Isothermal reactors with multiple reactions

By Hans Livbjerg

The stoichiometric equation for a reaction scheme with multiple reactions is:

(1)

where i j is the stoichiometric coefficient for component Aj in reaction no i. The sign of i j

follows the rules: positive for products, negative for reactants and zero for inerts.
The relation between the rates of the individual reactions Ri, i=1,...,q and the production rates for
the components rj is given as:

(2)

For the definition of measures of conversion for a multiple reaction scheme we start by choosing
one key reactant which must be present in the feed stream (flow reactors) or at the start of
reaction (batch reactors). Otherwise the key component can be chosen arbitrarily among all the
reaction components whether they appear in a single reaction or in several reactions. We now
define a conversion Xi for each of the q reactions by the following expression, in which A
(without index) denotes the key reactant:

(3)

The term, “moles of A at the basis condition” denotes NA0 for a batch reactor and FAo for a flow
reactor.
Following this definition we can express the contribution to the production of any component Aj
from the i'th reaction as:

1
 (4)

These terms are used to construct a stoichiometric table which provides expressions for the
reactor variables in terms of the conversions for all reactions, but otherwise is quite equivalent to
the stoichiometric table for a single reaction.
Finally note, that the choice of key reactant is completely arbitrary and may be considered a means
of normalising the X es to make them dimensionless. The conversions for multiple reactions, in
spite of their name, are primarily connected with the individual reactions and not with any specific
reaction component.

Example 1 Designing a reactor for two reactions

The following two elementary reactions occur in the liquid phase:

(a)

with rate constants: k1 = 0,01 m3/mol/s and k2 = 0,001s-1

We wish to compute the exit concentrations CAf, CBf, CCf, CDf from a CSTR reactor with a volume
of V = 5 m3 and a volumetric feed rate of v = 0,005 m3/s. The feed concentrations are CA0 = CB0
= 0,1 mol/m3.
Start by choosing A as key reactant. The production rate for Aj from the i'th reaction is then given
by FA0 Xi i j (cf. Eq. (4)) and we can now set up a stoichiometric table as previously, but now

including the contributions to the production rates from both reactions.

Table 1. Stoichiometric table for example 1:

Mole fraction
Aj Fj0 Produced Fj Cj = Fj /v
yj = FJ /FT

A FA0 -FA0 X1 -FA0 X2 FA0(1- X1 - X2) CA0(1- X1 - X2)

B FA0 -FA0 X1 FA0(1- X1) CA0(1- X1)

C 0 FA0 X1 FA0 X1 CA0 X1

D 0 2FA0 X2 2FA0 X2 2CA0 X2

2
Calculations:

Total mole rate at Xi=0: FT0 = 2 FA0 (sum of column 2)

Total mol rate for given Xi: FT = 2 FA0 - FA0 X1 + FA0 X2
= FT0(1 - yA0 X1 + yA0 X2) (sum of column 4)

Because the reaction mixture is a liquid we assume that the density and thereby v is constant: v
= vo. The composition of the reaction mixture can be expressed either by the mole fractions of the
components yj = Fj/FT or by their molar concentrations Cj = Fj /v. Both results are shown in the
two last columns of the table, in which we can see that all mole rates and concentrations are
uniquely determined by X1 og X2.

Reactor model:
The reactor model comprises the mole balances for two reaction components. The general rule
at this point is to set up the mole balances for q separate species, where q is the total number of
reactions. Choose species C and D for the mole balances. (These can be chosen arbitrarily but it
is possible to choose an unfortunate set which yields linearly dependent model equations which
cannot be solved. In that case a new set must be chosen). A secure way to choose mole balance
species is to select from each reaction a species which appears only in a single reaction. The
procedure, now, is to insert in the mole balance equations all molar flow rates and concentrations
from the expressions in the stoichiometric table so that the mole balance models eventually only
contain Xi, i=1,2,..,q as variables:

Mole balance for C:

0 FC + V rC = 0 FA0 X1 + V R1 = 0

and for D:
0 FD + V rD = 0 2FA0 X2 + 2V R2 = 0

Insertion of the rate expressions:

R1 = k1CACB = k1CA02(1 X1 X2)(1 X1)
and
R2 = k2CA = k2CA0(1- X1 - X2)
insertion of Ri into the balance equations:
FA0 X1 = V k1CA02(1- X1 - X2)(1- X1) (b)

FA0 X2 = V k2CA0(1- X1 - X2) (c)

Equation b and c provide a set to be solved for the unknown conversions, X1 and X2.
Divide (c) with (b):

3
and solve for X2:

(d)

Substitute into the mole balance (c) to obtain:

and:

This polynomial has two roots:

with corresponding values of X2 calculated from (d):

Although, mathematically there are two solutions to the problem, only one root pair i.e. (X1,X2)
= (0,2679, 0,3660) provides a physically acceptable solution. The other root pair yields negative
mole rates for some of the components which is immediately clear from the stoichiometric table.
This solution is therefore not a physically realisable one and must be discarded. With a knowledge
of X1 and X2 from the acceptable solution, we can determine all variables of technical significance
for our reactor from the stoichiometric table. We conclude the example by calculating the molar
concentrations in the reactor outlet by means of the last column of the table with (X1,X2) =
(0,2679, 0,3660):

CAf = 0,03660 mol/m3

CBf = 0,07321 -
CCf = 0,02679 -
CDf = 0,07321 -

4
The notion that there are both acceptable and non-acceptable solutions to the mole balance
equations deserves a further elaboration. For a solution to be a physically realisable state of the
reactor, all mole rates Fj must be non-negative. In the present example, the requirement that
Fj 0, after insertion of Fj from the stoichiometric table, yields the following set of inequalities:
FAf 0 1 - X1 - X2 0

FBf 0 1 - X1 0

FCf 0 X1 0

FDf 0 X2 0

Each inequality defines a line which divides the X1,X2-plane into an acceptable and a non-
acceptable half plane. Taken together the inequalities define a polygon - in this case a triangle
which encloses the acceptable area in the X1,X2-plane as shown on figure 1.

Figure 1. The polygon of acceptable steady state solutions

to example 1.

5
Generalized mole balance equations

The work of setting up mole balances for q components to derive the reactor model equations,
can conveniently be systematized by employing the mole balance equations below, which easily
can be derived from general mole balances. The equations are valid for any reaction scheme
consisting of q reactions. The equations correspond to the reactor models for a single reaction and
it is easy to ensure oneself that the reactor models for a single reaction emerge from the equations
below when the number of equations is set to q = 1.
To use the equations, one must set up a stoichiometric table to compute the concentrations in
terms of the conversions. These expressions for the concentrations can then be inserted into the
rate expressions Ri to transform the latter into unique functions of Xi.
The following equations apply for homogeneous reactors. The corresponding equations for a
catalytic reactor are obtained by substituting W (catalyst mass) for V (reactor volume) and R`i
(catalytic reaction rate, mol/kg/s) for Ri (homogeneous reaction rate, mol/m3/s).

Steady state CSTR reactor:

(5)

Steady state plug flow reactor:

(6)

6
Batch reactor:

(7)