Nomenclature Presentation

A Computational Fluid Dynamics Study of
Fluid Catalytic Cracking Cyclones
John W. McTernan and Issam Abu-Mahfouz

Buell Fisher-Klosterman, Penn State University Harrisburg
Lebanon, PA Middletown, PA
Fluidized Catalytic Cracking (FCC)
Reactor Regenerator
• Vaporized oil feedstock is • Coke is built up on the
exposed to a catalyst heated catalyst during the reaction.
to approximately 1,000 deg F • The catalyst is exposed to
- causing the cracking oxygen in the regenerator
reaction. igniting combustion.
• Usually takes place in a riser • The catalyst is
“regenerated” by the
pipe
combustion of the coke.
• Cyclones are used to • Cyclones are used to
separate the reacted oil separate the catalyst from
feedstock from the catalyst. the flue gas.
• Reaction is endothermic. • Reaction is exothermic.
FCC Refinery Flow Diagram
Chemisty of Cracking
• Alkanes (paraffins) are long chain single bonded

hydrocarbons.
• Cracking breaks the long chain into usable short
chain alkanes and alkenes.
o alkanes used for gasoline and fuel oil
o alkenes have a double covalent bond: ethanol, alcohol, and plastics
Reactor and Regenerator
• Considered the heart of a refinery.
• Reactor and Regenerator are thermally balanced.

• heat generated from the exothermic reaction in the regenerator, supplies the
endothermic reaction in the reactor.
• Catalyst used to facilitate the reaction is expensive and

abrasive.
• Shutdowns are required to repair cyclone lining from

erosion.
Cyclone
Nomenclature
FCC Cyclone
Installation
Cyclone Basics
2 Stage Cyclone System
Dense Flow Cyclones (High Mass Loading)

•Large quantities of particulate enter the
cyclone.
•1st Stage Cyclone.
Dilute Flow Cyclones (Low Mass Loading)

•Loading is approximately 5% of a
dense flow cyclone.
•2nd Stage Cyclone
Three Major Parameters of Cyclone Performance
1. Efficiency
• Percent of particles captured of the particles that enter.
• Typical ranges from 95% to 99%.
2. Pressure Drop
• Pressure loss due to cyclone operation.
• Cost to operate system.
3. Erosion
• Catalyst is highly abrasive.
• Maintenance costs.
• Downtime.
Methods of Cyclone Design
1) Laboratory Modeling
2) Algebraic Equations
3) Computational Fluid Dynamics (CFD)

Laboratory Modeling
• Laboratory modeling and empirical testing have been

the major methods for developing cyclone
technology.
• Scale models were created and results verified by

industrial performance.
o Particulate tests
o Pressure Loss
o Erosion
Algebraic Models
• Developed from
empirical data.
• Rigid equations that

can only be used for
narrow applications.
Computation Fluid Dynamics (CFD) Models
• Flow is represented by the Navier-Stokes equation.
o Coupled differential equations that must be solved

simultaneously.
• The Navier-Stokes equation is not closed for turbulent

flow.
o Approximations, averages, and constants from
testing are used to solve the equations for turbulent
flow.
Turbulent Flow Models
1. k-ε
• k – turbulent kinetic energy
• ε - dissipation rate of k
• Assumes isotropic turbulence
• Fast and consistent convergence
2. Renormalized Group (RNG) k-ε

• Adds an additional variable to the standard k-ε
model
Turbulent Flow Models
3. Reynolds Stress Model (RSM)
• Incorporates transport equations for each Reynolds
stress dissipation component
• Forgoes isotropic turbulence assumption
• 7 equations
• High computation costs
4. Large Eddy Simulation (LES)

• Transient
Cyclones
standard design modified outlet tube high efficiency cyclone

Cyclone Sizing and Dimensions
Model Cyclone 1 Cyclone 2 Cyclone 3

Description Variable
Input (inches) (inches) (inches)
Barrel
Diameter D D 50 50 66
Barrel Height M M/D 1.3 1.3 1.3

Cone Height L L/D 2.07 2.07 2.07
Hopper
N N/D 1.63 1.63 1.63
Height
Cone
Diameter T T/D 0.4 0.4 0.4
Vortex Finder
Diameter F F/D 0.535 0.35 0.356
Inlet Width A A/D 0.285 0.285 0.216

Inlet Height B B/D 0.64 0.64 0.4848
Vortex Structure
• Can provide useful information regarding erosion and
collection efficiency.
• There is not an exact apex or termination point of the

vortex.
• Subjective to how the profile is portrayed
• Vortex can dissipate after entry into the hopper
• Following the gas path can give clues about the particle
path.
• Vortex length is related to erosion in the hopper and

dipleg.
Axial velocity at (a) 40 fps inlet
Axial velocity at (b) 60 fps inlet.
Cyclone 2: Axial Velocity
40fps 60fps 80fps

Velocity: Cyclone 1
60fps Inlet
Velocity Magnitude Tangential Velocity Axial Velocity

Cyclone 1 at 40inches below cyclone roof, 60fps inlet velocity
Velocity magnitude – Tangential velocity – Axial velocity

Cyclone 1 Cyclone 2 Cyclone 2
Velocity magnitude at 40inches below cyclone roof, 60fps inlet

velocity
Pressure Loss
 Pressure loss of a cyclone is often inversely related to

cyclone efficiency.
 Increasing the inlet velocity of a cyclone will often

increase the efficiency, but at the cost of increasing
the pressure loss.
 It has been determined that approximately 80% of

the pressure loss is due to the viscous stresses.
 The remaining loss of pressure occurs at the inlet and

outlet of the cyclone and by frictional forces along
the cyclone wall.
Static pressure at (a) 40 fps
Static pressure at (b) 60 fps
Pressure Loss
Cyclones 1 and 3 show high pressure regions along the
walls. This implies that pressure loss due to drag is
dominant.
Cyclone 2 has a modified outlet tube – meaning the

outlet area is much less than the inlet area. This
increases the outlet velocity.
COMSOL produces pressure loss with an average percent

error of about 3% with the error decreasing with
increasing inlet velocity.
Conclusions
• The simple geometry of cyclones is deceptive.

• Turbulent flow is highly complex.
• Small changes in geometry can produce major effects on performance.
• Cyclones are highly efficient making efficiency

increases difficult to detect.
• Model is lacking the effects of anistropic turbulence

Conclusions
• Cyclones push the limits of CFD code and solvers.

• Anisotropic turbulence requires very robust and
computationally costly solvers.
• Need to model particle influence on the gas stream
• All parameters of cyclone performance must be

considered during design.

Nomenclature Presentation

Uploaded by

Copyright:

Available Formats

Nomenclature Presentation

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nomenclature Presentation

Uploaded by

Copyright:

Available Formats

A Computational Fluid Dynamics Study of

Fluid Catalytic Cracking Cyclones

John W. McTernan and Issam Abu-Mahfouz

• Alkanes (paraffins) are long chain single bonded

• Considered the heart of a refinery.

• Reactor and Regenerator are thermally balanced.

• Catalyst used to facilitate the reaction is expensive and

• Shutdowns are required to repair cyclone lining from

Dense Flow Cyclones (High Mass Loading)

Dilute Flow Cyclones (Low Mass Loading)

3) Computational Fluid Dynamics (CFD)

• Laboratory modeling and empirical testing have been

• Scale models were created and results verified by

• Rigid equations that

• Flow is represented by the Navier-Stokes equation.

o Coupled differential equations that must be solved

• The Navier-Stokes equation is not closed for turbulent

2. Renormalized Group (RNG) k-ε

4. Large Eddy Simulation (LES)

standard design modified outlet tube high efficiency cyclone

Model Cyclone 1 Cyclone 2 Cyclone 3

Barrel Height M M/D 1.3 1.3 1.3

Inlet Width A A/D 0.285 0.285 0.216

• There is not an exact apex or termination point of the

• Vortex length is related to erosion in the hopper and

40fps 60fps 80fps

Velocity Magnitude Tangential Velocity Axial Velocity

Velocity magnitude – Tangential velocity – Axial velocity

Velocity magnitude at 40inches below cyclone roof, 60fps inlet

 Pressure loss of a cyclone is often inversely related to

 Increasing the inlet velocity of a cyclone will often

 It has been determined that approximately 80% of

 The remaining loss of pressure occurs at the inlet and

Cyclone 2 has a modified outlet tube – meaning the

COMSOL produces pressure loss with an average percent

• The simple geometry of cyclones is deceptive.

• Cyclones are highly efficient making efficiency

• Model is lacking the effects of anistropic turbulence

• Cyclones push the limits of CFD code and solvers.

• All parameters of cyclone performance must be

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.