Unsteady State Operation in Trickle Bed Reactors

1
Unsteady State Operation in Trickle Bed
ReactorsModulation of input variables or
parameters to create unsteady state conditions to
achieve performance better than that attainable
with steady state operation Motivation and
Objectives

• Performance enhancement in existing reactors
• Design and operation of new reactors
• Lack of systematic experimental or rigorous
  modeling studies in lab reactors necessary for
  industrial application
• Experimentally investigate unsteady state flow
  modulation (periodic operation) for a test
  hydrogenation system
• Develop model equations incorporating multiphase,
  multicomponent transport that can simulate
  unsteady state operation

CREL
2
Strategies for Unsteady State Operation

• Flow Modulation (Gupta, 1985 Haure, 1990 Lee
  and Silveston, 1995)
• Liquid or gas flow
• Liquid/gas ON-OFF or HIGH-LOW flow
• Isothermal/non-isothermal/adiabatic conditions
• Composition Modulation (Lange, 1993)
• Periodic switching between pure or diluted
  liquid/gas
• Quenching by inert or product (adiabatic)
• Activity Modulation (Chanchlani, 1994 Haure,
  1994)
• Enhance activity due to pulsed component
• Removal of product from catalyst site
• Catalyst regeneration due to pulse

CREL
3
Possible Advantages of Unsteady State Operation

• Gas Limited Reactions
• Partial Wetting of Catalyst Pellets -Desirable
• Internal wetting of catalyst
• Externally dry pellets for direct access of gas
• Replenishment of reactant and periodic product
  removal
• Catalyst reactivation
• Liquid Limited Reactions
• Partial Wetting of Catalyst Pellets-Undesirable
• Achievement of complete catalyst wetting
• Controlled temperature rise and hotspot removal

CREL
4
Test Reaction and Operating Conditions
Alpha-methylstyrene hydrogenation to isopropyl
benzene (cumene)
Operating Conditions

• Superficial Liquid Mass Velocity 0.05-2.5
  kg/m2s
• Superficial Gas Mass Velocity
  3.3x10-3-15x10-3 kg/m2s
• Operating Pressure 30 -200 psig (3-15
  atm)
• Feed Concentration 2.5 - 30 (200-2400
  mol/m3)
• Feed Temperature 20-25 oC
• Cycle time, t (Total Period) 5-500 s
• Cycle split,s (ON Flow Fraction) 0.1-0.6
• Max. Allowed Temperature Rise 25 oC

L(peak)
(1-s)t
s
L(mean)
L (base)
t, (sec)
CREL
5
Comparison of Performance under Gas and Liquid
Limited Conditions
Gas Limited Conditions (g 20) Low Pressure,
High Liquid Feed Concentration
Liquid Limited Conditions (0.4 lt g lt2) High
Pressure, Low Liquid Feed Concentration
CREL
6
Effect of Cycle Split and Total Cycle Period on
Performance Enhancement
Gas Limited Conditions (g 20) Operating
Conditions Pressure30 psig Cycle Split (s)
Liquid ON Period/Total Cycle Period(t)
CREL
7
Effect of Liquid Mass Velocity and Total Cycle
Period on Unsteady State Performance
L (ON)
L (mean)
s
s
(1-s)
(1-s)
t, (sec)
L (ON)L(mean) /s
Enlargement of enhancement zone at lower mass
velocity
CREL
8
Effect of Liquid Reactant Concentration on
Performance Enhancement
L (ON)
L (mean)
s
s
(1-s)
(1-s)
t, (sec)
L (ON)L(mean) /s
Lower conversion at higher feed concentration
reduces enhancement even at lower liquid mass
velocity
CREL
9
Effect of Pressure on Steady and Unsteady State
Performance
L (ON)
L (mean)
s
s
(1-s)
(1-s)
t, (sec)
L (ON)L(mean) /s
g 3
g 24
At low mean liquid mass velocity, unsteady state
performance is higher than steady state even as
liquid limitation is reached
CREL
10
Effect of Cycling Frequency on Performance
Optimum cycling frequency depends upon feed
concentration, pressure and cycle split
CREL
11
Effect of Base-Peak Flow Modulation on
Performance Enhancement under Liquid Limited
Conditions
(1-s)t
L (mean)
s
L(peak)
L (base)
t, (sec)
L(mean) sL (peak)(1-s)L (base)
At high peak to base flow ratio, unsteady state
operation gives better performance even under
(near) liquid limited conditions (0.4ltg lt 2)
CREL
12
Phenomena occurring under unsteady state
operation with flow modulation in a trickle-bed
reactor
GOAL To Predict Velocity, Holdup, Concentration
and Temperature Profiles
CREL
13
The Model Structure
Bulk Phase Equations
Species
Energy
zL
CREL
14
Advantages of Maxwell-Stefan Multi-component
Transport Equations over Conventional Models

• Multicomponent effects are considered for
  individual component transport ks are
  matrices
• Bulk transport across the interface is
  considered
• Nt coupled to energy balance (non zero)
• Transport coefficients are corrected for high
  fluxes
• k corrected to ko kF
  exp(F)-I-1
• Concentration effects and individual pair binary
  mass transfer coefficients considered
• Thermodynamic non-idealities are considered by
  activity correction of transport coefficients
• Holdups and velocities are affected by interphase
  mass transport and corrected while solving
  continuity and momentum equations

CREL
15
Flow Model Equations
Momentum
uiL,uiG
Continuity
eL,eG,P
Pressure
Z
Staggered 1-D Grid
CREL
16
Stefan-Maxwell Flux Equations for Interphase Mass
and Energy Transport
Gas-Liquid Fluxes
Liquid-Solid and Gas-Solid Fluxes

• Bootstrap Condition for Multicomponent Transport
• Interphase Energy Flux for the Gas-Liquid
  Transport and Bulk to Catalyst
• Interface Transport
• Net Zero Volumetric Flux for Liquid-Solid and
  Gas-Liquid Interface for
• Intracatalyst Flux

CREL
17
Catalyst Level Equations
Approach I Rigorous Single Pellet Solution of
Intrapellet Profiles along with
Liquid-Solid and Gas-Solid Equations
CiCP
L
G
xc
Approach II Apparent Rate Multipellet Model
Solution of Liquid-Solid and Gas-Solid
Equations
CiCP
CiCP
CiCP
L
L
G
L
G
G
Type III Both Sides Externally Dry
Type I Both Sides Externally Wetted
Type II Half Wetted
CREL
18
Liquid Holdup and Velocity Profiles
Operating Conditions Liquid ON time 15 s, OFF
time65 s Liquid ON Mass Velocity 1.4
kg/m2s Liquid OFF Mass Velocity 0.067
kg/m2s Gas Mass Velocity
0.0192 kg/m2s
CREL
19
Transient Simulation Results Alpha-methylstyrene
Concentration Profiles
Alpha-methylstyrene Concentration during ON cycle
of flow modulation Feed Concentration 1484
mol/m3 Pressure 1 atm. Reaction
Conditions Gas Limited (g 25)
(Intrinsic Rate Zero order w.r.t. Alpha-MS)
CREL
20
Transient Cumene and Hydrogen Concentration
Profiles
Profiles show build up of cumene and hydrogen
concentration during the liquid ON part of the
cycle
CREL
21
Alpha-methylstyrene and Cumene Concentration
Profiles During Flow Modulation
Supply and Consumption of AMS and Corresponding
Rise in Cumene Concentration Operating
Conditions Cycle period40 sec, Split0.5
(Liquid ON20 s) Liquid ON Mass Velocity
1.01 kg/m2s Liquid OFF Mass Velocity
0.05 kg/m2s Gas Mass Velocity
0.0172 kg/m2s
CREL
22
Catalyst Level Hydrogen and Alpha-methylstyrene
Concentration Profiles During Flow Modulation
Concentration of Alpha-MS in previously dry
pellets during Liquid ON (120s, Wetted Catalyst
) and Liquid OFF(2040 s, Dry catalyst)
Concentration of Hydrogen during Liquid ON
(120s, Wetted Catalyst ) and Liquid OFF(2040
s, Dry catalyst) for negligible reaction test
case
CREL
23
Simulated Cycle Time and Cycle Split Effects on
Unsteady State Performance
Cycle Split and Cycle Period Effects Agree
Qualitatively with Experimental Results
CREL
24
Transient Fluid Dynamic Simulationusing CFDLIB
(Los Alamos)
2D-Test bed Dimensions 29.7x7.2 cm 33x8 (264
cells with preset porosity) Cycle period 60
s Cycle split 0.25 Liquid Velocity 0.1 cm/s
(central point source) Gas Velocity 10 cm/s
(uniform feed) Gas-Solid and Liquid-Solid Drag
Closure Two-Phase Ergun Equation
Bed Porosity (lighter areas higher porosity)
CREL
25
Liquid Holdup Comparison between Steady and
Unsteady Operation
t 15 s
t 25 s t40 s Steady
State Unsteady
State
CREL
26
Summary

• Performance enhancement was seen to be a strong
  function of the extent of
• reactant limitation
• Performance enhancement under gas limited
  conditions was found to be
• significantly dependent upon the cycle split,
  cycle period, liquid mass velocity
• and cycling frequency
• Performance enhancement under liquid limited
  conditions was observed only with
• BASE-PEAK flow modulation (to a lesser extent
  than under gas limited
• conditions)
• Rigorous modeling of mass and energy transport
  by Stefan-Maxwell equations
• and solution of momentum equations needed to
  simulate unsteady state flow,
• transport and reaction has been accomplished.
  Qualitative comparison with the
• experimental observations has been
  successfully demonstrated.
• The developed code can be used as a generalized
  simulator for any
• multicomponent, multi-reaction system and
  can be converted to a
• multidimensional code for large scale
  industrial reactors
• Fluid dynamic codes (CFDLIB) have been used to
  demonstrate better flow
• distribution under unsteady state operation.
  These codes would help achieve
• quantitative predictions when used in
  conjunction with the reaction transport

CREL
27
Recommendations for Future Work

• Downflow and Upflow Comparison
• Generalization of the conclusions obtained for
  complex reactions
• Steady and Unsteady State Models
• Implementation for multi-reaction problems and
  conversion to a flow
• sheet based package (ASPEN user model)
• Implementation for multi-dimensional test cases
  in the framework of
• CFD codes (CFDLIB or FLUENT)
• Unsteady State Experiments
• Testing of reaction networks for possible
  enhancement in selectivity via
• flow or composition modulation

CREL
28
Acknowledgements

• Advisors Prof. M. P. Dudukovic and Prof. M.
  Al-Dahhan
• Committee members Prof. B. Joseph, Prof. R. A.
  Gardner
• Dr. M. Colakyan (Union Carbide)
• Dr. R. Gupta (Exxon Research)
• CREL Industrial Sponsors
• Dr. Kahney, Dr. Chou, G. Ahmed (Monsanto)
• Dr. Patrick Mills (Du Pont)
• Engelhard, Eastmann Chemicals
• CREL Students and Research Associates
• Y. Wu, Y. Jiang
• Computer and Laboratory Support
• Dr. Y. Yamashita, Dr. S. Kumar,
• S. Picker, J. Krietler
• Parents, Roommates, and Friends

CREL