Chapter 8

The Batch and

Semibatch Reactors

Introduction
8.1 The Isothermal Batch Reactor
Example 8.1.A Example of Derivation of a Kinetic Equation from Batch
Data
Example 8.1.B Styrene Polymerization in a Batch Reactor
Example 8.1.C Production of Gluconic Acid by Aerobic Fermentation of
Glucose
8.2 The Nonisothermal Batch Reactor
Example 8.2.A Decomposition of Acetylated Castor Oil Ester
8.3 Semibatch Reactor Modeling
Example 8.3.A Simulation of Semibatch Reactor Operation (with L.H.
Hosten†)
8.4 Optimal Operation Policies and Control Strategies
8.4.1 Optimal Batch Operation Time
Example 8.4.1.A Optimum Conversion and Maximum Profit for a
First-Order Reaction
8.4.2 Optimal Temperature Policies
Example 8.4.2.A Optimal Temperature Trajectories for First-
Order Reversible Reactions
Example 8.4.2.B Optimum Temperature Policies for Consecutive
and Parallel Reactions
384 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

INTRODUCTION
Batch reactors are generally used for liquid phase reactions. When a solid
catalyst has to be kept in suspension or when there are two liquid phases, as in
the nitration of aromatics, for instance, an agitator is required. Agitation is also
necessary to equalize the temperature in the reactor and to keep it at the desired
level by efficient heat exchange through a jacket or an internal coil.
Consequently, the batch reactor is generally considered to be spatially uniform in
composition and temperature. The composition changes with time, however.
Temperature sequencing may be favorable for the selectivity or for achieving
complete conversion in a safe way.
In pure batch operation the reactants are completely fed into the reactor
at the beginning. For better control of temperature this type of operation may not

Figure 8-1
Batch reactor for the production of antibiotics by fermentation. Reprinted from S. Aiba,
A.E. Humphrey and N.F. Milis, ”Biochemical Engineering”, 2nd Ed., University of
Tokyo Press, Tokyo, 1973 and from J.E. Bailey and D.F. Ollis, ”Biochemical
Engineering Fundamentals”, McGraw Hill, New York, 1986.
 8.1 THE ISOTHERMAL BATCH REACTOR 385

be advisable and the reactant(s) may have to be added progressively to the

contents of the vessel. The reactor is then said to operate in the semibatch mode.
Or, a product may be withdrawn, for instance, water in an esterification subject
to equilibrium, to reach complete conversion of the reactant(s).
Batch and semibatch reactors are most often used for low production
capacities, where the cost of labor and dead time are only a small fraction of the
unit cost of the product. They are generally encountered in the area of specialty
chemicals and polymers and in pharmaceuticals, in particular, in plants with a
wide variety of products. Large batch reactors are used in fermentations for
antibiotic production, an example of which is given in Fig. 8-1.

8.1 THE ISOTHERMAL BATCH REACTOR

Because of the uniformity of concentration, the continuity equation for the key
reactant A may be written for the entire reactor volume:

= VrA (C A )
dN A
− (8.1-1)
dθ
where θ = residence time in the reactor. It is convenient to specifically represent
this residence time in the reactor by a special symbol — for completely batch
systems it is the same as “clock” time t, but in other applications the distinction
will be useful. For a general set of reactions, (8.1-1) can be extended to

dN j R
= V ∑ α ij ri ≡ VR j (8.1-2)
dθ i =1

These mass balances, for example (8.1-1), are often written in terms of
conversions:
dx A V
= rA (8.1-3)
dθ N A0

Then, (8.1-3) is readily put into integral form:

x Af
dxA
θ = N A0 ∫ (8.1-4)
x A0
VrA

Note that the batch residence time θ can be interpreted as the area from
xA0 to xAf under the curve of NA0 / V rA(xA) versus xA.
386 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

The above equations were already encountered in Section 1.1.2 of Chapter 1. The
equations defining reaction rates were written there for a “point”, a volume in
which the composition is uniform. This is also assumed here for the ideal, well
stirred batch reactor, whatever its volume.
The volume of the reaction mixture can change for two reasons: (1)
external means (e.g., filling a reaction vessel or adding a second reactant) and (2)
changes in densities of reactants or products (e.g., molar expansion of gases).
The first possibility is often termed “semibatch” operation. The second is usually
not very important for liquids, and is neglected. The proper formulation for gases
will be derived below, although batch gas-phase reactors are not commonly used
in industry because of the small mass capacity. A gas phase could be part of the
reaction mixture, however. Also laboratory gas-phase reactors have been utilized.
With no expansion, as for liquids, (8.1-4) becomes
x Af
dx A
θ = C A0 ∫ r (x ) (8.1-5a)
x A0 A A

C Af
dC A
=− ∫ r (C )
C A0 A A
(8.1-5b)

and for simple rate forms can be easily integrated analytically, as illustrated in
Section 1.3.
For reactions with the stoichiometry

aA + bB ... qQ + sS + ...

the following mole balance can be made at a given extent of reaction based on
conversion A:
NA = N A0 − N A 0 x A
b
N B = N B 0 − N A0 x A
a
q
N Q = N Q 0 + N A0 x A
a
s
N S = N S 0 + N A0 x A
a
NI = NI0 (inert)
_______________________________
q + s ... − a − b ...
N t = N t 0 + N A0 xA
a
 8.1 THE ISOTHERMAL BATCH REACTOR 387

Therefore, the total number of moles is given by

N t = N t 0 + N A 0δ A x A

from which
Nt
= 1 + ( y A 0 δ A )x A ≡ 1 + ε A x A (8.1-6)
Nt0

For gases and using the equation of state:

pt V = ZN t RT

V  Z T pt 0  Nt
= 
V0  Z 0 T0 pt  Nt0

 Z T pt 0 
=  (1 + ε A x A ) (8.1-7)
 Z 0 T0 pt 

The concentrations to be substituted into the rate equation, can be expressed as

N A N A0 (1 − x A )  Z 0 T0 pt 
CA = =   (8.1-8)
V V0 (1 + ε A x A )  Z T p t 0 

and the partial pressures,

p A = pt
NA N
= pt A0
(1 − x A ) (8.1-9)
Nt N t 0 (1 + ε A x A )

= p A0
(1 − x A )  pt 
 
(1 + ε A x A )  pt 0 
For an nth-order reaction e.g.,

N An 0 (1 − x A )
n

VrA = VkC An = k (const. T, pt)

V0n −1 (1 + ε A x A )n −1
and (8.1-4) becomes
(1 + ε A xA ) n −1
x Af
1
θ = n −1 ∫ dx A (const. T, pt)
kC A0 x A 0 (1 − x A )n

which in the absence of molar expansion, εA = 0, is the same as (8.1-5a), of

course.
388 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

EXAMPLE 8.1.A
EXAMPLE OF DERIVATION OF A KINETIC EQUATION FROM BATCH
DATA
The reaction A + B → Q + S is carried out in the liquid phase at constant
temperature. It is believed the reaction is elementary and, since it is bimolecular,
it is natural to first try second-order kinetics. The density may be considered
constant.
Let B be the component with highest concentration, while the most
convenient way to follow reaction is by titration of A. A batch-type experiment
led to the data given in Table 8.1.A-1.
If the hypothesis of second order is correct, the following relation
between the rate and the concentration of A and B will be valid for any time, and
therefore for any composition:

rA = rB = kC A C B (8.1.A-a)

and k has to have the same value for all levels of CA and CB. When the
differential method is used, equation (8.1.A-a) is the starting point. By
substituting the rate equation (8.1.A-a) in the material balances, (8.1-1) with CA =
NA / V,
dC A
rA = − = kC A C B (8.1.A-b)
dθ
This means rA may be obtained as tangent to the curve CA versus θ.
Substituting the corresponding CA and CB leads to k. The values of CB
follow from CB = CB0 – (CA0 – CA). Table 8.1.A-2 gives the values of k obtained
in this way. The variation of k is small and does not invalidate the second-order
hypothesis, especially as the precision of the method is getting smaller as the
reaction proceeds. A value of 61 × 10-2 m3/h kmol may be used for k.

TABLE 8.1.A-1
CONCENTRATION-VERSUS-TIME DATA

C B0 = 0.585 kmol/m 3
C A0 = 0.307 kmol/m 3
Time (h) CA (kmol/m3)
0 0.307
1.15 0.211
2.90 0.130
5.35 0.073
8.70 0.038
 8.1 THE ISOTHERMAL BATCH REACTOR 389

TABLE 8.1.A-2
COMPARISON OF k DETERMINED BY INTEGRAL AND DIFFERENTIAL
METHOD

k (m3/kmol h)
Time CA CB rA From (b) From (e)
(h) (kmol/m3) (kmol/m3 h)

1.15 0.211 0.489 6.52 × 10-2 63.1 × 10-2 61.2 × 10-2

2.90 0.130 0.408 3.32 × 10-2 62.5 × 10-2 61.9 × 10-2
5.35 0.073 0.351 1.49 × 10-2 58.4 × 10-2 62.0 × 10-2
8.70 0.038 0.316 0.72 × 10-2 59.6 × 10-2 60.8 × 10-2

The integral method is based on (8.1-5a). Before integration is possible,

CA and CB must be expressed as a function of one variable, the fractional
conversion xA. In this case
C A = C A0 (1 − x A )

 x 
C B = C B 0 1 − A 
 M

CB0
where M =
C A0

Equation (8.1-5a) becomes

xA
C A0 dx A
θ= ∫ (1 − x )(1 − x (8.1.A-c)
kC A0C B 0 0 A A /M)
or
M M (1 − x A )
C B 0 kθ = ln (8.1.A-d)
1− M M − xA
or, with concentrations,
1 C C
kθ = ln B 0 A (8.1.A-e)
C A0 − C B 0 C A0 C B

These equations also lead to a constant value for k, which confirms that the
reaction has second-order kinetics. Peterson [1962] has discussed further aspects
of differential versus integral fitting of data from batch reactor experiments.
▄
390 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

EXAMPLE 8.1.B
STYRENE POLYMERIZATION IN A BATCH REACTOR
The kinetics of free radical polymerization and the molecular weight distribution
of the polymer were already discussed in Section 1.6.2 of Chapter 1. To improve
the chemical and mechanical properties of the polymer great efforts were
undertaken a number of years ago to achieve narrow distributions. This is
possible with anionic or cationic — so-called living — polymerization, in which
chains can not terminate or transfer and grow at a rather uniform rate, thus
yielding a polymer with a polydispersity close to one. This type of
polymerization requires very special operating conditions and high purity of the
feed, however.
More recently it has become possible to achieve this goal also in free
radical polymerization by stabilizing propagating macroradical chains and make
them “dormant”. This has been called “living” free radical polymerization
(LFRP). It involves reducing the concentrations of the active chains, thus making
bimolecular terminations almost negligible. One way of doing this is by
scavenging the active chain by means of a nitroxide chemical that will simply be
called here TEMPO:
Rn + TEMPO Rn − T

The process is called NMLP (from nitroxide mediated “living” polymerization).

A second possibility is to regulate the equilibrium between active and dormant
chains by means of an exchange or redox reaction, involving a metal ion like Cu,
Mn, Fe and a halogen like Br:

Rn + Br − Cu II Rn − Br + Cu I

This is called “atom transfer radical polymerization” (ATRP).

In the following developments Rn-T and Rn-Br, the dormant chains and their
concentrations, will be represented by Dn.
Butté et al. [1999] applied a comprehensive mechanistic scheme
covering both NMLP and ATRP to the simulation of the polymerization of
styrene. The scheme is given in Table 8.1.B-1, together with the rate equations.
The nomenclature in Table 8.1.B-1 is completely in line with that of Section
1.6.2. The symbols represent both the species and their concentration. The
initiation by thermal decomposition of the monomer, not included in the scheme
of Section 1.6.2, was observed by Hui and Hamielec [1972].
 8.1 THE ISOTHERMAL BATCH REACTOR 391

TABLE 8.1.B-1
REACTION SCHEME AND RATE EQUATIONS FOR LFRP
(BOTH NMLP AND ATRP)
___________________________________________________
REACTION STEPS RATE EQUATIONS
Initiation
by monomer decomposition
3M 1 → 2 R0 r0 = k it M 13
by initiator decomposition
I → 2 R0 rid = k id I
Propagation
Rn + M 1 → Rn +1 r p = k p M 1λ0
Exchange steps
Scavenging
in NMLP Rn + T → Dn rs1 = k s1Tλ0
in ATRP Rn + X → Dn + Y rs 2 = k s 2 Xλ0
Reactivation
in NMLP Dn → R n ra1 = k a1 µ 0
in ATRP Dn + Y → R n + X ra 2 = k a 2Yµ 0

Termination
by combination
Rn + Rm → Pn+ m rtc = k tc λ 20
Transfer
with monomer
Rn + M 1 → R0 + Pn r fm = k fm M 1λ0
Dormant species decomposition
in NMLP Dn → Pn rdc1 = k dc1 µ 0
in ATRP Dn + X → Pn + 2 rdc2 = k dc2 Xµ 0
___________________________________________________

All the rate coefficients are assumed to be independent of the chain

length. The rates of the various steps are directly related to the moments of the
polymeric species, taken to be sufficiently representative of the complete
distribution of the species. λ 0 , µ 0 , and η 0 are the zero-th moments of the active,
dormant and dead chain distributions and also their total concentration. The
moments λ1 , µ1 , and η1 represent the total number of monomer units in the
active, dormant and dead chains, i.e., their degree of polymerization. The
392 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

moments λ 2 , µ 2 , and η 2 can be related to the variance around the mean of the
distributions and provide information on their width. λ1 / λ 0 , µ1 / µ 0 , and η1 / η 0
are the number average chain lengths of the various chains.
For the ATRP type of LFRP the continuity equations for the various
species can be written:
Initiator:
dI
− = rid
dt
Monomer:
dM 1
− = (3rit + rp + r fm )
dt
For X, Y and Z:
dX
− = rs 2 − ra 2 + rdc2
dt
dY
− = rs 2 − ra 2
dt
dZ
= rdc2
dt
Active chains:
These are radicals, for which pseudo-steady state may be assumed, so
that
dλ 0
= 0 = 2(r0 + rid ) − rs2 + ra2 − rtc
dt
dλ1 µ λ
dt µ0
( λ0
)
= 0 = ra 2 1 − ra 2 + rtc + rfm 1 + rp

dλ2 µ λ λ 
dt µ0
( λ0
) 
= 0 = ra 2 2 − rs 2 + rtc + rfm 2 + rp 1 + 2 1 
λ0 


Dormant chains:

dµ i λ µ
dt λ0
(
= rs2 i − ra2 + rdc i
µ0
) i = 0,1, 2

Dead chains (macromolecules):

dη 0 1
= rdc + r fm + rtc
dt 2
 8.1 THE ISOTHERMAL BATCH REACTOR 393

d η1 µ λ
= rdc 1 + (r fm + rtc ) 1
dt µ0 λ0
2
dη 2 µ λ λ 
= rdc 2 + (r fm + rtc ) 2 + rtc  1 
dt µ0 λ0  λ0 

The model consists of 11 ordinary differential equations and three algebraic

equations (the pseudo steady state equations for λ 0 , λ1 , λ 2 ). The unknowns are
the concentrations of M, X, Y, S, I2 and the moments λi , µ i ,η i (i = 0, 1, 2).
Butté et al. [1999] carried out experiments for both types of free radical
polymerization of styrene and simulated the results using kinetic parameters
obtained from the literature after adaptation accounting for their own
observations.
Only the results for ATRP are shown here. Figure 8.1.B-1 deals with the
polymerization of styrene at 110 °C with 1-phenlylethylbromide as an initiator
and addition of a Cu-complex. The model prediction and experimental results are
in reasonable agreement. The number average molecular weight increases with
the conversion of the monomer and the polydispersity Pw/Pn drops to a value
close to one from a relativity low conversion onwards. The evolution of the Cu-
concentration is also shown.
Also in LFRP of styrene, Matyaszewski et al. [1997] obtained values of
Pw/Pn of 1.1.

Figure 8.1.B-1
Atom transfer radical polymerization of styrene at 110°C [Butté et al., 1999].
▄
394 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

EXAMPLE 8.1.C
PRODUCTION OF GLUCONIC ACID BY AEROBIC FERMENTATION OF
GLUCOSE
Gluconic acid is produced on an industrial scale by aerobic fermentation of
glucose with the enzymes Aspergillus niger and catalase. The stoichiometry is
simple: glucose + ½ O2 yields gluconic acid. The fermentation is carried out at
30°C in a batch reactor and consists of three stages with different loadings of
concentrated solutions of the substrate, glucose. The charge also contains corn
steep liquor, MgSO4, K- and ammonium phosphate, urea, anti-foam and the
enzymes. The rate depends on the pH, controlled by the addition of NaOH,
except in the third stage, so that the product is really a mixture of gluconic acid
and Na-gluconate.
Reuss et al. [1986] modeled this process on the basis of a careful
experimental program. The kinetics of gluconic acid production were described
by means of the Michaelis-Menten equation (Section 1.5.1 of Chapter 1). At a
pH of 5.6 the constant KM in that equation amounted to 0.36 mmol/h and the
specific oxygen uptake rate (half of the specific production rate, γP) 65 mol
oxygen per kg dry weight and per hour.
The production rate is intrinsically very fast but limited by the oxygen
uptake of the liquid. The volumetric oxygen transfer coefficient kLav was also
affected by the increasing viscosity of the medium as the amount of biomass
increased.
Using the definitions and symbols of microbial kinetics the model
equations can be written:

dX
= µX
dt
dP
= γPX
dt
dS µX γ P X
− = + + mS X
dt yχ / S y P / S

dCL µX γ P X
= k L av (CL* − CL ) − − − mO X
dt yχ / O yP / O

X represents the biomass, P the product, and S the glucose (or substrate)
concentration (kg/m3). For a clear insight into the first equation in particular it is
useful to remember that µ is the specific biomass growth rate in h-1 (see Section
 8.1 THE ISOTHERMAL BATCH REACTOR 395

1.5.2). γP is the specific production rate of P in kg P per hour and kg biomass; yχ/S
is the biomass weight selectivity (kg biomasss produced per kg glucose
converted); yP/S is the product wt selectivity (kg gluconic acid/kg glucose
converted); mS is the glucose consumption per hour and referred to 1 kg dry
biomass. In the fermentation jargon it is called ”maintenance coefficient for
glucose” (0.048 kg glucose/kg dry weight·h). The last equation describes the
evolution with time of CL, the concentration of dissolved oxygen, with saturation
concentration C*L. The weight of biomass produced per mol oxygen converted is
represented by yχ/O and yP/O is the weight of P produced per mol of oxygen
converted. mO is the oxygen consumption per hour, again referred to 1 kg of dry
biomass and called “maintenance coefficient for oxygen” (1.61 mol O2/kg dry
weight·h).
Figure 8.1.C-1 shows the computed and experimental time histories for
biomass, glucose and product.
Reuss et al. [1986] applied this model to investigate optimal operating
policies, e.g., controlled continuous feeding of the glucose, or operating at a
higher pressure so as to increase the oxygen concentration and enhance the
global rate.

Figure 8.1.C-1
Batch fermentation of glucose into gluconic acid [Reuss et al., 1986].

▄
396 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

8.2 THE NONISOTHERMAL BATCH REACTOR

In practice it is not always possible, or even desirable, to carry out a reaction
under isothermal conditions. In this situation both the energy and mass balances
must be solved simultaneously:

dx A
rA ( x A , T )
V
= (8.1-3)
dθ N A0

= V (− ∆H )rA ( x A , T ) + qAk
dT
mt c p (8.2-1)
dθ
where (8.2-1) is the appropriate simplified heat balance and Ak is the heat
exchange surface from Section 7.2.4. The term qAk represents any addition or
removal of heat from the reactor. For adiabatic operation, q = 0, while for a heat
exchange coil it would have the form

q = U (Tr − T ) (8.2-2)

where Tr = temperature of heating or cooling medium. Equation (8.2-1) can be

combined with (8.1-3) to yield

− (− ∆H )N A0 A = qAk = 0
dT dx
mt c p adiabatic (8.2-3)
dθ dθ
Thus,
θ
mt c p (T − T0 ) − (−∆H ) N A0 ( x A − x A0 ) = ∫ qAk dθ (8.2-4a)
0

= qAkθ , q = const. (8.2-4b)

= 0, adiabatic (8.2-4c)
For the latter situation, the adiabatic temperature change, for a certain conversion
level, is
T − T0 =
(− ∆H )N A0 ( x − x A0 ) =
1
( x A − x A0 ) (8.2-5)
mt c p
A
λ

Therefore, in this case T can be substituted from (8.2-5) into (8.1-3), which then
becomes a single differential equation in xA [or xA can be substituted into (8.2-1)].
This is done by utilizing (8.1-4), where the integral is evaluated by choosing
increments of xA and the corresponding T(xA) from (8.2-5). Again, the reaction
 8.2 THE NONISOTHERMAL BATCH REACTOR 397

time θ can be represented by the area under the curve NA0 / V rA(xA, T(xA)) versus
xA.
Some analytical solutions are possible for simple-order rate forms —
they are given for the analogous situation for plug flow reactors in Chapter 9.
Finally, the maximum adiabatic temperature change occurs for xA = 1.0,
and then (for xA0 = 0):
(− ∆H )N A0
(∆T )ad = Tad − T0 = (8.2-6)
mt c p

Equation (8.2-4c) can be written in the alternate form

T = T0 + (∆T )ad x A (8.2-7)

More general situations require numerical solutions of the combined mass and
heat balances. Several situations can occur:
The temperature is constant or a prescribed function of time, T(θ) —
here, the differential equation (8.1-3) can be solved alone as

rA ( x A , T (θ ))
dx A V
=
dθ N A0

Also, (8.2-1) or (8.2-3) can then be solved to find the heating requirements:

dT (θ )
q(θ ) Ak = mt c p − (− ∆H )V rA ( x A (θ ), T (θ ))
dθ
Heat exchange is zero, constant, or a prescribed function of time. First,
(8.2-4) is used to compute T = T(xA, θ) and then substituted into the continuity
equation for A, (8.1-3), which can then be integrated:

rA ( x A , T ( x A ,θ ))
dx A V
=
dθ N A0

The temperature variation can be found by using the computed values of xA(θ):

T (θ ) = T ( x A (θ ),θ )

Alternatively, the combined equations (8.1-3) and (8.2-1) can be solved

simultaneously as coupled differential equations.
Heat exchange is given by q = U(Tr – T). Direct numerical solution of the
coupled mass and heat balances is performed.
If convergence problems arise in the numerical solutions, it is often
useful to use conversion as the independent variable. The mass balance links
398 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

increments of conversion to increments of time and through the heat balance

these yield increments of temperature. Iterations on the evaluations of the rates
are also often required.
For case 3 above, values of the heat transfer coefficient are required. The
factor U, appearing in (8.2-2), is a heat transfer coefficient, defined as follows:

1 1 d Ak 1 Ak
= + + ⋅ (8.2-8)
U α k λ Am α r Ar
where:

αk, Ak = heat transfer coefficient (kJ/m2 h °C) and heat transfer surface
(m2) on the side of the reaction mixture, respectively
αr, Ar = the same, but on the side of the heat transfer medium
Am = logarithmic mean of Ak and Ar
λ = conductivity of the wall through which heat is
transferred (kJ/m h °C)
d = wall thickness (m)

The literature data concerning αk and αr are not always in accordance. As a guide
the following relations are given.
For reactors in which heat is transferred through a wall, αk may be
obtained from the following dimensionless equation for stirred vessels:

 cpµ 
0.66 0.33
α k dr  µw   d 2 Nρ L
0.14

  = 0.36 s 
   (8.2-9)
λ  µ   µ   λ 
where:

dr = reactor diameter (m)

ds = propeller diameter (m)
µw = viscosity of the reaction mixture at the temperature of the wall
(kg/m h)
µ = viscosity of the reaction mixture at the temperature of the
reaction mixture
λ = thermal conductivity of the reaction mixture (kJ/m h °C)
N = revolutions per hour (h-1)
ρL = density of reaction mixture (kg/m3)

[From Chilton, Drew, and Jebens, 1944].

More extensive work by Chapman, Dallenbach, and Holland [1964] on a
batch reactor with baffles and taking into account the liquid height HL and the
propeller position Hs led to the following equation:
 8.2 THE NONISOTHERMAL BATCH REACTOR 399

−0.56
 cpµ 
0.65 0.33 0.4
αk dr  µw   d 2 Nρ L
0.24
  Hs   HL 
  = 1.15 s 
       (8.2-10)
λ  µ   µ   λ   ds   ds 

Further work on this subject has been done by Strek [1963].

For αr several cases are possible. When the reaction vessel is heated
(e.g., with steam), the Nusselt equation may be applied, provided film
condensation is prevailing.
For heat transfer through a coil, αk may be calculated from an equation
such as (8.2-9), but with a larger coefficient, due to the effect of the coil on the
turbulence. According to Chilton et al. [1944], this coefficient would be 0.87. It
is likely to depend also on the mixing intensity. Other literature sources also
mention a value of 1.01.
Values of αr may be obtained from the following equation, valid for
turbulent conditions:
 c p µs
0.4
α r dt
0.8
d G 
= 0.023 t    Φ (8.2-11)
λ  µ   λ 
where:

Φ = 1 + 3.5(dt / dc)
dt = inner diameter of the pipe (m)
dc = coil diameter (m)
µs = viscosity of the heat transfer medium at the surface of the coil
(kg/m h)

Equation (8.2-11) is an adaptation of the classical Dittus and Boelter equation for
straight pipes. Further information on this topic can be found in Holland and
Chapman [1966]. Various aspects of agitator selection are dealt with by Gates et
al. [1975].

EXAMPLE 8.2.A
DECOMPOSITION OF ACETYLATED CASTOR OIL ESTER
This example has been adapted from Smith [1970] and Cooper and Jeffreys
[1971]. The overall reaction for the manufacture of drying oil is

 acetylated 
 castor oil (l ) →  drying (l ) + CH COOH( g )
 ester   oil 
3
 

The charge of oil to the batch reactor is 227 kg and has a composition such that
complete hydrolysis gives 0.156 kg acid/kg ester. The initial temperature is T0 =
400 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

613 K. The physiochemical properties are as follows: cp = 2.51 kJ/kg K,

molecular weight of the acid = 60 kg/kmol, (–∆H) = –62.8 × 103 kJ/kmol. The
rate of acid production is first order with respect to the ester concentration,
expressed in kmol/m3 [Grummitt and Fleming, 1945].

 1   22450 
rA =  C A exp 35.2 −  kg acid/m3 s (8.2.A-a)
 60   T 

A constant heat supply is provided by an electrical heater and a final conversion

of 70 percent is desired.
This is an example of case 2 discussed above, and so (8.2-4) is utilized.
First, the adiabatic situation is computed using (8.2-5). The adiabatic curve is
linear in conversion and has a slope which is the adiabatic temperature change

Figure 8.2.A-1
Temperature-conversion progress for various rates of heat input. From Cooper and
Jeffreys [1971].
 8.2 THE NONISOTHERMAL BATCH REACTOR 401

from (8.2-6):

(∆T ) = (− 62.8 × 10 kJ/kmol )(0.156 kg/kg )(227 kg )

3
(8.2.A-b)
ad
(227 kg )(2.51 kJ/kg K )(60 kg/kmol )
= – 65 K

Thus,
T = 613 − 65 x A K (8.2.A-c)

as shown in Fig. 8.2.A-1. For this endothermic reaction the temperature drops
drastically with adiabatic operation, and heating needs to be considered.
Temperature-conversion curves for various heat inputs were calculated
by Cooper and Jeffreys, using (8.2-4b) to obtain T = T(xA, θ):

qAk
T = 613 − 65 x A + θ (8.2.A-d)
mt c p
For qAk = 52.8 kW,

qAk 52.8 kW
= = 0.0927 K/s
mt c p (227 kg )(2.51 kJ/kg K )

Substitution into (8.1-3) leads to:

 22450 
C A0 (1 − x A )exp 35.2 −
dx A 1
=  (8.2.A-e)
dθ 60C A0  T 

(1 − x A )  22450 
= exp 35.2 −  (8.2.A-f)
60  613 − 65 x A + 0.0927θ 

Fig. 8.2.A-1 shows the evolution of temperature with conversion for

various heat inputs. For a heat input of 105.6 kW, e.g., the temperature drops
until a conversion of 40 percent is reached.
It is also instructive to look at the conversion-time profile (Fig. 8.2.A-2).
For qAk = 52.8 kW the first half of the final conversion of 70 % is reached in
about 2.5 min, while the second half takes 5 min. This is even more pronounced
for the adiabatic case with its rapid temperature drop. The total reaction time
required for 70 percent conversion depends upon the heat input rate, as shown in
the table below.
One could also select the proper heater size to achieve 70 % conversion
in a given time — say, 20 min. In such a case equations (8.2.A-d) and (8.2.A-e)
402 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

Figure 8.2.A-2
Conversion versus time curve for adiabatic operation and a heat input rate of 52.8 kW.

Heat input rate (kW) θf for 70% conversion (min)

(Adiabatic, q = 0) (38.25)
5.28 23.64
52.8 7.48
105.6 4.72
158.9 3.55

would have to be solved iteratively for the unknown value of q. Alternatively, the
simulation results presented here yield by means of a simple interpolation qAk = 8
kW. ▄

8.3 SEMIBATCH REACTOR MODELING

When addition of reactants, together with removal of liquid and vapor, is also
considered, the equations become:
Continuity equation for species j:

d (C jV ) N
= F1'C j , 0 − F2'C j − F3', vC j ,v − V ∑ α ij ri (8.3-1)
dt i =1

in which F1' is the liquid feed rate and F2' is the liquid withdrawal rate (both in
m3/h), and F3' ,v is the vapor withdrawal rate (m3 vapor/h).
Energy equation:
N
= F1'c p ρT1 − F2'c p ρT − F3', v ∑ C j ,v ∆H v + Q + V ∑ ri (− ∆H i ) (8.3-2)
dT
ρc pV
dt i =1
 8.3 SEMIBATCH REACTOR MODELING 403

Accounting for the change in volume:

dV
= F1' − F2' − F3',1
dt
with
273 1 1
F3',1 = F3',v pt (8.3-3)
273 + T 22.4 C t ,1

where T is in degrees C and Ct,1 is the total liquid concentration. Equation (8.3-3)
has to be substituted into (8.3-1) and (8.3-2).

EXAMPLE 8.3.A
SIMULATION OF SEMIBATCH REACTOR OPERATION (WITH L.H. HOSTEN†)
Consider the following set of reactions:

with rate equations

for A D:

 C 
r1 = k1  C A − D  with (–∆H1) = 62805 kJ/kmol
 K1 

for A B+C :

 C C 
r2 = k 2  C A − B C  with (–∆H2) = –83740 kJ/kmol
 K2 
for C → E :

r3 = k 3C C with (–∆H3) = (–∆H2).

The rate coefficients are given by

 5033  -1
k1 = k3 = 10 exp 23.25 −  h
 T 
404 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

 5033  -1
k2 = 60 exp 23.25 −  h
 T 

and the equilibrium constants by

K1 = 0.5; K 2 = 2.92 kmol/m3

The reactor is a cylindrical vessel with a diameter of 0.8 m and is

operated in a semibatch mode. The initial load is 0.1826 m3 of pure A at a
concentration of 17.48 kmol/m3 and an initial temperature of 25°C. The copper
coil for heat exchange has an internal diameter of 0.015m, an external diameter
of 0.019 m, a length of 14 m, and an exchange surface of 0.748 m2. The thermal
conductivity of the copper is 1.402 kJ/m h K. The diameter of the coil is 0.35 m.
The paddle agitator has a diameter of 0.3 m and revolves at 150 rpm.
A mixture of A and D, containing respectively 7.1 and 31.89 kmol/m3, is
fed at 25°C during 3 hours at a rate of 0.1068 m3/h. Then, the reaction is
continued for another 3 hours. There is no liquid withdrawal. The operating
pressure is 1 bar.
Integration of the set of equations (8.3-1) to (8.3-3), for example by
means of a Runge-Kutta-Gill routine, yields concentrations, temperature, volume
of the reactor contents, and vapor flow rate as functions of time. Determining the
vapor flow rate requires an additional equation, expressing that, when the reactor
content is boiling, the sum of the partial pressures must equal the total pressure
above the liquid. When the liquid behaves in an ideal way, the vapor pressures
satisfy Raoult’s law:
pi = xi pi*

where the equilibrium pressures p*i are calculated from Riedel’s equation. The
flow rate F3' ,v is then calculated by means of a one-dimensional search
procedure.

TABLE 8.3.A-1
DATA FOR RIEDEL’S VAPOR PRESSURE EQUATION

Component Tb (K) Tc (K) Pc (bar)

A 454.1 594.4 57.1
B 351.5 516.2 63.0
C 340.3 523.2 37.8
D 373. 647.3 217.6
E 420. 520. 60.
 8.3 SEMIBATCH REACTOR MODELING 405

TABLE 8.3.A-2
PHYSICAL PROPERTIES OF THE VESSEL CONTENTS AND HEATING MEDIUM

Vessel Contents Heating Medium

Thermal conductivity (kJ/m h K) 1.0429 0.979
Viscosity (kg/m h) 1.287 5.4
Specific heat (kJ/kg K) 2.805 2.847
Density (kg/m3) 1000
Heat of vaporization (kJ/kmol) 35493

The boiling points under normal conditions and the critical properties of
the five components, required in the calculation of their vapor pressure according
to Riedel’s equation, are given in Table 8.3.A-1.
The Q in (8.3-2) is the heat exchanged through the coil (kJ/h):

Q = UA
(T '
in ) (
− T − Tex' − T ) (
= Wc ph Tin' − Tex' )
T −T
'
ln in'
Tex − T

Figure 8.3.A-1
Evolution of liquid volume and product concentrations.
406 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

Figure 8.3.A-2
Evolution of reactor temperature and of inlet and exit temperatures of the heat
exchanging medium.

Figure 8.3.A-3
Evolution of vapor-phase composition.
 8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES 407

where Tin' and Tex' represent the temperature of the heating medium at the inlet
and exit of the coil, respectively; W is the mass flow rate (kg/h) and cph is the
specific heat. The overall heat transfer coefficient U includes scale formation on
both sides of the coil surface.
The flow rate of the heating medium is 250 kg/h, and its inlet
temperature is 100°C.
Average physical properties of the reacting components and of the
heating medium are given in Table 8.3.A-2. The average heat of vaporization of
the reacting fluid is 35493 kJ//kmol.
Fig. 8.3.A-1 shows the concentrations in the liquid and the volume of the
latter as a function of time; Fig. 8.3.A-2 shows the temperature of the reactor
contents, T, and the inlet and exit temperatures of the heat exchanging medium,
Tin' and Tex' respectively and Fig. 8.3.A-3 the vapor-phase composition and vapor
withdrawal rate, F3' v . ▄

8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES

Two main types of situations are considered:

1. Optimal batch operation time for the sequence of operations in a given

reactor.

2. Optimal temperature (or other variable) variations during the course of

the reaction, aiming at minimizing the reactor size.

Only the main features will be discussed here — more extensive details are given
in Aris [1961]. To simplify the mathematical details only constant volume
reactors are considered, but for most practical situations this is not a serious
restriction.

8.4.1 Optimal Batch Operation Time

The discussion follows that of Aris [1961, 1965, 1969]. The price per kilomole of
chemical species Aj is wj, so that the net increase in value of the reacting mixture
is

W (θ ) = ∑ w j (N j − N j 0 )
N
(8.4.1-1)
j =1

N M
= ∑ w j ∑ α ij ξ i
j =1 i =1
408 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

M
= ∑ (∆W )i ξ i
i =1

where
N
(∆W )i ≡ ∑ α ij w j (8.4.1-2)
j =1

which is constant for a given stoichiometry and chemical costs. For a single
reaction, it is more common to introduce the conversion of the key species, A,
into (8.3.A-1):
N A0 x A
W (θ ) = (∆W ) (8.4.1-3)
αA

The cost of operation is usually based on four steps:

1. Preparation and reactor charging time θ P , with cost per unit time, of WP.

2. Reaction time θ R , with WR.

3. Reactor discharge time θ Q , with WQ.

4. Idle or “down” time θ 0 , with W0.

The total operation cost, then, is

WT = θ 0W0 + θ PWP + θ QWQ + θ RWR (8.4.1-4)

Since the reactor operation is the main point of interest here, all the other
times will be taken to be constant, and the main question is to determine the
optimal reaction time, with its corresponding conversion. The net profit is

W (θ R ) − WT (8.4-1-5)

and the optimum value of θR is found from

d
[W (θ R ) − WT ] = 0 (8.4.1-6)
dθ R
or
dW (θ R )
= WR (8.4.1-7)
dθ R
From (8.4.1-3),
dW (θ R ) N dx A
= (∆W ) A0
dθ R α A dθ R
 8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES 409

and combined with (8.1-3):

rA
= (∆W )V (8.4.1-8)
αA θR

At the optimum
rA ( x A )θ =
V WR
(8.4.1-9)
αA R
∆W

The actual optimum reaction time θR is calculated from (8.1-4), evaluated at xAR =
xA(θR) given by (8.4.1-9):
x AR
N dx A
θ R = A0 ∫ r (x ) (8.4.1-10)
V 0 A

Instead of the maximum net profit (8.4.1-5), the maximum of the net
profit per unit time may be desired:

W (θ R ) − WT
(8.4.1-11)
θT

where θT ≡ θ 0 + θ P + θ Q + θ R . Then, the optimum θ R is found from

d W (θ R ) − WT 
 =0 (8.4.1-12)
dθ R  θT 
or
dW (θ R ) W (θ R ) − WT
− WR = (8.4.1-13)
dθ R θT

Aris [1965] has provided a convenient graphical procedure for solving

(8.4.1-13) for the optimal value of θR. Fig. 8.4.1-1 illustrates a typical curve for
net profit. It is recognized that the right-hand side of (8.4.1-13) is precisely the
slope of the tangent line OL . From (8.4.1-13),

W (θ R ) − WT
=
d
[W (θ R ) − WT ]
θT dθ R

it is seen that the θR indicated in the figure is the time that satisfies (8.4.1-13),
and is the optimum value for maximum net profit per unit time. The point M and
corresponding θ R' give the optimum for maximum net profit, from (8.4.1-6).
410 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

Figure 8.4.1-1
Net profit curve. From Aris [1965].

EXAMPLE 8.4.1.A
OPTIMUM CONVERSION AND MAXIMUM PROFIT FOR A FIRST-
ORDER REACTION
For a simple first-order reaction, (8.1-4) gives
xA
−1
ln(1 − x A )
dx
θ = C A0 ∫ = (8.4.1.A-a)
0
kC A0 (1 − x ) k
or
x A = 1 − e − kθ (8.4.1.A-b)
Thus, from (8.4.1-3),
(
W (θ R ) = (∆W )N A0 1 − e − kθ R ) (8.4.1.A-c)

and the value of θ R for maximum profit is found from (8.4.1-7):

(
WR = (∆W )N A0 ke − kθ R ) (8.4.1.A-d)
or

(θ R )opt = 1 ln  (∆W )k N A0 
 
(8.4.1.A-e)
k  WR 
The optimum conversion is
 8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES 411

(xA )opt = 1 − WR
(8.4.1.A-f)
(∆W )k N A0
If the result (8.4.1.A-f) is substituted into the first-order rate form

V (rA )opt ,θ R = V k C A0 (1 − x A ) =
WR
∆W
which is (8.4.1-9) for this situation. ▄

8.4.2 Optimal Temperature Policies

This section considers two questions: (1) What is the best single temperature of
operation? (2) What is the best temperature progression during the reaction time,
or in other words the best trajectory? For single reactions, the results are
relatively straightforward. If the reaction is irreversible, and if the rate increases
with temperature the optimal temperature for either maximum conversion from a
given reactor operation, or minimum time for a desired conversion, is the highest
temperature possible. This highest temperature, Tmax, is imposed by other
considerations such as reactor materials and catalyst physical properties.
Similarly, for reversible endothermic reactions where the equilibrium conversion
increases with temperature (Efor. > Erev.), the highest allowable temperature is the
best policy.
The case of reversible exothermic reactions is more complicated, because
even though the rate may increase with temperature, once the equilibrium
conversion is reached, higher temperatures have an adverse effect on the
equilibrium conversion. Thus, there is an optimum intermediate temperature
where reasonably rapid rates are obtained together with a sufficiently high
equilibrium conversion. The precise value of the optimal temperature is obtained
from (8.1-4) at the final conversion xAf :
x Af
dxA
θ f = C A0 ∫ r (x ,T ) (8.4.2-1)
0 A

This can always be integrated for a constant value of temperature, and then the
best temperature found for a given conversion xAf. It can be shown that this is
exactly equivalent to the problem of choosing the optimal temperature for the
maximum conversion for a given reaction time θf.
412 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

EXAMPLE 8.4.2.A
OPTIMAL TEMPERATURE TRAJECTORIES FOR FIRST-ORDER
REVERSIBLE REACTIONS
For a first-order reversible reaction, the reaction rate is
1
A 2
S

rA = k1C A − k 2 C S (8.4.2.A-a)

= k1C A0 (1 − x A ) − k 2 C A0 x A (C S 0 = 0)

It is convenient to use dimensionless variables [e.g., Millman and Katz, 1967]

with
k1 = A1e − E1 / RT → u = A1θ f e − E1 / RT (8.4.2.A-b)

k 2 = A2 e − E2 / RT → β u α
where
θ f A2
β=
(θ f A1 )α
(8.4.2.A-c)

and
E2
α=
E1

θ
τ=
θf

Then, the mass balance (8.1-3) becomes

= u (1 − x A ) − βu α x A
dx A
(8.4.2.A-d)
dτ
The optimum value of u leads to a temperature

E1 / R
T= (8.4.2.A-e)
ln(θ f A1 / u )

For a given u, (8.4.2.A-d) can be easily integrated to give:

x Aeq 1 − ( x A0 / x Aeq )
τ= ln (8.4.2.A-f)
u 1 − ( x A / x Aeq )
 8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES 413

Figure 8.4.2.A-1
Dimensionless temperature versus parameter β. From Fournier and Groves [1970a].

Figure 8.4.2.A-2
Conversion versus parameter β. From Fournier and Groves [1970a].
414 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

where the equilibrium conversion is given by

(
x Aeq (u ) = 1 + βu α −1 )
−1
(8.4.2.A-g)

Equation (8.4.2.A-f) can be rearranged into

   x   − uτ / x 
x A = x Aeq 1 − 1 −  A0  e Aeq
 (8.4.2.A-h)
 
   Aeq  
x   

The value of the optimum T, or equivalently u, for maximum xAf with a given θf
can now be readily calculated from

 ∂x Af 
  = 0 (8.4.2.A-i)
 ∂u τ =1

It can be shown that for a single reaction this result is equivalent to the problem
of the optimum u for a minimum θf with a given xAf [Aris, 1961]. Fournier and
Groves [1970a] have provided useful charts based on equations (8.4.2.A-h) and
(8.4.2.A-i). With Figs. 8.4.2.A-1 and 8.4.2.A-2, both equivalent problems can be
readily solved by beginning with the known quantities: either α and β(θf) or α and
xAf. Other kinetic schemes have also been evaluated by Fournier and Groves
[1970a].
Even better results for the reversible exothermic reaction can be obtained
by choosing an optimal temperature variation with time. This type of operation is
also feasible in practice, using automatic control techniques. Qualitative
reasoning indicates that a high temperature at the beginning would be best, since
this increases the rate constant, and the equilibrium limitations are usually not
particularly important at this point. As the reaction progresses and approaches
equilibrium, it is important to have lower temperatures that favor higher
equilibrium conversions. Thus, the optimum temperature trajectory would be
expected to decrease with time. Also, the maximum overall rate, made up of the
sum of the instantaneous point rates, will be largest if each of the point rates is
maximized. This reasoning cannot be extended to multiple reactions, however,
since the overall optimum will be made up of the interactions of several rates.
▄
For a single reaction, the condition of optimality to be fulfilled in each
point is
∂rA
=0 (8.4.2-2)
∂T
 8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES 415

[A proof is given by Aris, 1965.] Only for simple cases analytical solutions are
possible. For the reaction
1
A+B 2
Q+S
the rate is
rA = k1C A C B − k 2 C Q C S

= A1e − E1 / RT C A2 0 (1 − x A )(M − x A ) − A2 e − E2 / RT C A2 0 x A2 (8.4.2-3)

where
C B0
M = and CQ 0 = 0 = C S 0
C A0

Then, the optimum temperature at each point is found from (8.4.2-2), with the
results
−1
 − R   A2 E2  x A2  
Topt =   ln    
 E1 − E2   A1E1  (1 − x A )(M − x A )  
−1
 1 
≡ ln[B2 B3 ] (8.4.2-4)
 (− B1 ) 
with
E1 − E 2 A2 E2
B1 = and B2 =
R A1E1

Fournier and Groves [1970b] have provided solutions for several other reaction
types. The definitions of B1 and B2 are the same for any single reversible
reaction, but a parameter B3 depends on the type of reaction:

Reaction B3
xA
A S
1 − xA
C A0 x A2
A Q+S
1 − xA
xA
A+B S
C A0 (1 − x A )(M − x A )

Other kinetic forms can be similarly handled. The calculation procedure

is then as follows. First, the result of utilizing (8.4.2-2), such as (8.4.2-4), is used
416 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

to determine Topt(xA) which is then used in the integration of the mass balance
(8.1-4) for θ(xA):
x Af
dx A
θ = C A0 ∫ r (x (8.4.2-5)
xA 0 A A , Topt ( x A ))

One complication that occurs can be seen from (8.4.2-4): for low
conversions, B3 may have a sufficiently small value to make B2B3 ≤ 1.0. Then,
(8.4.2-4) gives a value Topt → ∞ (or negative). In practice, of course, the
temperature will have to be limited to some value Tmax < Topt over a range of
conversion going from zero to some critical value xAc. This critical conversion,
xAc > 0, can be found by first using Tmax in (8.4.2-5).
A more general consideration of these problems involves the
optimization of some sort of objective function, like the equipment cost which
usually depends on outlet conversions and total residence time. It is difficult to
include all possible costs (e.g., safety), and so a simpler quantity, such as
selectivity, is often used instead.
Denbigh and Turner [1971] consider two major categories:

1. Output problems. These are concerned with the attainment of the

maximum output — the amount of reaction product(s) per unit time and
reactor volume.
2. Selectivity problems. These are concerned with maximizing the
selectivity — the fraction of reactant converted to the desired product.

The first type is most important for simple reactions with no side products and/or
very expensive reactors and catalysts. The second type occurs with complex
reactions, often producing harmful waste. Output problems are somewhat easier
to solve in general, since their simpler reaction schemes involve less
mathematical details.
The above case of single reversible exothermic reactions was an example
of an output problem. Intuitive logic led to the qualitative conclusion that the
optimum temperature profile was the one that maximized the rate at each point.
This was also the quantitative solution and led to the design techniques
presented. For selectivity problems, if the kinetics are not too complex, the
proper qualitative trends of the optimal temperature profiles can also often be
deduced by reasoning. However, the quantitative aspects must usually be
determined by formal mathematical optimization methods. Simple policies, such
as choosing the temperature for maximum local point selectivity, rarely lead to
the maximum final overall selectivity because of the complex interactions
between the various rates.
 8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES 417

A few examples of this qualitative reasoning are worth discussing.

Consider the scheme

If E 3 > E1 , E 2 > E1 , the optimum temperature trajectory is decreasing in time, as

for the simple reversible exothermic case, but not quite as high in the beginning,
to avoid excessive side reaction. If E3 > E1 > E 2 , the reversible reaction is
endothermic, and so a high temperature level is desirable; but if too high,
especially when CA and CB are large, too much side reaction occurs. The
optimum trajectory here is increasing. If E 2 > E1 > E 3 , a decreasing trajectory is
again to be preferred. Horn [see Denbigh and Turner, 1971, for references] has
worked out the mathematical details of these cases.
Another example is the familiar
1 2
A+ B → Q → S (Q desired)

If E 2 > E1 , the initial temperature should be large for a rapid first reaction, but
the temperature should be diminished as Q accumulates to preferentially slow
down the degradation reaction 2. Again, a decreasing trajectory is preferable.
An example that has two answers depending on whether it is looked at
from the output or yield viewpoint is the following:

From the output point of view, the optimum trajectory is an increasing one. At
the beginning of the reaction, the temperature should be low in order to promote
formation of Q rather than S; but at the end of the reaction time, the temperature
should be high to offset the otherwise low conversion rate — this gives more Q
even though it also results in more S. If the reactor cost is not important, as in the
selectivity problem, the temperature should be as low as possible throughout the
reaction time. This gives, relatively, the highest Q but requires a very large
reactor for significant conversion.
A final example is taken from Denbigh and Turner [1971]:
418 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

Here, the desired product S is formed through an intermediate, and both the feed
reactants and the intermediates can undergo side reactions. The four possible
cases from the point of view of yield are:

E1 > E 2 , E3 > E4 uniform high temperature

E1 < E 2 , E3 < E4 uniform low temperature
E1 < E 2 , E3 > E4 increasing temperature trajectory
E1 > E 2 , E3 < E4 decreasing trajectory.

Denbigh gave an example in which under isothermal conditions a yield of 25

percent was obtained, against 60 percent with an optimum temperature trajectory.

EXAMPLE 8.4.2.B
OPTIMUM TEMPERATURE POLICIES FOR CONSECUTIVE AND
PARALLEL REACTIONS
The two basic coupled reaction schemes, consecutive and parallel, were
considered in an interesting and useful simple way by Millman and Katz [1967],
and illustrate the computation of optimum temperature trajectories. The details
are expressed in dimensionless form, as above:
For consecutive reactions:
1 2
A→Q→S (Q desired)

dxQ
= u (1 − x A ) = u (1 − x A ) − βu α xQ
dx A
and
dτ dτ
For parallel reactions:

= (u + βu α )(1 − x A )
dxQ
= u (1 − x A )
dx A
and
dτ dτ
where
 8.4 OPTIMAL OPERATION POLICIES AND CONTROL STRATEGIES 419

Figure 8.4.2.B-1
Temperature histories for consecutive reaction: α = 2, β = ½. Yields: best isothermal,
0.477; best proportional, 0.489; optimal, 0.491. From Millman and Katz [1967].

Figure 8.4.2.B-2
Temperature histories for parallel reaction: α = 2, β = ½. Yields: best isothermal, 0.535;
best proportional, 0.559; optimal, 0.575. From Millman and Katz [1967].
420 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

C A0 − C A CQ θ
xA = xQ = τ=
C A0 C A0 θf

E2 θ f A2
u = θ f A1e E1 / RT α= β=
E1 (θ f A1 )α

The rigorous optimization could be performed with several mathematical

techniques — see Beveridge and Schechter [1970], and for a concise discussion
of the Pontryagin maximum principle, see Ray and Szekely [1973]; also see Aris
[1961] for specific chemical reactor examples.
The basic idea of a simple technique devised by Millman and Katz
[1967] was to assume that the temperature trajectory to be determined could be
approximated as a linear function of the desired product concentration to be
maximized; specifically,
u (τ ) ≈ c0 + c1 xQ (τ )

Then, the two parameters c0 and c1 are determined for the optimal condition
max {x Q (τ )} . This still requires a search technique to obtain the values of c0 and
c1, but it was found that these computations were much simpler than the
completely rigorous optimization. Actually, further terms in dxQ / dτ and
∫ xQ dτ gave better results than the linear function, and are based on standard
three-mode process controller actions.
Two typical results are shown in Figs. 8.4.2.B-1 and 8.4.2.B-2, and it is
seen that the best proportional (simple linear) results are close to the true optimal
values. Note that this appears to be true, even though the temperature curves u(τ)
have some differences between them — apparently the final yield is not
particularly sensitive to all the details of the curves. ▄

PROBLEMS
8.1 The esterification of butanol with acetic acid, using sulphuric acid as a
catalyst, was studied in a batch reactor:

O
H2SO4
C4H9OH + CH3COOH C4H9—O—C—CH3 + H2O

The reaction was carried out with an excess of butanol. The following
data were collected [C.E. Lejes and D.F. Othmer, I & E. C., 36, 968
(1945)]:
 PROBLEMS 421

Time (h) Acetic Acid

Concentration (mol/l)
0 2.327
1 0.7749
2 0.4514
3 0.3151
4 0.2605

Set up a suitable kinetic model of the homogeneous type.

8.2 The following data on the conversion of hydroxyvaleric acid into

valeroacetone were collected:

Time (min) 0 48 76 124 204 238 289

Acid Concentration 19.04 17.6 16.9 15.8 14.41 13.94 13.37
(mol/l)

Derive a suitable kinetic model by means of both the differential and

integral method of kinetic analysis.

8.3 The batch saponification of ethyl acetate,

CH3COOC2H5 + NaOH CH3COONa + C2H5OH

was carried out in a 200-ml reactor at 26°C. The initial concentrations of

both reactants were 0.051 N.
(a) From the following time-versus-concentration data, determine
the specific rate and tabulate as a function of composition of the reacting
mixture:
Time (s) NaOH (mol/l)
30 0.0429
90 0.0340
150 0.0282
210 0.0240
270 0.0209
390 0.0164
630 0.0118
1110 0.0067

(b) Determine a suitable reaction rate model for this reaction.

422 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

8.4 A daily production of 50,000 kg (50 tons metric) of ethyl acetate is to be

produced in a batch reactor from ethanol and acetic acid:

C2H5OH + CH3COOH CH3COOC2H5 + H2O

(A) (B) (R) (S)
The reaction rate in the liquid phase is given by

rA = k (C A C B − C R C S / K )
At 100°C,
k = 7.23 × 10-6 m3 / kmol s
K = 2.93

A feed of 23 percent by weight of acid, 46 percent of alcohol, and no

ester is to be used, with a 35 percent conversion of acid. The density is
essentially constant at 1020 kg/m3. The reactor will be operated 24h per
day, and the time for filling, emptying, and the like, is 1 h total for
reactors in the contemplated size range. What reactor volume is
required?

8.5 A gas-phase decomposition A → R + S is carried out with initial

conditions of T0 = 300 K, pt = 5 bar, and V0 = 0.5 m3. The heat of
reaction is –6280 kJ/kmol, and the heat capacity of A, R and S are 125.6,
104.7, and 83.7 kJ/kmol K, respectively. The rate coefficient is

 10,000 
k = 1014 exp −  h −1
 T 

(a) Compute the conversion-time profile for isothermal conditions.

Also, determine the heat exchange rates required to maintain
isothermal conditions.
(b) Compute the conversion-time profile for adiabatic conditions.

8.6 A desired product P is made according to the following reactions

scheme:
 PROBLEMS 423

Discuss qualititatively the optimum temperature profile for the two

cases:
(a) E 2 > E3 > E1 ; (b) E 2 > E1 > E3 . Describe your reasoning
carefully.

8.7 One method of decreasing the large initial heat release in a batch reaction
is to utilize “semibatch” operation. Here, the reactor initially contains no
reactant and is filled up with the reacting liquid — thus, there is an
inflow but no outflow, and the reacting volume continuously changes.
The mass balances are:
Total:
dV
= F0'
dt
Reactant A:
d
(VC A ) = F0'C A0 − kVC A
dt
(a) Show that the reactant concentration at any time is, with
isothermal operation,

CA (
F ' 1 − e − kt
= 0
)
(
C A0 k V0 + F0' t )
where V0 = initial volume.
(b) Derive an expression for the rate of heat release and sketch the
curve.

8.8 In a batch reactor with a volume V = 5 m3, an exothermic reaction A → P

is carried out in the liquid phase. The rate equation is

rA = kC a
with
 7900 
k = 4 × 10 6 exp −  s −1
 T 

The initial temperature T0 of the reaction mixture is 20°C and the

maximum allowable reaction temperature is 95°C. The reactor contains a
heat exchanger with area Ak = 3.3 m2, and it can be operated with steam
(Tr = 120°C, U = 1360 W/m3°C) or with cooling water (Tr = 15°C, U =
1180 W/m2°C). The times required for filling and emptying the reactor
are 10 and 15 min, respectively. Other physicochemical data are:
424 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

∆H = –1670 kJ / kg;
ρcp = 4.2 × 106 J / m3 °C;
MA = 100 kg / kmol;
CA0 = 1 kmol / m3.

The desired conversion is xAf ≥ 0.9, and the batch reaction and complete
reaction cycle times along the steam and water consumption rates are to
be determined for the following policies of operation:

(a) Preheat to 55°C, let the reaction proceed adiabatically, start

cooling when either T = 95°C or xA = 0.9 occurs, and cool down
to 45°C.
(b) Heat to 95°C, let the reaction proceed isothermally until xA = 0.9
occurs, cool down to 45°C. [See H. Kramers and K.R.
Westerterp, Elements of Chemical Reactor Design and
Operation, Academic Press, New York (1963).]

8.9 The reversible reaction

1
A 2
R

has the following rate parameters:

A1 = 7 s-1 E1 = 41,868 kJ / kmol;

A2 = 5000 s-1 E2 = 83,736 kJ / kmol.

The reaction is to be carried out in a batch reactor with a maximum

allowable temperature of Tmax = 800 K. For a conversion of xAf = 0.8,
determine the optimum isothermal operating temperature, and the
resulting batch holding time. Also determine the heat exchange rate
required.
If an optimum temperature profile is to be utilized, determine this as a
function of conversion and a function of processing time. Determine the
heat exchange rates required.

Additional data:

Density of liquid = 100 kg/m3;

Heat capacity = 4.187 kJ/kg °C;
Initial mole fraction of reactant A = 0.5;
Molecular weights = 100 for A
= 20 for solvent.
 REFERENCES 425

8.10 In Example 8.4.2.B, the dimensionless equations for a parallel reaction

were derived:

The initial conditions are xA = xQ = 0 at τ = 0.

(a) Derive an expression for the optimal temperature for maxu {xQ(τ =
1)}.
(b) For the parameters α = 2, β = 1/2, what is uopt? If E1 = 83736 kJ /
kmol, what is Topt?

8.11 An endothermic third-order reaction 3A → 2B + C is carried out in a

batch reactor. The reaction mixture is heated to 400°C. The reaction then
proceeds adiabatically. During the heating period, 10 mol-% of A is
converted. From this instant on, what is the time required to reach a
conversion of 70 percent?

(–∆H) = –104670 kJ / kmol;

V = 1 m3 = constant;
cpm = 2.47 kJ / kg K;
mt = 950 kg;
NA0 = 10.2 kmol;
10000
ln k = − + 5 [k in (m / kmol A) / s].
3 2

RT

REFERENCES
Aris, R., The Optimal Design of Chemical Reactors, Academic Press, New York (1961).
Aris, R., Introduction to the Analysis of Chemical Reactors, Prentice-Hall, Englewood
Cliffs, N.J. (1965).
Aris, R., Elementary Chemical Reactor Analysis, Prentice-Hall, Inc., Englewood Cliffs,
N.J. (1969).
Beveridge, G.S.G., and Schechter, R.S., Optimization Theory and Practice, McGraw-
Hill, New York (1970).
Butté A., Storti G., and Morbidelli, M., Chem. Eng. Sci., 54, 3225 (1999).
Chapman, F.S., Dallenbach, H., and Holland, F.A., Trans. Inst. Chem. Eng., 42, T398
(1964).
Chilton, T.H., Drew, T.B., and Jebens, R.H., Ind. Eng. Chem., 36, 510 (1944).
Cooper, A.R., and Jeffreys, G.V., Chemical Kinetics and Reactor Design, Prentice-Hall,
Englewood Cliffs, N.J. (1971).
Denbigh, K.G., and Turner, J.C.R., Chemical Reactor Theory, 2nd ed., Cambridge
University Press, London (1971).
Fournier, C.D., and Groves, F.R., Chem. Eng., 77, No. 3, 121 (1970a).
Fournier, C.D., and Groves, F.R., Chem. Eng., 77, No. 13, 157 (1970b).
Gates, L.E., Henley, T.L., and Fenic, J.G., Chem. Eng., 110, No. 26, 110 (1975).
Grummitt, O., and Fleming, F., Ind. Eng. Chem., 37, 4851 (1945).
426 CHAPTER 8: THE BATCH AND SEMIBATCH REACTORS

Holland, F.A., and Chapman, F.S., Liquid Mixing and Processing in Stirred Tanks,
Reinhold, New York (1966).
Hui, A.W., and Hamielec, A.E., J. Appl. Pol. Sci., 16, 749 (1972).
Matyaszewski K., Patten, T.E., and Xia, J., J. Am. Chem. Soc., 119, 674 (1997).
Millman, M.C., and Katz, S., Ind. Eng. Chem. Proc. Des. Dev., 6, 447 (1967).
Peterson, T.I., Chem. Eng. Sci., 17, 203 (1962).
Ray, W.H., and Szekely, J., Process Optimization, Wiley, New York (1973).
Reuss, M., Fröhlich, S., Kramer, B., Messerschmidt, K., and Pommerening, G.,
Bioprocess Engineering, 1, 79 (1986).
Reuss, M., Fröhlich, S., Kramer, B., Messerschmidt, K., and Pommerening, G., Bioproc.
Eng., 1, 79 (1986).
Smith, J.M., Chemical Engineering Kinetics, McGraw-Hill, New York (1970).
Strek, F., Int. Chem. Eng., 3, 533 (1963).