Advanced Reaction Engineering and Catalysis

Fluidised Bed Reactors

A/Prof Einar O. Fridjonsson

Department of Chemical Engineering
University of Western Australia
[email protected]
 Fluidisation
When a fluid is passed at sufficient velocity through granular
material the solids bed will behave like a dynamic liquid bed.

Fluid-like behaviour of solids:

(A) (B) (C) (D) (E)
 Fluidised Beds
1922- Fritz Winkler developed the first use of fluidisation for gasification of
fine-grain coal.

1940s- Fluidised catalytic cracking (FCC) process developed to convert heavy

petroleum into desired products.

1960s- Development of more active catalysts resulted in the invention of up-

flow fast fluidised bed reactors.

Current applications include:

- Cooling/heating - Drying - Sublimation-desublimation

- Adsorption-desorption - Coating - Granulation
- Combustion - Catalytic and non-catalytic fluid-reactions
 Fluidised Catalytic Cracking (FCC) examples

TReactor = 480 to 540oC

TRegenerator = 570 to 590oC

Riser cracking process (UOP system)

a) Reactor b) Stripper
Bubbling Fluidised process (KO system) c) Riser d) Slide valve
e) Air grid f) Regenerator
a) Reactor b) Regenerator
 Fluidised bed combustion (FBC) example
(Lurgi process)

CaCO3 → CaO + CO2

SO2 + CaO + 1/2O2 → CaSO4

a) Circulating fluidized-bed reactor e) Convective pass

b) Recycling cyclone f) Dust filter
c) Siphon g) Turbine
d) Fluidized-bed heat exchanger h) Stack
 Fluidised Bed Reactors
(Advantages)
- Easy handling and transport of solids due to fluidisation.

- Uniform temperature distribution:

Eliminates the development of local hot and cold spots in exothermic
and endothermic reactions. Since over-heating can cause product
degradation FBR is ideal for exothermic reactions.

- Large solid-gas exchange area (due to small solids grain size).

- High heat-transfer coefficients between bed and immersed heating or

cooling surfaces.

- Uniform Particle Mixing: A result of the fluidization of the solid particles,

eliminating concentration gradients which allows for better fluid-solid
contact.

- Continuous Operation: eliminates downtimes associated with batch

processes.
 Disadvantages:
- Expensive solids separation/gas purification due to particle entrainment.

- Broadening of residence-time distribution for reaction gas due to bubble

development.

- High velocity of solids results in erosion of internals and attrition of solids.

- Increased Reactor Vessel Size due to expansion of solid bed.

- High pumping requirements: Fluidization requires high velocities which causes

higher pumping (energy) costs.

- Pressure loss during highly exothermic reactions can cause runaway reactions
when the heat transfer is severely restricted due to compacted bed.

- Difficulty in scale-up.
 Fluidisation (1)
Flowing gas through packed bed exerts a drag on the solid particles.

Fgravity FDrag

At low velocities the pressure drop corresponds to a Packed bed (e.g. Ergun Equation)

Δ𝑃 (1 − ϵ) 150 1 − 𝜖 7
= 𝜌𝑔 𝑈 2 +
𝐻 𝜓𝑑𝑝 𝜖 3 𝑅𝑒𝑝 𝜓 4

At high enough velocity the total drag on the particles will equal their weight (𝑊𝑠 ),
i.e. particles begin to lift (but are not fluidized).

The pressure drop due to the weight of the solids force balance is described by:

Δ𝑃
= 1 − 𝜖 𝜌𝑠 − 𝜌𝑔 𝑔
𝐻

𝜌𝑠 : density of solids (catalyst), 𝜌𝑔 : density of gas, 𝜖: porosity of expanded bed, 𝐻 height of bed.
𝑈 = superficial velocity 𝜓 is the particle sphericity.
 Fluidisation (2)
Fluidisation states with increasing fluid velocity (U):
Circulating fluidised beds (CFB)
U
Bubbling Turbulent Fast Pneumatic
Minimum
Fixed bed fluidization Slugging fluidization Conveying
fluidization fluidization

Lean

Streaks Dense
and
Channels

Gas or liquid Gas or liquid Gas Gas Gas Gas Gas or liquid

a b c d e f g
 Fluidisation (3)

a
b
c

Umf = minimum fluidisation velocity Umb = minimum bubbling velocity

Ums = minimum slugging velocity Uc = velocity with maximum pressure fluctuations
Uk = minimum velocity of turbulent fluidisation Utr = maximum fluidisation velocity
 Fluidisation (4)

The mean pressure drop of a typical fluidised bed:

 Minimum fluidisation velocity (𝑈𝑚𝑓 )
Particle weight:
Δ𝑃
= 1 − 𝜖 𝜌𝑠 − 𝜌𝑔 𝑔
𝐻

Particle drag (Ergun Equation):

Δ𝑃 1 − 𝜖 150 1 − 𝜖 7
= 𝜌𝑔 𝑈 2 +
𝐻 𝜓𝑑𝑝 𝜖 3 𝑅𝑒𝑝 𝜓 4
Where 𝜓 the shape factor of catalyst particle, sometimes called the sphericity.
Setting the pressure drop from both equations equal, we obtain:
1−𝜖 150 1−𝜖 7
1 − 𝜖 𝜌𝑠 − 𝜌𝑔 𝑔𝐻𝐴𝑠 = 𝜌𝑔 𝑈 2 + 𝐻𝐴𝑠 Eqn.(1)
𝜓𝑑𝑝 𝜖 3 𝑅𝑒𝑝 𝜓 4

𝜌𝑔 𝑑𝑝 𝑈
For 𝑅𝑒𝑝 < 10 (𝑅𝑒𝑝 = ) we can solve the expression to obtain,
𝜇
2
3
𝜓𝑑𝑝 𝜖𝑚𝑓
𝑈𝑚𝑓 = 𝑔(𝜌𝑠 − 𝜌𝑔 ) Eqn.(2)
150𝜇 1 − 𝜖𝑚𝑓

Note that 𝑈 = 𝑈𝑚𝑓 , ϵ = 𝜖𝑚𝑓 .

 Minimum fluidisation velocity(2)
(sphericity and void fraction)

The sphericity (𝜓) measures the particle non-ideality in roughness and shape,
1/3 2/3 T.1
𝐴𝑠𝑝ℎ𝑒𝑟𝑒 𝜋 6𝑉𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒
𝜓= =
𝐴𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒 𝐴𝑝𝑎𝑟𝑡𝑖𝑐𝑙𝑒
𝜓 for common particles shown in Table (T.1).
𝜓 = 0.6 is a typical value for granular solids.

The void fraction at minimum fluidisation (𝜖𝑚𝑓 ),

A good correlation (less than 10% error if particles are small):

0.029 0.021
𝜇2 𝜌𝑔
𝜖𝑚𝑓 = 0.586𝜓 −0.7
𝜌𝑔 𝑔(𝜌𝑠 − 𝜌𝑔 )𝑑𝑝3 𝜌𝑠

If 𝜖𝑚𝑓 < 0.40 it should be considered suspect.

 Geldart Diagram
Fluidisation of gas-solids systems depends on the powders used. These are classified by
groupings proposed in 1973 by Prof. D. Geldart. These groupings are extensively used for
design purposes.

𝝆𝒇 : fluid density [g/cm3]

𝝆𝒔 : solid density [g/cm3]
dp: particle diameter [m]

Group A particles:
Small size particles (ca. 0.1mm) or low bulk density (e.g. catalysts used in FCC).
Group B particles:
Moderate size particles, moderate bulk density (e.g. sands with mean diameter 0.06 to 0.5mm).
Group C particles:
Very-fine grained, cohesive powders (e.g. flour, fines from cyclones and electrostatic filters).
Group D particles:
Large size particles, high bulk density (e.g. sands with mean diameter > 0.5mm).
 Minimum bubbling velocity (𝑈𝑚𝑏 )

Minimum Bubbling velocity (𝑈𝑚𝑏 ) is the fluidisation velocity when gas

bubbles first appear.

Geldart & Abrahamsen (1978) observed that

𝑈𝑚𝑏 41250𝜌𝑓 0.1 𝜇0.9

=
𝑈𝑚𝑓 𝜌𝑠 − 𝜌𝑓 𝑔𝑑𝑝
𝑈𝑚𝑏
< 1 is generally true for Geldart Group B and D particles.
𝑈𝑚𝑓
𝑈𝑚𝑏
> 1 is generally true for Group A particles.
𝑈𝑚𝑓
Minimum slugging velocity (𝑈𝑚𝑠 )
Gas slugging can occur at high enough velocities (𝑈 ≥ 𝑈𝑚𝑠 )

Slugging only occurs when Height/Diameter of bed > 2

2
and bubbles grow to D. Otherwise the bed will go from
3
bubbling to turbulent fluidization directly.

It is a undesirable regime due to poor mixing

characteristics and large pressure fluctuations.
Stewart and Davidson proposed the slugging criteria:

𝑈𝑚𝑠 = 𝑈𝑚𝑓 + 0.07 𝑔D

For bed heights less than 30cm,

Baeyens and Geldart (1974) found it hard to produce
slugs, and proposed a modified criteria:

2
𝑈𝑚𝑠 = 𝑈𝑚𝑓 + 0.07 𝑔D + 0.0016 60D0.175 − 𝐻𝑚𝑓
 𝑼𝒄 , 𝑼𝒌 and 𝑼𝒕𝒓
Transition from bubbling to turbulent regime (𝑈𝑐 and 𝑈𝑘 )
Yerushalmi and Cankurt (1979 ) showed that,
𝑈𝑐 : Point of maximum stdev. in pressure fluctuations, due to bubbles/slugs
reaching maximum diameter.
𝑈𝑘 : Point when pressure fluctuation reach steady state with increasing 𝑈 as
large bubbles break into smaller bubbles.
Horio (1990) suggested using:
0.936𝜇𝐴𝑟 0.472 1.46𝜇𝐴𝑟 0.472 𝑔𝑑𝑝3 𝜌𝑠 − 𝜌𝑓
𝑈𝑐 = 𝑈𝑘 = where 𝐴𝑟 = 2
𝑑𝑝 𝜌𝑓 𝑑𝑝 𝜌𝑓 𝑣 𝜌𝑓
Maximum fluidization velocity (𝑈𝑡𝑟 ) Kunii and Levenspiel (1991) :
𝑔(𝜌𝑠 −𝜌𝑔 )d2𝑝
𝑈𝑡𝑟 = for Rep < 0.4
𝜇
1/3
0.0178𝑔2 (𝜌𝑠 −𝜌𝑔 )2
𝑈𝑡𝑟 = 𝑑𝑝 for 0.4 < Rep < 500
𝜌𝑔 𝜇

𝜌𝑠 𝑈𝑚𝑓 𝑑𝑝
where Rep = is the particle Reynolds number.
𝜇
 Transitions between circulating
fluidised bed (CFB) regimes:
Turbulent bed (TB) to fast fluidised bed (FF) (Bi et al. 1995):

𝜌𝑠 − 𝜌𝑔 𝑔𝑑𝑝
𝑈𝑇𝐵−𝐹𝐹 = 1.53
𝜌𝑔

Fast fluidised bed (FF) to Pneumatic conveying (PC) (Bi and Fan 1991):

0.315
0.542 1/3
𝐺𝑠 𝜌𝑔 𝜌𝑠 − 𝜌𝑔 𝑔
𝑈𝐹𝐹−𝑃𝐶 = 21.6 𝑔𝑑𝑝 𝑑𝑝
𝜌𝑔 𝑈𝐹𝐹−𝑃𝐶 𝜇2

where 𝐺𝑠 is the fixed solids flux from the reactor.

 Reh’s Fluidised-bed State Diagram
Used to estimate the bed operating range and gas velocity needed to create a given fluidized
state:

a) Circulating fluidised bed

b) Fluidised-bed roaster
c) Bubbling fluidised bed
d) Shaft furnace
e) Moving bed
Ordinate:
3 𝑈2 𝜌𝑓
4 𝑔𝑑𝑝 𝜌𝑠 −𝜌𝑓

𝑈𝑑𝑝
Abscissa: 𝑅𝑒𝑝 =
𝑣

Auxilliary grid lines:

𝑈3 𝜌𝑓
𝑀=
𝑔𝑣 𝜌𝑠 − 𝜌𝑓

𝑔𝑑𝑝3 𝜌𝑠 − 𝜌𝑓
𝐴𝑟 = 2
𝑣 𝜌𝑓
𝜖 is the mean void fraction.
𝑣 is the kinematic viscosity.
Bubble Development & Solids Mixing
 Bubble Development
For physical operations and non-catalytic reactions, the state of the fluidised bed can
often be described by a single quantity averaged over the whole bed (e.g. the mean
bed porosity 𝜖).

For designing catalytic reactions in fluidised beds details of the fluid-flow conditions
must be known.

The development of bubbles causes the existence of two types of gas phase:
Emulsion phase: where gas and solids form a homogeneous-like emulsion.
Bubble phase: where the gas is separated in bubbles from the solids.
Elemental reactor volume

Emulsion
Frequency and size of Bubbles is mainly governed
Bubble by the liquid surface tension and buoyancy forces.
Bubble At high gas velocities the liquid inertia becomes
more important than the liquid surface tension.
 Bubble Development (2)

Davidson’s [1960s] theoretical development and experimental verification:

1) 𝑈𝑏𝑟 (rise velocity of bubble) depends only on the bubble size.
2) Gas behaviour in vicinity of the bubble depends only on the relative
velocity of rising bubble (𝑈𝑏𝑟) and gas rising in the emulsion (𝑈𝑒).

Two extremes can be observed for an isolated bubble (𝜖 = 𝜖𝑚𝑓 ):

𝑈𝑏𝑟
𝛽= 𝜖
𝑈𝑚𝑓 𝑚𝑓
𝛽 > 1 is typical of Geldart groups A & B
𝛽 < 1 is typical of Geldart group D

Cloud

𝛽<1 𝛽>1
 Design : Bubble Development (3)
𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑜𝑓 𝑐𝑙𝑜𝑢𝑑 𝑈𝑒
≅
𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟 𝑜𝑓 𝑏𝑢𝑏𝑏𝑙𝑒 𝑈𝑏𝑟

Every rising gas bubble drags behind a wake of solids.

𝑣𝑜𝑙. 𝑜𝑓 𝑤𝑎𝑘𝑒
𝛼=
𝑣𝑜𝑙. 𝑜𝑓 𝑏𝑢𝑏𝑏𝑙𝑒

𝛼 varies between 0.2 and 2.0.

Emulsion (solids phase)

Cloud (intermediate between solid and gas phase)

Bubble (gas phase)

Wake (Solids trailing bubble)

 Wake
The wake fraction (𝛼) gives the fraction of bubble volume occupied by the wake
(i.e. the solids rich proportion of the bubble). The following figure can be used
as a guide:

Adapted from D. Kunii and O. Levenspiel,

Fluidisation Engineering, 2nd. Ed., (Stoneham, Mass.; Butterworth-Heinemann 1991).
 Bubble development(4)
Patterns of bubble coalescence and solids circulation in large industrial fluidised beds:

Bubble growth:
1) Effective hydrostatic pressure
decreases towards the top.
2) In vertical direction, the trailing
bubbles catch-up to the
leading bubbles and coalesce.
3) In horizontal direction,
neighbouring bubbles coalesce.
 Solids Mixing in FBR’s
Vigorous mixing is achieved through motion of gas phase (bubbles) through
the solid bed.
Main mechanisms are bubble drift and wake transport

Upward motion of solids with bubbles (mainly by wake transport), causes the
surrounding solids (emulsion phase) to move downward.

Particle drift Wake transport

Ex. Radial distribution of bubble-gas flow: 𝒖𝒃 r

Cloud

Wake
Fluidised Bed Reactor
Design
 Design : Gas distribution (1)
On the laboratory scale:
The common gas distribution geometry are porous plates of glass, ceramics.

In pilot-plants and full-scale fluidized beds

Many different designs are used e.g.

Perforated plate Nozzle plate Bubble-cap plate

 Design : Gas distribution (2)
Requirements:
1) Uniform fluidisation over entire cross section of the bed
∆𝑃𝑑
– Usually achieved if  0.1 to 0.4, where ∆𝑃𝑑 is the pressure drop across
∆𝑃𝑓𝑏
the distribution plate, ∆𝑃𝑓𝑏 is the pressure drop across the fluidised bed.
𝜌𝑜 𝐶𝐷 𝑈𝑜2
– ∆𝑃𝑑 = , where all are obtained at the orifice (o).
2
– 𝑉 = 𝑁𝑜 𝐴𝑜 𝑈𝑜 , where 𝑉 is the gas flow rate, 𝑁𝑜 is the number of orifices, 𝐴𝑜 is
the cross-sectional area of the orifice, 𝑈𝑜 is the velocity of gas in the orifice.
– Δ𝑃𝑑 ≥ 2.5𝑘𝑃𝑎 (Rule of thumb for commercial sized grids)

2) Provide complete fluidisation, without dead-spots.

3) The design pressure drop (∆𝑃𝑑 ∝ 𝑈 2 ) must exceed the pressure fluctuations or (if
not economical), a design must tolerate short periods of gas flow reversal without
solids entering the orifice (e.g. bubble-cap design).
 Design : Gas distribution (3)
For perforated & bubble cap type grids:
∆𝑃𝑔𝑟𝑖𝑑 = 𝐾𝑔𝜌𝐵𝑒𝑑 𝐿𝐵𝑒𝑑

𝐾 = 0.3 for upward or lateral gas entry.

𝐾 = 0.1 for downward gas entry.

0.1
Orifice velocity (𝑈𝑜 )

𝐶𝑑 𝐿𝑜 /𝑑𝑜
2∆𝑃𝑔𝑟𝑖𝑑
𝑈𝑜𝑟𝑖𝑓𝑖𝑐𝑒 = 𝐶𝑑
𝜌𝑔𝑎𝑠,𝑜𝑟𝑖𝑓𝑖𝑐𝑒

Use figure to find 𝐶𝑑 (the grid hole

discharge coefficient).

t/𝑑𝑜
 Design : Gas distribution (4)
Hole designs: Rules of thumb:
Number of grid holes 𝑁𝑜 > 20/m2
to minimise stagnation zones.

𝒅𝒐 𝒅𝒐

𝑳𝒐 𝑳𝒐
1
𝐿𝑜 = 1
𝑁𝑜 𝑠𝑖𝑛60𝑜 𝐿𝑜 =
𝑁𝑜

Plenum designs:
 Design : Gas distribution (5)
Gas jets can form at the orifice or where gaseous reactants are added to the bed. The
depth of their penetration (𝐿) is critical for bed operation and is a critical design
parameter i.e.

These jets are strongly corrosive and the placement of internals (such as
heat exchanger tubes) must not be located within their range.

Multiple empirical correlation available e.g.

For vertical gas jets [Merry, 1975]:

𝟎.𝟑 𝟎.𝟐
𝑳 𝝆 𝒇 𝒅𝒐 𝒖𝟐𝒐
= 𝟓. 𝟐 𝟏. 𝟑 −𝟏
𝒅𝒐 𝝆 𝒔 𝒅𝒑 𝒈𝒅𝒐

For horizontal gas jets [Merry, 1971]:

𝟎.𝟐 𝟎.𝟒
𝑳 𝝆 𝒇 𝒅𝒑 𝝆𝒐 𝒖𝟐𝒐
= 𝟓. 𝟐𝟓 − 𝟒. 𝟓
𝒅𝒐 𝝆 𝒔 𝒅𝒐 𝟏 − 𝜺 𝝆𝒔 𝒈𝒅𝒑

Karri [1990] suggested the following approximation:

Lup ≈ 2Lhor ≈ 3Ldown
 Design of Solids Recycle System
Solids carried out of the bed by fluidised gas are
generally collected in cyclones.

For Bubbling Beds the solids are returned through

the standpipe of the cyclone which dips directly into the bed.

For Circulating fluidised beds,

due to large solids volumes,
a very large cyclone is generally needed
which is arranged beside and outside the
bed, with valves needed to connect
it to the bed vessel.
These valves are commonly either of
Syphon (A) or L-Shaped (B) type
 Solids (Catalyst) Attrition:
Normally occurs due to surface abrasion (catalyst particles become rounded),
fragmentation can also occur, especially when fresh catalyst is added to the bed.

q3: mass density

dp: particle size

For a new candidate catalyst extensive experiments are usually preformed in small
laboratory scale FBR’s to measure the attrition rate 𝑟𝑎 ,
1 ∆𝑚𝑎
𝑟𝑎 = 𝑚𝑎 : elutriated mass, 𝑚𝑏 : bed mass.
𝑚𝑏 ∆𝑡
relatively attrition-resistant fluidised-bed catalysts 𝑟𝑎 ~ 0.2%/day.
 Solids (Catalyst) Attrition(2):
In Industrial fluidised-bed reactors the sources of attrition are typically:

1) Jet attrition at gas distribution grid openings and nozzles.

2) Bubble attrition, due to solids motion caused by bubbles.
3) Attrition in cyclones.
4) Attrition in pneumatic conveying lines (e.g. between reactor and
regenerator beds)

To address these different sources a variety approaches exist.

- Jet attrition can be controlled by specially designing the distributor (e.g.

using bubble caps).
- Bubble attrition can be lowered by limiting bubble growth (which
corresponds to lowering the bubble velocity).
- Attrition in cyclones, can be prevented by replacing them with filters or
moving the cyclone above the TDH to lower the cyclone load which also
helps address the pneumatic attrition.
 Transport Disengagement Height (TDH)
For Geldart group A solids, the Transport Disengagement Height (TDH) can be
estimated using the following graph:

For Geldart group B solids,

TDH = 18.2𝑑𝑣

where 𝑑𝑣 is the size of bubbles bursting at the top of the bed.

This is a bench-top model and may or may not scale well to plant-size equipment.
 Solids (Catalyst) Attrition(3):
Dynamic attrition model:

𝑚𝑙𝑜𝑠𝑠
Gwyn Equation: = 𝐾𝑎 ∙ 𝑡 −𝑏
𝑚𝑏𝑒𝑑,0
Solids
𝑏 is constant for a catalyst, 𝐾𝑎 decreases with mean particle size. Loss flow

Cyclone
Steady state attrition model: Freeboard Cyclone-attrition

Fluidized bed

Catalyst
feed
bubble-attrition
Return line

Catalyst jet-attrition
discharge

Balance model for fluidized bed catalyst attrition:

 Scale-up
There are no generalised rules to guide scale-up.
Different flow regimes will scale differently.
Scale models must be chosen carefully, and simulations are required.
Pilot plant test needed before process is made commercial scale.

Some observations:
Bed Diameter: bubble velocity increases with bed diameter reducing residence
time and conversion.

Grid Design: Industrial distributors generally operate worse than laboratory

porous plates leading to lowered conversion.

Internals: Laboratory scale reactor usually operate without internals; internals

increase mixing and can offset some loss in conversion.
 Scale-up (2)
Catalyst: Bubble growth is influenced by particle-size distribution - due to
wall effects this effect is not fully developed until commercial scale
equipment.
Lateral mixing of Reactants: Laboratory scale equipment generally have
uniform distribution across the bed, while commercial equipment
usually have uneven distributions (due to multiple feed points).

Secondary reactions in Freeboard: Walls rapidly cool reactants in small scale

equipment. Commercial scale equipment has more entrained
catalyst and less wall cooling which can cause significant secondary
reactions in the freeboard - this can severely affect selectivity.

Catalyst attritions: These are minimal on the laboratory scale and much
greater on the commercial scale.
Modelling Fluidised Bed Reactors
 Evolution of FBR modelling

Two phase model

K
V1 V2

m1 Bubble Emulsion 𝒎𝟐
phase phase

(D/uL)1 (D/uL)2

v1 v2
 CFD Modelling
With increasing computational power the tendency has been to use first principles to
model the fluid mechanics. CFD’s model solve each scale (e.g. particle-particle or
fluid-particle interactions).

Larger geometry

Lattice Boltzmann Discrete Particle Continuum Discrete Bubble

Model Model Model Model

Fluid-particle Particle-particle Particle-particle Large scale motion

interaction interaction Interaction Industrial size
(Bubble behaviour)

Larger scale phenomena

Bubbling bed model (Example)
The Kunii-Levenspiel (K-L) model:
 K-L Model Assumptions:
1) The bubbles are all one size

2) The solids in the emulsion phase flow smoothly downward, essentially in plug flow.

3) The emulsion phase exists at minimum fluidizing conditions. The gas occupies the
same void fraction in this phase that it had in the entire bed at the minimum
fluidization point. In addition, since the solids are flowing downward, the
minimum fluidizing velocity refers to the gas velocity relative to the moving solids,
𝑈𝑚𝑓
𝑈𝑒 = − 𝑈𝑠
𝜖𝑚𝑓
The velocity of the moving solids (𝑈𝑠 ) is in the downward direction, the velocity of
the gas in the emulsion (𝑈𝑒 ) is in the upward direction.

4) In the wakes, the concentration of solids is equal to the concentration of solids in

the emulsion phase, and therefore the gaseous void fraction in the wake is also the
same as in the emulsion phase. Since the emulsion phase is at the minimum fluidizing
condition, the void fraction in the wake is equal to 𝜖𝑚𝑓 . The wake, however, is quite
turbulent, and the average velocities of both solid and gas in the wake are assumed to
be the same and equal to the upward velocity of the bubbles.
 Fluid-mechanics parameters
To calculate bubble velocity (𝑼𝒃 ), we need:
Porosity at minimum fluidization, 𝜖𝑚𝑓
Minimum fluidization velocity, 𝑈𝑚𝑓
Bubble size, 𝑑𝑏

To calculate mass transport coefficient (𝑲𝒄𝒆 , 𝑲𝒃𝒄 ), we need:

Porosity at minimum fluidization,𝜖𝑚𝑓
Minimum fluidization velocity, 𝑈𝑚𝑓
Bubble velocity, Ub
Bubble size, 𝑑𝑏

To calculate reaction-rate parameters, we need:

Fraction of the total bed occupied by bubbles, 𝛿
Fraction of the bed consisting of wakes, 𝛼𝛿
Volume of catalyst (solids) in the bubbles, clouds, and emulsion, 𝛾𝑏 ,𝛾𝑐 𝛾𝑒
𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑠𝑜𝑙𝑖𝑑 𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡 𝑖𝑛 𝑝ℎ𝑎𝑠𝑒 𝑖
Where 𝛾𝑖 =
𝑣𝑜𝑙𝑢𝑚𝑒 𝑖𝑛 𝑏𝑢𝑏𝑏𝑙𝑒𝑠
𝑖 = 𝑏, 𝑐 & 𝑒 (i.e. bubbles, clouds & wakes and emulsion)
 𝑈𝑏 , 𝑈𝑠 and 𝑈𝑒
For a single bubble, Davidson found that bubble rise velocity is
1/2
𝑈𝑏𝑟 = 0.71 𝑔𝑑𝑏

The bubble velocity:

𝑈𝑏 = 𝑈 − 𝑈𝑚𝑓 + 𝑈𝑏𝑟
1/2
𝑈𝑏 = 𝑈 − 𝑈𝑚𝑓 + 0.71 𝑔𝑑𝑏
Bed fraction in bubbles (𝛿):
𝑈 − 𝑈𝑚𝑓 𝑈𝑏𝑟 𝑈
𝛿= =1− , for 𝑈𝑏 ≫ 𝑈𝑚𝑓 𝛿 =
𝑈𝑏 𝑈𝑏 𝑈𝑏
𝛼𝛿𝑈𝑏
Velocity of emulsion solids: 𝑈𝑠 =
1 − 𝛿 − 𝛼𝛿

𝑈𝑚𝑓
Velocity of emulsion gas: 𝑈𝑒 = − 𝑈𝑠
𝜖𝑚𝑓
 Mass transport coefficients

Gas interchange between bubble and cloud (Davidson’s theory):

𝑈𝑚𝑓 𝒟 0.5 𝑔0.25
𝐾𝑏𝑐 = 4.5 + 5.85
𝑑𝑏 𝑑𝑏 1.25

Gas interchange between cloud-wake and emulsion (Higbie’s penetration theory):

0.5
𝜖𝑚𝑓 𝒟𝑈𝑏𝑟
𝐾𝑐𝑒 = 6.77
𝑑𝑏 3

𝒟 is the gas diffusion coefficient.

Note that,
𝐾𝑏𝑐 = 𝐾𝑐𝑏
𝐾𝑐𝑒 = 𝐾𝑒𝑐
 Fluidised bed mass transfer
a) Transport between gas and solid
There is enough separation between solid particles to consider this mass
transfer only between a gas and a single sphere (Use Fröessling Eqn):

𝑺𝒉 = 𝟐. 𝟎 + 𝟎. 𝟔 𝑹𝒆 𝟏/𝟐 𝑺𝒄 𝟏/𝟑
Because of small particles size and high surface area per volume in fluidized
beds this transport seldom limits the reaction.
A B
b) Transport between bubbles and the clouds
𝑊𝐴𝑏𝑐 = 𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐
𝑊𝐵𝑐𝑏 = 𝐾𝑐𝑏 𝐶𝐵𝑐 − 𝐶𝐵𝑏
𝑾𝑨𝒃𝒄
𝑊𝐴𝑐𝑒 = 𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶𝐴𝑒 𝑾𝑩𝒄𝒆
𝑊𝐵𝑐𝑒 = 𝐾𝑒𝑐 𝐶𝐵𝑒 − 𝐶𝐵𝑐
Bubble
𝑾𝑨𝒄𝒆
Cloud

Emulsion
 Reaction Mechanisms
Consider an 𝑛th order constant volume catalytic reaction,
𝑛
In the bubble phase, 𝑟𝐴𝑏 = −𝑘𝑏 𝐶𝐴𝑏
𝑛
In the cloud, 𝑟𝐴𝑐 = −𝑘𝑐 𝐶𝐴𝑐
𝑛
In the emulsion phase, 𝑟𝐴𝑒 = −𝑘𝑒 𝐶𝐴𝑒

𝑛−1
𝑐𝑚3 𝑐𝑚 3 where k’ is the reaction rate per catalyst
𝑘𝑐𝑎𝑡 = 𝜌𝑠 𝑘’
𝑐𝑎𝑡∙𝑠 𝑚𝑜𝑙 weight - this is normally obtained empirically.
𝑘𝑏 = 𝛾𝑏 𝑘𝑐𝑎𝑡
𝑘𝑐 = 𝛾𝑐 𝑘𝑐𝑎𝑡
𝑘𝑒 = 𝛾𝑒 𝑘𝑐𝑎𝑡
𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑠𝑜𝑙𝑖𝑑 𝑐𝑎𝑡𝑎𝑙𝑦𝑠𝑡 𝑖𝑛 𝑝ℎ𝑎𝑠𝑒 𝑖
Where 𝛾𝑖 = , 𝑖 = 𝑏, 𝑐 & 𝑒 (i.e. bubbles, clouds & wakes
𝑣𝑜𝑙𝑢𝑚𝑒 𝑖𝑛 𝑏𝑢𝑏𝑏𝑙𝑒𝑠
and emulsion)

Typically 𝛾𝑏 ~ 0.005, 𝛾𝑐 ~ 0.2 and 𝛾𝑒 ~ 1.5.

𝑈𝑚𝑓
𝛾𝑏 = 0.001 to 0.010 3
𝜖𝑚𝑓
1−𝛿 𝛾𝑐 = 1 − 𝜖𝑚𝑓 +𝛼
𝛾𝑒 = 1 − 𝜖𝑚𝑓 − 𝛾𝑐 − 𝛾𝑏 𝑈𝑚𝑓
𝛿 𝑈𝑏𝑟 −
𝜖𝑚𝑓
 Material Balance: A B

𝑾𝑨𝒃𝒄
𝑾𝑩𝒄𝒆
Bubble
𝑾𝑨𝒄𝒆
Cloud

Emulsion
In by flow - out by flow - out by mass transport - generation = 0
n
𝑈𝑏 𝐴𝑐 𝐶𝐴𝑏 𝛿 𝑧 − 𝑈𝑏 𝐴𝑐 𝐶𝐴𝑏 𝛿 𝑧 + Δ𝑧 − 𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐 𝐴𝑐 Δ𝑧𝛿 − 𝑘𝑏 𝐶𝐴𝑏 𝐴𝑐 Δ𝑧𝛿 = 0

Divide by 𝐴𝑐 Δ𝑧𝛿 and take limit as Δ𝑧 → 0, to obtain:

𝑑𝐶𝐴𝑏 n
𝑈𝑏 = −𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐 − 𝑘𝑏 𝐶𝐴𝑏
𝑑𝑧
Perform same material balance on the cloud and emulsion phases (basing all terms on
the bubble volume): 𝑈𝑚𝑓
3 𝜖𝑚𝑓 𝑑𝐶𝐴𝑐 𝑛
𝑈𝑏 𝛿 +𝛼 = 𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐 − 𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶𝐴𝑒 −𝑘𝑐 𝐶𝐴𝑐
𝑈𝑚𝑓 𝑑𝑧
𝑈𝑏𝑟 − 𝜖𝑚𝑓

𝑈𝑚𝑓 𝛼𝛿𝑈𝑏 𝑑𝐶𝐴𝑒 𝑛

− = 𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶𝐴𝑒 − 𝑘𝑒 𝐶𝐴𝑒
𝜖𝑚𝑓 1 − 𝛿 − 𝛼𝛿 𝑑𝑧
 Material balance(2)
The K-L model simplifies these equations by assuming that the derivatives on the LHS
of the cloud and emulsion material balances are negligible compared with the RHS
𝑧
terms. By also letting 𝑡 = we obtain:
𝑈𝑏

𝑑𝐶𝐴𝑏 𝑛
= − 𝑘𝑏 𝐶𝐴𝑏 − 𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐
𝑑𝑡
𝑛
𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐 = 𝑘𝑐 𝐶𝐴𝑐 + 𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶𝐴𝑒
𝑛
𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶𝐴𝑒 = 𝑘𝑒 𝐶𝐴𝑒

Plug in reaction rate in-terms of catalyst (𝑘𝑚 = 𝛾𝑚 𝑘𝑐𝑎𝑡 , where 𝑚 = 𝑏, 𝑐, 𝑒) to obtain:

𝒅𝑪𝑨𝒃
(Bubble balance): = − 𝜸𝒃 𝒌𝒄𝒂𝒕 𝑪𝑨𝒃 − 𝑲𝒃𝒄 𝑪𝑨𝒃 − 𝑪𝑨𝒄
𝒅𝒕

(Cloud balance): 𝑲𝒃𝒄 𝑪𝑨𝒃 − 𝑪𝑨𝒄 = 𝜸𝒄 𝒌𝒄𝒂𝒕 𝑪𝒏𝑨𝒄 + 𝑲𝒄𝒆 𝑪𝑨𝒄 − 𝑪𝑨𝒆

(Emulsion balance): 𝑲𝒄𝒆 𝑪𝑨𝒄 − 𝑪𝑨𝒆 = 𝜸𝒆 𝒌𝒄𝒂𝒕 𝑪𝒏𝑨𝒆

 Example: 1st Order Reaction (𝒏=1)
The three balance equations can be combined into one expression,

Rearrange Emulsion balance 𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶𝐴𝑒 = 𝛾𝑒 𝑘𝑐𝑎𝑡 𝐶𝐴𝑒 to obtain,

𝐾𝑐𝑒
𝐶𝐴𝑒 = 𝐶
𝛾𝑒 𝑘𝑐𝑎𝑡 + 𝐾𝑐 𝐴𝑐
𝑛
Substitute into the Cloud balance 𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐 = 𝛾𝑐 𝑘𝑐𝑎𝑡 𝐶𝐴𝑐 + 𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶𝐴𝑒 to
obtain,
𝑛 𝐾𝑐𝑒
𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐 = 𝛾𝑐 𝑘𝑐𝑎𝑡 𝐶𝐴𝑐 + 𝐾𝑐𝑒 𝐶𝐴𝑐 − 𝐶
𝛾𝑒 𝑘𝑐𝑎𝑡 + 𝐾𝑐 𝐴𝑐

𝐾𝑏𝑐
Rearrange to obtain, 𝐶𝐴𝑐 = 𝐾𝑐𝑒 𝛾𝑒 𝑘𝑐𝑎𝑡
𝐶𝐴𝑏
𝛾𝑐 𝑘𝑐𝑎𝑡 + + 𝐾𝑐𝑒 + 𝐾𝑏
𝛾𝑐 𝑘𝑐𝑎𝑡
𝑑𝐶𝐴𝑏
Substitute into the Bubble balance = − 𝛾𝑏 𝑘𝑐𝑎𝑡 𝐶𝐴𝑏 − 𝐾𝑏𝑐 𝐶𝐴𝑏 − 𝐶𝐴𝑐
𝑑𝑡

𝑑𝐶𝐴𝑏 𝐾𝑏𝑐 𝐶𝐴𝑏

= 𝛾𝑏 𝑘𝑐𝑎𝑡 𝐶𝐴𝑏 + 𝐶𝐴𝑏 −
𝑑𝑡 𝐾 𝛾𝑘
𝛾𝑐 𝑘𝑐𝑎𝑡 + 𝐾𝑏𝑐 + 𝑐𝑒 𝑒 𝑐𝑎𝑡 + 𝐾𝑐𝑒
𝛾𝑐 𝑘𝑐𝑎𝑡
 Overall transport coefficient
(𝒏=1)

𝑑𝐶𝐴𝑏 𝐾𝑏𝑐 𝐶𝐴𝑏

= 𝛾𝑏 𝑘𝑐𝑎𝑡 𝐶𝐴𝑏 + 𝐶𝐴𝑏 −
𝑑𝑡 𝐾 𝛾𝑘
𝛾𝑐 𝑘𝑐𝑎𝑡 + 𝐾𝑏𝑐 + 𝑐𝑒 𝑒 𝑐𝑎𝑡 + 𝐾𝑐𝑒
𝛾𝑐 𝑘𝑐𝑎𝑡

Rearrangement results in,

𝑑𝐶𝐴𝑏
− = 𝑘𝑐𝑎𝑡 𝐾𝑅 𝐶𝐴𝑏
𝑑𝑡

where the overall transport coefficient (𝐾𝑅 ) is,

1
𝐾𝑅 = 𝛾𝑏 +
𝑘𝑐𝑎𝑡 1
+ 1
𝐾𝑏𝑐
𝛾𝑐 +
1 𝑘
+ 𝑐𝑎𝑡
𝛾𝑒 𝐾𝑐𝑒
 Design Equation:
(𝑛 = 1 Rxn.)

Expressing 𝐶𝐴𝑏 as a function of conversion (𝑋), we obtain,

𝐶𝐴𝑏 = 𝐶𝐴0 1 − 𝑋

𝑑𝐶𝐴𝑏
By substituting this expression into Eqn. (− = 𝑘𝑐𝑎𝑡 𝐾𝑅 𝐶𝐴𝑏 ) we obtain,
𝑑𝑡

𝑑𝑋
= 𝑘𝑐𝑎𝑡 𝐾𝑅 (1 − 𝑋)
𝑑𝑡

Integration yields the fluidized bed design equation for 1st order reaction,

1
𝐿𝑛 = 𝑘𝑐𝑎𝑡 𝐾𝑅 𝑡
1−𝑋
 Bed height and catalyst weight
(𝑛 = 1 Rxn.)
The bed height (𝐻) necessary for this conversion is 𝐻 = 𝑡𝑈𝑏

𝑈𝑏 1
𝐻= 𝐿𝑛
𝑘𝑐𝑎𝑡 𝐾𝑅 1−𝑋

The catalyst weight (𝑊) is,

𝑊 = 𝐻𝜌𝑠 𝐴𝑠 (1 − 𝛿) 1 − 𝜖𝑚𝑓

𝜌𝑠 𝐴𝑠 𝑈𝑏 1 − 𝜖𝑚𝑓 1 − 𝛿 1
𝑊= 𝐿𝑛
𝑘𝑐𝑎𝑡 𝐾𝑅 1−𝑋
 Conclusion
Tutorial: Fluidisation