ISOTHERMAL REACTOR DESIGN

In Chapter 1 & 2, we discussed balances on batch & flow reactors. In

Chapter 3, we discussed rxns. Here, we will combine rxns and
reactors.
Topics
Part 1: Mole Balances in Terms of Conversion
1. Algorithm for Isothermal Reactor Design
2. Applications/Examples of CRE Algorithm
3. Reversible Reactions
4. ODE Solutions to CRE Problems
5. General Guidelines for California Problems
6. PBR with Pressure Drop
7. Engineering Analysis
Part 2: Measures Other Than Conversion
1. Measures Other Than Conversion
2. Membrane Reactors
3. Semibatch Reactors

Part 1: Mole Balances in Terms of Conversion

1. Algorithm for Isothermal Reactor Design
The algorithm for the pathway of interest can be summarized as:
1. Mole Balance and Design Equation (choose reactor type)

2. Rate Law (choose rxn type; gas or liq. phase)

3. Stoichiometry
4. Combine
5. Evaluate
The Evaluate Step can be carried out
A. Graphically (Plots)
B. Numerically (Quadrature Formulas)
C. Analytically (Integral tables)
D. Using Software Packages

1
Elementary gas phase reaction in different reactor types
CSTR
The elementary gas phase reaction 2A + B → C
takes place in a CSTR at constant temperature (500 K) and constant pressure
(16.4 atm). The feed is equal molar in A and B. x = 0.9, k= 10dm6/mol2s
VCSTR =?

Mole balance

Rate Law

Stoichiometry gas phase, isothermal (T = T0), no pressure drop (P = P0)

Derivation

2
Combine

Evaluate

PFR and Batch Reactors

Elementary Gas Phase Reaction: 2A + B → C
PFR
Mole Balance

Rate Law

gas phase, isothermal (T = T0), no pressure

Stoichiometry
drop (P = P0), CAo=CBo (Θ=1), v=vo(1+εX)

3
Combine

CAo=0.2, v=vo=25 dm3/s, k=10 dm6/mol2 s, ε=-0.5,

Parameter Evaluation
X=0.9

V=227 dm3

Deriving CA and CB

Remember that the reaction is:

For a gas phase system:

If the conditions are isothermal (T = T0) and isobaric (P = P0):

We must divide by the stoichiometric coefficient of our basis of

calculation yielding:

4
 And if the feed is equal molar, then:

This leaves us with CA as a function of conversion alone:

Similarly for CB:

Batch Reactor Constant Volume, V=Vo and the pressure changes.

Mole Balance

Rate Law

Stoichiometry

Combine

5
Parameter Evaluation CAo=0.2, k=10 dm6/mol2 s,

6
Scale – Up of Liquid – Phase Batch Reactor Data to Design of a CSTR:
Scale – Up a lab experiment to pilot – plant operation of full – scale
production. Find k from experimental data and use it to design a full – scale
flow reactor.
Batch Operation:
Liquid phase: (density change is small → V = V0)
Gas phase with constant volume V = V0.

1 ⎛ dN A ⎞
⋅⎜ ⎟ = rA
V ⎝ dt ⎠
1 ⎛ dN A ⎞ 1 dN A dN A / V0 dC A
⋅⎜ ⎟= ⋅ = = = rA
V ⎝ dt ⎠ V0 dt dt dt

Since conc’n is a measured quantity in liquid – phase rxns:

dC A
− = −rA
dt

Let’s calculate the time necessary to achieve a given conversion X for

the irreversible second order rxn:

A→B
dX [1]
N A0 ⋅ = −rA ⋅ V0
dt
[2]
− rA = k ⋅ C A
2

C A = C A0 ⋅ (1 − X ) [3]
Combine [1], [2] and [3];
dX
= k ⋅ C A0 ⋅ (1 − X ) 2
dt
Rearrange: dX
= k ⋅ C A0 ⋅ dt
(1 − X ) 2
@ t = 0; X = 0; T = T0 (isothermal) → k → constant
⎛ X ⎞
t X
1 dX 1
∫t =
0
k ⋅ C A0 ∫ (1 − X )
0
2
⇒ t= ⋅⎜ ⎟
k ⋅ C A0 ⎝ 1 − X ⎠

7
It is important to know the reaction time, tR, to achieve a certain
conversion.
Flow reactors use characteristic rxn times, tR.
The time for a total cycle is much longer than tR, as one must account for
the time to fill (tF), heat (th) and clean (tC).
t = tf + th + tC + tR
Design of Continuous Stirred Tank Reactors
CSTRs are usually used for liquid phase rxns:

FA0 ⋅ X v0 ⋅ C A0 ⋅ X
V= = [1] (Design Eq’n for CSTR)
(− rA ) exit − rA
Divide by v0: V C A0 ⋅ X [2]
τ= =
v0 − rA
Volumetric flow rate

A single CSTR:
First order irreversible rxn:

− rA = k ⋅ C A [3] Combine [2], [3] and [4]

C A = C A0 ⋅ (1 − X ) [4]

1 ⎛ X ⎞
τ = ⋅⎜ ⎟ [5]
k ⎝1− X ⎠
τ ⋅k [6]
X=
1+τ ⋅ k
C A0
Combine [4] and [6] CA =
1+τ ⋅ k

rxn Damköhler #,
Da: a dimensionless # that for a first order rxn

says the degree of conversion that can be achieved in cont. flow reactor.

8
 − rA0 ⋅V rate of rxn of A
Da = :
FA0 rate of convective transport of A t =0

For a second order rxn:

− rA0 ⋅ V k ⋅ C A0 ⋅V
2
Da = = = τ ⋅ k ⋅ C A0
FA0 v0 ⋅ C A0

Da ≤ 0.1 X < 0.10

Da ≥ 10 X > 0.90

Da
X= First order liq-phase rxn (Eq’n [6])
1 + Da

CSTRs in Series
CA1, X1

CA0 For first order rxn (v = v0)

CA2, X2 The effluent conc’n of reactant
A from the first CSTR is
-rA1 -rA2
C A0 V1
V1 V2 C A1 = where τ 1 =
1 + τ 1 ⋅ k1 v0
From the mole balance on reactor 2:
FA1 − FA2 v0 (C A1 − C A2 )
V2 = =
− rA2 k2 ⋅ C A2
Solving for CA2:
C A1 C A0
C A2 = =
1 + τ 2 ⋅ k 2 (1 + τ 2 ⋅ k 2 )(1 + τ 1 ⋅ k1 )

9
 If both reactors are of equal size, (τ 1 = τ 2 = τ )
and operate at the same T (k1 = k2 = k)
C A0
C A2 =
(1 + τ ⋅ k ) 2
for n equal sized CSTR system
C A0 C A0
C An = =
(1 + τ ⋅ k ) n
(1 + Da) n

Substituting for CAn in terms of conversion

C A0
C A0 ⋅ (1 − X ) = ⇒
(1 + Da) n
1 1
X = 1− = 1−
(1 + Da) n
(1 + τ ⋅ k ) n

CSTRs in Parallel:
Equal sized reactors are placed in parallel rather than in series:
FA01 FA02 FA03 FA0n

X1 X2 X3 Xn

Individual volume is given by

⎛ X ⎞
V = FA0i ⋅ ⎜⎜ i ⎟⎟ [1]
⎝ − rAi ⎠
Since Vi = Vj, then Xi = X; → -rA1 = -rA2 = ...... = -rA
V [2]
Vi =
n
FA0 [3]
FA0i =
n
Substitute [2] & [3] into [1]

V FA0 ⎛ X i ⎞ X
= .⎜ ⎟ = FA0 ⋅ [4]
n n ⎜⎝ − rAi ⎟⎠ − rA

10
Example: 2nd order rxn, v = v0, CA = CA0 (1-X); FA0 X = v0 CA0 X
Combine –rA = k CA2 and V = FA0 X / -rA

FA0 ⋅ X v0 ⋅ C A0 ⋅ X
V= =
k ⋅ CA k ⋅ C A0 ⋅ (1 − X ) 2
2 2

Divide by v0:

V X
τ= =
v0 k ⋅ C A0 ⋅ (1 − X ) 2
(1 + 2 ⋅τ ⋅ k ⋅ C A0 ) − (1 + 2 ⋅τ ⋅ k ⋅ C A0 ) 2 − (2 ⋅τ ⋅ k ⋅ C A0 ) 2
X=
2 ⋅τ ⋅ k ⋅ C A0
(1 + 2 ⋅ Da) − 1 + 4 Da
X= Since τ ⋅ k ⋅ C A0 = Da
2 ⋅ Da for a second order rxn.

Since X can not be greater than 1.0, ( - ) sign is chosen.

Example: The elementary liquid phase reaction

is carried out isothermally in a CSTR. Pure A enters at a volumetric flow

rate of 25 dm3/s and at a concentration of 0.2 mol/dm3.
What CSTR volume is necessary to achieve a 90% conversion when k = 10
dm3/(mol*s)?

Mole Balance

Rate Law

Stoichiometry liquid phase (v = vo)

11
 Combine

Evaluate at X = 0.9,

V = 1125 dm3

Space Time

2. Applications/Examples of the CRE Algorithm

12
Applying the algorithm to the above reaction occuring in Batch, CSTR, PFR

13
 3. Reversible Reaction
To determine the conversion or reactor volume for reversible
reactions, one must first calculate the maximum conversion that can
be achieved at the isothermal reaction temperature, which is the
equilibrium conversion. (See Example 3-8 in the text for additional
coverage of equilibrium conversion in isothermal reactor design.)

Equilibrium Conversion, Xe

Calculate Equilibrium Conversion (Xe) for a Constant Volume System

Example: Determine Xe for a PFR with no pressure drop, P = P0
Given that the system is gas phase and isothermal, determine the reactor
volume when X = 0.8 Xe

First calculate Xe: C Be

Equilibrium constant Kc is; KC = 2
C Ae
C Ae = C A0 (1 − X e )
⎛1 ⎞
C Be = C A0 ⎜ X e ⎟
⎝2 ⎠

Xe = 0.89
X = 0.8Xe = 0.711

14
Deriving The Equilibrium Constant (KC) and Equilibrium
Conversion (Xe) for a Non-Constant Volume System:
The reversible reaction:

which takes place in gas phase PFR. Since gas phase reactions almost
always involve volume changes, we will have to account for volume
changes in our calculations. The equilibrium constant, KC, for this
reaction is:

A is the limiting reactant A ⇔ B/2

where CAe and CBe are:

Substituting for CAe and CBe gives us:

Substituting known values (CA0 = 0.2 mol/dm3 and KC = 100 dm3/mol),

and realizing that:

Solving for the equilibrium conversion, Xe, yields:

Xe = 0.89

15
Batch Reactor With a Reversible Reaction

The following reaction follows an elementary rate law

Initially 77% N2, 15% O2, 8% inerts are fed to a batch reactor where 80%
of the equilibrium conversion (Xe = 0.02) is reached in 151 µ s. What is
the specific reaction rate constant k1?

Additional Information

16
For 80% of equilibrium conversion X = 0.8 Xe = 0.016

Use Simpson's three point formula to integrate

with ∆X = 0.016/2 = 0.008

17
Tubular Reactors:
Gas-phase rxns are usually carried out in tubular reactors where the gas
is generally turbulent.

Reactants Products

No radial variation in velocity, conc’n, T or –rA.

dX
Design Eq’n: FA0 ⋅ = − rA
dV
In the absence of pressure drop or heat exchange, integral form of the
design equation is used:
X
dX
V = FA0 ⋅ ∫
0
− rA

Pressure Drop in Reactors

In liquid-phase rxns, the conc’n of reactants is not effected by the
changes in pressure.
However, in gas phase rxns, conc’n is propotional to total pressure.
Pressure Drop & Rate Law
Example: Analyze the following second order gas phase reaction
that occurs isothermally in a PBR:

Mole Balance
Must use the differential form of the mole balance to separate variables:

Rate Law
Second order in A and irreversible:

18
 Stoichiometry

Isothermal, T = T0

Combine

Need to find (P/P0) as a function of W (or V if you have a PFR).

Pressure Drop in Packed Bed Reactors

Ergun Equation

P = pressure (kPa) • •

Dp = diameter of particle (m) m = m (st − st )

0

Φ = porosity = volume of void / total bed volume ρ 0 ⋅ v0 = ρ ⋅ v

1 – Φ = volume of solid / total V
gC = 1 (metric)
µ = viscosity of gas (kg / m s)
z = length down the packed bed (m)
u = superficial velocity = v0 (volumetric flow) / A2 (cross-area)
ρ = gas density (kg/m3)
G = ρ u = superficial mass velocity (kg/m2 s)

Ergun, S., 1952. Fluid flow through packed columns. Chemical Engineering Progress 48, pp. 89–94.

19
 Variable Density

Catalyst Weight

Then

We will use this form for multiple reactions:

Since ⎛ F ⎞
FT = FT 0 + FA0δX = FT 0 ⎜⎜1 + A0 δX ⎟⎟
⎝ FT 0 ⎠
FT ⎛ F ⎞
= 1+ ε ⋅ X ⎜⎜ ε = y A0 ⋅ δ = A0 ⋅ δ ⎟⎟
FT 0 ⎝ FT 0 ⎠

20
 when ε<0; ∆P will be less than ε = 0.
when ε>0; ∆P will be greater than ε = 0.

Isothermal Operation

Recall that

Notice that

The two expressions are coupled ordinary differential equations. We can solve
them simultaneously using an ODE solver such as Polymath. For the special
case of isothermal operation and epsilon = 0, we can obtain an analytical
solution.

Analytical Solution , [e], PFR with

dy
2y ⋅ = −α
dW
y = 1 ( P = P0 ) @ W = 0
y 2 = 1 − α ⋅W

CAUTION: Never use this form if

21
 Combine

Solve

Could now solve for X given W, or for W given X.

For gas phase reactions, as the pressure drop increases, the concentration
decreases, resulting in a decreased rate of reaction, hence a lower
conversion when compared to a reactor without a pressure drop.

Effect of presure drop on P,rate, con’n, and x.

22
Pressure Drop in Pipes:
Pressure drop for gases flowing through pipes without packing can be
neglected. For flow in pipes, ∆P is given by:

dP du 2 ⋅ f ⋅ G 2
= −G ⋅ −
dL dL g⋅D
Where
G = ρ u (g/cm3s) = mass velocity is constant along L.
u = average velocity of gas, cm/s.
f = fanning friction factor
f = fnc (Re, pipe roughness

P ⎡ 4 ⋅ f ⋅ G 2 ⋅V ⎤
= ⎢1 − ⎥ = (1 − α pV )
1/ 2

P0 ⎣ ρ 0 ⋅ P0 ⋅ AC ⋅ D ⎦

Example p 187

23
Optimum Particle Diameter

Laminar Flow, Fix P0, ρ0, Φ

ρ0 = P0(MW)/RT0
ρ0P0 ∼ P02

Cross-sectional area

Superficial mass velocity = ρ u

24
 Increasing the particle diameter descreases the pressure drop and increases
the rate and conversion.

However, there is a competing effect. The specific reaction rate decreases as the
particle size increases, therefore so deos the conversion.

k ∼ 1/Dp

DP1 > DP2

k1 > k2

Higher k, higher conversion

25
The larger the particle, the more time it takes the reactant to get in and out
of the catalyst particle. For a given catalyst weight, there is a greater
external surgace area for smaller particles than larger particles. Therefore,
there are more entry ways into the catalyst particle.

7. Engineering Analysis - Critical Thinking and Creative Thinking

We want to learn how the various parameters (particle diameter,
porosity, etc.) affect the pressure drop and hence conversion. We need
to know how to respond to "What if" questions, such as:
"If we double the particle size, decrease the porosity by a factor of 3, and
double the pipe size, what will happen to D P and X?"

To answer these questions we need to see how a varies with these

parameters

Turbulent Flow

26
Compare Case 1 and Case 2:
For example, Case 1 might be our current situation and Case 2 might be
the parameters we want to change to.

For constant mass flow through the system = constant

Laminar Flow

Part 2: Mole Balances in terms of Conc’n & Molar Flow Rates

In some cases, it is more convenient to deal with number of moles or
molar flow rates rather than conversion.
Membrane reactors and multiple rxns taking place in gas phase are
examples.
The main difference in molar flow rates, you have to write the mole
balance on each and every species.
Membrane reactors can be used to increase conversion when the rxn is
thermodynamically limited as well as to increase the selectivity when
multiple rxns are occuring.

A. Membrane reactors
B. Multiple reaction
Liquids: Use concentrations, i.e., CA

27
1. For the elementary liquid phase reaction carried out in a
CSTR, where V, vo, CAo, k, and Kc are given and the feed is pure A, the
combined mole balance, rate laws, and stoichiometry are:

There are two equations, two unknowns, CA and CB

Gases: Use Molar Flow Rates, I.E. FI

2. If the above reaction, ,carried out in the gas phase in a

PFR, where V, vo,CAo,k, and Kc are given and the feed is pure A, the
combined mole balance, rate laws, and stoichiometry yield, for isothermal
operation (T=To) and no pressure drop (∆P=0) are:

28
• Microreactors

Are characterized by their high surface area to volume ratios (due to many
micro-tubes and channels). Dchannel = 100µm, Lch= 2 cm

Control of heat and mass transfer resistance!!

They are used for highly exothermis rxns, for rxns with toxic or explosive
intermediates. ALso,i for the productions of speciality chemicalsi
combinatorial chemical screening, chemical sensors.

In modeling, we assume they are PFR. dFa/dV = rA

29
Membrane Reactors
Membrane reactors can be used to achieve conversions greater than the original
equilibrium value. These higher conversions are the result of Le Chatelier's
Principle; you can remove one of the reaction products and drive the reaction to
the right. To accomplish this, a membrane that is permeable to that reaction
product, but is impermeable to all other species, is placed around the reacting
mixture.

Example: The following reaction is to be carried out isothermally in a membrane

reactor with no pressure drop. The membrane is permeable to Product C, but it is
impermeable to all other species.

For membrane reactors, we cannot use conversion. We have to work in

terms of the molar flow rates FA, FB, FC.

Mole Balances

Rate Laws

30
Stoichiometry
Isothermal, no pressure drop

Combine Polymath will combine for you

Parameters

Solve Polymath

3. Semibatch Reactors
Semibatch reactors can be very effective in maximizing selectivity in liquid
phase reactions.

A + B ⎯⎯→
kD
D ( product )
A + B ⎯⎯→
kU
U (undesired )
rD = k D ⋅ C A ⋅ C B
2

rU = kU ⋅ C A ⋅ C B
2

kD ⋅ C A
Selectivity ( D / U ) =
kU ⋅ C B

The reactant that starts in the reactor is always the limiting reactant

31
 Three Forms of the Mole Balance Applied to Semibatch Reactors

If you have multiple reactions, use concentrations to make mole balances!!

ODE Solutions to CRE Problems

dX − rA
=
dV FA0
⎧⎪⎡ (1 − X ) ⎤ ⎡ C A0 ⋅ X
2
⎤ ⎫⎪
rA = − k ⎨⎢C A0 ⋅ ⎥ −⎢ ⎥⎬
⎪⎩⎣ 1 + ε ⋅ X ⎦ ⎣ 2(1 + εX ) ⋅ K C ⎦ ⎪⎭

[V, X] = ode45(@vdp1, [0 500], [0]) Matlab solution

32
Example: Elementary Irreversible Reaction
Consider the following irreversible elementary reaction

-rA = kCACB
The combined mole balance, rate law, and stoichiometry may be written
in terms of number of moles, conversion, and/or concentrati

Polymath Equations

33
Equilibrium Conversion in Semibatch Reactors with Reversible
Reactions
Consider the following reversible reaction:

Everything is the same as for the irreversible case, except for the rate law

Where:

At equilibrium, -rA=0, then

34
function f=volume(V,x)
Ca0=0.2;
Kc=100;
Fa0=5;
k=2;
epsilon=-0.5;
ra=k*(((Ca0*(1-x(1))/(1+epsilon*x(1)))^2)-
(Ca0*x(1)/(2*Kc*(1+epsilon*x(1)))));
f(1)=ra/Fa0;

>> [V x]=ode45('volume',[0 500],[0]);

>> plot(V,x)
>> grid
>> xlabel('Volume')
>> ylabel('conversion')

0.9

0.8

0.7

0.6
conversion

0.5

0.4

0.3

0.2

0.1

0
0 50 100 150 200 250 300 350 400 450 500
Volume

35
 function f = packedbed(W,x)
alpha = 0.0002;
k = 10;
epsilon = -0.5;
FA0=2.5;
CA0 = 0.2;
f = zeros(2,1);
CA = (CA0*(1-x(1))*x(2))/(1 + epsilon*x(1));
ra = k*(CA.^2);
f(1) = ra / FA0;
f(2) = -alpha*(1 + epsilon*x(1))/(2*x(2));

>> [W x]=ode45('packedbed', [0 1000], [0;1]);

>> plot(W,x(:,1))
>> xlabel('W, kg')
>> ylabel('X')
>> plot(W,x(:,2))
>> xlabel('W, kg')
>> ylabel('y')

1.001

0.999

0.998
y

0.997

0.996

0.995

0.994
0 10 20 30 40 50 60 70 80 90 100
W , kg

0.9

0.8

0.7

0.6

0.5
X

0.4

0.3

0.2

0.1

0
0 10 20 30 40 50 60 70 80 90 100
W , kg

36
Combining mole balances, rate laws and
stoichiometry

37
Pressure Drop in Packed Bed Reactors

Ergun Equation

P = pressure (kPa) • •

Dp = diameter of particle (m) m = m (st − st )

0

Φ = porosity = volume of void / total bed volume ρ 0 ⋅ v0 = ρ ⋅ v

1 – Φ = volume of solid / total V
gC = 1 (metric)
µ = viscosity of gas (kg / m s)
z = length down the packed bed (m)
u = superficial velocity = v0 (volumetric flow) / A2 (cross-area)
ρ = gas density (kg/m3)
G = ρ u = superficial mass velocity (kg/m2 s)

Ergun, S., 1952. Fluid flow through packed columns. Chemical Engineering Progress 48, pp. 89–94.

For gas phase reactions, as the pressure drop increases, the concentration
decreases, resulting in a decreased rate of reaction, hence a lower
conversion when compared to a reactor without a pressure drop.

Effect of presure drop on P,rate, con’n, and x.

38
Part 2: Mole Balances in terms of Conc’n & Molar Flow Rates
In some cases, it is more convenient to deal with number of moles or molar flow rates
rather than conversion.

Membrane reactors and multiple rxns taking place in gas phase are examples.
Microreactors with multiple reactions

39