Mody and Marchildon: Chemical Engineering Process Design
Chapter 19 CHEMICAL REACTION P:/CEPDtxt2007/CEPDtextCh19
A chemical engineer may expect, at some point in her or his career, to have to arrange for
the conduct of a chemical reaction. In considering this challenge we will use a structure
based on the physical state of the components (reactants, catalysts) that may be involved.
However, first there are two general principles that the designer should follow in any
situation.
1. Find out the chemical characteristics of the reaction.
 is it reversible?
 what is the order of the reaction with respect to each of the reactants?
 what is the rate constant? What is the equilibrium constant?
 is it a simple reaction, following some simple path such as
A -----> B
or are there intermediate species between the feed material and the desired product?
 is the desired product subject to further, undesired reaction?
 are there other (undesirable) reactions, parallel to the reaction that produces the
desired product?
Ascertaining all of these facts may be very difficult. For a new reaction it may require
extensive experiments, which hopefully and ideally can be carried out at the laboratory
bench, i.e., at small scale. Understanding the chemistry is absolutely vital to successful
reactor design. The scale of a whole plant, for instance, can depend on the value of a
single kinetic constant.
2. Find out what are the physical phenomena that accompany the chemical reaction.
 is the reaction significantly endothermic, so that heat must be continuously supplied
in order to sustain its progress? Is there a volatile product or by-product that must be
continuously removed by vaporization?
 is the reaction significantly exothermic, so that heat must be continuously removed in
order to maintain temperature and to prevent run-away?
 is there a significant resistance to mass transfer of a reactant or product from one
phase to another?
 if a porous catalyst is involved, is the diffusion of components to and from the
reactive sites a limiting rate?
It is not unusual for the design of a particular reactor design to be focussed more on heat,
mass and momentum transfer, rather than on chemistry.
Seven reaction situations are examined:
1.
2.
3.
4.
5.
6.
7.
Gas phase
Liquid phase
Gas-liquid
Immiscible liquids
Fluid and solid, non-catalytic
Solid-catalyzed
Bio-reactions
Cusack R W (1999 October) ‘A Fresh Look at Reaction Engineering’, Chemical
Engineering p.134-146.
Cusack R W (1999 December) ‘Reaction Engineering - Part 2: Choosing the Right
Reactor’, Chemical Engineering p.80-85.
Cusack R W (2000 February) ‘Reaction Engineering - Part 3: Optimize Design and
Operation’, Chemical Engineering p.88-95.
Dutta S and Gualy R (2000 October) ‘Build Robust Reactor Models’, Chemical
Engineering Progress p.37-51.
Worstell J H (2000 June) ‘Succeed at Reactor Scale-Up’, Chemical Engineering Progress
p.55-60.
Worstell J H (2001 March) ‘Don’t Act Like a Novice about Reaction Engineering’,
Chemical Engineering Progress p.68-72.
19.1. Gas-Phase Reactions
We are asked to design a reactor to react a gas or a mixture of gases. Likely the choice
will be to operate at
 high temperature, because reaction occurs faster and because there is no need or
desire, as there would probably be in liquid-phase processing, to keep materials below
their boiling point, and
 high pressure, because reaction occurs faster when the volumetric concentrations of
the components are higher.
Gas-phase reactions are generally done in continuous plug-flow reactors, as opposed to
back-mixed vessels. Some possibilities are
1. a simple long tube, operated in turbulent flow in order to approximate plug flow and to
achieve good mixing and heat transfer,
2. a simple tube fully or partially filled with static mixers, to achieve even better
pluginess, mixing and heat transfer,
3. a vertical vessel filled with a packing (but the packing uses up reaction volume),
4. a vertical vessel fitted with periodic sieve-plate baffles, to achieve a staged, close-toplug-flow state,
5. a multi-tube vertical vessel to achieve more surface area if heat rejection is a major
concern, which it often is with gaseous reactions.
For reaction in any phase, plug flow is attractive because it makes the most efficient use
of reactor volume compared with any other continuous-flow configuration. Comparing,
for instance, with a back-mixed reactor, for a simple reaction
d[X] / d(time)
= - k x [X]
where the consumption of component "X" is first order in "X" and irreversible, then the
following conversions are achieved:
Kinetic constant x
Residence time
k x time
0
0.2
0.5
1
2
infinity
Plug-flow reactor
CONVERSION
1 - Exp(-k x time)
0
0.181
0.393
0.632
0.865
1
Fully back mixed reactor
CONVERSION
k x time / (1 + k x time)
0
0.167
0.333
0.5
0.667
1
Unfortunately perfect plug flow is an ideal that is never fully attained in practice. In the
flow in a tube there is always some degree of axial dispersion. This is obvious for
laminar flow because of the parabolic velocity profile; axial dispersion due to molecular
difffusion is very small. For turbulent flow, which has a much flatter velocity profile,
axial dispersion is due to the turbulent fluctuations themselves and may significantly
detract from the efficiency of a nominally plug-flow tubular reactor. Cusack (2000
February) provides the graphical relationships of Levenspiel(1993) that show how a tracer
test can be used to determine this loss of efficiency.
Levenspiel O (1999) Chemical Reaction Engineering, 3rd Edition, Wiley, New York
Streiff F A and Rogers J A (1994 June) ‘Don’t overlook static mixer reactors’, Chemical
Engineering p.76-82.
19.2. Liquid-Phase Reactions
One or more miscible liquids are to be reacted. The product or a by-product may be a
vapour but for purpose of this discussion the system is still considered liquid-phase. Also,
the case of immiscible liquids is considered later. Broadly the reactor choices are batch,
continuous plug-flow, and continuous back-mixed. These options are discussed here.
19.2.1. Batch Reactor and Fed-Batch Reactor
A batch reactor operates in a time-sequence of events: an amount of reactant is supplied,
it is converted chemically to product, and the product is discharged. Most new chemical
technologies start out as batch processes before ever being converted to continuous
processing. Batch processes have advantages, real or potential:
 relatively low capital cost
 ease of incremental expansion (build more reactors)
 flexibility for different products
 transparency and ease of control
and they have real or potential disadvantages
 generally higher labour cost (which may be relieved by modern sequencing control)
 batch-to-batch non-uniformity of product
 internal non-uniformity if sufficient agitation is not supplied for adequate mixing of
reactants or for adequate suspension of solid reactants or catalyst particles.
Batch reactors are sometimes operated in fed-batch mode, where additions are made to
the batch during the course of the process. The designation particularly applies to the
case where a significant amount of a reacting component is added. Some situations
where this approach is applicable are
 a reacting component is susceptible to an undesirable side-reaction or to volatilization
 the reaction is a large producer or consumer of heat, so it is desired to moderate its
rate
 a reaction by-product is continuously volatilized from the vessel, making room for
more liquid.
19.2.2. Continuous Plug-Flow Reactor
For the advantage and with the considerations explained in Section 3.10.2.1, plug-flow or
staged plug-flow is a candidate for liquid-phase as well as gas-phase reactions. If a
volatile product or by-product results from the reaction, it can be carried along with the
liquid or move ahead of it: prediction of the relative velocities of the two phases and
prediction of the in situ volume fraction occupied by liquid (and therefore available for
reaction) is the province of two-phase flow relationships..
19.2.3. Continuous Back-Mixed Reactor
The back-mixed reactor or Continuous Stirred Tank Reactor (CSTR) is very common in
chemical processing. It has advantages
 simple fluid dynamics, dependent only on the ability to be agitated
 easy to scale up
 uniformity, absence of hot spots
and disadvantages
 inefficient use of reactor volume because reactant concentration are all at the final
values and therefore low, minimizing the driving force
 possibility of inadequate mixing and therefore channeling and zones of stagnation.
The ‘Stirred’ in CSTR tends to imply a mechanical agitator or stirrer, which is frequently
the case. Such an arrangement, with its requirement for a rotating seal, may be
inadvisable when the pressure differs significantly from atmospheric and/or when it is
imperative that the contents of the vessel be isolated from the atmosphere. In these cases
an external pump-around loop, flowing at a rate of 10 (or more) times the net flow
through the vessel, or the introduction of a sparging gas may produce adequate mixing.
A very common configuration in chemical processing is the back-mixed reactor followed
by a separator which removes product and recycles un-reacted material to the reactor.
There are four circumstances where this configuration should be considered:
1. the reaction must be taken to a very high degree of conversion. This is difficult in a
CSTR, where the driving force for reaction becomes very low when the starting reactants
have been almost all consumed. With separation and recycle, the CSTR can be operated
at a modest conversion and still achieve high conversion by the system as a whole;
2. the reaction is reversible. If product is not removed then the reaction reaches a standstill;
3. the desired product is susceptible to further, un-wanted reaction;
4. the product causes poisoning of a catalyst.
One of the problems in scale-up of reactors is the provision (or removal) of heat. At the
bench scale or in a small pilot plant, the wall of the vessel may provide sufficient surface
area for heat transfer. As the scale gets larger, the ratio of wall surface area to reactor
volume gets less. In this case two options for increased heat transfer are
 provide an internal heating coil or a bank of heating coils
 provide an external heat exchanger with liquid pumped through it or with a boiling
mixture being circulated by gravitational force with no need for a pump.
Simply cranking up the temperature of the heating system may not be possible and it may
be undesirable because the hotter surfaces may degrade the product
Hairston D (2003 November) ‘Tweaking Chemical Reactors’, Chemical Engineering
p.25-31.
Parkinson G (2001 December) ‘A small reactor with a prodigious output’, Chemical
Engineering p.15
19.3. Gas-Liquid Reactions
Not uncommonly one reactant is a liquid and the other is a gas. It may be impractical to
raise or lower the temperature (and pressure) to bring all reactants into the same phase.
Assuming that reaction takes place in only the liquid phase, it is necessary to transfer the
gaseous component into this phase: that is, mass transfer becomes a consideration. The
relative roles of mass transfer and of chemical reaction in dictating the overall speed of
the process is illustrated in the diagram.
R3
M2
Reaction
Rate
R1
M1
Mass
Transfer
Rate
M3
0
R2
0
[C]
[C]vle
In this simple example, the component C is being transferred from the gas to the liquid
and the rate of reaction depends on the liquid-phase concentration [C] in some simple
manner, as shown by line ‘R1’. The rate of transfer of C into the liquid is proportional to
the difference between the concentration [C]vle that would be in vapour-liquid
equilibrium with the vapour and the actual liquid-phase concentration [C], as shown by
the line ‘M1’. The rate of transfer and the rate of reaction have to be equal at steadystate, so the concentration [C] and the overall rate of the process are determined by the
point at which the two lines cross. In the case shown, both reaction and mass transfer are
exerting a significant influence and neither phenomenon can be ignored.
By contrast, the combination of lines ‘R2’ and ‘M2’ show a case where mass transfer has
a negligible effect and the combination of lines ‘R3’ and ‘M3’ show a case where mass
transfer has the dominating effect.
Some considerations are
 for the particular system it is necessary to establish what is the actual relation between
the chemical rate and the transfer rate. This requires reference to the literature and
also careful pilot-plant study
 on scale-up the chemical rate stays the same but the mass transfer coefficient or rate
factor may change
 very fast reaction rates may actually enhance mass transfer, by continually lowering
[C] at the interface to zero (or by continually raising it if [C] is being transferred out
of the liquid phase)
 in cases where mass transfer effects are significant there is an opportunity to use
agitation intensity to speed up the process
 in cases where chemical reaction rate is a significant limitation, it may be profitable to
look for a better catalyst
Fair J R (1967 July 3) ‘Designing Gas-Sparged Reactors’, Chemical Engineering p.67-74.
19.4. Reaction of Immiscible Liquids
Two common situations are
1. the reactants themselves are immiscible and react only at their interface, and
2. a non-reacting immiscible liquid is added the the reaction mixture in order to extract
the product as it is made. This removal of product may be desirable to protect it from
further (or reverse) reaction.
If the reaction is being carried out in a batch vessel or in a continuous stirred tank reactor,
the usual procedure is to provide high-intensity agitation to break up the dispersed phase
and produce lots of surface area. If the reaction is being carried out in a pipeline then
high-intensity, high-pressure-drop static mixers will maximize the dispersion. In either
case, energy must be provided, either to rotate an agitator or to pump the liquid through
the static mixers.
The interplay between reaction rate and mass transfer is similar to that in a gas-liquid
reactor.
19.5. Fluid-Solid Reactions, Non-Catalytic
This type of system occurs mainly in combustors and in ore smelting. The overall rate of
reaction depends, again, on a combination of mass transfer and chemical reaction rate, the
individual actions being
1. diffusion of fluid (gas or liquid) from the bulk of the fluid to the surface of the
particles,
2. diffusion through any crust that exists at the surface,
3. diffusion within the particle,
4. reaction within the particle,
5. diffusion of gaseous reaction products back out of the particle.
The diameter of the particles is a major determinant of rate. The surface area of a fixed
total mass varies inversely as the diameter of the particles into which it is divided. The
external mass transfer coefficient tends to vary inversely as diameter. The internal
diffusion rate varies inversely as diameter squared. All effects therefore point to the need
to subdivide the solid mass as finely as possible.
Fluid and solid reactants combinations are typically brought together in rotating drums or
in fluidized beds in order to enhance surface contact between the two phases. Of course
if the bulk of the mass transfer resistance lies within the particles, the enhanced contact
will be ineffective and a packed or slowly moving bed is appropriate.
19.6. Solid-Catalyzed Reactions
Solid or heterogeneous catalysts are used in many industrially important reactions. The
fluid medium may be gas or liquid or a mixture of both. The chemical reaction occurs at
the solid surface and, in order to provide as much surface area as possible in a given
volume of catalyst, the catalyst particles are porous. Pore diameter is typically in the
range 0.001 to 0.1 micron, providing a surface area of 10,000 to 100,000 square metres
(20 football fields) per kilogram of catalyst.
Just as in the case of non-catalyst solids (section 3.10.5), the overall reaction speed is
governed by a combination of diffusion and actual chemical reaction. In this case the
principal diffusion mechanism is the movement of reactive molecules up the pores and
the movement of product molecules back out of the pores. The ratio of actual reaction
rate to the rate that would be achieved if there were no pore diffusion resistance (i.e., the
purely chemical rate) is called the effectiveness factor, . Design of a solid-catalyzed
system requires some knowledge of the chemical rate and of the effectiveness factor and
of the conditions that affect them, e.g., temperature, pressure, concentration of reactants.
This is knowledge that comes from the catalyst manufacturer and from laboratory studies.
Liquid systems tend to have higher diffusional resistance than gas because diffusivity is
lower and viscosity is higher.
Other considerations are the selectivity of the catalyst for the desired reaction and also the
speed and nature of catalyst de-activation and/or poisoning. Also, the heat of reaction
may be such as to make the temperature of the interior of the catalyst significantly
different from the bulk temperature. A factor in scale-up is that external mass transfer
around the catalyst pellet may also be a limitation: pilot plant studies should address this
effect.
Reactor type is usually a choice between fixed bed and fluidized bed.
A fixed bed tends to be used if
1. the particles are large
2. the catalyst has a long life
3. plug flow is desired
4. heat effects are small.
A fluidized bed tends to be used if
1. the particles are small (e.g., less than 0.2 mm diameter)
2. significant heat must be removed
3. the catalyst requires frequent regeneration
4. plug flow is not required
5. the catalyst is not prone to attrition.
A variant of the fixed bed reactor is the moving bed, where the pellets are cycled into,
through, and out of the reactor so that they can be regenerated off-line.
Bartholomew C H and Hecker W C (1994 June) ‘Catalytic Reactor Design’, Chemical
Engineering p.70-75.
Loffler D G (2001 July) ‘Avoid Pitfalls in Evaluating Catalyst performance’, Chemical
Engineering Progress p.74-77.
19.7. Bio-reactions
In this situation we wish to either
1. grow a living organism in a controlled manner, or
2. use an organism to convert one chemical species into another.
These are the reactions that are becoming more and more important in the chemical
processing industries.
Much of the technology of ordinary chemical reactions carries over to bio-reactions.
Some aspects that are specific to or more important for bio systems are the following.
1. At present most bio-reactions are carried out batch-wise. Batch time is often long
(days) so labour requirements are not a factor.
2. Avoidance of contamination is vital. Only the desired organisms must be present, not
the variety that float around in the atmosphere at large.
3. Control of conditions, e.g., pH, temperature, oxygen concentration, is frequently
important in order to achieve the desired products and in desired yield.
4. Agitation is sometimes a challenge: the reaction mixture must be kept uniform (to
avoid lack of reaction or undesirable reactions in stagnant zones) but at the same time the
agitation must not damage cells. Agitator requiring seals should be avoided but this is
generally not possible.
5. If the product is a pharmaceutical, a proven record of the time-wise profile of each
batch must be kept for certification purposes.
6. If oxygen is required it must be universally distributed.
Williams J A (2002 March) ‘Keys to Bioreactor Selections’, Chemical Engineering
Progress p.34-41.