EQ1021 Chemical Reaction Engineering

Lecture 5. Continuous Stirred Tank Reactor (CSTR)

1 Design Equations
2 Single CSTR
3 CSTR in series

1
Lecture 5. Continuous Stirred Tank Reactor (CSTR)

1 Design Equation. Material balance in a volume system.

2 Single CSTR.
2.1 Space-time, Space-velocity, Residence-time.
2.2 What are the feed parameters: cA0, cB0, FA0, FB0, QV0?
2.3 Graphical Representation of the Mole Balance
2.4 Design problems: calculating V, QV0 or the outlet stream composition.
2.4.1 Analytical solutions
2.4.2 Numerical solutions
2.5 Optimizing selectivity
2.6 Maximizing production
2.7 Process economics: optimizing benefit or cost.

3 CSTRs in series.
3.1 Mole balance to a CSTR in the Series.
3.2 Analytical Solution
3.3 Numerical Solution. Single or Multiple Reaction Systems.
3.4 Graphical Solution.

2
1. Design Equations
Material Balance in a System Volume

3
2. Single CSTR
Why? http://encyclopedia.che.engin.umich.edu/Pages/Reactors/Reactors.html

Advantages:
(1) Relatively cheap to build;
(2) Ease of control of temperature in each stage, since each operates in a stationary state; heat transfer
surface for this can be easily provided;
(3) Can be readily adapted for automatic control in general, allowing fast response to changes in
operating conditions (e.g., feed rate and concentration);
(4) Relatively easy to clean and maintain;
(5) With efficient stirring and viscosity that is not too high, the model behavior can be closely
approached in practice to obtain predictable performance.

Disadvantages:
• The most obvious disadvantage in principle stems from the fact that the outlet stream is the same as
the contents of the vessel. This implies that all reaction takes place at the lowest concentration (of
reactant A, say, cA) between inlet and outlet. For normal kinetics, in which rate of reaction (-RA)
decreases as cA decreases, this means that a greater volume of reactor is needed to obtain a desired
conversion. (For abnormal kinetics, as in autocatalytic reactions, the opposite would be true.)
• Not used for industrial gas-phase reactions (can be used in laboratory)
4
 2. Single CSTR

dN j
 Fj 0  Fj  G j
dt
QV 0
dN j
Fj  Fj 0   Gj Gj  Rj V
dt
no uniformity of
QV accumulation properties in
of materials the volume
composition

 
changes with time
Fj  Fj 0  R j exit
V Eq. 5-1

• Let’s suppose a single reaction system.

Applying the mole balance to the key component (A):

V
Fk 0 ( 1  X )  Fk 0  Rk exit V Fk 0 X   Rk exit V ck 0 X   Rk exit the ratio V/QV0 has
QV 0
 units of time

• Spatial time, : is the natural performance measure for CSTR, analogue to reaction time (tR) for a batch
reactor (in constant density systems). 5
2. Single CSTR
Space-time, Space-velocity, Residence-time

• Thus, a space-velocity of 5 hr-l means that five reactor volumes of feed at specified conditions are being fed
into the reactor per hour. A space-time of 2 min means that every 2 min one reactor volume of feed at
specified conditions is being treated by the reactor.
• In a CSTR, usually space-time and space-velocity are calculated at the conditions of the inlet (reactor feed)
• The reactor volume is taken as the volume of the reactor physically occupied by the reacting fluids. It does not
include the volume occupied by agitation devices, heat exchange equipment, or headroom above liquids.

6
2. Single CSTR
Space-time, Space-velocity, Residence-time

• The mean residence time () is defined as the average time that any macroscopic portion of the reactant
mixture remains in the reactor, from the moment it enters the reactor until it leaves. It is, therefore, the true
reaction time.

• For a CSTR, it is defined as:

V

Qv

where QV is the volumetric flow rate at the reactor outlet.

 For a constant density system, QV = QV0 and then CSTR =  CSTR.

 For a reactant gas system, at constant P and T,  is calculated as follows:

V CSTR
CSTR  
Qv 0 ( 1   k  X ) ( 1   k  X )
7
2. Single CSTR
What are the feed parameters: cA0, cB0, FA0, FB0, QV0?

• Two feed streams, one containing 2.8 mol A/liter and the other containing 1.6 mol B/liter, are to be
introduced at equal volumetric flow rates into the reactor,

QV 0  Q
streams
V 0 stream

QVA=QVB=2 liter/min
Fj 0  F
streams
j 0 stream

c QV 0stream
j 0 stream
c j0  streams

Qstreams
V 0 stream

Combine ALL the feed streams into a SINGLE ONE and calculate the parameters!!
8
 2. Single CSTR
Graphical Representation of the Mole Balance

QV 0 ck 0 X   Rk exit  Eq. 5-2

for liquid or gaseous systems,

any k
QV
 gas species
k  xk0
 k

• In liquid systems (QV0=QV=constant),

Eq. 5-1 becomes:

QV c k  QV 0 c k 0  Rk exit V

QV 0
Eq. 5-2
Eq. 5-3

ck 0  c k   Rk exit  Eq. 5-3

Graphical representation of the design

equations for a mixed flow reactor. 9
2. Single CSTR
Design Problems

What do we know?
 Inlet conditions: Qv0, T0 (not always necessary), cj0, Fj0, etc.
 Reaction rate parameters: A’, A, E, E’.
 Reaction stoichiometry and rate law

Design Problems
1. Finding reaction rate
2. Calculating reactor volume or spatial time for a given conversion
3. Determining conversion for a given spatial time
4. Maximizing selectivity
5. Maximizing production
6. Maximizing benefit

10
2. Single CSTR
Design Problems. Finding Reaction Rate.

CSTR gives differential Reaction Rate data!!!!

ck  c k 0  Rk exit 

 cA RA cB RB 
1 c A1 R A1 c B1 R B1 
2 c A2 R A2 cB 2 RB 2 
3 c A3 RA3 cB 3 RB 3 
     

 RA
1. Change volume flowrate, QV0, to the reactor (change ).
2. Measure the concentraction of any of the chemical species
that participate in the reactions.
k
3. Graph Rj versus different functions of cj and determine rate
law and rate parameters.
4. The process is the same for single and multiple reactions
systems.
cA ·cB 2 11
2. Single CSTR
Design Problems

12
2. Single CSTR
Design Problems. Calculating volume for a specific conversion.

• Let’s consider a single CSTR and see which volume is necessary in order to obtain a certain conversion,
in a single reaction system.

• First, graph the reaction rate inverse,  1

vs conversion, X (Levenspiel plot).
Rk
 1 
 
• Then, calculate the volume of reactor needed: 
1
  Rk  exit , X  3 X 1
Rk

1 V 
X    V


 Rk exit Fk 0 ck 0 knowing FA0  1



Area   Rk  exit , X  2 X 1
(rectangle) X1 → V1
2 X1 → 3.71V1  1 
 
3X1 → 11.97V1   Rk  exit , X  X 1

X
• The reactor volume increases very quickly as the X1 2 X1 3 X1

reaction progresses. Usually there is not a single reactor

but a cascade of reactors in series to decrease reaction
volume.
13
2. Single CSTR
Design Problems. Calculating volume for a specific conversion.

• If we compare the spatial time (volume) of a CSTR with the reaction time in a BR:
CSTR  1  Batch
 
1
   R A  exit , X  3 X 1
R Ak
tR/cA0
1
 RA
 1 
 
  R A  exit , X  2 X 1 X dX
 tR

  R X  c
 1  0 A A0
    
  R A  exit , X  X 1 Area

X
X1 2 X1 3 X1
0 X
 1
X 
c A0

 R A exit
Area

the time necessary for the same conversion in the CSTR is higher. The reason is that CSTR is always
operating at the lowest reaction rate. The BR, on the other hand, starts at a high rate and gradually
decreases to the final rate, thereby requiring less time. 14
Fogler. Essentials of Chemical Reaction Engineering.

We need to find an X such that

the area of the CSTR rectangle
is 25 dm3.

This is a trial and error

procedure.

VCSTR= 50 · 0.5 = 25 dm3

X = 0.5
15
 2. Single CSTR
Design Problems. Analytical Solutions. Irreversible First-Order Reaction

QV 0

QV  QV 0

• Mole Balance to the key component, A:

FA  FA0  R A V  c A  c A0   k c A 

1
 Design problem 1: known cA calculate V (or  ):   c A0  c A 
 RA
analytical solution
c A0
 Design problem 2: known V or  calculate cA : cA 
1k 
Eq. 5-4 16
 2. Single CSTR
Design Problems. Analytical Solutions. Reversible First-Order Reaction.

QV 0

QV  QV 0 A↔B Known Keq

• Mole Balance to the key component, A: c A  c A0    kc A  k cB  

1
 Design problem 1: known cA calculate V (or  ):   c A0  c A 
 RA
 Design problem 2: known V or  calculate cA :

c A  c A0   k c A  k c A0  c A    k c A0  k  k  c A

analytical solution

Eq. 5-5 cA  

c A0 1  k   c A0 1  k / K eq 
1  k  k     1 

1  1 
K  k
 eq  17
 2. Single CSTR
Design Problems. Analytical Solutions. Parallel First-Order Reactions.

A 
k1
B
QV 0 k2
A  C
k3
A  D
QV  QV 0 Writing the mole balances:
c A  c A0   k1  k 2  k 3 c A
cB  cB0  k1c A 
cC  cC 0  k 2 c A 
cD  cD 0  k 3 c A
• These equations can be solved sequentially
to give:
c A0
cA 
1  k1  k 2  k 3 
k1c A0 
cB  c B 0 
1  k1  k 2  k 3 
Eq. 5-6
k 2 c A0 
cC  cC 0 
1  k1  k 2  k 3 
k 3 c A0 
cD  c D 0  18
1  k1  k 2  k 3 
 2. Single CSTR
Design Problems. Analytical Solution. Consecutive First-Order Reactions.

QV 0 A 
k1
B 
k2
C 
k3
D

Writing the mole balances:

QV  QV 0 c A  c A0   k1c A 
cB  cB0  k1c A  k 2 cB  
cC  cC 0  k 2 cB  k 3 cC  
cD  cD 0  k 3 cC 

• These equations can be solved sequentially to give:

c A0
cA 
1  k1 
cB 0 k1 c A 0 
cB  
1  k 2  1  k1 1  k 2  
Eq. 5-7
cC 0 k 2 cB 0  k1 k 2 c A 0 2
cC   
1  k 3  1  k1 1  k 2   1  k1 1  k 2   1  k 3  

cD  cD 0  c A0  c A   cB0  cB   cC 0  cC 
from the stoichiometric 19
relationship
2. Single CSTR
Design Problems. Analytical Solution. Consecutive First-Order Reactions.

• Class Exercise: Find the value of  that maximizes cB , when cB0 =0.

cB 0 k1cB0 
cB  
1  k 2  1  k1  1  k 2  

dcB k1cB0 1  k1  1  k 2    k1cB 0  k1 1  k 2    k 2 1  k1  

0  0
d 1  k1   1  k2  
2 2

1  k1  1  k2     k1 1  k 2    k 2 1  k1    0

1  k1   k 2   k1 k 2  2  k1   k1 k 2  2  k 2   k1 k 2  2  0

1

k1 k 2

20
2. Single CSTR
Design Problems. Numerical Solutions: knowing , calculate cj

• Competitive first-order reactions and a few other simple cases can be solved analytically, but any
reasonably complex kinetic scheme will require a numerical solution to obtain the outlet
concentrations for a certain .

• Mathematics programs such as Mathematica, Mathcad, and MatLab offer nearly automatic solvers for
sets of algebraic equations.

• Class scientific calculators as CASIO Classpad 330, HP 50g, HP Prime or several Texas Instrument
models can be programmed to solve systems of non-linear equations

• The system converges quickly provided your initial guesses for the unknowns are good, but it will take
you to never-never land when your initial guesses are poor.

• A more robust method of solving the design equations for multiple reactions in a CSTR is given in the
next section (false transients)

21
2. Single CSTR
Design Problems. Numerical Solutions: Calculating Concentrations ( is known)

• Example 1: Suppose the competing, elementary reactions:

c A0  c A
 k I c AcB  k II c A  0

c B 0  cB
 k I c Ac B  0

cC 0  cC
 k I c Ac B  0

c D 0  cD
 k II c A  0


Data:
cA0=2 mol/l Solution:
cB0= 3 mol/l
cA= 0.36603
cC0= 0.2 mol/l
cB= 1.73205
cD0= 0.2 mol/l
cC= 1.46795
kI= 0.5 l/mol·min
cD= 0.56605
kII= 0.25 min-1
=4 min 22
2. Single CSTR
Design Problems. Numerical Solutions: Calculating Concentrations ( is known)

Example 2: Suppose the following reaction network is occurring in a constant-density CSTR:

AB r1  k1c 1A/ 2  k 1cB

BC r2  k 2 cB2
BD E r3  k 3 cBcD

The rate constants are: k1=0.6 mol1/2/(m3/2·h), k-1=0.8 h-1 , k2=0.1 mol/(m3·h) and k3=0.06 mol/(m3·h).
Solve for cA0= 3 mol/m3, cD0= 3 mol/m3 and cB0 = cC0 = cE0 = 0 and  = 1 h.

c A0  c A 3  c A  0.6 c 1A/ 2  0.8cB  0

 k1c A1/ 2  k 2cB  0

 cB  0.6 c 1A/ 2  0.8cB  0.1cB2  0.06 cBcD  0
cB 0  cB
 k 1c A1/ 2  k 2cB  k 3cB2  k 3cBcD  0  cC  0.1cB2  0

cC 0  cC 3  cD  0.06 cBcD  0 Solution:
 k 3cB2  0
 cE  cE 0  c D 0 c D 
cD 0  cD cA= 2.43434
 k 3c B c D  0
 cB= 0.46310
cE  cE 0  c D 0 c D 
cC= 0.02145
cD=2.91889
cE=0.081
23
2. Single CSTR
Design Problems. Numerical Solution. The Method of False Transients.

• The method of false transients converts a steady-state problem into a time-dependent problem. Mole
balances govern the steady-state performance of a CSTR.

How does a reactor reach the steady state?

• There MUST BE a startup transient that eventually evolves into the steady state, and a simulation of
that transient will also evolve to the steady state.
The simulation need not be physically exact. Any startup trajectory that is mathematically convenient
can be used, even if it does not duplicate the actual startup. It is in this sense that the transient can be
false.
• Suppose at time t =0 the reactor is instantaneously c10 , c20 , c 30 ,...
filled to volume V with fluid of initial concentrations
QV 0
c'10, c'20, . . . : The initial concentrations are usually
set equal to the inlet concentrations, c10, c20, . . . ,
but other values can be used. c' 10
QV 0
c' 20
c' 30 ,...
• The simulation begins with QV0 set to its steady-state
value. For constant-density cases, QV is set to the
same value.
changing with time 24
2. Single CSTR
Design Problems. Numerical Solution. The Method of False Transients (cont.).

c10 , c20 , c30 ,...

dc A c A0  c A
QV 0
  R A c A , cB , , cS ,T 
dt 
dcB cB 0  cB
c10   RB c A , cB , , cS ,T 
QV 0 dt 
c20 
c30 ,... dcS cS 0  cS
  RS c A , cB , , cS ,T 
dt 

changing with time

• This set of first-order ODEs is easier to solve than the algebraic equations where all the time
derivatives are zero. The initial conditions are that cA=c'A0, cB=c'B0, …, cS=c'S0 at t=0.
• The long-time solution to these ODEs will satisfy the mole balances provided that a steady-state
solution exists and is accessible from the assumed initial conditions. There may be no steady state.
• Stirred tank reactors can also exhibit oscillations or more complex behavior known as chaos. It is also
possible that the reactor has multiple steady states, some of which are unstable. Multiple steady
states are fairly common in CSTRs when the reaction exotherm is large.
• The method of false transients will go to a steady state that is stable but may not be desirable.
25
2. Single CSTR
The Method of False Transients (cont.). Calculating concentrations

Naumann, Ex 4.2
• Example 1: Suppose the competing, elementary reactions:
Inlet: cA0=2 mol/l cB0= 3 mol/l
dc A c A0  c A
  k I c AcB  k II c A cC0= 0.2 mol/l cD0=0.2 mol/l
dt 
dcB cB 0  cB kI= 0.5 l/mol·min
  k I c Ac B kII= 0.25 min-1
dt 
dcC cC 0  cC  = 4 min
  k I c AcB
dt 
dcD cD 0  cD cA/cA0
  k II c A
dt 
cC/cA0
dc A 2  cA
  0.5 c AcB  0.25 c A
dt 4
dcB 3  cB cD/cA0
  0 . 5 c Ac B
dt 4
dcC 0.2  cC
  0. 5 c Ac B cB/cA0
dt 4
dcD 0 .2  c D
  0.25 c A t/
dt 4

• The method can also be used to find the equilibrium concentrations resulting from a set of batch
chemical reactions. To do this, formulate the ODEs for a batch reactor and integrate until the
concentrations stop changing. 26
Hydrodealkylation of Mesitylene

The production of m-xylene by the hydrodealkylation of mesitylene over a Houdry Detrol catalyst involves the
following reactions, where the subscripts are: M = mesitylene, X = m-xylene, T = toluene, C = methane, and H =
hydrogen (H2):

+ H2 + CH4 r1  13.81cM cH0.5 ( kmol ·m 3 ·h 1 )

M X C
m-Xylene can also undergo hydrodealkylation to form toluene:

+ H2 + CH4 r2  7.553 c X cH0.5 ( kmol ·m 3 ·h 1 )

X T C
The second reaction is undesirable, because m-xylene sells for a higher price than toluene (65 cents/lb vs. 11.4
cents/lb). Thus we see that there is a significant incentive to maximize the production of m-xylene.
The hydrodealkylation of mesitylene is to be carried out isothermally in gas phase, at 833 K and 35 atm in a CSTR with
catalytic pellets inside. The feed is 66.7 moI% hydrogen and 33.3 mol% mesitylene.
Plot the concentrations of hydrogen, mesitylene, and xylene as a function of space-time. Calculate the space-time for
which xylene concentration is maximum.
27
 Hydrodealkylation of Mesitylene (cont.)

tau cH CM cX cT cMe cM+cX+cT

0 0.3412 0.1706 0 0 0
0.05 0.2874 0.1245 0.03833 0.00776 0.05384696 0.17059
0.1 0.2521 0.1007 0.05065 0.01921 0.08903292 0.17056
0.15 0.2268 0.0859 0.05503 0.02969 0.11443365 0.17062
0.2 0.2074 0.07556 0.05631 0.03874 0.13378117 0.17061
0.25 0.192 0.06789 0.0562 0.04650 0.14920395 0.17059
0.3 0.1793 0.06194 0.05545 0.05320 0.16186419 0.17059
0.4 0.1596 0.0532 0.0532 0.06421 0.18161441 0.17061
0.5 0.1448 0.04703 0.05071 0.07287 0.19644594 0.17061
0.7 0.1237 0.03878 0.0461 0.08572 0.21757546 0.17060
1 0.1033 0.03136 0.04062 0.09861 0.23780121 0.17059
1.3 0.08992 0.02672 0.03648 0.10741 0.25125755 0.17061
0.40 1.6 0.08025 0.0235 0.03325 0.11383 0.26092627 0.17058
2 0.07079 0.02043 0.02992 0.12025 0.27038694 0.17060
cH
0.35
cj CM
0.30 cX
cT
0.25 cMe

0.20

0.15

0.10

0.05

0.00
0 0.5 1 1.5 2 2.5 28
h)
2. Single CSTR
Optimizing selectivity.

Selectivity: for multiple reaction systems. In CSTRs, differential selectivity.

Rdesired product
Sdesired/byproducts =
Rbyproducts

k1 mol
A  X r1  k1 k1  0.02
l ·min
k2c A
k2
A  B r2  k 2 c A k 2  0.06 min 1 SB / XY 
k1  k 3 c A2
k3 l
A  Y r3  k 3 c A2 k 3  0 .3
mol ·min

• We act on concentration and temperature to favor reaction 2 over 1 and 3.

• Optimal selectivity doesn’t imply optimal production
 sometimes selectivity is higher at low T (reaction rate is slower at low temperatures)
 sometimes selectivity is higher at low concentration (reaction rate is slower at low cj)
29
2. Single CSTR
Optimizing selectivity.

Fogler, Prob. 6.6 Consider the following system of gas-phase reactions:

mol
A 
k1
X r1  k1 k1  0.002
l·min
A 
k2
B r2  k 2 c A k 2  0.06 min 1
l
A 
k3
Y r3  k3c A2 k3  0.3
mol·min
B is the desired product, and X and Y are foul pollutants that are expensive to get rid of. The specific
reaction rates are at 27°C. The reaction system is to be operated at 27°C and 4 atm. Pure A enters the
system at a volumetric flow rate of 10 dm3/min.
a) Sketch the instantaneous selectivities (SBX, SBY, and SB/XY =RB/(RX+RY) as a function of the
concentration of cA at 27°C. Which representation gives the most information about
selectivity?
b) At which concentration is the selectivity in B maximum at 27°C? Derive an expression to
calculate maximum selectivity with concentration.
c) Suppose that E1 = 10,000 cal/mol. E2 = 20,000 cal/mol, and E3 = 30,000 cal/mol. What is the
influence of temperature on selectivity?
d) If 90% conversion of A is desired, what temperature would you recommend for a single CSTR?
Which should the reactor space time and volume be?
30
2. Single CSTR
Optimizing selectivity.

1.4
14 1.4
SBX

SBY 1.2
SBX 12 1.2
SB/XY
SBXY
SBY 10 1 SB/XY 1.0

8 0.8 0.8

6 0.6 0.6

4 0.4 0.4

2 0.2 0.2
295°C
K
310°C
K

0 0 330°C
K
0.0
0 0.05 0.1 0.15 0.2 0.00 0.05 0.10 0.15 0.20
cA cA

Variation of selectivities with the Variation of SB/XY with the concentration of A

concentration of A at 27°C (300 K) at different temperatures

31
2. Single CSTRs
Process Economics. Optimizing Benefit or Cost.

• Let’s take the following multiple reaction system as example

A  kI
B 
k II
C
where A is the raw material, B is the desired product, and C is an undesired by-product. The process
flow diagram is given in the following figure.
• For simplicity, the separation step has been assumed to give clean splits of the three components
without material loss. Pure A is recycled back to the reactor. As a practical matter, the B and C streams
must be clean enough to sell, so the asumption is quite logical.

FB (pure)
QV0 QV cB = 4.5 kmol/m3
FA0 (pure)

separation
step
QV FC (pure)
FAi
FA, FB

reactor FC FA (pure)
cA = 4.5 kmol/m3
34
2. Single CSTR
Process Economics. Optimizing Benefit or Cost.

• The goal is an optimized process design that includes the costs of product recovery, in-process
recycling, and by-product management.

• In general, benefit (B) or profit in €/ h, is defined as the difference in value between the product and
the reactants, minus the cost of the plant (fixed charges) and the operation cost.

 reactor   reactor   separation step   separation step 

B  €h    F j w j   F j 0 w j        
  fixed ch arg es   operating cos t   fixed ch arg es   operating cos t 
difference in
production value

• Maximum Benefit:
dB dB
0 or 0
dX dV

• Minimum Cost:

dC dC
0 or 0
dX dV
35
3. CSTRs in series
Lecture (Missen, Mims, Saville, 14.4 Multistage CSTR)

Are there advantages in using a multistage CSTR consisting of two or more relatively small tanks
rather than a single tank of relatively large size to achieve the same result?
(A multistage CSTR consists of two or more stirred tanks in series).

Recall that a major disadvantage of a CSTR is that it operates at the lowest concentration between inlet
and outlet.

For a single CSTR, this means that operation is at the lowest possible concentration in the system, and
for normal kinetics, the required reactor volume is the largest possible. If two tanks (operating at the same
T) are arranged in series, the second operates at the same concentration as the single tank above, but the
first operates at a higher concentration, so that the total volume of the two is smaller than the volume of
the single vessel. If more than two tanks are used, the volume advantage becomes even greater.

Furthermore, there is an opportunity to adjust temperature and/or concentration (or conversion)

between stages to achieve an optimal situation with respect to total volume for a given number of stages
(N). If N is chosen as a parameter and allowed to vary, there is a practical limit to the number of stages
chosen (usually about five), since the cost of the interconnecting piping and valving increases.

Also, although the sizes of the N stages may not be equal for an optimal V, there is an advantage in the
cost of fabrication in making them all the same size. Thus, a minimum value of V is not necessarily the best
answer to the design of a CSTR, but it is one to consider.
36
3. CSTRs in series (in cascade or multistage CSTR)

• Many times reactors are connected in series so that

the exit stream of one reactor is the feed stream for another reactor.
When this arrangement is used it is often possible to speed calculations
by defining conversion in terms of location at a point downstream
rather than with respect to any single reactor.

• That is, the conversion X is the total number of moles of A

that have reacted up to that point per mole of A fed to the first reactor.
However, this definition can only be used provided that
there are no side streams withdrawn and the feed stream
enters only the first reactor in the series.
ck ( i 1 ) , X i ck ( N 1 ) , X N
ck 0 , X 1  0 ck 1 , X 2 ck 2 , X 3 cki , X ( i 1 )
Fk 0 , QV 0

ck ,N , X N

ck1 = ck0 (1-X1) ck2 = ck0 (1-X2) cki = ck0 (1-Xi) ckN = ck0 (1-XN)
Fk1 = Fk0 (1-X1) Fk2 = Fk0 (1-X2) Fki = Fk0 (1-Xi) FkN = Fk0 (1-XN) 37
3. CSTRs in series (in cascade or multistage CSTR)
Mole balance to a CSTR in the series

ck ( i 1 ) , X i ck ( N 1 ) , X N
ck 0 , X 1  0 ck 1 , X 2 ck 2 , X 3 cki , X ( i 1 )
Fk 0 , QV 0

ck ,N , X N

• For reactor i and species Ak: Fk ,i  Fk ,i   Rk X i ,T exit Vi Eq. 5-8

where Fk ,i is the molar flowrate of Ak at reactor i inlet stream and Fk ,i at the reactor outlet stream.

1 1
• Referring Fk ,i , Fk ,i to the feed to the first reactor, Fk0:  R k  X ,T   Rk X i ,T exit

Vi

   
Vi 1 Fk 0
Fk 0 X i  X i   Rk X i ,T exit Vi   Xi  Xi
Fk 0  Rk X i ,T exit
 Xi Xi X
area of a rec tan gle
38
Eq. 5-9
 3. CSTRs in series (in cascade or multistage CSTR)
Why?

The total volume is usually SMALLER for a series of CSTRs

(not always, depends on the shape of –Rk)

Type (a) curve Type (b) curve

(ordinary reactions) (autocatalytic reactions or exothermic
a cascade of CSTRs (red area) requires LESS
reactions in adiabatic conditions)
volume than a single one (blue area) a cascade of CSTRs (red area) can require MORE
volume than a single one (blue area)

1 1
 Rk  Rk V1 CSTR
V1 CSTR Fk 0
Fk 0 1
V3  f X  V1
1
Fk 0
 Rk Fk 0  f X 
V2 continuously  Rk
Fk 0 growing with a minimum
V1 V2 V3
Fk 0 Fk 0 Fk 0

X1 X2 Xf X1 X2 Xf 39
3. CSTRs in series (in cascade or multistage CSTR)
Volume optimization in a curve type b (with a minimum)

• 1 CSTR up to Xmin (or up to a slightly higher X1), a series of CSTRs afterwards.

1
 Rk

1
 f X 
 Rk
V2 V3 with a minimum
V1
Fk 0 Fk 0
Fk 0

X min X 1 X2 Xf

40
3. CSTRs in series
Analytical Solution

• There is a neat analytical solution for the special case of first-order reactions. The outlet concentration
from the nth reactor in the series of CSTRs is

c k ,n
c k ,n  Eq. 5-9
1  kn n

where ck ,n is the concentration of component Ak at the inlet of the nth reactor, kn is the rate constant
at the reactor operating temperature and n is the mean residence time
(n = 1, 2, ... ,N)
Applying Equation (5.9) repeatedly gives the outlet concentration for the entire train
of reactors:
ck 1
c k ,N  Eq. 5-10
1  k1 1 1  k2  2  1  k N  N 

• When all the kn are equal (i.e., the reactors are at the same temperature) and all the n are equal (i.e.,
the reactors are the same size),
ck 1
c k ,N  Eq. 5-11

1k  
N
41
3. CSTRs in series
Numerical Solution. Single or multiple reaction system.

Numerical calculations are the easiest way to determine the performance of CSTRs in series. Simply
analyze them one at a time, beginning at the inlet.

• In the case of a single reaction, the easiest way is to solve in X:

 knowing or supposing 1, calculate X1 from the mole balance to the first reactor, then for a known
or fixed 2, calculate X2 for the second reactor, etc.

• In the case of multiple reactions: fix  for every reactor and solve for concentrations (cj), not for
conversion. For every reactor, as many non-linear equations (mole balances) as chemical species
participate in the rate laws may have to be solved simultaneously. It depends though, on the
information we wish to obtain from the system.
Example:
R A   k 1 c A

A  B  C RB  k1 c A  k 2 cB  k 2 cC
k1 k2
 k2
R  k c  k  c
 C 2 B 2 C

 to calculate cA: solve mole balance equation for A

 to calculate cB: solve mole balance equations for A, B and C
 to calculate cC: solve mole balance equation for A, B and C 42
3. CSTRs in series
Graphical Solution: single reaction systems.
Equal or variable-size reactors: Jones’ method.

There are available a certain number of reactors of different sizes

and we have to optimize operation.

1.  can be different for every reactor.The mole

balance is written as:
ck 0 F   k rn
 k0  Eq. 5-12
n Vn Xn  Xn

2. Draw the curve –k·r =f[X]

3. From (0.0) draw a line of slope Fk0/V1 until it

intersects the curve at a point of abscissa X1
(output of the first reactor). From (X1, 0) draw
a second line of slope Fk0/V2 that intersects
the curve again in a point of abscissa X2
As Fk0 is the same for all the
reactors, we KNOW the slope
4. As many lines as reactors are drawn until of the lines
the last line of slope Fk0/VN intersects the curve
at a point of abscissa Xn = XN. 43
3. CSTRs in series
Graphical Solution : single reaction systems.
Equal or variable size reactors: Jones’ method.

• Reactors can also have the same volume:

 In case we need to optimize a series of CSTRs, one near-optimal criterion is chosing all the reactors
of the same size
 In this case, the method works the same way, but all slopes are equal.

fixing XF and  → N fixing N and  → XF

44
3. CSTRs in series
And what about several reactors in parallel?

• We divide the feed among all the reactors. The optimum is always obtained when  is the same for all
reactors

QV 0 ·V1
QV 0 ,1 
V1  V2  V3

V1 V  V2  V3
1   1 1
QV 0 ,1 QV 0 
R Ak

QV 0 ·V2
QV 0 ,2  all reactors work at the
V1  V2  V3
QV 0 same conversion
V2 V  V2  V3 and at the
2   1
QV 0 ,2 QV 0 same reaction rate

X
X
QV 0 ·V3
QV 0 , 3 
V1  V2  V3
V3 V  V2  V3
3   1
QV 0 ,3 QV 0

45
Example.

Levenspiel 6-11
FA0 Q V 0 ·c A0  150 mol / s
0.5 L/s
RA
cA=5 mol/L
0.035

0.03
QV0=1.5 L/s 0.5 L/s
0.025
cAf
0.02
0.5 L/s
0.015

0.01 slope= FA0/V=0.025/.5

V=300 liters
0.005 3 CSTRs in parallel

0
0 0.2 0.4 0.6 0.8 1
46
X
3. CSTRs in series (in cascade or multistage CSTR)
Why not a PFR?

For the great majority of reaction schemes, piston flow is optimal. Thus, the reactor designer normally
wants to build a tubular reactor and to operate it at high Reynolds numbers so that piston flow is closely
approximated.

But this may not be possible. There are many situations where a tubular reactor is infeasible and where
continuous-flow stirred tank reactors must be used instead. Typical examples are reactions involving
suspended solids and autorefrigerated reactors where the reaction mass is held at its boiling point.

There will usually be a yield advantage, but a cost disadvantage, from using several CSTRs in series.

47