3.

M ole B alances
Reactor Technology Fundamentals

Chemical Engineering, R Rawatlal

A real reactor may exhibit complex internal and overall shapes, and irregular flow patterns. This influences the temperature
and concentration within the reactor, and complicates our effort to estimate the reaction rate and hence write the design
equations. However, by classifying ideal flow patterns, we begin to make progress in developing our designs.

3.1 Translating between concentration and conversion

We established the following design equations for the three reactor types as follows

B atch C ontinuous stirred P lug Flow

𝑑𝑐𝑖 𝑐𝑖 − 𝑐𝑖,𝑖𝑛 𝜕𝑐𝑖
= 𝑟𝑖 = 𝑟𝑖 = 𝑟𝑖
𝑑𝑡 𝜏 𝜕𝜏
𝑑𝑋𝑖 𝑟𝑖 𝑋𝑖 𝑟𝑖 𝑑𝑋𝑖 𝑟𝑖
=− =− =−
𝑑𝑡 𝑐𝑖,0 𝜏 𝑐𝑖,𝑖𝑛 𝑑𝜏 𝑐𝑖,𝑖𝑛

To solve these equations, we must establish the reaction rate as a function of the
concentration or in terms of the conversion. In fact, the 𝑟𝑖 is usually known as a function of
concentration; this is it’s default representation. There are times when it is represented in
terms of partial pressure, but a partial pressure can easily be translated to a concentration or
vice versa.

In the case of an elementary reaction, the reaction rate can easily be established as the
product of the reactants raised to the stoichiometric coefficient:
𝜈
𝑟𝑖 = 𝑘Πi 𝑐𝑖 𝑖

For example, in the case of the sulphur dioxide reaction equation:

2. 𝑆𝑂2 + 𝑂2 → 2. 𝑆𝑂3

..we established that the reaction rate is given by

2
−𝑟𝑆𝑂2 = 𝑘. 𝑐𝑆𝑂 .𝑐
2 𝑂2

Also from the earlier discussion, we have that

𝑟𝑆𝑂2 𝑟𝑂2 𝑟𝑆𝑂3
= =
−2 −1 2
The conversion concept helps us relate the changes in concentration to a reference
component rather than requiring that we develop a separate balance for each component.

1|Page
3.2 A pplication of conversion
We have so far observed that conversion may be defined in terms of the initial or the inlet
concentration of the component. It is important to understand the difference between these
two types; in the same reactor, it is of course possible to define both initial amounts and
inlet amounts, and for these to be different values. Consider the CSTR below:

The inlet stream is at a concentration of 𝑐𝑖,𝑖𝑛 , and this quantity is generally different from the
initial concentration in the reactor 𝑐𝑖0 . It is fairly clear that in the CSTR case, 𝑐𝑖,𝑖𝑛 is the
appropriate basis for the conversion.

When developing the conversion expression, we shall use the term 𝑐𝑖,𝑟𝑒𝑓 for the reference
amount, and define to which quantity it refers for application to the reactor of interest.

Let’s also generalise our reaction equation to the following

2𝐴 + 𝐵 → 2𝐶

Let 𝑋𝐴 be conversion of component 𝐴:

𝑁𝐴,𝑟𝑒𝑓 − 𝑁𝐴 𝑁𝐴
𝑋𝐴 = =1−
𝑁𝐴,𝑟𝑒𝑓 𝑁𝐴,𝑟𝑒𝑓

If the volume is the same before and after the reaction, clearly

𝑁𝐴,𝑟𝑒𝑓 − 𝑁𝐴 𝑐𝐴
𝑋𝐴 = =1−
𝑁𝐴,𝑟𝑒𝑓 𝑐𝐴,𝑟𝑒𝑓

We may then write

𝑐𝐴 = 𝑐𝐴,𝑟𝑒𝑓 (1 − 𝑋𝐴 )

We may also note that the change in concentration of 𝐵 is twice the change in concentration
of 𝐴 as follows:

2(𝑐𝐵,𝑟𝑒𝑓 − 𝑐𝐵 ) = 𝑐𝐴,𝑟𝑒𝑓 − 𝑐𝐴 = 𝑐𝐴,𝑟𝑒𝑓 . 𝑋𝐴

Or

𝑋𝐴
𝑐𝐵 = 𝑐𝐵,𝑟𝑒𝑓 − 𝑐𝐴,𝑟𝑒𝑓 .
2

2 |Page
We also recognise that the amount of 𝐶 formed is equal to the amount of 𝐴 converted:

𝑐𝐶 − 𝑐𝐶,𝑟𝑒𝑓 = 𝑐𝐴,𝑟𝑒𝑓 − 𝑐𝐴 = 𝑐𝐴,𝑟𝑒𝑓 . 𝑋𝐴

or

𝑐𝐶 = 𝑐𝐶,𝑟𝑒𝑓 + 𝑐𝐴,𝑟𝑒𝑓 . 𝑋𝐴

We propose also to define

𝑐𝑖,𝑟𝑒𝑓
Θ𝑖 =
𝑐𝐴,𝑟𝑒𝑓

Generally speaking, we have then

𝜈𝑖
𝑐𝑖 = 𝑐𝐴,𝑟𝑒𝑓 (Θ𝑖 − 𝑋 )
𝜈𝐴 𝐴

This relationship is generally true for any component of any reaction. (Test whether it is for
the example at hand.) We may now write

3 𝑋𝐴
−𝑟𝐴 = 𝑘. 𝑐𝐴2 . 𝑐𝐵 = 𝑘. 𝑐𝐴,𝑟𝑒𝑓 (1 − 𝑋𝐴 )2 (Θ𝐵 − )
2
We have then, for the Batch Reactor:

𝑑𝑋𝐴 1 3 (1
𝑋𝐴
=− 𝑘. 𝑐𝐴0 − 𝑋𝐴 )2 (Θ𝐵 − )
𝑑𝑡 𝑐𝐴0 2

For the CSTR:

𝑋𝐴 1 3 𝑋𝐴
=− 𝑘. 𝑐𝐴,𝑖𝑛 (1 − 𝑋𝐴 )2 (Θ𝐵 − )
𝜏 𝑐𝐴,𝑖𝑛 2

For the PFR:

𝑑𝑋𝐴 1 3 𝑋𝐴
=− 𝑘. 𝑐𝐴,𝑖𝑛 (1 − 𝑋𝐴 )2 (Θ𝐵 − )
𝑑𝜏 𝑐𝐴,𝑖𝑛 2

We can then solve these equations to obtain the conversion of 𝑋𝐴 in the reactor. We note that
Θ𝐵 is required data; this is typically given to us with the problem or design specification (along
with 𝑐𝐴,𝑟𝑒𝑓 ). If we want to know the concentrations of the other species, now that we know
𝑋𝐴 , we can simply use
𝜈𝑖
𝑐𝑖 = 𝑐𝐴,𝑟𝑒𝑓 (Θ𝑖 − 𝑋 )
𝜈𝐴 𝐴

We can therefore always translate between the concentration and the conversion, and we can
exploit this relationship to obtain a simpler design equation.

Exercise: Obtain the concentrations for all three species for the 𝑆𝑂2 problem.

3 |Page
3.3 G as phase reactions
We have to far used moles and conversion somewhat interchangeably by assuming that the
volume is constant. In a gas phase reaction, however, even if the reaction volume is constant,
the pressure may change. For example, in the example

2𝐴 + 𝐵 → 2𝐶

the pressure will reduce over time as three moles are increasingly converted to two. In the case
of the batch reactor, this pressure drop would be experienced as a loss of concentration; in the
case of the CSTR and PFR, this would affect the pressure in the gas and hence the system
hydrodynamics and flow patterns.

B A TC H R EA C TOR
Let’s formulate the mole fraction as a function of conversion. Let the initial concentrations of
all the species be 𝑐𝑖0 and the total reactor volume be 𝑉. The initial number of moles is then
𝑁𝑖0 and, assuming that 𝐴 is the reference component, the initial mole ratio is Θ𝑖 = 𝑁𝑖0 /𝑁𝐴0 .

We note that since 𝐴 is the reference component, Θ𝐴 = 1. Also, for species which are not
initially present (e.g. products), Θ𝑖 = 0.

Initial C hange Final

𝐴 𝑁𝐴0 = Θ𝐴 . 𝑁𝐴0 −𝑋𝐴 𝑁𝐴0 𝜈𝑖
𝑁𝐴0 (Θ𝐴 − 𝑋𝐴 ) = 𝑁𝐴0 (Θ𝑖 − 𝑋 )
𝜈𝐴 𝐴
𝐵 𝑁𝐵0 = Θ𝐵 . 𝑁𝐴0 1 𝑋𝐴 𝜈𝑖
− 𝑋𝐴 𝑁𝐴0 𝑁𝐴0 (Θ𝐵 − ) = 𝑁𝐴0 (Θ𝑖 − 𝑋𝐴 )
2 2 𝜈𝐴
𝐶 𝑁𝐶0 = Θ𝐶 . 𝑁𝐴0 +𝑋𝐴 𝑁𝐴0 𝜈𝑖
𝑁𝐴0 (Θ𝐴 + 𝑋𝐴 ) = 𝑁𝐴0 (Θ𝑖 − 𝑋𝐴 )
𝜈𝐴
Total 𝜈𝑖 𝜈𝑖
𝑁𝐴0 ∑ Θ𝑖 −𝑁𝐴0 ∑ 𝑋 𝑁𝐴0 ∑ (Θ𝑖 − 𝑋𝐴 )
𝜈𝐴 𝐴 𝜈𝐴
𝑖 𝑖 𝑖

Generally then, the mole fraction for any component is given

𝜈
(Θ𝑖 − 𝜈 𝑖 𝑋𝐴 )
𝐴
𝑦𝑖 = 𝜈𝑖
∑𝑖 (Θ𝑖 −
𝜈𝐴 𝑋𝐴 )

So this all proceeds naturally from the generalised mole balance and this result could be
anticipated without the details shown here. For flow reactors, however, the situation is not
quite so simple.

4|Page
C STR , P FR
In the case of the gas-phase flow reactors, the exit flowrate is driven by the pressure. Since the
gas phase reaction rate influences the pressure, we have to pay more attention to the flow
model, and recognise that the inlet and exit flowrate may need elaboration.

Constant pressure

If we assume that the pressure across the reactor remains constant, then any increase or
decrease in pressure due to the reaction will tend to change this pressure. The starting point
to this problem is to estimate the pressure at the inlet point. For the pressure is less than 15
atm, in a batch system, we may write

𝑝. 𝑉 = 𝑛. 𝑅. 𝑇

For a flow system, at the inlet and exit points, we write instead

𝑝. 𝑄𝑖𝑛 = 𝐹𝑖𝑛 . 𝑅. 𝑇 = ∑ 𝐹𝑖,𝑖𝑛 . 𝑅. 𝑇𝑖𝑛

𝑖

𝑝. 𝑄𝑜𝑢𝑡 = 𝐹𝑜𝑢𝑡 . 𝑅. 𝑇 = ∑ 𝐹𝑖,𝑜𝑢𝑡 . 𝑅. 𝑇𝑜𝑢𝑡

𝑖

Noting that for a flow system, we may write the equivalent flowrate (compared with the Batch
case) as
𝜈𝑖
𝐹𝑜𝑢𝑡 = 𝐹𝐴,𝑖𝑛 ∑ (Θ𝑖 − 𝑋 )
𝜈𝐴 𝐴
𝑖

If the pressure is the same at the inlet and exit points, we have

𝐹𝑖𝑛 . 𝑅. 𝑇 𝐹𝑜𝑢𝑡 . 𝑅. 𝑇
𝑝= =
𝑄𝑖𝑛 𝑄𝑜𝑢𝑡

We may write this instead as

𝐹𝑜𝑢𝑡 . 𝑅. 𝑇𝑜𝑢𝑡
𝑄𝑜𝑢𝑡 = 𝑄𝑖𝑛
𝐹𝑖𝑛 . 𝑅. 𝑇𝑖𝑛

Under the isothermal assumption / case, 𝑇𝑜𝑢𝑡 = 𝑇𝑖𝑛 . We may also substitute our expressions
for the molar flows to obtain
𝜈 𝜈𝑖
𝐹𝐴,𝑖𝑛 ∑𝑖 (Θ𝑖 − 𝜈 𝑖 𝑋𝐴 ) ∑𝑖 (Θ𝑖 −
𝜈𝐴 𝑋𝐴 )
𝐴
𝑄𝑜𝑢𝑡 = 𝑄𝑖𝑛 = 𝑄𝑖𝑛
𝐹𝐴,𝑖𝑛 ∑𝑖 Θ𝑖 ∑𝑖 Θ𝑖

We may therefore identify the volume expansion factor as

𝜈𝑖
𝑄𝑜𝑢𝑡 ∑𝑖 (Θ𝑖 − 𝜈𝐴 𝑋𝐴 )
=
𝑄𝑖𝑛 ∑𝑖 Θ𝑖

..and so modify our development for non-equimolar gas-phase reactions.

5|Page
3.4 Liquid phase reaction volum e change
Since liquids are consider approximately incompressible under the conditions of most chemical
reactors, and because the reaction itself doesn’t change the liquid volume significantly in most
cases, volume change in the liquid case is not considered when the flowrates are equal.

When liquid flowrates are not equal, however, the reaction volume will change, and this will
influence the material balance. It is important to note that this is not really a concern at the
design phase of a reactor, since the design is for average flowrates and averaged operations, so
a single flowrate value may be adopted.

In real reactors being operated, however, the situation is very different. Flowrates are moving
all the time as operators and control systems attempt to maintain particular control targets
or respond to changes in product specification. This topic is explored in more detail in the
controls course. For now, it is still instructive for us to consider how to perform the material
balance under conditions of flowrates not being equal and/or changing with time.

Clearly, the mass in the reactor is influenced by these flowrates, and we may generally write

𝑑𝑚
= 𝑚̇𝑖𝑛 − 𝑚̇𝑜𝑢𝑡
𝑑𝑡
In terms of density

𝑑𝜌𝑉
= 𝜌𝑖𝑛 . 𝑄𝑖𝑛 − 𝜌𝑜𝑢𝑡 . 𝑄𝑜𝑢𝑡
𝑑𝑡
Assuming that the densities are the same in the reactor and in the inlet and exit streams:

𝜌 = 𝜌𝑖𝑛 = 𝜌𝑜𝑢𝑡

Hence

𝑑𝑉
= 𝑄𝑖𝑛 − 𝑄𝑜𝑢𝑡
𝑑𝑡
We now return to the CSTR balance assuming that only perfect mixing:
𝑑𝑐𝑖 . 𝑉
= 𝑐𝑖,𝑖𝑛 . 𝑄𝑖𝑛 − 𝑐𝑖 . 𝑄𝑜𝑢𝑡 + 𝑟𝑖 . 𝑉
𝑑𝑡
This time we must acknowledge that the volume is changing with time, and that we can’t
simply divide left and right by 𝑉. Instead, we expand using the product rule as follows:

𝑑𝑉 𝑑𝑐𝑖
𝑐𝑖 +𝑉 = 𝑐𝑖,𝑖𝑛 . 𝑄𝑖𝑛 − 𝑐𝑖 . 𝑄𝑜𝑢𝑡 + 𝑟𝑖 . 𝑉
𝑑𝑡 𝑑𝑡
We can now substitute the overall mass balance as follows

𝑑𝑐𝑖
𝑐𝑖 (𝑄𝑖𝑛 − 𝑄𝑜𝑢𝑡 ) + 𝑉 = 𝑐𝑖,𝑖𝑛 . 𝑄𝑖𝑛 − 𝑐𝑖 . 𝑄𝑜𝑢𝑡 + 𝑟𝑖 . 𝑉
𝑑𝑡

6 |Page
We can then group the flow terms as

𝑑𝑐𝑖
𝑉 = (𝑐𝑖,𝑖𝑛 − 𝑐𝑖 ). 𝑄𝑖𝑛 + 𝑟𝑖 . 𝑉
𝑑𝑡
And so

𝑑𝑐𝑖 𝑄𝑖𝑛
= (𝑐𝑖,𝑖𝑛 − 𝑐𝑖 ). + 𝑟𝑖
𝑑𝑡 𝑉
Therefore, in the case of the unsteady flow, we must solve the following balances
simultaneously:

𝑑𝑉
= 𝑄𝑖𝑛 (𝑡) − 𝑄𝑜𝑢𝑡 (𝑡)
𝑑𝑡
𝑑𝑐𝑖 𝑄𝑖𝑛 (𝑡)
= (𝑐𝑖,𝑖𝑛 − 𝑐𝑖 ). + 𝑟𝑖
𝑑𝑡 𝑉(𝑡)

As you will find in any controls course, we must then deduce the functions 𝑄(𝑡) which meet
the control objectives that we define.

7 |Page