Part Three

REACTION
TECHNOLOGIES

Most operations in refining and petrochemistry alike utilize chemical changes.

The aim is to transform the molecules o f a crude oil - which vary widely and
depend on the crude origin - into well defined molecular species to obtain the
products in quantity and quality corresponding to market demand. This is essen-
tially what is accomplished by the various, usually catalytic, refining processes
described in Volume 3. In Chapter 7 o f this volume, the different techniques
required to carry out a chemical change are described initially in general terms.
Following this, an illustration is given, with certain important refining and petro-
chemical processes discussed more in detail.
 Chemical Reactor Technology

Pierre Trambouze

The term chemical reactor is used for any equipment allowing a chemical
reaction to be performed, i.e. transform molecular species into other
molecular species. Here however, the term is confined to cases in which the
chemical change is carried out to produce one or more specific chemical
species (olefins, for example) or t o eliminate one or more compounds from a
mixture (desulfurization, for example). The chemical reactor as defined is
therefore the heart of most refining units such a s hydrotreating, catalytic
cracking, catalytic reforming, etc. In addition, reactors are also found in pollu-
tion control facilities implementing chemical processes to purify a number of
effluents.

7.1 An Introduction to Reactors

If, during a visit to a refinery, an effort is made to identify the equipment where
chemical changes are carried out, the variety of their shapes and sizes may
seem surprising. For instance, furnaces, tanks, columns, drums, mixers and
simple tubes can be seen. Accordingly, at first glance any attempt at classify-
ing them might be considered vain and each type of equipment encountered
in actual practice could be viewed as a special case. However, all this equip-
ment has a s common denominator the fact that they accomplish a chemical
change. Therefore, by referring to the main characteristics of a chemical reac-
tion (see Vol. 3, Chapter 2), it should be possible to define a number of crite-
ria t o serve as a basis for classification. In fact, no systematic analysis can be
made until they have been classified in a number of properly characterized
types with which industrial equipment can be identified.
392 Chaoler 7 CHEMICAL REACTOR
TECHNOLOGY

7. I. 1 Reactor Classification
In Volume 3, Chapter 2, the basic principles governing a chemical change were
covered along with the basis for classifying reactors. The classification uses
two criteria: the type of phases present and the operating mode (batch or con-
tinuous). This general classification will be employed here, adapted however
in order to take petroleum industry specificity into account. The discussion
will therefore be confined to continuous systems, virtually the only ones uti-
lized due t o the large capacity requirements in refining. Furthermore, opera-
tions in which one of the reactants or one of the products constitutes a solid
phase will also be disregarded, as they are relatively infrequent in a refinery
(coking is one of the exceptions). However, particular attention will be paid to
systems in which the solid is a catalyst, as catalytic processes account for
most of the chemical processes in refining.
When a continuous operation is considered, there is necessarily a flow of
the different phases in the reactor from their inlet t o their outlet. This internal
flow may be of different types, which obviously depend on the structure of the
equipment. Thus, well characterized types of flow are defined, with which the
flows observed in actual practice can be compared and sometimes identified.
These typical flows, often termed ideal models, are of two sorts (Fig. 7.1):
plug flow, such as the flow when a low-viscosity fluid circulates at high
velocity in a tube,
flow through a stirred tank.

-
Figure
7.1 Continuous operation: the two ideal flow models.
-
 Chapter 7 CHEMICAL REACTOR
TECHNOLOGY 393

These two types of flow are widely dissimilar in structure and influence the
progress of a chemical change inside the equipment in quite different ways.
The evolution of a chemical species entering each of these model reactors can
be imagined. In the tubular reactor, the chemical species moves at the same
time as its environment all along the tube and changes gradually. A continuous
and decreasing concentration profile is established for the reactants between
the reactor inlet and outlet. This is why this type of reactor is sometimes
called a concentration gradient reactor. In a stirred vessel in contrast, the
reactant entering the reactor is suddenly mixed with an environment (reaction
medium) whose composition is identical to the stream exiting the reactor.
There is therefore a concentration discontinuity in the incoming stream at the
reactor inlet. Meanwhile, the different chemical species concentrations are
uniform throughout the volume of the reaction medium.
An intermediate case between these two ideal models is what is called the
staged reactor, comprising a series of mixed zones. Instead of a sudden con-
centration discontinuity at the reactor inlet, there is a series of lesser discon-
tinuities at the inlet of each of the mixed zones.
The continuous models defined above will be utilized in the classification
in this chapter:
tubular,
perfectly stirred tank,
staged.
The classification table (Table 7.1) where the reactor types are listed is the
basis for a more detailed description with reference to the most important
refining processes. These processes were described in Volume 3, so here only
the technological aspect of the reactors used in the processes will be empha-
sized. Going beyond the framework of a simple refinery, the discussion will
include certain units often associated with a refinery in a petrochemical com-
plex. For instance, some elements will be given on steam cracking for olefin
production and steam reforming for hydrogen production.
When there are several phases circulating in a reactor, adequate contact
must be assured among them. Contact among the phases is what allows the
transfer of chemical species without which the reaction could not take place.
The reactant(s) must reach the reaction phase and certain reaction products
must often be eliminated from it. Thus the same concepts are used in analyz-
ing the operation of these reactors as during the study of equipment accom-
plishing physical separations by mass transfer between phases: the interfacial
area between two phases and the mass transfer coefficient associated with
this interface.
The same as for separations, there is a choice between different phase cir-
culation modes: cocurrent, cross-current or countercurrent. Although the
last circulation mode is generally the only one selected for separations, the
other solutions may also be implemented for reactions. In fact, the leading role
played by the chemical change means that the counter-current arrangement is
no longer necessarily optimum as was true for a purely physical process. This
394 Chapter 7 CHEMICAL TECHNOLOGY
REACTOR

Phases Perfectly
Tubular Staged
present stirred tank

Gas Steam cracking

Liquid Visbreaking Olefin
dimerization
Gas + Liquid Clauspol process Absorption in

,
plate columns
Liquid + Liquid HF alkylation Sulfuric acid
alkylation
Gas + Catalyst Steam Catalytic FCC regenerator
reforming reforming
G + L + Catalyst Hydrotreating Hydrotreating Hydrotreating
I
Table

Ll! Reactor classification with the examples chosen for technological description.

is also why phase hold-up, particularly reaction phase hold-up,is an important

parameter in designing a multi-phase reactor.
Another characteristic of the reaction often has a decisive influence on
reactor structure and its outside appearance: the thermal effect inevitably
linked with any chemical change. Many industrial reactions have considerable
heats of reaction, whether exothermic or endothermic. For these cases, the
reactor must be provided with thermal exchange surfaces capable of trans-
ferring the corresponding heat in the right direction.
The following two special cases are often distinguished:
isothermal reactor,
adiabatic reactor.
Comparing a practical case to an isothermal reactor is sometimes an
acceptable approximation. The adiabatic reactor, often encountered industri-
ally, is in contrast simply a special case of the general case, corresponding to
a zero exchange surface. Note that the differentiation in reactor appearance
mentioned earlier is often due to the location of the thermal exchange sur-
face. Thermal exchange can take place:
inside the reactor,
through the reactor walls,
outside the reactor.
If the temperature level at which the exchange occurs is also taken into
account, the result is a wide variety of equipment types. In them, heat can be
exchanged for example by a flame radiating toward the reaction tube, a heat
carrier circulating in an internal exchanger or the reaction medium circulating
in an external exchanger whose size may be much greater than the reactor
itself. This is why reactors exist in the form of furnaces, drums or heat
exchangers.
 Chapter 7. CHEMICAL TECHNOLOGY 395
REACTOR

The following is a review of some of the essential characteristics of reac-

tors involving one or more phases. These characteristics are important when
industrial reactors are being designed. For further details, the reader is
advised to refer to specialized works (Trambouze et al., 1988).

7.1.2 Homogeneous Reactors

7.1.2.1 Tubular Reactor
As its name indicates, this reactor consists of a tube in which the reaction
medium circulates (Fig. 7.1). The heat exchange necessary either for heat
input to, or removal from, the system occurs through the tube wall. Depending
on the temperature at which the reaction takes place, the technologies can be
quite different:
at moderate temperature (< 500°C): a system of tubes whose outside is
kept in contact with a heat carrier,
at high temperature (> 500°C): the tube is placed in a furnace heated by
burners.
Polymerization of ethylene and steam cracking of petroleum cuts to obtain
ethylene can be mentioned respectively as examples of the two technologies.
The reactor volume is obtained from molar balances and the energy bal-
ance. For the plug flow assumption to be usable, the following constraints
must be complied with:

L
- 3 100
dT
with:
G superficial mass velocity, i.e. the total mass flow rate divided by the
tube cross-section
L total reactor length
Re Reynolds number
p dynamic viscosity of the reaction medium
dT tube inside diameter
As an indication, in a steam cracking furnace where the tube has a diame-
ter of around 8 cm, the Reynolds number is greater than lo6, while the resi-
dence time is shorter than one second. The tubular reactor is obviously well
suited to accomplishing rapid reactions, especially in the gas phase.
During the design of a tubular reactor, the thermal transfer at the tube wall
must also be considered in addition to the concept of residence time related
to the reaction rate and the desired conversion. Thermal transfer dictates the
temperature profile all along the tube.
396 Chapter 7 CHEMICAL TECHNOLOGY
REACTOR

The transfer coefficients between the flowing fluid inside the tube and the
wall are calculated conventionally. Outside the tube, the way in which the
exchange occurs must be taken into account:
transfer by forced convection for concentric tubes (see Chapter 6),
transfer by radiation for a furnace (see Chapter 5).
The pressure drop calculation is often necessary in the gas phase and is
performed based on the Fanning equation.

7.1.2.2 Continuous Stirred Tank Reactor

For this type, it is assumed that there are identical concentrations and a uni-
form temperature (perfect mixing) in the whole reactor volume. The reactor
works continuously, however, i.e. it is supplied with fresh feed and the reaction
products are withdrawn continuously. A mechanically agitated tank can be
used, like the one schematically shown in Figure 7.1. The choice of agitator and
of its working conditions is very important (see Chapter 9). For the perfect
mixing assumption to be satisfied, it is necessary to have a mixing time tM,
resulting from the agitation, at least ten times shorter than the mean residence
time 8 required to accomplish the chemical change. For large volume vessels,
it is difficult for tM to be less than one minute. This therefore limits strict
accomplishment of the perfect mixing assumption t o reaction systems requir-
ing residence times 8 longer than 10 min.
Accordingly, a system with external recirculation (Fig. 7.2) is better suited
for large reaction volumes, especially if the reaction is significantly exo- or
endothermic. Perfect mixing is then approximated by having a recirculation
flow rate equal t o at least ten times the feed flow rate Q. The recirculation flow
rate will most frequently be greater than lOQ t o maintain a small temperature
difference between the exchanger inlet and outlet.
There is theoretically no limit on the volume of a continuous stirred tank
reactor, particularly in the version with a circulation loop. However, it may be
advantageous to place several identical reactors in series (staged reactor) to
reduce the overall cost of the system.

7.1.3 Reactors with lbvo Fluid Phases

Reactors where either a gas and a liquid phase or two immiscible liquid
phases are present come under this heading.
Numerous industrial applications correspond to the first category, such
as liquid phase oxidation and hydrogenation or gas absorption by a liq-
uid with a chemical reaction. The following can be mentioned:
- amine treatment of natural gas to eliminate H,S and CO,,
- elimination of CO, from synthesis hydrogen by washing with caustic
soda or amines.
 Chapter 7 CHEMICAL TECHNOLOGY 397
REACTOR

Distributor ___ A3

_ _ _ + s3

Heat _ _ _ A2
exchanger

Anti-vortex _ _ _ + s2
device

1 7.2 I Diagram o f a continuous stirred tank reactor with a continuous recirculation

1loop:
A,, A,, A3. three possibilities for reactant feed.
S,, S,, S; three possibilities for withdrawal from the reaction medium.

Liquid-liquid systems with a chemical reaction are undoubtedly less fre-

quent in industry. Nevertheless, aromatics nitration or sulfonation and
paraffin alkylation can be cited a s typical examples.
The presence of two fluid phases is generally related t o the fact that one of
the reactants present in one phase must be transferred to the other phase
where the reaction takes place. There is most often only one reaction phase
and it is liquid. For example, the most common pattern is a s follows:
reactant A, present in phase I,
transfer of A, to phase 11,
reaction of A, in phase 11.
However, there are cases when reactions can occur simultaneously in the
two fluid phases with transfers toward each of the phases. Situations also exist
in which a reaction product is eliminated by transfer from the reaction phase.
Before describing the equipment that can be used in gas-liquid and liquid-
liquid systems in turn, a few theoretical concepts will be given concerning the
phenomenon of mass transfer with a chemical reaction. These concepts,
398 TECHNOLOGY
Chaoter 7, CHEMICAL REACTOR

which are in fact valid whatever the fluid phases present, will be used later on
as guidelines in selecting equipment to solve a practical problem.

7.1.3.1 Mass Transfer with a Chemical Reaction

Various regimes are differentiated and may be characterized by referring to
the value of the Hatta number (Fig. 7.3). For a second-order reaction that takes
place in phase I1 and involves a reactant A, from phase I (by transfer) and a
reactant A, present in phase 11:
A, A2
-I -+k A3
the Hatta number is expressed as follows:

with:
C2,,(kmo1.m-3) concentration of A, in phase I1
9,,, (rn2.s-') diffusivity of A, in phase I1
k (m3.kmol-1s-1) second-order chemical kinetic constant
k,, ( m d ) transfer coefficient of compound A, on the phase I1 side
Depending on the value of the Hatta number, the following regimes can be
differentiated:
Ha 6 0.3: slow reaction. The reaction takes place in all of phase 11 and
mass transfer is not influenced by the reaction.
0.3 < Ha < 5: moderate reaction rate. The reaction occurs in all of phase
11, but mass transfer is accelerated. The mass transfer coefficient is mul-
tiplied by an acceleration factor E whose maximum value is equal to
Halt h(Ha) .
Ha > 5: rapid reaction. The reaction takes place only in the fraction of
phase I1 close to the interface (film). The mass transfer acceleration fac-
tor E can reach a value equal to Ha.
If the reaction is very rapid (very large, but unknown k value), the regime
may be termed an instantaneous reaction, with the reaction occurring in
a fraction of the film close to the interface.
Referring to the theory of the two films, these different regimes can be
schematically shown by the concentration profiles of compounds A, and A, in
phase I1 in the vicinity of the interface (Fig. 7.3). Although the values of the
Hatta number given as characteristic of the different regimes are strictly valid
only for a pseudo-first order reaction (reactant A, in great excess), general
conclusions can, however, be drawn which are usable in actual practice.
For a system involving a slow reaction, the reaction occurs in all the reac-
tion phase 11. It is therefore advantageous to have maximum hold-up in phase
11. Moreover, mass transfer must assure a virtual saturation of phase I1 by com-
pound A,. It is then pointless to try and increase mass transfer by consuming
 Chapter 7. CHEMICAL REACTORTECHNOLOGY 399

I Phase I
A-
Phase II Phase I
A-
Phase II

I I
I

I
I *
0 6" x
Fluid film Fluid film
a. Slow b. Moderate rate

0 0 61, x
Fluid film Fluid film
c. Rapid d. Instantaneous

-
Figure
7.3 Mass transfer between two fluid phases with a chemical reaction in phase
various regimes.

more mechanical energy than the quantity that allows this quasi-saturation to
be achieved.
If in contrast the reaction is very rapid, it takes place in the film and it is
pointless to have significant hold-up in phase 11. In this case, the aim will be to
increase the interfacial contact area between the two phases to a maximum.
The main equipment that can be used to carry out reactions requiring two
fluid phases will be described below, with a discussion of gas-liquid and then
liquid-liquid systems in turn.
400 Chapter 7 CHEMICAL REACTOR
TECHNOLOGY

7.1.3.2 Gas=LiquidReactors
Given what has just been stated for reactor selection criteria it seems logical
to use a classification based on the liquid phase hold-up, broken down a s fol-
lows:
reactors with significant hold-up, such as bubble columns and stirred
tanks;
reactors with medium hold-up, among them counter-current columns
(with plates or packing) and cocurrent tubular reactors (empty or
packed tube);
reactors with low liquid hold-up,which can be extremely varied, such a s
spray columns (scrubbers), ejector or falling film reactors.
Only the ones that are most frequently found in refineries will be consid-
ered here.

a. Bubble Columns
As its name suggests, the base of this vertical cylindrical column filled with liq-
uid is provided with a device t o disperse gas. The device produces bubbles
which rise through the liquid phase. The gas fraction that has not been
absorbed by the liquid exits at the top of the column (Fig. 7.4). The
height/diameter ratio is generally between 3 and 10. Gas is injected by means
of perforated tubes or plates, or less frequently by using sintered materials
(plates or candles).
The columns are usually sized on the basis of the superficial velocity u of
the gas phase (u = (Qc/S) with Qc the gas flow rate and S the column cross-sec-
tional area). The superficial velocity will most frequently be chosen less than
30 cm/s. The liquid properties (viscosity and interfacial tension) must be
taken into account, a s they have great influence on bubble size. In particular,
if the liquid is prone to foaming, the superficial gas velocity must be reduced
accordingly.
The possible entrainment of liquid phase droplets should not be ignored.
In some cases a demister will need to be placed at the top of the column (see
Chapter 4).
Starting with the basic column described above, several variations can be
contemplated, among them the following:
- By incorporating a guide cylinder to direct the gas bubbles (inside or out-
side t h e cylinder), an air lift is obtained. The liquid is thus circulated
methodically yielding a more homogeneous liquid phase.
- External liquid circulation can be added to the bubble column, in order
to accomplish a heat transfer in particular. Inside the column in this ver-
sion there is either a gas-liquid cocurrent or a counter-current.
All these reactors have the following characteristics in common:
Large liquid phase hold-up, generally greater than 70%. Two different bub-
ble flow regimes can be observed:
 Chapter 7 CHEMICAL TECHNOLOGY 401
REACTOR

Gas

Liquid ~ Liquid

t
Liquid

a. Simple coluum b. Air lift C . Coluum with ejector

system to disperse
the gas
C: guide cylinder
. ....optional liquid circulation I
-
~

Figure
7.4 Differenttypes o f bubble columns.

- heterogeneous flow, consisting of fairly large irregular bubbles,

- homogeneous flow, composed of small bubbles, allowing higher gas hold-
up ratios t o be achieved (up to 20%). Gas hold-up E~ is related to the
superficial gas velocity uGby a relation of the following type:
EG = ue
For homogeneous flow, n ranges between 0.7 and 1, it is close t o 1 if the liq-
uid is a hydrocarbon or an electrolyte solution. For heterogeneous flow, R
ranges between 0.4 and 0.7, often close t o 0.6. Homogeneous flow is obtained
with gas-liquid systems for which bubble coalescence is very low, such as for
aqueous electrolyte solutions.
Fairly small volume interfacial area: 100 to 500 m-l. The interfacial area A
can be estimated from the injected power by the gas flow rate:
A=.(;)" (7.4)

with n ranging roughly between 0.65 and 0.7 and K close to 3.

402 Chapter 7 CHEMICAL REACTOR
TECHNOLOGY

Pis expressed based on the gas flow rate Qc and the pressure loss
P=Q,xAP (7.5)
Backmixing which is considerable for the liquid phase and appreciable for
the gas phase. The larger the column diameter, the more backmixing occurs.
Given the significant backmixing, a bubble column will often be equivalent to
one or more in-series stirred tank reactors.
Relatively high pressure drop for the gas phase flow, in an initial approxi-
mation equal to the liquid column height. Bubble columns can be used as
semi-continuous (captive liquid phase), or continuous (of the stirred tank or
staged reactor type) reactors. They are well suited to carrying out operations
in which the liquid phase chemical reaction is relatively slow. In actual prac-
tice, bubble columns are used for example in oxidations (cumene, cyclohex-
ane, etc.), chlorinations and fermentations.
b. Plate Columns
This type of column is universally used for distillations. It can also be selected
as a gas-liquid reactor. Each plate can be considered as a contact and reaction
stage and the liquid and gas phases circulate counter-currently (see
Chapter 1). Of all the types of plates used in distillation, those with weirs are
more particularly adapted to carrying out reactions. It may be desirable to
increase the liquid height on each plate so as to increase liquid hold-up in the
column. Liquid hold-up of 70 to 80% can be obtained on a plate.
The interfacial area per unit of liquid phase volume on a plate can range
from 100 to 500 m-l. Each plate can be provided with a coil for thermal
exchange, and this allows reactions with an appreciable heat of reaction (for
example, absorption of nitrous vapors) to be carried out.
c. Packed Columns
These are also counter-current gas-liquid contactors and are often used in dis-
tillation (see Chapter 2). The diameter of the column is calculated according
to the gas and liquid flow rates that will be handled. Generally, the gas super-
ficial velocity ranges from 60 to 80% of the flooding velocity. Under these con-
ditions the interfacial area varies between 50 and 80%of the total packing sur-
face area. Liquid phase hold-up is usually small, from 5 to 15%.
The packed column is therefore suited for carrying out rapid reactions,
requiring only little liquid phase. It will often be necessary to seek the maxi-
mum interfacial area per unit of column volume (see Chapter 2). An optimum
needs to be found in order to have maximum interfacial area, while avoiding
the flooding phenomena related to excessively small size. The final choice
must also take packing cost into account. A good compromise often leads to
selecting a packing size close to 3/4" or 1".
Given the low backmixing for the two phases (especially for the gas phase),
the packed column, like the plate column, allows relatively high conversion.
This is mainly true if the reactant(s) are carried by the gas phase (for example,
elimination of CO, and/or H,S from a gas by scrubbing with an alkaline solu-
tion).
 Chapter 7 CHEMICAL TECHNOLOGY 403
REACTOR

7.1.3.3 LiquideLiquid Reactors

For the gas-liquid reactors described above, the problem of phase separation
after contacting does not usually condition their design. This is not the case
for reactors involving two liquid phases. In this type of reactor, two contra-
dictory objectives are sought:
disperse one phase in the other as finely as possible in order to promote
transfer;
separate the two phases rapidly by settling.
A compromise is therefore required, which depends greatly on the physical
behavior of the two phases. In particular, this means how readily the disper-
sion can be achieved without creating an emulsion and especially how rapidly
settling is operated. Under the heading liquid-liquid reactor there are those
used as extractors (see Chapter 3). They can be classified into different cate-
gories, depending on whether outside energy contributes or not either to pro-
duce the dispersion or to accelerate settling.

a. Without Outside Energy Input

In this category, reactors are fairly similar to those discussed in the section on
gas-liquid systems. Here however, since the two phases have relatively similar
physical properties, either one can fulfill the function of dispersed phase. As
such, the interface only needs to be controlled in the high or low position
depending on whether the light or the heavy phase is to be dispersed. These
reactors have rather low efficiency as extractors, but their advantage is sim-
plicity. They can be used advantageously to carry out slow reactions by select-
ing the phase where the reaction takes place as the continuous phase.

b. With Outside Energy Input

The simplest of this type of reactor is the mixer-settler, which consists of a
stirred tank and a vessel for settling (Fig. 7.5). The apparatus can be reduced
to the stirred tank if it operates batch-wise. Here the reaction and settling
operations are carried out consecutively in the reactor, with and then without
stirring. As a continuous reactor, the mixer-settler is often integrated as an ele-
ment in a cascade of stages working counter-currently.
The great advantage of the mixer-settler is that it dissociates the two func-
tions: dispersion with reaction and settling of phases. Each part of the appara-
tus can therefore be adapted to the problem that is being dealt with. The layout
in a counter-current cascade leads to fairly complex systems however, with
intermediate pumps and interface controls. This is why different types of extrac-
tion columns with stirred compartments have been designed (see Chapter 3).

7.1.4 Heterogeneous Catalytic Reactors

This section deals with the various reactors used for a chemical reaction cat-
alyzed by a solid when the reactants can be present in one or two fluid phases.
404 Chapter 7 CHEMICAL REACTOR
TECHNOLOGY

Light phase I
b

- Heavy phase II

-44
Light phase I Heavy phase II

Figure
7.5 Liquid-liquidreactor of the mixer-settler type.

7.1.4.1 Guidelines for Catalyst Implementation

The aim of a solid catalyst is to accelerate a specific chemical reaction.
Because of this, during implementation the catalyst, the chemical reaction and
the appropriate operating conditions must all be considered as a whole.
Additionally, care must be taken during industrial application to preserve the
essential characteristics of the catalyst to the extent possible. These charac-
teristics are:
its activity,
its selectivity,
its stability.
Some fundamental considerations can be employed to define the require-
ments for ideal catalyst implementation. First of all, diffusional phenomena
must be prevented from altering the intrinsic qualities of the catalyst.
Diffusional phenomena can in fact occur inside or outside the catalyst particle.
Inside the particles, it is necessary to facilitate reactant access to the
active sites distributed on the surface of the catalyst pores as much as
possible. The most effective way to do this is by reducing the diameter
and increasing the porosity of the catalyst particles.
Outside the particles, it is necessary to minimize the differences in con-
centrations due to mass transfer between the fluid phase and the outside
surface of the catalyst particle. In addition to fluid phase turbulence, cre-
ating good transfer conditions requires large exchange surfaces. This
once again emphasizes the advantage of small particles.
 Chapter 7. CHEMICAL TECHNOLOGY 405
REACTOR

Although these considerations at the catalyst particle level are crucial,

they are not sufficient to select the type of catalyst implementation. Other fac-
tors will involve the reaction characteristics, for example the reaction scheme,
kinetics and enthalpy of reaction.
Kinetics: the reaction rate is generally seen to decrease as the reactants
are used up. If the catalyst activity is to be exploited to the fullest, theory
suggests that a continuous tubular type of reactor should be chosen. The
selectivity criterion will often be at least as important as the activity cri-
terion, so the choice will not be as easy. This is because the overall per-
formance of different possible types of reactor will have to be compared.
Heat of reaction: when it is not negligible it can have a decisive influence
on the choice of reactor type, considering reaction thermodynamics and
kinetics as well as catalyst stability.
The selected technology should be such that temperature control is possi-
ble, making sure that heat transfers are sufficient between the fluid(s) and the
catalyst particles as well as between the reaction mixture and the outside of the
reactor. However, the second type of heat transfer can in some cases be elimi-
nated, resulting in an adiabatic reactor which has the advantage of simplicity.
Lastly, catalyst implementation can be defined by referring to the following
three characteristics:
the possibility of thermal exchanges;
the catalyst particles are stationary (fixed bed) or moving (moving bed,
fluid bed or suspended catalyst);
the reactants and products form a single phase (gas or liquid) or two
phases (gas + liquid).
Combining these different characteristics leads to a number of practical
possibilities that will be described rapidly so as to outline their respective
advantages and drawbacks.

7.1.4.2 One0Phase Catalytic Reactors

a. Fixed Beds
The fixed bed consists of a compact stationary mass of catalyst particles
inside a tube. The catalyst particles usually have an equivalent diameter rang-
ing from 1to 5 mm. When the mass of catalyst is contained in a single tube con-
sisting of the outside shell of the reactor, the result is an adiabatic fixed bed as
shown in Figure 7.6.

Sizing the catalytic bed

The fixed catalytic bed reactor is sized in the same way as a one-phase
reactor, on the basis of mass and enthalpy balances, incorporating the kinet-
ics of the relevant reaction. For the simple case of a first order reaction
(A, + A2) and assuming isothermal operation, integrating the molar balance
leads to the following equation:
406 ChaDler 7 CHEMICAL REACTOR
TECHNOLOGY

c Reactant inlet

Adiabatic
catalytic bed
Sketch o f an adiabatic
J fixed bed catalytic

-
-
reactor.

$71 Product outlet

support grate

C,, concentration in reactant A, at the reactor inlet (kmol/m3)

C,, concentration in the same component at the outlet
A, outside surface area of catalyst particles per unit of bed volume (m-')
K , kinetic constant related to the outside surface area of the catalyst par-
ticles and relative to component A, (m/s)
Q volume flow rate of the reaction fluid (m3/s)
S cross-section of the catalytic bed (m?
L length of the catalytic bed (m)
V volume of catalyst = SL (m4.
The expression above involves the parameter V/Q, which has the dimen-
sion of a time and is sometimes inappropriately called the contact time. Q/Vis
usually called the space velocity W H (volume of feed per volume of catalyst
per hour). For reasons of convenience, it is often preferable to express Q at
inlet conditions. The concept of WHSV, weight hourly space velocity (weight
flow rate of feed over the weight of the bed), is often used instead of the space
velocity.
 Chapter 7 CHEMICAL TECHNOLOGY 407
REACTOR

The relationship between the two is written a s follows:

PPH = W H X pF/pc
pF fluid density under identical conditions as for the definition of space
velocity
pc catalyst bed density
Note that Eq. 7.6 can be used to determine the volume of the bed or the
mass of catalyst to be placed in it, but not the shape of the bed.
Once the volume or mass of the bed has been determined, the reactor
dimensions should be identified (height and diameter). Accordingly, flow phe-
nomena must be taken into consideration the same as for homogeneous reac-
tors. Plug type flow is generally sought in order to eliminate backmixing, short
circuits and dead zones. For fixed bed reactors where there is only one fluid
phase this is fairly easy to accomplish.
Backmixing can be taken into account by utilizing the concept of turbulent
diffusion. This phenomenon can be estimated by means of a dimensionless
number, called the axial mass Peclet number, Pe,,, which is itself a function of
the Reynolds number:

Pe,, = -
S'F d p
Re=-vSF d p PF
(7.7)
9 A PF
with: VsF = superficial velocity of the fluid = Q/S (m/s).
As soon a s the value of Re reaches 10, Pe,, approaches an extreme value
close to 2. The axial diffusivity, can be estimated. The plug flow assump-
tion can then be verified by calculating the dimensionless group 9A/VsFL,
which must be lower than By using the extreme value for the axial mass
Peclet number, VSFdp/9bA = 2, it can be noted that the ratio between bed length
and particle diameter must be greater than 50:
L/d, > 50 (7.8)
This condition is generally fulfilled for industrial reactors.
Moreover, any influence of mass transfer can be detected by using a prac-
tical rule stated by Satterfield: if for a gaseous fluid phase a bed length of over
100 d, is necessary to reach 90%conversion of the main reactant, the reaction
can be considered a s not limited by mass transfer. If the fluid phase is a liquid,
the situation is different. Here a much higher extreme value of L/d, is obtained,
approximately lo4. It is thus much more difficult to get rid of the influence of
extragranular diffusion phenomena when the fluid phase is a liquid.
Furthermore, flow without short circuits is frequently considered possible
only if the pressure drop per unit of bed length exceeds a certain threshold.
The threshold is often set at approximately 2500 Pa/m and it is frequently
reached without any special attention.
Generally recommended ratios for the bed are as follows: (length/diame-
ter) greater than 0.5 and (reactor diameter/diameter of catalyst particles)
greater than 20, t o reduce wall effects.
408 Chapter 7 CHEMICAL TECHNOLOGY
REACTOR

Calculating pressure drops in a fixed bed catalytic reactor

Pressure drops through a granular bed can be estimated by means of the
Ergun equation:

with:
E~ fraction of volume occupied by the solid
VsF superficial velocity of the fluid
pF fluid dynamic viscosity
pF fluid density
dp equivalent diameter of catalyst particles
The coefficients A and B are close to: A = 150, B = 1.75 for spherical parti-
cles.
According to the equation above, the fraction of volume occupied by the
solid has a great influence on pressure drop. Generally, E,, varies between 0.55
and 0.65. Dense loading of the catalyst tends to increase this value. For a mov-
ing bed, E~ is lower and corresponds to the loose bed density.
The Ergun equation can be adapted to beds consisting of dissimilar diame-
ter spheres or of non-spherical particles. To do this, the pressure drop is mea-
sured by means of any gas (generally air) in a tube of diameter D, such that
m = D/dp is greater than 20. This is sufficient to fix the parameters A and B, or
even the group parameters:

However, care is required in evaluating E,, correctly, since the pressure loss
is very sensitive to this parameter. If this laboratory measurement is impossi-
ble, an estimate can always be made on the following bases:
A normally compacted bed of particles exhibits a fraction of void (1 - E ~ )
that usually ranges from 0.38 to 0.42.
When the catalyst is loaded, (1 - EJ is closer to 0.42.
After a certain operating time, the fraction of void approaches 0.38, often
representing the maximum compaction of the bed. Recently, loading systems
have appeared that allow a compacted catalytic bed to be obtained from the
beginning. With these systems, the catalyst is poured in a shower so that:
each particle finds its equilibrium position before being hindered by oth-
ers;
sloping hills are prevented from forming, since slopes cause particle seg-
regation (the largest particles roll down to the bottom of the slopes) as
well as channeling, which is always detrimental to satisfactory bed oper-
ation.
 Chapter 7 CHEMICAL TECHNOLOGY 409
REACTOR

These systems often use rotating distributors to produce a regular shower

of particles. A similar result can be obtained by loading the particles through
a perforated plate. Perforations are 10 to 15 mm in diameter depending on par-
ticle size and each hole irrigates some 150 cm2 of bed.

Thermal problems
The most common technology in industry is the fixed adiabatic bed
(Fig. 7.6) because of its great simplicity. The adiabatic technique is even used
for reactions which are appreciably exo- or endothermic. The catalytic bed is
then divided into several sections with the reaction fluid temperature read-
justed between them.
Inside an adiabatic catalytic bed, the temperature T varies in relation with
the reaction advancement 5. The variation can be shown in the form of a graph
5 = f ( T ) (adiabatic operating line), which roughly gives a segment of a straight
line (AB, PQ) (Fig. 7.7). In certain cases, the chemical change can be limited by
the thermodynamic equilibrium. Thermal control then aims to bring the tem-
peratures of the various catalytic bed sections back into a zone where the
reaction rate is not excessively hindered by thermodynamics.
For an exothermic reaction, the fluid temperature between catalytic sec-
tions can be readjusted either by thermal exchange or by injecting a colder
fluid. These two possibilities have been schematically represented in
Figure 7.7. The cold fluid injected may be either a gas or a liquid, but one of
the reactants taking part in the reaction is preferred. Thus in hydrogen treat-
ments of petroleum cuts (hydrocracking, hydrodesulfurization), the tempera-
ture inside the reactor is kept within acceptable limits by injecting recycle
hydrogen between the catalytic beds. In contrast, liquid water is injected in
the CO conversion step during hydrogen production.
For endothermic reactions, the fluid phase must be heated between cat-
alytic beds. Depending on the operating temperature range, exchangers or fur-
naces will be used. As an example, 3 or 4 in-series reactors are used in the cat-
alytic reforming process, each one preceded by a heating or reheating furnace
(see Section 7.2.6).
For highly exo- or endothermic reactions, adiabatic operation is virtually
impossible. Here, heat must be removed (or added) in the midst of the cat-
alytic zone. The catalytic bed must often be provided with exchange surfaces.
The technology that is often selected is termed multi-tube. Here the catalyst is
placed in numerous tubes in parallel and immersed in a fluid that assures the
thermal exchange. When the operating temperature is high, the tubes are
placed in a furnace and heated directly by burners. A typical example of the
use of this type of technology is naphtha and natural gas steam reforming (see
Section 7.2.5).
The thermal exchange is calculated in much the same way as for heat
exchangers or furnaces. The difficulty lies in correctly estimating the heat
transfer coefficientinside the catalytic tube. In actual practice, it is not advis-
able to have tubes with a diameter greater than 60 mm, otherwise consider-
able radial gradients can be established.
410 Chapter 7. CHEMICAL REACTOR
TECHNOLOGY

C
5 Thermodynamic
equilibrium

n I / I ,
0
To T2 T3 Tl Temperature
-
Figure
7.7 Temperature control in a fixed bed reactor.
A. External thermal exchanger.

b. Moving Beds
The moving bed is also made up of a mass of catalyst particles inside a tube.
Here however, the catalyst is given a translation motion from top to bottom
due to the force of gravity. There is a choice between three possibilities for the
flow direction of the fluid phase:
 Chapter 7 CHEMICAL REACTORTECHNOLOGY 411

Catalyst inlet Catalyst inlet

I
A
;r
Catalyst outlet 1
Gas

Gas

~-
Overall
movement
of catalytic
bed

outlet

1 Figure
7.8 Simplified sketch o f a co-
current moving bed reactor.
Figure
outlet

Simplified sketch of a cross-

current moving bed reactor.

from top t o bottom (cocurrent, Fig. 7 . Q

from bottom to top (counter-current),
horizontally (cross-current, Fig. 7.9).
The case of cross-current flow used in continuous catalytic reforming will
mainly be discussed. The added complication of the system due to catalyst cir-
culation is offset only by the possibility of regenerating the catalyst continu-
ously.
Plug type flow is required for reactants and catalyst alike. It is important
that neither the bed nor its accessories have any stagnant zones. As will be
seen later on, this means particularly careful design of internals. Provided this
precaution is taken, the moving bed offers all the features of a fixed bed, in par-
ticular the advantages of a gradient reactor. It can achieve high conversions
under good selectivity conditions. Additionally, the moving bed offers the pos-
sibility of withdrawing catalyst continuously or batch-wise in order to dis-
patch it to off-site regeneration and then introduce it back into the circuit.

Moving bed reactor design

When the reactor has been sized a s a fixed bed (considering constant activ-
ity over time, but variable depending on the height along the bed), it is impor-
tant t o design all the accessories required for mobility of the solid.
412 Chapter 7 CHEMICAL TECHNOLOGY
REACTOR

First of all, the means of introducing the feed and withdrawing the effluent
must be defined for the reactor itself. With cross- current flow, the feed is intro-
duced and the products withdrawn via vertical cylindrical grates.

Pressure drop related to flow through the granular medium

Pressure drops can be calculated using the Ergun relation. Accordingly,
(1 - EJ should be correctly estimated: the fraction ot void in a moving bed in
dense flow is practically equal to that of a fixed bed made up of the same non-
compacted particles. This value must be measured experimentally for each
catalyst, along with the angle of repose and the angle of internal friction, which
are important characteristics for solid flow.
Figure 7.10 schematically shows a cylindrical vessel with a solid circulating
through it. The characteristic angles a (of repose) and y (of internal friction)
are indicated on the diagrams and show the importance of these angles in the
design of the upper and lower parts of the vessel.

A B

D 1
i
Figure
7.10 Characteristic angles
for a granular solid:
a = angle o f repose,
y = angle o f internal
friction.

":/ + p g r
internal friction
i
 Chapter 7. CHEMICAL TECHNOLOGY 413
REACTOR

Catalyst circulation
The catalyst is generally circulated by pneumatic transport. At the base of
the reactor or of the regenerator, which are constantly full of catalyst, the gran-
ular solid is introduced into a lift pot (Fig. 7.11). A current of so-called sec-
ondary gas pushes the catalyst toward a rising stream where the catalyst is
entrained by a primary gas carrier current. In this way the catalyst is carried
in the dilute phase to the upper catalyst tank from where it comes back down
by gravity to the reactor. The overall catalyst flow rate is therefore controlled
by the flow rate of the vertical pneumatic transport system.

Simplified diagram of
a catalyst liH pot and
riser.
-1
Figure
7.1 1

, II Secondary a 9

Primary air
.
-J

For a specified catalyst flow rate, which is always easy to achieve by pneu-
matic transport, the dense phase catalyst flow must never become a bottle-
neck. The large reactor and regenerator capacities will not be limiting on the
flow process. In contrast, the flow piping between apparatuses and the with-
drawal orifices can accommodate only limited flow rates. The connecting legs
will therefore have to be sized in number and diameter so that they can han-
dle the specified flow rate.
The solid flow rate through orifices and piping is given by the following
expressions:
414 ChaDler 7 CHEMICAL TECHNOLOGY
REACTOR

Q in m3 of non-compacted catalyst per second

do,dt orifice or tube diameters in meters
n constant t o be determined experimentally, approximately m
These correlations are valid above d = 0.05 m and a s long as the stand-by
height above the orifice or tube inlet is greater than 5d. The flow rate Q in a
tube decreases when a fluid circulates counter-currently t o the solid. In par-
ticular, the flow may come t o a standstill if at some point in the tube the bal-
ance between antagonistic forces (gravity and pressure) is reached. In vertical
pipe, blocking of the flow will occur for:

(Y),,,,, = gpc
(7.1 1)

It is very important to control this phenomenon properly as the pipe

throughput will be reduced long before blocking occurs. For a pressure drop
equal to half the allowable maximum, the catalyst flow rate will already have
dropped by 20% as compared to the flow rate without back pressure.
In calculating connecting legs and the scavenging flow rates of the fluids
passing through them, the following is generally taken:

(7.12)

in order to keep a safety margin.

Since the catalyst flow must be a s even as possible over the whole reactor
cross-section, large diameter vessels must be provided with a large number of
drains for the catalyst. These drains come together in one or more collectors
and the catalyst is thus carried by gravity to the lower vessel. This problem is
very directly related t o that of solid particle distribution residence time which
must be as close a s possible to plug flow distribution. The shape of equipment
bottoms must be designed t o prevent any stagnant zones. Moreover, cross cur-
rents of fluid and solid cause an added difficulty: the transverse fluid current
tends to press the catalyst against the vertical disengagement grate of the
fluid.

c. Fluidized Beds
This type is a bed of solid particles kept in suspension by the rising circulation
of one or more fluids. According to the type of fluid, the following kinds of beds
can be distinguished:
gas-solid,
liquid-solid,
gas-liquid-solid (termed ebullating beds).
The random particle motion generated by fluid circulation can be obtained
only if the rising velocity of the fluid exceeds a certain limit. This limit is gen-
erally called the minimum fluidization velocity (V,,),. Above this velocity the
 Chapter 7. CHEMICAL TECHNOLOGY 415
REACTOR

bed begins moving and expands. Below it, the relative particle positions are
unchanged and the bed remains fixed. The choice of a fluidized bed reactor to
implement a catalyst depends on a number of qualitative considerations that
can be classified into advantages and drawbacks.
Advantages:
- Solids can be withdrawn and reintroduced continuously. This is espe-
cially appreciated in cases where the catalyst becomes rapidly deacti-
vated.
- All intraparticular phenomena are eliminated due to the particle size.
- Internal and external concentration and temperature gradients are elimi-
nated.
- Radial gradients are eliminated.
- Axial gradients are attenuated.
- Risks of hot spots and thermal instability are minimized. The fluidized
bed is therefore well suited to exothermic reactions.
- Additionally, experience shows that heat transfer coefficients are high for
surfaces immersed in the bed.
Drawbacks:
- Bed expansion causes an increase in reactor volume for the same cata-
lyst weight.
- The random motion and resulting mixing make the reactor behave more
like a stirred tank reactor than a plug flow reactor. For most reaction sys-
tems, this leads to an increase in the reaction volume and a loss in selec-
tivity.
- Since the hydrodynamics and modelling of these reactors are very com-
plex, scaling up poses serious problems. This means that their use is
reserved for applications that justify significant research and develop-
ment efforts, such as for very high tonnage production units or very
widespread processes.
- The inevitable entrainment of solid particles requires incorporating a
large disengagement zone, or freeboard, and a fines recycle system
(cyclones).
- lnternal parts are eroded.
The following section deals in turn with the problems posed by the hydro-
dynamics and modelling of catalytic fluidized beds.

Fluidized bed hydrodynamics

Generally speaking, two fluidization regimes can be identified:
Particulate fluidization: here the bed exhibits an average particle den-
sity that is practically constant from one point to another. On the macro-
scopic scale the bed can be considered homogeneous, there are no zones
without solids and no bubbles. This type of fluidization is obtained
mainly when the fluid is a liquid or when gas fluidization is accomplished
with fine particles.
416 Chapter 7 CHEMICAL TECHNOLOGY
REACTOR

Aggregative fluidization: in this case there is occurrence of bubbles

practically without solids. This is most frequently observed in gas-solid
fluidization and for ebullating beds.
Before beginning a study of a fluidized bed reaction, the feasibility of the
operation must first be verified. D. Geldart classified powdered solids under
four headings according to their fluidizing aptitude (Fig. 7.12). The parameters
taken into account for this classification are: apparent bulk density of the par-
ticles (pp - p 3 and the equivalent average diameter dp.

10
pP-pG
(g/cm3)5
at ihe surface

as soon as:
2 -

0.5 -

0.2 - Difficult
\
fluidization
0.1 I

-
Figure
7.12 Classification of particles according to their fluidizing aptitude.

According to Geldart, particle size distribution does not constitute any

obstacle to satisfactory fluidization, provided distribution is approximately
normal, i.e. Gaussian. The four groups he determined are as follows:
Class C: fluidization is practically impossible due to very high surface
forces between particles,
Class A: fluidization is easy and of the particulate type,
Class B: fluidization is easy and of the aggregative type,
Class D: fluidization is possible but of poor quality.
The classification established by Geldart does not take into account other
factors that may have an impact in some special cases: tendency to cake in a
humid atmosphere or to aggregate near the melting point.
 Chapter 7 CHEMICAL TECHNOLOGY 417
REACTOR

Once the feasibility of fluidization has been verified, three hydrodynamic

factors must be determined:
the minimum fluidization superficial velocity (VsF),,,,
the superficial velocity causing entrainment of at least part of the solid,
bed expansion versus VsF.
Two comments should be made:
(1) When designing a fluidization installation, optimum use of the reaction
volume requires operating with minimum expansion which assures both good
solid mobility and good mixing. This is normally obtained for a moderate
expansion of approximately 10 to 25%, using superficial velocities that lead to
greater expansion is generally not justified. However, operating flexibility may
require working at VsF values that are quite different from those anticipated
during design. In this case the response of the bed to these variations will have
to be determined.
(2) In the conventional fluidized bed concept, the solid remains confined
in the bed and entrainment is prevented by the use of cyclones and/or devices
to retain entrained particles. Some time ago, more particularly during the
development of fluidized bed cracking processes (FCC), the reaction zone was
shifted from the fluidized bed as such to the solid transfer lines. As a result,
the cracking reactor has become a tube where the reactants and the catalyst
circulate cocurrently.
Minimum fluidization superficial velocity (VsF),,,
When a fluid circulates from bottom to top through a bed of solid particles
and the pressure drop is measured versus the superficial velocity of the fluid,
a curve is obtained which qualitatively has the shape shown in Figure 7.13. As
long as the fluid superficial velocity remains lower than (VSF)m, the pressure
drop will increase with VsF.
The recorded pressure drop corresponds to the conventional pressure
drop resulting from fluid flow through granular material in a fixed bed. It can
be estimated as discussed earlier with the Ergun equation (see Section
7.1.4.2.a “Calculating pressure drops in a fixed bed catalytic reactor”).
It is obvious that the force exerted on the bed by fluid friction, with pres-
sure drop being the direct effect, increases with the fluid flow rate. The time
comes when this force reaches the apparent weight of the bed. The bed, which
hitherto exhibited a certain degree of cohesion, begins to be lifted causing a
loss of cohesion and a rearrangement of particles. In fact, it evolves toward
something like the situation occurring at the end of solid particle sedimenta-
tion in a fluid.
When the particles start to move in relation to one another, the bed starts
fluidizing and the pressure drop becomes roughly independent of the superfi-
cial fluid velocity. For each increase in fluid flow rate there is a corresponding
increase in bed expansion, and therefore in the volume fraction occupied by
the fluid in the bed. This leads to the stabilization of the pressure drop.
418 Chapler 7 CHEMICAL REACJOR
TECHNOLOGY

It therefore would seem logical to estimate the minimum fluidization super-

ficial velocity from the balance between friction and gravity forces:

with:
E~~volume fraction of bed occupied by particles at the fluidization
threshold
pp particle density

. . ~

1
H

I
...

t
vSF,< (vSFhn
I
I v,,-u, Of
I I the smallest
I
I I particles
I
I
I

I Apparent weight of bed

I
I (p)mf = Bed cross section
I I
I

I
I

Entrainment
begins

0.1

-____
Figure
7.13 Pressure drop through the bed versus the fluid superficial velocity The dashed
curve corresponds to the initial decompacting of the bed.
 Chapter 7. CHEMICAL TECHNOLOGY 419
REACTOR

This expression can easily be made dimensionless, yielding:

1
Ar=A Epm Re,+B (7.14)
(1 - & p a 3 (1 - Ep,33

Ar = PF(Pp- Archimedes number

IG
Re, = dp (vSF)rn PF Reynolds number
PF
It has been observed that:
Epm is frequently close to 1650,
the value of A
(1 - ~ p d 3

1
the value of B is frequently close to 24.5.
(1 - ~prn)’

The expression of (V,,), can finally be obtained:

The parameter dp appears in the expression of (VsF),.

Contrary to fixed bed reactors for which particle size distribution is rela-
tively narrow and their geometry well defined, here the particles exhibit a
more or less narrow distribution depending on the case. An average value of
dp is used, taken from a cumulative weight distribution curve.
Terminal falling velocity of particles - Entrainment velocity
When the fluid flow rate continues to increase through the fluidizing bed,
the smallest particles start being carried along in the gas current. This elutri-
ation phenomenon is often utilized to remove the fines from a batch of cata-
lyst. However, catalyst entrainment is seldom tolerated and the extent of the
phenomenon must therefore be assessed, if only to properly design the
cyclones that send the entrained catalyst back to the bed. By likening a parti-
cle to a sphere with a diameter dp, it can be shown that the terminal falling
velocity is the result of a force balance:
Apparent weight of the particle = drag force:

(7.16)

The drag force expression involves ut, the terminal falling velocity:

CD = drag coefficient.
 Next Page

420 Chapter 7. CHEMICAL REACTOR

TECHNOLOGY

At low Reynolds number values (Re < 1):

and Stokes’ law is obtained:

(7.18)

At high Reynolds number values (Re > lo3):

CD 0.43
and the result is Newton’s law:

(7.19)

In between the two, i.e. 1 < Re < lo3, the following analytic expression can
be used:
69.43
In C, = -5.50 + (7.20)
In (Re) + 7.99

Remember that these relations are valid only if the particles under consid-
eration can reasonably be likened to spheres.
In order to avoid excessive particle entrainment fluidization must be car-
ried out in such a way that:
(‘SF), < S‘F < Ut (7.21)
However, the minimum fluidization superficial velocity is calculated on the
basis of an average diameter dp for the whole bed, whereas for ut only the
average diameter of the fraction corresponding to the smallest particles is
taken into consideration. If the aim is an entrained fluidized bed, an attempt is
evidently made to use a VsF value that is at least equal to ut corresponding to
the biggest particles that are due to be entrained.
Expansion in a fluidized bed
In a bed of motionless particles, the volume fraction of the bed occupied
by the solid is E,,. When a fluid is circulated from bottom to top through the
particle bed with the superficial velocity increasing, no variation in bed height
is observed until a threshold value where the bed decompacts suddenly. The
total bed pressure drop then stabilizes out at a value corresponding to the
apparent weight of the column of catalyst. When the superficial velocity is
increased further, bed expansion can be expected. This is seen by an increase
in the bed height and a decrease in E ~ A. difference should be made, however,
depending on whether the fluid is a liquid or a gas.