\ Conversion of A or 'per-pass conversion’ of A (dimensionless)
* a ,r ; Conversion of A for an individual reactor i (dimensionless)
Xa Overall conversion of A for a reactor with recycle
(dimensionless)
уa Mole fraction of A (dimensionless)
У в/ а Yield o f В in terms o f A fed to the reactor (dimensionless)
a Fraction recycled (dimensionless)

5 Dirac delta function
0 b/ a Local selectivity for В with reference to A consumed
(dimensionless)
v Stoichiometric coefficient (dimensionless)
в Total residence time of a batch reactor (s)
r M ean residence time of a continuous reactor, residence time of a
P F R (s)
£ Extent o f reaction (dimensionless)
ДE Activation energy (J mol-1)
ДЯя H eat o f reaction (J mol-1)
ДU Change in internal energy (J mol-1)
Subscripts
Inlet conditions to a reactor with recycle; intermediate stream
i z'th reaction; z'th species; z'th reactor
e Exit conditions
0 Inlet conditions or time t = 0
Superscripts
Denotes equilibrium
Further reading
This book is designed to serve as an introductory text to the subject of
chemical reaction engineering. The following texts are suggested to
broaden the reader’s perspective and to develop more advanced aspects
of the subject (reference [4], in particular, is more advanced).
[1] ‘Chemical reaction engineering’, Levenspiel, O., 2nd ed. (1972).

[2] ‘Chemical reactor theory: an introduction’, Denbigh, K. G. and
Turner, J. C. R., 3rd ed. (1984).
[3] ‘Elements o f chemical reaction engineering’, Fogler, H. S., 2nd ed.
(1992).
[4] ‘Chemical reactor analysis and design’, From ent, G. F. and
Bischoff, К . B., 2nd ed. (1990).
Index
activation energy 36, 49, 50 autocatalytic reaction 48 heat generation 60-2

autocatalytic reaction 48 CSTRs in series 21-6 endothermic reaction 61
adiabatic reactor CSTR in series with PFR 26-8 heat of reaction 49, 53
CSTR 54 design equation 11 heat removal 60-2
multistage PFR 62-5 energy balance 52-4, 65, 66 hot spot 56, 57
PFR 55, 57, 58, 66 mean residence time from
temperature versus conversion 55 RTD 69
Arrhenius equation 36, 49 optimum temperature 51, 52, 66 instability 60, 61
axial mixing 8 optimum volumes in a series of instantaneous selectivity
CSTRs 24-6 definition 3, 34
order of CSTRs in series, 24, 41 parallel reactions 35
batch reactor 5-8 (see also parallel reactions 37, 40 parallel reactions of different
comparison of reactor parallel reactions of different order 40
performance) orders 40, 42, 43, 48 interstage cooling 3, 63-5
constant volume 14, 31 residence time distribution 67, 68 isothermal reactor, see batch
constant pressure 14-16, 31 series reactions 47, 48 reactor, CSTR, PFR
design equation 5 single isothermal reaction 19, 20,
energy balance 58-9 31
parallel reactions 35-8 steady-state multiplicity 59-62 kinetics 1-3,
parallel reactions of different thermal effects in 59-62 temperature dependence 49-52
order 40
series reactions 45-6
single isothermal reaction 13-16, 31 design equations 3
bypassing 67 batch reactor 6 local selectivity
CSTR 11 definition 3, 34
PFR 9 parallel reactions 35
С curve, 68 deviations from ideal flow 67 parallel reactions of different
chemical kinetics, see kinetics Dirac delta function 68 order 40-5
CFSTR, see CSTR
‘Cold-shot’ 3, 63-5
comparison of reactor E curve 68-9 material balance equation, general
performance, endothermic reactions form 5
batch reactor versus PFR 10, 11 energy balance 3, 4 mean residence time 21, 69, 70
CSTRs in series versus PFR 3, batch reactor 58, 59 mixing, see perfect mixing
22, 23 CSTR 52-4, 65, 66 multiple reactions 4
CSTR versus PFR 12, 20, 21 PFR 54-8, 65, 66 see also parallel reactions, series
CSTR versus PFR, parallel energy of activation, see activation reactions
reactions of different energy multiplicity of steady states, see
order 41-5 equilibrium 2, 3, 50 steady-state multiplicity
CSTR versus PFR, series equilibrium constant 3, 50 multistage adiabatic PFR 62-5
reactions 47-8 extent of reaction 7
recycle reactor versus CSTR 3, 30
consecutive reactions, see series non-ideal reactors 67-71
reactions fraction recycled 29-31
Conversion 7
CSTR 5, 11, 12 (see also operating line, conversion versus
comparison of reactor heat balance, see energy balance temperature for batch reactors
performance) heat capacity of a stream 53 and PFRs 55
optimum reactor type, see pressure drop 18 series reactions 3, 45-8, 61

comparison of reactor preexponential factor 36 semi-continuous, semi-batch
performance reactor 1, 3
optimum temperature 50 -2, 66 space time 3, 17
order of reaction 2 space velocity 3, 17
rate constant 2
choice of reactor 21, 43 start-up, of a CSTR 61
reaction order, see order of reaction
order of reactors 23-8 steady-state multiplicity 59-62
reactor, see batch reactor, CSTR,
overall selectivity stirred tank, see CSTR
PFR
definition 3, 33 stoichiometric coefficients 2, 33, 34
recycle reactor 28-31, 32
parallel reactions 34-7
see also comparison of reactor
parallel reactions of different
performance
order 41-5
recycle ratio 28 temperature
residence time 17 optimum 3, 4, 50-2
residence time distribution 67-71 profiles in a PFR 56-67
parallel reactions
reversible endothermic reactions tracer 67
of same order 34-40, 48
dependence of rate on tubular reactor 8
of different order 40-5, 48
temperature 49, 50
perfect mixing 3, 5, 11, 67
in a PFR 56
per-pass conversion 28-31
reversible exothermic reactions
plug-flow 3, 8-11, 67
dependence of rate on variable volume batch reactor, see
PFR 5, 8 (see also comparison of
temperature 50 batch reactor; constant pressure
reactor performance)
multiple steady states in a volume change due to reaction 15,
CSTR in series with PFR 26-8
CSTR 61 17
design equation 9
multistage PFR 62-6
energy balance 54-8, 66
optimum temperature 3, 50-2, 66
‘mean’ residence time from
RTD 69 PFR 56, 57
Reynold’s number 8 yield
optimum temperature 51 definition 3, 33
parallel reactions 35-8, 48 parallel reactions 34
parallel reactions of different parallel reactions of different
order 40-5, 48 second-order reaction, CSTRs in order 42
residence time distribution 68 series 27
series reactions 45, 48 selectivity, see instantaneous
single isothermal reaction 16-19, selectivity; local selectivity;
31 overall selectivity zero-order reaction 24, 26
CO_
CQ tr С
Э 2< «к CD . G
о
OL . 4 CD С
>> <n X 8 ^ ^ oi S X s со ОС i j I ? ! !
r“ ^ ф ПЗ Ф«*-
CO ^
_• 0) bо
s | A » CLО
138.906 140 115 140 908 144 24 (144.917) 150.36 151.965 157.25 158.925 162.50 164.93 167.26 168.934 173.04 174.967
227.028 232.038 231.036 238.029 237.048 (244.064) (243.061) (247.070) (247.070) (251 080) (252.083) (257.095) (258.10) (259.101) (262.11)
00 CO 3
<
103
S9 CO
O_ § 4 - 8 .1 2 си
Lu
71
in ,
- l8
Lr
>>
CT) U. 8 * COCD
in U „ ® я 0100
00
CO £ ся *- E с <- о>
IliU I V H lu w a )
с э о
CNJ & i с сл
л 00 о | d l- 3 |
102
No
Yb
70
00 О 1 $2(/> О Ф о -* 2 ® « 5
oo CL об ■2 со ф jC сд
о
CNJ iriE 2®
r*- о о oi . ®-Э i
CVJ с $ « S' ф
r - D s COZT 00
Tm
л m< ^
101
Md
Ш O) j ®и £ £
69
Я « S
—> > о CO< „
г** Л(Л5
oo CQ go
? . a-Q
i fз Ь т■зсw
Itlllll I I I H A Ii
о
CO CNJ ^ i Q.O S.™
♦L 2 з Е o n
C0
._ 0O (Л Я О с i
Q) с з >•
Fm
100
CM © S о С ^ О) ®О ош>
68
S!#P ^
Er
(D O °
CO
cnj
со СЭ c\i I m CO 00 oo OL о
Г4' ! T-
&Uhi
.I CNJ
00
СП ■*” *С
ОCСN
О 00
J О) с 00 Il.n l Е
Ho
67
Es
99
in m op ЙN ^ ^ ooh Tf ° со с 3 > ч ^
о СО
CNJ « о «СО f- о
CNJ
— <в 3 -р . .
§® 8о gэ a
с §
g. §
и
cn
CDCD T> 5 о 0 >£ CNJ
Dy
66
98
o C n
Cf
юO
con C О C\j ® I 8 T- T- N CNJ : £ 1 3 . 9 ^ х> ®
«•!«•••••«. w v l g l i l t
<N та ~ з 8 1 м
CO I CO « з я 2 « w
CO
_Uf со o i3 ^ ▼ - , C « м 8 я § Я>
COCDl r-N
J
Bk
Tb
65
97
fsa jj t - < g *- C NJ « ® ® шI i
CO О«
D > ^я ~Q) ? -S
I-
COj CNJ с Z ®8 ф Е
«
о 00= $ ICDX> 3 z s : sф ^- ^w. сео еЕ £®
Cm
Gd
64
96
>> ^ j«Q. g I (СфО 0) О Q
w Ih h i h m I
оо си о _о та
о a) w 2 *“ С—
COI X ф 8
с 9? ’ф
к- »ф 9^ 2
ф
rs. О 5 !in «С a) Am > £ фта о. « m
Eu
63
95
5 ■*(£ Ы ^
« тс а* | Е о^о
со — Ь ш
I
и|««1
фW
Й 10 Ф
Й® 5 о%
фТ® .О
<m I''
«о I00 0) in
— Ф ^ mж О.9л *П-
cd 4»3 « 3 9 _jg о i5 *®
Sm
(O CO ~
Pu
62
94
э == гс С
*О_^ С О
£ 1иЛIft
f5
> > m ? 0C5 N 0 8 2 l S я с йз .
« с г Ф3 о с
СЯФ
< CD f IЯ“i >s
Pm
О
Np
93
61
s i| £NflC
, 2 °<o “o £OS
s I I ? 2- ° о.-ё
>> А ел CO — ^ со U.я»
U]
О фD фл
и! lit*
iiets-if
9
J
iSffl
Nd
60
ss l-S ? 4 e i f
in 8-g?b .<§
И « |||м 1 ||1 |яЫ«
<o ® 2 с ф .5 -g
00 ^ о 2 о —ф о
uf) >< >CD T- JO! 1П и " я ф с i sq 2
Pa
59
со в S
91
Pr
H о . Я1я
8 CO 2 0 Я
Ф . с 14 ^
^ г-$ •=;Ф
i ° <о С сл$ о о
r о Ф § 8 в !
< CD CN
NJJJb-
—I
Ce
58
90
Th
'T >> C ? й | i°0C

o> !«S % £.sg
£ Э О - ф « Ш
* J f ^ Си
C
0
O CTgjE 5
08 L 8 j£ y§®Sh * |
La
Ac
57
I 89
o S | г: .У oi : <£ > . o> |

™CO .» I« ^ 00 Ir- !Е
•е . £ з g'CD Ь
4 ! со j in 100 S ДО
■_ О- со ° Я I» о
5 хГ 8 £® 5« ф£ -S
ф
m 00 CNJ —CNJ | _ in CNJ | £ h- i5 сл я Е со
0» - cvi 0)0 о « - W — CO CD та COI00 «# О I с сл >s ® ^
^ m s o l *CO ( / ) ^ »n flQ COI со DC C cb ; S С Я® Ф ® ^
*- CN
NJJ ф 2с .£
Ф i ТI ®
ф о
§ 1
00 00 о ш ® Ч ~ ^ :д ®
я S) 05 o> ° 1h- ^ < Оe *щ ®'°
J Z C7 10*1» ф ф ё 5 ЯС |Ф
ф О
cn
<° i ^ CNJ
C\J
* s
CD
CO
to a: us
00
« O wmCOo Io l i - nrvi sQ S.
Ф w• »
—W® ” £
_J_ Z ^ S wT со
Q iX F О R D
I H E M I S T R~Y
M E RS =
SERIES ED ITO R S
STEPHEN G. DAVIES This series of short texts provides accessible accounts of a range of essential topics
in chemistry and chemical engineering. W ritten with the needs of the student in
RICHARD G. COMPTON
mind, the Oxford Chemistry Primers offer just the right level of detail for
JOHN EVANS undergraduate study, and will be invaluable as a source of material commonly
presented in lecture courses yet not adequately covered in existing texts. All the
LYNN F. GLADDEN
basic principles and facts in a particular area are presented in a clear and
straightforward style, to produce concise yet comprehensive accounts of topics
covered in both core and specialist courses.
This book covers the material required for a basic understanding of chemical
reaction engineering. Such material would normally be taught in a first chemical
reaction engineering course in a university chemical engineering department. The
principals of reaction engineering are simply and clearly presented; simple
illustrative problems are used to demonstrate how these principles are practically
applied. Further problems, with solutions, based on exam questions are supplied.
The book is written in a way that it could be used as a self-study guide and would
be useful for undergraduate chemical engineers early in their degree as well as
engineers and scientists of other disciplines interested in acquiring some knowledge
of reaction engineering outside of a formal teaching environment
Ian S. Metcalfe is Professor of Chemical Engineering, University of Edinburgh.

Chemical Reaction Engineering: A First Course

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chemical Reaction Engineering: A First Course

Uploaded by

Copyright:

Available Formats

ENGINEERING

Great Clarendon Street, Oxford 0 X 2 6DP

Oxford is a registered trade m ark o f Oxford University Press

© Ian S. Metcalfe, 1997, reprinted 2000, 2003, 2005, 2006

The m oral rights o f th e author have been asserted

All rights reserved. No p art o f this publication may be reproduced,

1.2 Starting knowledge required

1.3 The rate expression

Reaction rate is usually expressed in terms of the concentrations or partial

where va and vb are stoichiometric coefficients. The rate of disappearance of

ъ. г "Aс "в — t , f ncr Dn°

(a) the difference between batch, semi-batch, and continuous modes of

‘Constitutes a basic understanding of the material covered.

2.1 Batch reactors

3. The temperature will also be uniform throughout the reactor—however,

A first-order irreversible reaction,

From eqn 2.4,

{=_ f N- dNA = _ l ]n Na_

2.2 Plug flow reactors (PFRs)

• Homogeneous reactions (tube contains only the reactant and product

In tubular reactors, there is a steady movement of reagents in a chosen

In continuous operation at steady state in a PFR, the composition changes with Vr

A cross-sectional area (m2)

Dividing by the cross-sectional area and remembering that Adl = dV,

reactor length or volume is increased. Whereas in the case of a batch reactor

Example 2.2 A first-order irreversible reaction,

Solution From eqn 2.10,

2.3 Continuous stirred tank reactors (CSTRs)

reaction rate is the same at every point, and time independent.

The design equation for a CSTR is not a differential equation as it is in the

Example 2.3 A first-order irreversible reaction,

Solution From eqn 2.15,

Mean residence time, г = V/vj = 42.9 s

We will now apply the material balances or design equations derived in

3.1 Residence time of batch reactors

In general, for PFR and batch reactors, where we need to perform an

Differentiating eqn 2.5 gives eqn 2.6,

This can be substituted into eqn 2.7,

3.1.1 Constant volume batch reactor

and this expression can be easily integrated.

t= - J ln(l - xA) = \ In — - — - and (3.2)

3.1.2 Constant pressure batch reactor

For a constant pressure batch reactor, the reactor volume is no longer

% = л »го+ ( ',в + ^ ~ ‘|> ) ^ а

Example 3.1 Consider the gas-phase reaction,

Solution Na = Nao —NaoXa, = Nm + -N ao^a

Л'т | , 3 A^ao ,,3

PV = NTRT\ PVо = Л^то/гг

Putting everything in terms of one common variable, xA, by using

«С - ПСо Ч------ ПА0Х А

the reactor contains a catalyst.) Therefore, the volumetric flow rate is

v i = i>to (1 + £a*a) (3.8) residence tim e will be?

18 Calculation o f reactor volume and residence time

f AdnA f AdnA f *AnAOdxA

■-here I is the length variable and L is the total reactor length.

for the limits of integration remember that, formally, when I — L, xA = xAe)

ЗАР nA0 *Ае 3(0.4) (1.2) (0.9)

For a gas-phase reaction we use eqn 3.8,

The volume of the CSTR is given by, (a) outlet inlet

(a) PFR Solution

Сдо — CAl Vr Cai — Сде

Total volume = Vi + V2 + 4.32 m3

Fig. 3.6 Plot of inverse of reaction rate

Xa = 1 ---- — = 1 —(1 —XAiR i ) ( l —A,R2) • • • (1 ~ A,Rn) or,

a = '2 e x p ( - ^ ) N A 2 = r ~ r (.NA0-NA) (4.22)

N r = N r0 + j —^ T j^ N A °[l - exP (—(i + 2>}]

3 = 2.1 x 102 exp^ ) m°l s~ т - 3Ьаг-1

rA = - Л-° ~ ПЛ = (i + 2) — (4.25)