PERFORMANCE STUDIES OF TRICKLE BED REACTORS

1
PERFORMANCE STUDIES OF TRICKLE BED REACTORS

• Mohan R. Khadilkar
• Thesis Advisors M. P. Dudukovic and M. H.
  Al-Dahhan
• Chemical Reaction Engineering Laboratory
• Department of Chemical Engineering
• Washington University
• St. Louis, Missouri

CREL
2
Objectives and Accomplishments

• Examined the current state of the art in
  experimentation and
• modeling of trickle beds critically and
  successfully answered
• the often asked questions
• 1. Do upflow and downflow differ? When
  and Why ?
• 2. How to get reproducible scale-up data
  from small scale
• reactors independent of flow mode?
• 3. To what extent can current models
  predict the observed
• behavior?
• Formulated the rigorous approach to trickle bed
  modeling on
• pellet and reactor scale and illustrated the
  effectiveness of this
• approach for prediction of steady and
  unsteady state performance
• Examined experimentally and via models, unsteady
  state operation
• in trickle beds and identified regions of,
  and, reasons for
• performance enhancement

CREL
3
Performance of Trickle Bed Reactors
Chemical Kinetics
1. Comparison of Trickle Bed and Upflow
Effect of Bed Dilution Model Evaluation 2.
Rigorous Steady State Model 3. Unsteady State
Performance Experiments 4. Rigorous Unsteady
State Model 5. Transient Fluid Dynamic
Simulation
Fluid Dynamics
Phase Interaction Contacting
Transport Coefficients
Reactor Design Scale-up
Performance
CREL
4
Trickle Bed Reactors
Catalyst Wetting Conditions in Trickle Bed Reactor
Cocurrent Downflow of Gas and Liquid on a Fixed
Catalyst Bed
Operating Pressures up to 20 MPa Operating Flow
Ranges High Liquid Mass Velocity (Fully Wetted
Catalyst) (Suitable for Liquid Limited
Reactions) Low Liquid Mass Velocity (Partially
Wetted Catalyst) (Suitable for Gas Limited
Reactions)
Limiting Reactant criterion
Gas limited reaction if
Liquid limited reaction if
Flow Map (Fukushima et al., 1977)
CREL
5
FLOW REGIMES AND CATALYST WETTING
EFFECTSDOWNFLOW (TRICKLE BED REACTOR)
UPFLOW (PACKED BUBBLE COLUMN)
CREL
6
Motivation

• To understand the differences between downflow
  and upflow operation. Are upflow reactors
  indicative of trickle bed performance under
  different reaction conditions?
• To understand the effects of bed dilution with
  fines on reactor performance
• To develop guidelines for scale-up/scale-down of
  reactors for gas or liquid reactant limited
  reactions

Objectives

• Experimentally investigate the performance of
  downflow (Trickle Bed) and upflow (Packed Bubble
  Column) reactors for a test hydrogenation
  reaction
• Study the effects of pressure, feed
  concentration, gas velocity and bed dilution on
  the performance of both modes of operation
• Evaluate available reactor models in comparison
  with experimental data

7
Reaction Scheme
Catalyst 2.5 Pd on Alumina
(cylindrical 0.13 cm dia.) Fines Silicon
carbide 0.02 cm Range of Experimentation
Alpha-methylstyrene cumene
B (l) A(g) P(l)

• Superficial Liquid Velocity (Mass Velocity)
  0.09 - 0.5 cm/s (0.63-3.85 kg/m2s)
• Superficial Gas Velocity (Mass Velocity)
  3.8 -14.4 cm/s (3.3x10-3-12.8x10-3 kg/m2s)
• Feed Concentration 3.1 - 7.8 (230-600
  mol/m3)
• Operating Pressure 30 - 200 psig (3-15
  atm)
• Feed Temperature 24 oC

Limiting Reactant criterion
Gas limited reaction if
Liquid limited reaction if
CREL
8
Experimental Setup
CREL
9
Downflow and Upflow Experimental Results under
Gas and Liquid Limited Conditions without Fines
Downflow outperforms upflow due to partial
external wetting and improved gas reactant
access to particles
Upflow outperforms downflow due to more
complete external wetting and better transport
of liquid reactant to the catalyst
CREL
10
Downflow and Upflow Experimental Results under
Gas and Liquid Limited Conditions with Fines
ABOUT EQUAL PERFORMANCE DUE TO COMPLETE WETTING
Fines Packing Procedure Vol. of Fines Void
volume (Al-Dahhan et al. 1995)
CREL
11
Effect of Pressure and Gas Velocity on Performance
Transition to Liquid Limited Conditions
Negligible Effect of Gas Velocity
CREL
12
Slurry Kinetics
CREL
13
El- Hisnawi (1982) model

• Reactor scale plug flow equations
• Liquid phase gas reactant concentration
• Constant effectiveness factor
• Modified by external contacting efficiency
• Allowance for rate enhancement on
• externally dry catalyst
• Direct access of gas on inactively wetted
  pellets.

CREL
14
Beaudry (1987) model

• Pellet scale reaction diffusion equations
• For fully wetted and partially wetted slabs
• Effectiveness factor weighted based on
• contacting efficiency
• Overall effectiveness factor changes along
• the bed length
• Evaluation of overall effectiveness with change
  in
• concentration and contacting
• Overall Effectiveness factor at any location

CREL
15
Upflow and Downflow Performance at Low Pressure
(Gas Limited Condition)Experimental Data and
Model Predictions
CREL
16
Upflow and Downflow Performance at High Pressure
(Liquid Limited Conditions) Experimental Data
and Model Predictions
CREL
17
Summary

• DOWNFLOW OUTPERFORMS UPFLOW AT LOW PRESSURE.
• (Hydrogenation of alpha-methylstyrene is a gas
  limited reaction.
• Partial wetting is helpful in this situation.)
• UPFLOW OUTPERFORMS DOWNFLOW AT HIGH PRESSURE.
• (Hydrogenation of alpha-methylstyrene becomes a
  liquid limited reaction. Complete wetting is
  beneficial to this situation.)
• THE PREFERRED MODE FOR SCALE-UP (UPFLOW OR
  DOWNFLOW) DEPENDS ON THE TYPE OF REACTION SYSTEM
  AS WELL AS ON THE RANGE OF OPERATING CONDITIONS
  THAT AFFECT CATALYST WETTING.
• FINES NEUTRALIZE PERFORMANCE DIFFERENCES DUE TO
  MODE OF OPERATION AND REACTION SYSTEM TYPE ,
  DECOUPLE HYDRODYNAMICS AND KINETICS, AND HENCE
  ARE TO BE PREFERRED AS SCALE-UP TOOLS.
• THE TESTED MODELS PREDICT PERFORMANCE WELL
• (although improvements in mass transfer
  correlations are necessary)

CREL
18
Drawbacks of Evaluated Models

• Isothermal Operation
• Liquid Volatility Effects not Considered
• Coupling of Hydrodynamics and Transport Ignored
• Fully Internally Wetted Pellets Assumed
• Single Component, Dilute Solution Transport
  Assumed
• Multicomponent Effects not Considered

Simplified Models Accounting for Some of the
Above Effects

• Pellet Scale (Level I) Model (Kim and Kim, 1981
  Harold,1988)
• Reactor Scale (Level II) Model (LaVopa and
  Satterfield,1988 Kheshgi et al., 1992)

Extended Steady State Rigorous (Level III) Model
Test Reaction System Hydrogenation of
Cyclohexene
CREL
19
L-III Model Reactor Scale Equations (1-D)
Continuity
Species
Momentum
Energy
Fluxes Modeled by Multicomponent Stefan-Maxwell
Formulation
CREL
20
L-III Model Catalyst Scale Equations (Extension
of Harold,1988)
Liquid Filled Zone
Fully Externally Wetted, Partially Liquid Filled
Pellet
Gas Filled Zone
Intra-catalyst G-L Interface Continuity of
temperature, mass and energy fluxes, and
equilibrium relations between compositions
CREL
21
Simulation Results Multiplicity Effects

• Hysteresis Predicted
• Two Distinct Rate Branches Predicted
• (As Observed by Hanika, 1975)
• Branch Continuation, Ignition and
• Extinction Points
• Wet Branch Conversion (30 )
• Dry Branch Conversion (gt 95 )

• Wet Branch Temperature Rise (10-15 oC)
• Dry Branch Temperature Rise (140-160 oC)

CREL
22
Wet Branch Simulation
Reactor Scale Hydrodynamics
Reactor Scale Species Concentrations
Pellet Scale Multicomponent Flux Profiles
CREL
23
Dry Branch Simulation
Pellet Scale Pressure Profiles
Reactor Scale Hydrodynamics
Reactor Scale Species Concentrations
Pellet Scale Temperature Profiles
CREL
24
Intra-reactor Wet-Dry Transition

• Abrupt drop in liquid flow
• Temperature rise after liquid-
• gas transition

• Abrupt change in catalyst wetting
• Cyclohexene and cyclohexane
• mole fraction shows evaporation
• and reaction

CREL
25
Summary

• The reactor scale variation of phase holdups and
  velocities, multiplicity, and temperature rise
  were successfully simulated by the developed
  (LIII) model more rigorously than other models
  developed earlier
• Both wet and dry branch temperature, conversion,
  and corresponding fluxes were successfully
  modeled by the set of equations developed
• Intra-pellet reaction-transport equations for the
  wet and dry zones in presence of multicomponent
  interactions, evaporation, and condensation were
  successfully modeled
• The wet to dry transition prediction requires
  robust numerical techniques to yield stable
  solution at pellet scale (for the LIII model),
  but depicts the observed abrupt transition with a
  reactor scale (LII) model

CREL