𝑒: Stream

! / 2nd_Year / RE1 / 1_Isothermal_Reactors

exiting the
reactor

1. Isothermal Reactors 0 : Stream

entering the
𝑛𝐴,0 − 𝑛𝐴,𝑓 system
Total Conversion of the
𝜒=
System 𝑛𝐴,0
Equations 𝑓 : Stream
exiting the
system
Name Equation Notes

𝑉𝑅 : Reactor
volume

𝑁𝐴 : Moles of
Introduction to Reaction Engineering
Design Equation for an Ideal 1 𝑑𝑁𝐴
𝑟𝐴 = − A in reactor
Batch Reactor 𝑉𝑅 𝑑𝑡
Whenever looking at any process, a mass balance has to be made:
𝑟𝐴 : Rate of
reaction for "
reactant A
Input - Output + Production - Consumption = Accumulation

𝑛𝐴0 − 𝑛𝐴 𝑛𝐴 : Molar flow

Design Equation for an Ideal
𝑉𝑅 = rate of A in
CSTR 𝑟𝐴 Generally speaking, the input and output terms will just be how much is initially being fed into and out of the
reactor reactor; the consumption term will depend on the reaction kinetics and whether the species is a reactant or
product. It is important to remember that each term in this equation must be in mol/time. Although the
equations required to model these systems are normally quite difficult, a few simplifications can usually be
Design Equation for an Ideal 𝑑𝑛𝐴
𝑟𝐴 = − made:
PFR 𝑑𝑉𝑅

For all the reactors below, it is assumed that they are isothermal and isobaric (which will keep
Reciprocal is the rate constant k constant). A few assumptions for the isothermal reactors are:
Residence Time / Space 𝑉𝑅
𝜏= the space
Time 𝑣𝑇 1. Not consider any mass transfer limitations or non-ideal flow patterns (for now);
velocity
2. Reactants fully charged and ready for reaction. Reaction starts instantaneously at t =
The latter 0 with the reaction conditions. Does not consider time required to load reactants or heat

𝑛𝐴,0 − 𝑛𝐴 𝑁𝐴,0 − 𝑁𝐴 𝐶𝐴,0 − 𝐶𝐴 assume " up the reactor.

Conversion 𝑥𝐴 = = = constant
𝑛𝐴,0 𝑁𝐴,0 𝐶𝐴,0 3. Neglect heat release or uptake by the reaction.
volumetric
flow rate 4. Neglect heat transfer between parts of the reactor and/or heat exchange with
surroundings.

𝑖 : Stream
entering the
𝑛𝐴,𝑖 − 𝑛𝐴,𝑒 reactor
𝑥𝐴, pass =
Conversion Per Pass 𝑛𝐴,𝑖
These will apply to homogeneous reactions or reactions with average reaction rates. " 1 𝑑𝑁𝐴
𝑟𝐴 = −
𝑉𝑅 𝑑𝑡

Mass Balance for a Batch Reactor

A batch reactor is where we simply load a reactor up with the reactants (and sometimes other components The negative sign ensures r is always positive. It is also possible to derive this from the product convention
like inerts) and allow the reaction to go on up to a certain point. It is assumed that the reactor is uniform at by simply changing the sign:
all times, meaning that the same amount is reacted and produced at all points (strictly speaking, this is not
true as dead spaces do occur).
1 𝑑𝑁𝐵
𝑟𝐵 =
𝑉𝑅 𝑑𝑡

It is also important to remember that, for more complex reaction kinetics, rA and rB must include all
reactions that either A or B is involved in (if they are consumed, give it a -, otherwise it is +). This will be true
for all reactors.

Mass Balance for a Continuous Stirred Tank Reactor (CSTR)

1. Batch Reactor

Looking at the mass balance, we realise that there are no inputs or output in the reactor, thus the mass
balance for reactants simplifies to:

Production - Consumption = Accumulation

Apply the mass balance in the volume of reactor VR , filled with reactant A, over the time period (t) to (t +
Δt):

(𝑀𝐴 𝑟𝐴 𝑉𝑅 ) Δ𝑡 = (𝑀𝐴 𝑁𝐴 )𝑡 − (𝑀𝐴 𝑁𝐴 )𝑡+Δ𝑡

2. CSTR
Where 𝑀𝐴 is the molecular weight of 𝐴 ( 𝑔 𝑚𝑜𝑙−1 ), 𝑁𝐴 is the number of moles of reactant 𝐴 within the
This is a reactor where a feed comes into the reactor, reacts, and comes out at the outlet. In principle, the
system at time 𝑡 (𝑚𝑜𝑙 ) and 𝑟𝐴 is the net reaction rate of 𝐴 (𝑚𝑜𝑙 𝑚−3 𝑠−1 ). Dividing by 𝑀𝐴 and take the
composition should change throughout the reactor. However, this presents some mathematical difficulties,
limit Δ𝑡 approaches 0 :
thus it is assumed that the reactor has perfect mixing: The moment the feed enters the reactor, it is mixed
instantly and reacts immediately, giving the tank a uniform concentration distribution.
𝑑𝑁𝐴
𝑟𝐴 𝑉𝑅 = −
𝑑𝑡
# CSTRs always operate at outlet conditions (concentrations, temperature etc.)

Rearranging gives the design equation for the batch reactor immediately:

Also, to further simplify the calculations, steady state operation is assumed. This is reasonable since we
Design equation for an ideal batch reactor: are only interested in steady state when considering the design and sizing of the reactor as the conditions
reactors mostly operate at will be steady state. Thus our mass balance becomes:

Input - Output - Consumption = 0

𝑀𝐴 𝑛𝐴0 − 𝑀𝐴 𝑛𝐴 − 𝑀𝐴 𝑟𝐴 𝑉𝑅 = 0
3.2. PFR infintesimal slice

With an inlet molar flowrate nA,0 and outlet nA . Divide through by MA : Taking an infinitesimal slice, we can assume that it does behave at steady state. Thus, the mass balance
becomes:
𝑛𝐴,0 − 𝑛𝐴 − 𝑟𝐴 𝑉𝑅 = 0
Input - Output - Consumption = 0
Design equation for an ideal CSTR:

" 𝑛𝐴,0 − 𝑛𝐴 (𝑀𝐴 𝑛𝐴 )𝑉 − (𝑀𝐴 𝑛𝐴 )𝑉 +Δ𝑉 − 𝑀𝐴 𝑟𝐴 Δ𝑉𝑅 = 0

𝑉𝑅 =
𝑟𝐴
Dividing by MA and taking the limit as ΔV approaches 0:

𝑑𝑛𝐴
This equation is much simpler than the batch since it does not carry any differentials. However, it can still − − 𝑟𝐴 = 0
𝑑𝑉𝑅
become complex depending on the reaction kinetics. Also, when looking at it from the product convention,
simply multiply it by -1.
Rearranging:

Mass Balance for a Plug Flow Reactor (PFR) Design equation for an ideal PFR:

This is where the feed enters at one end of the reactor and passes through (maybe a catalyst), gradually
" 𝑑𝑛𝐴
reacting throughout. This makes it quite difficult as we cannot treat it as steady state since at each point, 𝑟𝐴 = −
the composition varies: 𝑑𝑉𝑅

Integrating this expression gives the reactor volume and change in the reactant composition.

3. PFR
Reaction Kinetics
A convenient way of looking at PFR is to visualize each plug as traveling through the tubular
The classic first-order reaction can be written as:
reactor as a small batch reactor. Over any cross-section normal to the reactor axis, the mass
flowrate, flow velocity, pressure, temperature and compositions are uniform and constant
# (batch reactor assumptions). Each plug of fluid is assumed to go through the same 𝑟𝐴 = 𝑘𝐶𝐴
temperature, pressure and concentration histories. Residence time in PFR is exactly equal to
the reaction time in batch reactor. Where k is the rate constant in s−1 . Sometimes the rate equation is expressed in a partial pressure form:
 𝑛𝐴 Changing Volumetric Flow Rates
𝑟𝐴 = 𝑘𝑃𝐴 , 𝑃𝐴 = 𝑃
𝑛𝑇 𝑇
There are many ways that this can occur, primarily in the gas phase reactions as liquids tend to be
incompressible, thus volume will remain constant. We can see this in the ideal gas law:
Where k is in mol m−3 s−1 Pa−1 . This is normally used when dealing with gas-phase reactions where one
can invoke the ideal gas laws.
𝑃 𝑣𝑇 = 𝑛𝑇 𝑅𝑇

Units are incredibly important through Reaction Engineering, especially for the rate constant
The volumetric flowrate is constant for reactions in the liquid phase (incompressibility), but
k. The units of k gives vital information about the reaction kinetics and its nature (order of
$ # may vary for gaseous reactions.
reaction, whether reaction is carried in gaseous phase etc). Always check the units of k
before answering the problems.

Assuming constant P and T, we can see if the number of moles increases, so will the volume. A simple
case of this would be:
One issue with either of these forms is that they are not expressed in NA or nA . There will also be times
when a particular conversion is required rather than an outlet flow rate:
𝐴 → 2𝑆
𝑛𝐴,0 − 𝑛𝐴 𝑁𝐴,0 − 𝑁𝐴 𝐶𝐴,0 − 𝐶𝐴
𝑥𝐴 = = =
𝑛𝐴,0 𝑁𝐴,0 𝐶𝐴,0 We shall look at this from both the CSTR and PFR’s perspectives as these are the only reactors which
involve flow rates. In both cases, as shown above, the total molar flow rate needs to be considered:
The last definition is only valid if the volumetric flow rate vT in the reactor is constant. Because of this and
depending on the reactor, a few transformations are required: 𝑛𝑇 = 𝑛𝐴 + 𝑛𝐼 + 𝑛𝑆

Batch CSTR PFR

Where A is the reactant, S is the product and I is an inert. Substitute in conversions:

Conversion to 𝑥𝐴 𝑁𝐴 = 𝑁𝐴,0 (1 − 𝑥𝐴 ) 𝑛𝐴 = 𝑛𝐴,0 (1 − 𝑥𝐴 ) 𝑛𝐴 = 𝑛𝐴,0 (1 − 𝑥𝐴 ) 𝑛𝑇 = 𝑛𝐴,0 (1 − 𝑥𝐴 ) + 2𝑥𝐴 𝑛𝐴,0 + 𝑛𝐼,0 = 𝑛𝐴,0 (1 + 𝑥𝐴 ) + 𝑛𝐼,0

Converting the
𝑑𝑁𝐴 = −𝑁𝐴,0 𝑑𝑥𝐴 − 𝑑𝑛𝐴 = −𝑛𝐴,0 𝑑𝑥𝐴 𝑛𝑇 = 𝑛𝑇 ,0 + 𝑛𝐴,0 𝑥𝐴
differentials of 𝑥𝐴

Generally, nA,0 and nT,0 will be known.

Conversion to
𝑁𝐴 = 𝐶𝐴 𝑉𝑅 𝑛𝐴 = 𝐶𝐴 𝑣𝑇 𝑛𝐴 = 𝐶𝐴 𝑣𝑇
concentration
Changing Volumetric Flow Rates in a PFR
Converting the
differentials of 𝑑𝑁𝐴 = 𝑉𝑅 𝑑𝐶𝐴 − 𝑑𝑛𝐴 = 𝑣𝑇 𝑑𝐶𝐴 From the design equation:
concentration
𝑑𝑛𝐴
𝑟𝐴 = −
𝑑𝑉𝑅
It is important to remember that the above conversions assume that the reactor volume VR and the
volumetric flowrate vT are constant throughout the reactors, which are not always the case.
Consider a simple first order reaction, where the rate is given in reactant concentrations:
 𝑟𝐴 = 𝑘𝐶𝐴
Changing Volumetric Flow Rates in a CSTR
Substitute in nA for CA : From the design equation:

𝑘𝑛𝐴 𝑛𝐴,0 − 𝑛𝐴
𝑟𝐴 = 𝑉𝑅 =
𝑣𝑇 𝑟𝐴

Thus: The presence of the changing volumetric flow rate may not be apparent outright in the equation until we
look at the reaction rate. Consider a first order reaction, where the rate is given in reactant partial pressure:
𝑘𝑛𝐴 𝑑𝑛𝐴
=− 𝑛𝐴
𝑣𝑇 𝑑𝑉𝑅 𝑟𝐴 = 𝑘𝑃𝐴 = 𝑘 𝑃
𝑛𝑇 𝑇
However, vT is dependent on nA by the ideal gas law:
Always start with design equation, carefully convert the terms and check if any term is
$ dependent upon the reaction coordinate.
𝑛 𝑅𝑇
𝑣𝑇 = 𝑇
𝑃

As shown previously, nT can be expressed as a function of conversion: The equation derived previously for nT still holds in this case:

𝑛𝐴
𝑛𝑇 = 𝑛𝑇 ,0 + 𝑛𝐴,0 𝑥𝐴 𝑟𝐴 = 𝑘 𝑃
𝑛𝑇 ,0 + 𝑛𝐴,0 𝑥𝐴 𝑇

𝑅𝑇 Convert all the variables to the conversion convention, and simplify the denominator by accounting for how
𝑃 (
∴ 𝑣𝑇 = 𝑛𝑇 ,0 + 𝑛𝐴,0 𝑥𝐴 ) much A was in the initial feed (yA,0 ):

𝑛𝐴,0 (1 − 𝑥𝐴 )
𝑘𝑃 𝑛𝐴 𝑑𝑛𝐴 𝑟𝐴 = 𝑘 𝑃𝑇
𝑅𝑇 𝑛𝑇 ,0 + 𝑛𝐴,0 𝑥𝐴
=−
𝑑𝑉𝑅 𝑛𝑇 ,0 (1 + 𝑦𝐴,0 𝑥𝐴 )

Thus:
However, we now have x~A ~ on the LHS. Convert all the variables to the conversion convention:

𝑘𝑃 𝑛𝐴,0 (1 − 𝑥𝐴 ) 𝑑𝑥𝐴 𝑛𝐴,0 𝑥𝐴 𝑛𝑇 ,0 𝑥𝐴 (1 + 𝑦𝐴,0 𝑥𝐴 )

= 𝑛𝐴,0 𝑉𝑅 = =
𝑅𝑇 𝑛𝑇 ,0 + 𝑛𝐴,0 𝑥𝐴 𝑑𝑉𝑅 𝑟𝐴 (1 − 𝑥𝐴 ) 𝑘𝑃𝑇

With a bit of rearranging:

𝑉𝑅
𝑅𝑇 𝑥 𝐴,𝑒
𝑛𝑇 ,0 + 𝑛𝐴,0 𝑥𝐴 Residence Time
∫0 𝑘𝑃 ∫0
𝑑𝑉𝑅 = 𝑑𝑥𝐴
1 − 𝑥𝐴
One key parameter used when evaluating reactors is the residence time, which is essentially
This will give us the volume of our PFR once we integrate both sides.
 how long it takes for the component to move through the reactor:
"
𝑉𝑅
𝜏=
𝑣𝑇

However, changing volumetric flow rates causes a bit of a problem as the value will be
different throughout the reactor. To resolve this, space time is defined:

" 𝑉𝑅
𝜏=
𝑣𝑇 ,0

The reciprocal of this is known as the space velocity.

5. Reciprocal of Rate of Reaction against Conversion

Looking at the design equations (with some slight rearrangement):

𝑛𝐴,0 − 𝑛𝐴 𝑛𝐴,0
Comparison Between PFR and CSTR CSTR : 𝑉𝑅 =
𝑟𝐴
= 𝑥
𝑟𝐴 𝐴,𝑒

Both being continuous reactors, it is often important to decide which is best suited to a particular scenario.
To start, it would be best to look at how the rate of reaction changes with the conversion, depending on the 𝑥 𝐴,𝑒
𝑛𝐴,0
∫0
reaction order (for a simple reaction): PFR : 𝑉𝑅 = 𝑑𝑥𝐴
𝑟𝐴

We realise that the volume of the CSTR is represented by the red square shown above, whilst the PFR is the
area below the curve. Clearly, for positive orders of magnitude, the volume of the PFR will always be less
than the one for the CSTR (comes close for lower conversions).

This is because, with a same positive order reaction and same volumetric flowrate, since in
PFR the reaction starts out at high concentration, the driving force is higher than that of the
CSTR whose concentration is the same as outlet concentrations. The average rate in PFR is
higher than the CSTR and outperforms it. The only case where the CSTR beats the PFR is in
# negative rate orders as the curve then flips such that the area under the curve is less than the
area of the square. However, CSTRs are still used in industry where rate orders tend to be
positive; to minimise the total volume multiple reactors are used (sometimes a combination
of CSTRs and PFRs).

4. Rate of Reaction against Conversion at Various Reaction Orders

As expected, the higher the conversion, the slower the reaction for all cases. However, depending on the
reaction order, they do so with different decreasing rates, given that higher orders indicate higher
dependencies on the reactant. To better analyse it in the context of reactors, it would be best to take the CSTR Cascade
reciprocal of the rate and multiply it by the initial flow rate (same shape for all of them):
Two CSTRs in Series
𝐶𝐴,1 𝐶𝐴,0
𝐶𝐴,2 = =
(1 + 𝑣 𝑇 ) (1 𝑣𝑇 )
Generally, a desired conversion by the end of the process (xAe ) will be given. The only thing we have left to 𝑘𝑉𝑅2 𝑘𝑉𝑅1 𝑘𝑉𝑅2
1+ 𝑣𝑇 +
set is the exit conversion of the first CSTR which will become the entrance conversion of the next:

This is essentially our equation for two CSTR reactors.

N CSTRs in Series
Imagine a scenario where the total volume of all the reactors is VR and they are all evenly sized:

6. Two CSTRs in Series Configuration

8. N CSTRs of the Same Volume in Series Configuration

Apply the same equation as shown above:

𝐶𝐴,0 𝐶𝐴,1 𝐶𝐴,𝑁−1

𝐶𝐴,1 = 𝑘𝑉𝑅
; 𝐶𝐴,2 = 𝑘𝑉𝑅
; 𝐶𝐴,𝑁 = 𝑘𝑉𝑅
1+ 𝑁𝑣𝑇 1+ 𝑁𝑣𝑇 1+ 𝑁𝑣𝑇
7. Total Reactor Volume with (left) and without (right) CSTR Series Configuration
Do the same substitution as above but all the way to the N𝑡ℎ equation:
From the graphical construction, if instead of one CSTR, two in series is used, the volume required
decreases. This is because CSTR operates at outlet concentration. The concentration of reactant A in the
first CSTR is higher, so the average rate for the cascade is higher and total volume required is lower. 𝐶𝐴,0
𝐶𝐴,𝑁 =
(1 + )
𝑘𝑉𝑅 𝑁
For ease of analysis, we will assume isothermal operation with a first order reaction and constant 𝑁𝑣𝑇

volumetric flow rate. Thus, from the design equation:

Interestingly, as N goes to infinity, since that:
𝑛𝐴,0 − 𝑛𝐴 𝐶𝐴,0 − 𝐶𝐴,1 𝐶𝐴,0
𝑉𝑅1 = = 𝑣𝑇 → 𝐶𝐴,1 =
𝑟𝐴 𝑘𝐶𝐴,1 𝑘𝑉 𝑥 𝑁
1 + 𝑣𝑇𝑅1 𝑒𝑥 = lim (1 + )
𝑁→∞ 𝑁
Similarly, for the second reactor:
Thus:
 Thus, the total volume is:

𝑒 ( )
𝑘𝑉 𝑅
− −𝑘𝜏
𝐶𝐴,𝑒 = 𝐶𝐴,0 𝑣𝑇
= 𝐶𝐴,0 𝑒

𝑘 [ 1 − 𝑥𝐴1 ]
𝑣𝑇 𝑥𝐴1 𝑛𝐴,0
𝑉𝑇 = 𝑉𝑅1 + 𝑉𝑅2 = + (1 − 𝑥𝐴1 ) − 1
𝑛𝐴,𝑒
This is exactly the same equation we would get if we had a PFR for a first order reaction:

We want to figure out the value of xA1 that will minimise the total volume i.e. find the local minima:
𝑑𝑛𝐴 1 𝑣𝑇
𝑟𝐴 = − → 𝑑𝑉𝑅 = − 𝑑𝑛𝐴 = − 𝑑𝐶
𝑑𝑉𝑅 𝑟𝐴 𝑘𝐶𝐴 𝐴
𝑑𝑉𝑇
=0
𝑑𝑥𝐴1
Integrating gives:

Thus:
𝑣𝑇 𝐶𝐴,𝑒 −
𝑘𝑉 𝑅

𝑉𝑅 = − ln → 𝐶𝐴,𝑒 = 𝐶𝐴,0 𝑒 𝑣𝑇 = 𝐶𝐴,0 𝑒−𝑘𝜏

𝑘 𝐶𝐴,0

[ 𝑛𝐴,𝑒 ]
𝑣𝑇 1 𝑛𝐴,0
− =0
𝑘 (1 − 𝑥𝐴1 )2
Which would make sense as, dividing the volume into many infinitesimally small CSTRs is essentially how
we model the PFR.

𝑛𝐴,𝑒
= 1 − 𝑥𝑇 = (1 − 𝑥𝐴1 )2
𝑛𝐴,0
Optimising CSTR Volumes
When we have multiple CSTRs, we would like to select the respective volumes such that they minimise the Here, xT is the total desired conversion (given). Since xA1 must be less than 1:
total volume. This cannot be easily seen graphically:

𝑥𝐴1 = 1 − √‾1‾‾‾‾‾
− 𝑥𝑇‾

Substituting into each of the equation for the reactor volumes:

𝑣𝑇 1 − √‾1‾‾‾‾‾
− 𝑥𝑇‾
𝑉𝑅1 =
𝑘 √‾1‾‾‾‾‾
− 𝑥𝑇‾

𝑣𝑇 √‾1‾‾‾‾‾
− 𝑥𝑇‾ √‾1‾‾‾‾‾
− 𝑥𝑇‾ √‾1‾‾‾‾‾
− 𝑥𝑇‾√‾1‾‾‾‾‾
− 𝑥𝑇‾
𝑘 [ 1 − 𝑥𝑇 1 − 𝑥𝑇 ] 𝑘 [ √‾1‾‾‾‾‾ − 𝑥𝑇‾ ]
9. Optimising Total CSTR Volume Graphically 1 − 𝑥𝑇 𝑣𝑇
𝑉𝑅2 = − = −
− 𝑥𝑇‾√‾1‾‾‾‾‾
− 𝑥𝑇‾ √‾1‾‾‾‾‾
− 𝑥𝑇‾√‾1‾‾‾‾‾
However, we can find it analytically. Restarting from our design equations (where xA1 is the conversion in
the first reactor and nA,e is the desired outlet flowrate of A):

𝑣𝑇 1 − √‾1‾‾‾‾‾
− 𝑥𝑇‾
𝑛𝐴,0 − 𝑛𝐴 𝑛𝐴,0 𝑥𝐴 𝑣𝑇 𝑥𝐴1 𝑉𝑅2 =
𝑉𝑅1 = = = 𝑘 √‾1‾‾‾‾‾
− 𝑥𝑇‾
𝑟𝐴 𝑘𝐶𝐴 𝑘 1 − 𝑥𝐴1

Thus the optimal volume is when it is evenly distributed between either reactor. This would make sense as,

𝑘 [ 𝑛𝐴,𝑒 ]
𝑛𝐴 − 𝑛𝐴,𝑒 𝑣𝑇 𝑛𝐴,0 (1 − 𝑥𝐴1 ) − 𝑛𝐴,𝑒 𝑣𝑇 𝑛𝐴,0 when it comes to first order reactions, the rate of reaction decreases linearly. However, when it comes to
𝑉𝑅2 = = = (1 − 𝑥𝐴1 ) − 1
𝑘𝐶𝐴,𝑒 𝑘 𝑛𝐴,𝑒 different orders of reaction:
 11. PFR with Recycle

However, when dealing with recycles, a few new terms need to be introduced:

The recycle ratio is defined as:

Stream recycled
10. Ideal Optimum CSTR Sizing Sequence at Various Reaction Orders 𝛼=
Outlet stream of reactor

As for higher orders of reactions, a gradually increasing volume will mean that we minimise The single pass conversion (how much was reacted with the reactor along). This is
the volume required as higher rates of reaction will be reached at lower conversions. The essentially the conversion over the black square:
# inverse is true for lower orders of reaction where we need decreasing volume sizes for the
opposite reason. However, in practise, CSTRs come in standard sizes, making it difficult to
get the required CSTR sizes.
" 𝑥𝐴, pass =
𝑛𝐴,𝑖 − 𝑛𝐴,𝑒
𝑛𝐴,𝑖

The total conversion for the system is the conversion for the whole system (the green
square):

PFR with Recycle 𝑛𝐴,0 − (1 − 𝛼)𝑛𝐴,𝑒 𝑛𝐴,0 + 𝛼𝑛𝐴,𝑒 − 𝑛𝐴,𝑒 𝑛𝐴,𝑖 − 𝑛𝐴,𝑒
𝜒= = =
𝑛𝐴,0 𝑛𝐴,0 𝑛𝐴,0
Often, a single run through a PFR does not yield a very high conversion. Introducing a recycle stream allows
higher conversion rates.

When we take a mass balance at the junction where the feed and the recycle meet, we realise that the
above substitution for nA,i is possible. To make calculations easier, we need to find a link between the total
and single pass conversions:

𝜒 𝑛𝐴,0 = 𝑥𝐴,pass 𝑛𝐴,𝑖

 𝑛𝐴,0 𝑣𝑇 ,0 1 1
⟹ 𝑥𝐴,pass = 𝜒 𝑉𝑅 = ln
𝑛𝐴,𝑖 𝑘 1 − 𝛼 1 − 𝑥𝐴, pass

𝑛𝐴,0 𝑉𝑅 𝑘 1 1
∴ 𝑥A, pass = 𝜒 = 𝑁𝑅 = ln
𝑛𝐴,0 + 𝛼𝑛𝐴,𝑒 𝑣𝑇 ,0 1 − 𝛼 1 − 𝑥𝐴, pass

Here, we introduce 𝑁𝑅 as the Damkohler number (the comparison between reaction and convection),
𝑛𝐴,0 (1 − 𝜒)
𝑛𝐴,𝑒 = (rearranging the equation for 𝜒) which is system specific.
1−𝛼

Damkohler number:
𝑛𝐴,0 1
∴ 𝑥A, pass = 𝜒= 𝜒
𝑛𝐴,0 + 𝛼 ( "
1−𝛼 )
𝑛𝐴,0 (1−𝜒) 𝛼(1−𝜒) 𝑘𝑉𝑅
1+ 1−𝛼 𝑁𝑅 ≡
𝑣𝑇 0

The relationship between total and single pass conversions:

With a bit of rearranging:

(1 − 𝛼)𝜒
𝑥A, pass =
(1 − 𝛼𝜒)
" 𝑥𝐴, pass = 1 − 𝑒−𝑁𝑅 (1−𝛼)
This is derived purely based upon material balances and we have not considered any reaction
kinetics yet, so it always apply regardless. Single pass conversion is always smaller than the Substituting into our expression relating the single pass to the overall and rearranging:
global pass, in extreme case equal.

1 − 𝑒−𝑁𝑅 (1−𝛼)
𝜒=
1 − 𝛼𝑒−𝑁𝑅 (1−𝛼)
This will make it very easy to determine the single pass conversion and, thus, every other quantity.

If we look at our two limiting cases:

Single Pass Conversion as a Function of Recycle Ratio No recycle (𝛼 = 0)

From the design equation for an ideal PFR (taking this to be a first order reaction): The expression for the total conversion simplifies to:

𝑣𝑇 𝑥 𝐴, pass
1 𝑣𝑇 ,𝑖 1 𝑣𝑇 ,0 + 𝛼𝑣𝑇 ,𝑒 1 𝑛𝐴,0 − 𝑛𝐴,𝑒
𝑘 ∫0
𝑉𝑅 = 𝑑𝑥 = ln = ln 𝜒= = 1 − 𝑒−𝑁𝑅
1 − 𝑥𝐴 𝐴 𝑘 1 − 𝑥𝐴, pass 𝑘 1 − 𝑥𝐴, pass 𝑛𝐴,0

Given the reaction is in liquid phase, vT,e is equivalent to vT,i :

𝑛𝐴,𝑒 = 𝑛𝐴,0 𝑒−𝑁𝑅

1
𝑣𝑇 ,𝑒 = 𝑣 This is just the expression that we would normally get for a PFR without recycle (which we would expect).
1 − 𝛼 𝑇 ,0
Complete recycle (𝛼 = 1)
Thus:
 " In an exam, you may immediately treat a PFR with complete recycle as a CSTR Unsteady State
Unsteady State CSTR
The expression for the total conversion simplifies to:
Say we suddenly increase the inlet feed of the CSTR, whilst maintain volumetric flowrate (which can be
done by introducing an inert to compensate for changes in mass):
𝑛𝐴,0 − 𝑛𝐴,𝑒 1−1
𝜒= =
𝑛𝐴,0 1−1

We have a slight issue as this is undefined. However, using L’Hopital’s rule:

1 − 𝑒−𝑁𝑅 (1−𝛼) 𝑁𝑅 𝑒−𝑁𝑅 (1−𝛼) 𝑁𝑅

lim = lim =
𝛼→1 −
1 − 𝛼𝑒 𝑅𝑁 (1−𝛼) 𝛼→1 𝑒 −𝑁 𝑅 (1−𝛼) + 𝑁𝑅 𝑒−𝑁𝑅 (1−𝛼) 1 + 𝑁𝑅

With a bit of rearranging, we will realise that this is the expression we would get in the CSTR:

𝑛𝐴,0 − 𝑛𝐴,𝑒 𝑘𝜏
= → (𝑛𝐴,0 − 𝑛𝐴,𝑒 ) (1 + 𝑘𝜏) = 𝑘𝜏 𝑛𝐴,0
𝑛𝐴,0 1 + 𝑘𝜏 13. Step Change in Inlet Feed for a CSTR

The old design equation for the CSTR cannot be used since steady state assumption is invalid. Thus we
𝑛𝐴,0 − 𝑛𝐴,𝑒 must re-derive the design equation:
𝑛𝐴,0 − 𝑛𝐴,𝑒 − 𝑘𝜏 𝑛𝐴,𝑒 = 0 → 𝑉𝑅 =
𝑘𝑛𝐴,𝑒
𝑘𝑛𝐴 (𝑡) 𝑑𝑁𝐴
𝑛𝐴,0 − 𝑛𝐴 (𝑡) − 𝑉𝑅 =
Plot the conversion against recycle ratio: 𝑣𝑇 𝑑𝑡

𝑉𝑅
Use 𝑁𝐴 = 𝑣𝑡 𝑛𝐴 = 𝜏 𝑛𝐴 . This is a first order ODE:

+ 𝑛𝐴 (𝑡) [ + 𝑘] =
𝑑𝑛𝐴 1 𝑛𝐴,0
𝑑𝑡 𝜏 𝜏

Define a new constant 𝛽 :

𝛽 ≡ [ + 𝑘] ,
1 𝑑𝑛𝐴 𝑛𝐴0
+ 𝛽𝑛 𝐴 =
𝜏 𝑑𝑡 𝜏

The ODE can be solved by re-arranging to make 𝑡 the subject and integrating the RHS with respect to 𝑑𝑛𝐴 .
This gives:
12. Single Pass Conversion and Total Conversion against Recycle Ratio for a PFR with Recycle

1 𝑛𝐴0
𝑡 = − [ln( − 𝛽𝑛𝐴 )]
𝑛𝐴 (𝑡)
As we can see, although recycling means that we will have a lower conversion, it comes with the benefit
that we do not have to have too much of a high conversion in the reactor itself. 𝛽 𝜏 𝑛𝐴,𝑠𝑠
Thus:

𝑛𝐴 (𝑡) = (𝑛𝐴,𝑠𝑠 −
𝛽𝜏 )
𝑛𝐴0 𝑛𝐴0
𝑒−𝛽𝑡 +
𝛽𝜏

This equation does show the two competing factors in the exponent being the flow and the reaction. In
addition, if we look at the next steady state (time goes to infinity), we find that:

𝑛𝐴,0 15. Concentration Profile and Outlet Concentration for a PFR Operating at Steady State
𝑛𝐴 =
(1 + 𝑘𝜏) As the change moves along the reactor, since assuming no axial mixing, Ce is still measured:

This is the design equation we would get for a feed nA,0 , which is in line with all the calculations we have
done.

16. Concentration Profile and Outlet Concentration Before the Residence Time

After the residence time 𝜏 , the new concentration profile reaches the outlet and a new 𝐶𝑒𝐼 will be detected
instead.

14. Response of Outlet with Step Change in Inlet

Note that the time taken to reach this state is not necessarily infinite as, after just 5 residence times, the
exponent would have decreased to 0.007 without the reaction.

Unsteady State PFR

17. Concentration Profile and Outlet Concentration After the Residence Time
This is slightly more complex and will only be done qualitatively. If the inlet of the PFR is suddenly
increased, this will have a ripple effect throughout the PFR, until it reach the outlet.

At steady state, the outlet concentration is measured to be Ce all the time:

Review Questions
1. What are the considerations involved in designing a reactor?

2. What are the considerations involved in choosing batch or continuous operation?

3. How are reaction rates defined and how do you determine the reaction kinetics?

4. What are the underlying assumptions for ideal isothermal reactors?

 5. What is the mass balance in an isothermal batch reactor?

6. How do you determine the reaction time in a batch reactor?

7. What is the steady state mass balance in an isothermal CSTR?

8. What is the steady state mass balance in an isothermal PFR?

9. How do you size the continuous reactor from mass balance?

10. How is conversion defined and why is it important?

11. What is the main difference between reactions carried out in gaseous phase and liquid phase?

12. How do changes in molar flowrate affect the volumetric flowrate? Does this also affect the mass
balance equations of continuous reactors?

13. How is residence time defined and why is it important?

14. Compare the required sizes of CSTR and PFR under various reaction orders.

15. What are the advantages of having CSTRs in series and why? How can this be related to the nature of
PFRs?

16. For N CSTRs in series, what are the limiting behaviors of the system when N = 1 and N tends to infinity,
respectively?

17. How should the CSTRs be sequenced under various reaction orders? Is it favorable in practice?

18. What is the difference between conversion per pass and total conversion? How can they be related?

19. For PFR with recycle, what are the limiting behaviors of the system under no recycle and complete
recycle, respectively?

20. How is the CSTR affected by a step change in inlet concentration?

21. How is the PFR affected by a step change in inlet concentration?

Powered by Wiki.js