Trickle Bed Reactor

3292 Ind. Eng. Chem. Res.
1997, 36, 3292-3314
High-Pressure Trickle-Bed Reactors: A Review

Muthanna H. Al-Dahhan,*,† Faical Larachi,‡ Milorad P. Dudukovic,† and
Andre Laurent§
Chemical Reaction Engineering Laboratory, Department of Chemical Engineering, Washington University,
Campus Box 1198, One Brookings Drive, St. Louis, Missouri 63130, Department of Chemical Engineering,
Laval University & CERPIC, Québec, Canada G1K 7P4, and Laboratorie des Sciences du Genie Chimique,
1 Rue Grandville BP, 451-54001 Nancy Cedex, France
A concise review of relevant experimental observations and modeling of high-pressure trickle-

bed reactors, based on recent studies, is presented. The following topics are considered: flow
regime transitions, pressure drop, liquid holdup, gas-liquid interfacial area and mass-transfer
coefficient, catalyst wetting efficiency, catalyst dilution with inert fines, and evaluation of trickle-
bed models for liquid-limited and gas-limited reactions. The effects of high-pressure operation,
which is of industrial relevance, on the physicochemical and fluid dynamic parameters are
discussed. Empirical and theoretical models developed to account for the effect of high pressure
on the various parameters and phenomena pertinent to the topics discussed are briefly described.
Introduction are their impracticality for reactions with rapidly

deactivating catalysts, such as in heavy oil hydrotreat-
Among the three-phase gas-liquid-solid reaction ing processes, and the possibility of liquid maldistribu-
systems encountered in industrial practice, trickle-bed tion, which may give rise to hot spots and reactor
reactors (TBRs) are the most widely used. They are runaway (McManus et al., 1993). The later, as shown
employed in petroleum, petrochemical, and chemical for example by McManus et al. (1993), could be brought
industries, in waste treatment, and in biochemical and about by in-pore heat generated within liquid-filled
electrochemical processing, as well as other applications. catalyst pellets whence heat withdrawal is curbed by
Table 1 lists some of the processes carried out in TBRs. the surrounding dry conditions. In order to exploit TBR
The economic impact of how well these reactors operate advantages in the presence of relatively slow catalyst
is considerable, since in the petroleum sector alone the deactivation, special reactor designs employing piston-
TBR annual processing capacity for various hydrotreat- like through-flow catalyst moving beds are used (Euzen
ments (e.g., hydrodesulfurization, hydrocracking, hy- et al., 1993).
drorefining, hydrodemetallization, hydrodenitrogena-
Most commercial TBRs normally operate adiabati-
tion, etc.) is estimated at ca. 1.6 billion metric tons
cally at high temperatures and high pressures and
(Trambouze, 1991). With the current market evolution
generally involve hydrogen and nonwaterlike liquids
toward increasing demand for light oil products, such
with superficial gas and liquid velocities up to 30 and 1
as middle distillates, and the decreasing needs for heavy
cuts, the refiners will have to keep improving their cm/s, respectively. Kinetics and/or thermodynamics of
processing units for upgrading heavy oil and residual reactions conducted in TBRs require high temperatures,
feedstocks (Trambouze, 1993). Any advance in TBR which in return increase gas expansion and impede the
technology will thus represent substantial savings, and gaseous reactant from dissolving sufficiently into the
this stimulates the continued research efforts aimed at liquid. Therefore, elevated pressures (up to 30 MPa)
improving TBR operation and performance. are necessary to improve the gas solubility and the
mass- and heat-transfer rates, to handle large gas
A trickle-bed reactor (TBR) consists of a fixed bed of
volumes at less capital expense, and to slow down the
catalyst particles contacted by a cocurrent downward
catalyst deactivation which may be triggered by hydro-
gas-liquid flow carrying both reactants and products.
gen starvation of the catalyst. Due to complexities
When the gas and liquid are fed cocurrently upward
associated with transport-kinetics coupling in TBRs,
through the catalyst bed, the system is called a flooded-
general scale-up and scale-down rules for the quantita-
bed reactor (FBR) or upflow reactor. The upflow con-
tive description of transport phenomena in TBRs work-
figuration is used sparingly in industrial practice where
ing under realistic conditions remain elusive. This is
TBRs prevail. Owing to a motionless catalyst bed,
especially true since the majority of the literature is
nearly plug flow is achieved in TBRs, and in that respect
concerned exclusively with nearly atmospheric condi-
they are superior to other three-phase reactors where
tions. Table 2 lists the papers on trickle-bed reactors
the catalyst is either slurried or fluidized. For instance,
which review various aspects of TBRs and summarize
TBRs’ high catalyst loading per unit volume of the liquid
and low energy dissipation rate make them preferable the progress in a specific area related to TBR perfor-
to slurry reactors. However, the disadvantages of TBRs mance. However, only the three most recent entries in
Table 2 made reference to the behavior of pressurized
TBRs (Gianetto and Specchia, 1992; Martinez et al.,
* To whom correspondence should be addressed. 1994; Saroha and Nigam, 1996), whereas all the others
† Washington University. Phone: (314)-935-7187, (314)-935-
dealt with TBR research at atmospheric conditions. In
6021. Fax: (314)-935-4832. e-mail: muthanna@ view of the rapid advances that are being realized in
wuche.wustl.edu, dudu@wuche.wustl.edu.
‡ Laval University. Phone: (418)-656-3566. Fax: (418)-656- the area of high-pressure TBRs, it is deemed appropri-
5993. e-mail: flarachi@gch.ulaval.ca. ate to supplement the information contained in the
§ Laboratorie des Sciences du Genie Chimique. Phone: 33- papers referenced in Table 2 with the description of the
3-83-35-21-21. Fax: 33-3-83-32-29-75. e-mail: alaurent@ensic.u- new developments on high-pressure trickle-beds, and
nancy.fr. this is the objective of this review. Although this review
S0888-5885(97)00082-1 CCC: $14.00 © 1997 American Chemical Society
Ind. Eng. Chem. Res., Vol. 36, No. 8, 1997 3293
Table 1. Some Examples of Three-Phase Reactions Carried out in Trickle-Bed Reactors
residuum and vacuum residuum desulfurization for the production of low-sulfur fuel oils (Meyers, 1996)
catalytic dewaxing of lubestock cuts to produce fuel or lube products for extremely cold conditions (Meyers, 1996)
sweetening of diesel, kerosene, jet fuels, heating oils (Meyers, 1996)
hydrodemetallization of residues (Trambouze, 1993; Euzen et al., 1993)
hydrocracking for production of high-quality middle-distillate fuels (Meyers, 1996)
hydrodentrification (Meyers, 1996)
isocracking for the production of isoparaffin-rich naphtha (Meyers, 1996)
production of lubricating oils (Meyers, 1996)
selective hydrogenation of butadiene to butene, of vinylacetylene to butadiene, of acetylene; of alkylanthraquinone to
hydroquinone for the production of hydrogen peroxide (Charpentier, 1976; Shah, 1979)
other catalytic hydrogenations: of nitro compounds, of carbonyl compounds, of carboxylic acid to alcohols,
of benzene to cyclohexane, of aniline to cyclohexylaniline, of glucose to sorbitol, of coal liquefaction extracts,
of benzoic acid to hydrobenzoic acid, of caprolactone to hexanediol, of organic esters to alcohols (Germain et al., 1979)
synthesis of butynediol from acetylene and aqueous formaldehyde (Gianetto and Speechia, 1992)
immobilized enzyme reactions (Belhaj, 1984)
VOC abatement in air pollution control (Diks and Ottengraf, 1991; Chen and Chuang, 1992)
wet air oxidation of formic acid, acetic acid, and ethanol (Goto and Smith, 1975; Levec and Smith, 1976;
Goto and Mabuchi, 1984; Baldi et al., 1985)
oxidation of SO2 (Hartman and Coughlin, 1971; Mata and Smith, 1981; Haure et al., 1990; Kiared and Zoulalian, 1992);
oxidation of glucose (Tahraoui, 1990); absorption of lean CO2 and H2S in caustic alkali solution;
absorption of lean NH3 in an aqueous solution of H2SO4 and H3PO4 (Charpentier, 1976)
biochemical reactions, fermentations (Bailey and Ollis, 1986)
is not exhaustive, all the recent experimental and tension (σ), gas solubility (C*), and Henry’s constant (H)
theoretical research on TBRs operated at elevated are the most sensitive to pressure. In the pressure
pressure is summarized. range relevant to TBR operation, i.e., subcritical and
The primary focus is on the developments realized above atmospheric conditions, it is anticipated that
during the past decade, mostly by the research groups pressure mainly modifies the gas-phase properties,
of the University of Twente in The Netherlands, of the whereas the liquid remains sensitive only to tempera-
Chemical Reaction Engineering Laboratory at Wash- ture. Without loss of generalization, we will illustrate
ington University, St. Louis, MO, and of Nancy in in what follows how hydrogen properties evolve with
France. The experimental conditions, the parameters pressure; gases other than H2 should, in principle,
measured in high-pressure TBRs, and with other rel- qualitatively follow similar trends. Figure 1 shows the
evant studies regarding high-pressure TBRs are sum- variation of properties of H2 in the pressure range [1-20
marized in Table 3. The following topics, in which MPa] and a temperature around 410 K (L’Air Liquide,
advances have been made, will be discussed: (1) pres- 1976; Sebastian et al., 1981; Reid et al., 1987). Estima-
sure effect on physicochemical properties; (2) phenom- tion of these properties as a function of pressure is
enological analysis of two-phase flow; (3) flow regime discussed in Reid et al. (1987) where a number of
transition; (4) single-phase pressure drop; (5) two-phase appropriate correlations are recommended. Apart from
pressure drop; (6) liquid holdup; (7) gas-liquid inter- FG, DG, C*, and H, the remaining properties are practi-
facial area and mass-transfer; (8) catalyst wetting cally independent of pressure. Gas density increases
efficiency; (9) catalyst dilution with inert fines in almost proportionally to pressure (curve 1). It can also
laboratory-scale high-pressure TBRs; (10) liquid-limited be varied by using gases of different molecular weight.
and gas-limited reactions in high-pressure TBRs; (11) The effect of increasing pressure is to reduce the gas-
evaluation of TBR models. side molecular diffusivity. Whether this decrease is
It is hoped that this paper will stimulate additional sharp or not depends on the value of the reduced
research on high-pressure trickle-bed reactors, which mixture temperature Tmix,r of the gas phase (Reid et al.,
is needed to further advance our understanding of this 1987), DP ≡ Pn. For Tmix,r around 2.5, which is the case
reactor type, and that such research will ultimately in our example for H2 (curve 3), the diffusivity is
result in more advanced fundamentally based TBR inversely proportional to pressure (n ) 0). For Tmix,r <
models. Professor Froment and his co-workers have 2.5, n is less than 0, whereas n is slightly positive for
advocated for many decades the fundamental approach Tmix,r > 2.5 (Reid et al., 1987). The decrease in the mole
in the description of catalytic kinetics and have suc- fraction of H2 dissolved/mole fraction of H2 in the vapor
ceeded in demonstrating the advantages of such an phase with pressure (curve 5) implies that the solute
approach. A number of the catalytic chemistries that (i.e., hydrogen) solubility in the liquid also increases as
they studied have to be conducted in trickle-beds, such the solute mole fraction in the vapor phase increases.
as hydrodesulfurization (Froment et al., 1994; Vanrys- Curve 6 (mole fraction of H2 dissolved in tetralin/H2
selberghe and Froment, 1996), and, hence, it is neces- partial pressure vs pressure) illustrates that the solubil-
sary to bring the TBR models to the same level that ity of the solute increases almost linearly as the solute
the kinetic description of the catalytic processes occur- partial pressure increases. In summary, we can state
ring in them has achieved. We hope that this paper that among the gas-phase properties which are likely
will provide a motivation for such an endeavor. to be most affected by pressure are gas density and
diffusivity, gas solubility in the liquid, and Henry’s
1. Pressure Effect on the Fluids constant.
Physicochemical Properties
2. Analysis of Two-Phase Flow Phenomena in
Pressure affects the physicochemical properties of High-Pressure TBRs
gases and liquids and through this influences the fluid
dynamics and transport in fixed-bed reactors. Density In an attempt to interpret the effect of pressure, or
(F), molecular diffusivity (D), dynamic viscosity (µ), gas density, on the hydrodynamics of trickle beds, two
thermal conductivity (κ), heat capacity (Cp), surface types of models have been advanced. In model I,
3294 Ind. Eng. Chem. Res., Vol. 36, No. 8, 1997
Table 2. Articles on Trickle-Bed Reactors That Contain Reviews of Some Aspects of Trickle-Bed Operation
ref title
Østergaard (1968) Gas-Liquid-Particle Operations in Chemical Reaction Engineering
Satterfield (1975) Trickle-Bed Reactors
Schwartz et al. (1976) A New Tracer Method for Determination of Solid-Liquid Contacting Efficiency in
Trickle-Bed Reactors
Van de Vusse and Wesselingh (1976) Multiphase Reactors
Weekman (1976) Hydroprocessing Reaction Engineering
Charpentier (1976) Recent Progress in Two-Phase Gas-Liquid Mass Transfer in Packed Beds
Goto et al. (1977) Trickle-Bed Oxidation Reactors
Dwivedi and Upadhyay (1977) Particle-Fluid Mass Transfer in Fixed and Fluidized Beds
Dudukovic (1977) Catalyst Effectiveness Factor and Contacting Efficiency in Trickle-Bed Reactors
Hofmann (1977) Hydrodynamics, Transport Phenomena, and Mathematical Models in Trickle-Bed Reactors
Hofmann (1978) Multiphase Catalytic Packed-Bed Reactors
Gianetto et al. (1978) Hydrodynamics and Solid-Liquid Contacting Effectiveness in Trickle-Bed Reactors
Satterfield et al. (1978) Liquid-Solid Mass Transfer in Packed Bed with Downward Cocurrent Gas-Liquid Flow
Specchia et al. (1978) Solid-Liquid Mass Transfer in Cocurrent Two-Phase Flow through Packed Beds
Charpentier (1978) Gas-Liquid Reactors
Dudukovic and Mills (1978) Catalyst Effectiveness Factors in Trickle-Bed Reactors
Baker (1978) Determination of the Extent of Catalyst Utilization in a Trickle-Flow Reactor
Dirkx (1979a,b) De Trickle-Bed Reactor
Charpentier (1979) Hydrodynamics of Two-Phase Flow Through Porous Media
Shah (1979) Gas-Liquid-Solid Reactor Design
Van Landeghem (1980) Multiphase Reactors: Mass Transfer and Modeling
Baldi (1981a) Design and Scale-up of Trickle-Bed Reactors. Solid-Liquid Contacting Effectiveness
Baldi (1981b) Heat Transfer in Gas-Liquid-Solid Reactors
Baldi (1981c) Hydrodynamics of Multiphase Reactors
Koros (1981) Scale-up Considerations for Mixed Phase Catalytic Reactors
Turek and Lange (1981) Mass Transfer in Trickle-Bed Reactors at Low Reynolds Number
Morsi et al. (1981) Hydrodynamics and Gas-Liquid-Solid Interfacial Parameters of Cocurrent Downward
Two-Phase Flow in Trickle-Bed Reactors
Charpentier (1982) What’s New in Absorption with Chemical Reaction
Morsi et al. (1982) Hydrodynamics and Interfacial Areas in Downward Cocurrent Gas-Liquid Flow
through Fixed Beds. Influence of the Nature of the Liquid
Tan and Smith (1982) A Dynamic Method for Liquid-Particle Mass Transfer in Trickle Beds
Herskowitz and Smith (1983) Trickle-Bed Reactors: A Review
Ramachandran and Chaudhari (1983) Three-Phase Catalytic Reactors
Crine and L’Homme (1983) Recent Trends in the Modelling of Catalytic Trickle-Bed Reactors
Germain (1983) Industrial Applications of Three-Phase Catalytic Fixed Bed Reactors
Hofmann (1983) Fluid Dynamics, Mass Transfer and Chemical Reaction in Multiphase Catalytic Fixed
Bed Reactors
Tarhan (1983) Catalytic Reactor Design
Mills and Dudukovic (1984) A Comparison of Current Models for Isothermal Trickle-Bed Reactors. Application of a
Model Reaction System
Gupta (1985) Handbook of Fluids in Motion
Rao and Drinkenburg (1985) Solid-Liquid Mass Transfer in Packed Beds with Cocurrent Gas-Liquid Downflow
Hanika and Stanek (1986) Operation and Design of Trickle-Bed Reactors
Dudukovic and Mills (1986) Contacting and Hydrodynamics in Trickle-Bed Reactors
Charpentier (1986) Mass Transfer in Fixed Bed Reactors
Gianetto and Berruti (1986) Modelling of Trickle-Bed Reactors
Ramachandran et al. (1987) Recent Advances in the Analysis and Design of Trickle-Bed Reactors
Ng and Chu (1987) Trickle-Bed Reactors
Lemcoff et al. (1988) Effectiveness Factor of Partially Wetted Catalyst Particles: Evaluation and Application
to the Modeling of Trickle Bed Reactors
Zhukova et al. (1990) Modeling and Design of Industrial Reactors with a Stationary Bed of Catalyst and
Two-Phase Gas-Liquid FlowsA Review
Levec and Lakota (1992) Liquid-Solid Mass Transfer in Packed Beds with Cocurrent Downward Two-Phase Flow
Wild et al. (1992) Heat and Mass Transfer in Gas-Liquid-Solid Fixed Bed Reactors
Gianetto and Specchia (1992) Trickle-Bed Reactors: State of Art and Perspectives
Martinez et al. (1994) Three-Phase Fixed Bed Catalytic Reactors: Application to Hydrotreatment Processes
Saroha and Nigam Trickle-Bed Reactors
changes in parameters such as liquid holdup, wetting forces acting on liquid flow. The driving forces consist
efficiency, and gas-liquid interfacial area have been of the pressure gradient (∆P/Z), the gas-liquid inter-
highlighted using bed-scale force considerations (Al- facial drag force (τLGa), and the gravitational force (FLg).
Dahhan and Dudukovic, 1994, 1995; Wammes and A simple overall force balance on the gas phase shows
Westerterp, 1991; Wammes et al., 1991b) . In model that the pressure gradient is proportional to the gas-
II, pore-scale force considerations have been used (Lara- liquid interfacial drag; consequently, the dimensionless
chi et al., 1992; Lara-Marquez et al., 1992; Cassanello pressure gradient ∆P/FLgZ solely determines whether
et al., 1996). These models can be briefly described as liquid flow will be gravity-driven or not. The pressure
follows: gradient depends, besides the bed characteristics, on the
Model I. The effect of high pressure and gas flow velocities of both phases and on the physicochemical
rate on the energy dissipation is the central idea based properties of the flowing fluids. As indicated in the
on which model I is formulated. It can be used to analysis presented above, regarding the fluids physico-
explain the changes brought by pressure in liquid chemical properties, mainly gas density is influenced
holdup, wetting efficiency, and gas-liquid interfacial by pressure. Thus, for given gas and liquid velocities,
area. The energy dissipation is due to the resisting a higher gas density produces a higher interfacial drag
frictional forces at the packing surface and the driving force or equivalently a higher pressure gradient. The
Table 3. Studies of the High-Pressure Cocurrent Downflow and Upflow Gas-Liquid Packed-Bed Reactors
investigators liquid/gas bed properties pressure (MPa) flow rates parameter
Abbott et al. (1967)V C4H10 packing diameter (mm): 1.16 2.41-3.79 not given βnc
N2 column diameter/height (mm): not given
Saada (1975)v H2O packing diameter (mm) 1.36 UL: 0.001-0.4 ∆P/Z; βe
N2 GB: 0.514; 0.974; 2.0640 UG: 0.004-1.8
column diameter/height (mm): 45.2/400
Kohler and Richarz (1984, 1985)V H 2O packing diameter (mm) 1 ReL: 0.1-8 βt; β3; βc
{organic solution GB: 1.5 ReG: 0-10
H2 Al2O3 sphere and cylinder: 1.5; 3
{N2 column diameter/height (mm): 30/800
Hasseni et al. (1987)V C6H12 packing diameter (mm) 0.1-10.1 L: 0.7-17 ∆P/Z; trickle-to-pulse flow
H2O GB: 1.3; 2; 3 G: ≈0-11 transition UGtrans(UL,FG)
gas oil column diameter/height (mm): 23/1000
ETG
{
N2
Ellman et al. (1990)V Hasseni and Larachi high-pressure data + ca. 4000 atmospheric data ∆P/Z; βnc
Ruecker and Agkerman (1987)V biphenyl not given 5.2 (295 °C) QL: 0.6 mL/s ηCE
H 2, N 2 QG: <10 std mL/s
van Gelder and Westerterp (1990)v CH3OH packing diameter (mm) 0.2-1.2 UL: 0.02 × 10-3-0.15 × 10-3 β t; D z
H2 glass cylinder: 3.8 × 4.8 UG: 0-0.0145
column diameter/height (mm): 65/500
Oyevaar et al. (1989)v H2O + DEA packing size (mm) 0.15-1.3 U L: 0 kLa; a; βe
CO2 + N2 glass cylinder: 3.8 × 4.8 UG: 0.01-0.09
column diameter/height (mm): 85.5/630
Wammes et al. (1990, 1991)V EtOH packing diameter (mm) 0.2-7.5 L: up to 7 ∆P/Z; βnc; βs; a; trickle-to-pulse
H2O GB: 3 G: up to 3 & pulse-to-dispersed bubble
H2O + 40% ETG Al2O3 cylinder: 3.2 × 3.2 flow UGrans(UL,FG)
H2O + 2 mol/L DEA column diameter/height (mm): 51/530-2620
H2O + 1.5 mol/L DEA + 40% ETG
{ He
N2
{ N2 + CO2
Ring and Missen (1991)V heavy oil + DBT packing diameter (mm) Lt; La; Li; Lmc; ηCE
H2 HDM catalyst: 2.17 10 (330-370 °C) L: 0.092-0.557
column diameter/height (mm): 25/510 G: 53 × 10-5-32 × 10-4
Tsamatsoulis and Papayannakos sulfur-containing HVGO packing size (mm) 5 (350-400 °C) L: 0.08-0.8 Le; La; Li; Dz
(1994, 1995), Papayannakos et al.{ sulfur-free HVGO silicate sand: 0.315-0.4 G: 0.003
(1992), Thanos et al. (1996)V,v ceramic extrudate: 1.1; 2.5
SiC: 0.25-2.38
glass cylinder: 2 × 4
Goto and Smith (1974)V H2O + HCOOH packing diameter (mm) 4.0 (212-240 °C) QL: 0.45-1.80 mL/s X
air 82.5% CuO and 16.5% ZnO: 0.541, 2.91 QG: <4 std mL/s
Levec and Smith (1976)V H2O + C3HCOOH packing diameter (mm) 6.7, 7.25 (252-286 °C) QL: 0.38-1.32 mL/s X
air Fe2O3: 0.541, 2.38 QG: <5.9 std mL/s
Al-Dahhan et al. (1994, 1995, 1996), H 2O packing diameter (mm) 0.31-5 L: 0.42-5 ∆P/Z; Lnc; Le; ηCE; X
Wu et al. (1996), H2O + H2O2 silica shell: 1.52 G: 0.0033-4.03
Khadilkar et al. (1996)V,v C6H14 Al2O3 + 0.5% Pd cylinder: 1.57 × 4.3
C6H14 + R-methylstyrene Al2O3 + 2.5% Pd cylinder: 1.3 × 5.6
{ H2 GB: 1.14 silicon carbide: 0.2
He trilobe: Al2O3 + 23% CuCr: 1.3 × 5.6
{ N2 column diameter/height (mm): 22/500
∆P/Z; βe; a; kLa; trickle-to-pulse flow &

flow & foaming-to-foaming
pulsing UGtrans(UL,FG)
X; Dz; Rt; kLa

∆P/Z; S; X
UL: 3 × 10-4-2 × 10-3

UG: 0.025-0.25
Figure 1. Influence of pressure on some characteristic properties
G: 0.17-0.53
of gaseous hydrogen. Temperature ≈ 410 K. Curve 1: density, kg/

G: ≈0.03-3
L: 1.4-24.5
L: 2.9-8.9
m3 (L’Air Liquide, 1976; experimental data). Curve 2: heat

capacity at constant pressure, kcal/kg/K (L’Air Liquide, 1976;
experimental data). Curve 3: molecular diffusivity in the gas,
cm2/s × 102 (equimolecular H2-n-C4H10) (Reid et al., 1987;
correlation). Curve 4: dynamic viscosity, Pa‚s × 106 (L’Air Liquide,
1976; experimental data). Curve 5: Henry constant (mole fraction
(160-180 °C)
of H2 dissolved in tetralin/mole fraction of H2 in the vapor phase)

(Sebastian et al., 1981; experimental data). Curve 6: mole fraction
of H2 dissolved in tetralin/H2 partial pressure, bar-1 × 103
0.2-8.1
0.1-0.4
(Sebastian et al., 1981; experimental data). Curve 7: thermal

0.65
conductivity cal/cm/s/K × 105 (L’Air Liquide, 1976; experimental

data).
gravitational force depends on liquid density and is not

column diameter/height (mm):

polypropylene extrudate: 3.4
affected by pressure in the usual operating range of

TBRs (e30 MPa). Therefore, the effect of gas on the
packing diameter (mm)
packing diameter (mm)
0.5% Pd/Al2O3: 3 × 3.2
pressure drop can be split into an effect of the superficial

55; 105/1080-1560
gas velocity and another due to gas density. Increased

GB: 0.85; 1.2; 2; 3
packing size (mm)

carbon pellet: 1.6
gas density may result either from increased reactor

Al2O3 + Pd: 2.2
23/400-1000
Al2O3 sphere: 2
operating pressure or from use of gases of higher

molecular weights. In either case, increased gas density
25/1500
leads to increased gas-liquid interaction and higher

pressure drop. Based on Al-Dahhan and Dudukovic’s
(1994, 1995) and Al-Dahhan’s (1993) phenomenological
analysis, five limiting cases can be deduced, each of
which describes the effect of reactor pressure and gas
superficial velocity on the pressure gradient, liquid

holdup, catalyst wetting, gas-liquid interfacial area,
etc. Parts a-c of Figure 2 illustrate these cases for
ETG + 0.05 mol/L DEA
pressure drop, liquid holdup, and catalyst wetting

H2O + 1.5 mol/L DEA
H2O + 40% C12H22O11
1,5,9-cyclododecatriene
efficiency. The criteria (the values of the reactor pres-

sure or gas density and superficial gas velocity) listed
H2O + 1% EtOH
below that represent these cases are based on the

EtOH; H2O
experimental observations of Al-Dahhan and Dudukovic

liquid C4-cut
N2 + CO2
PC; ETG
(1994, 1995), Larachi et al. (1990, 1992), Wammes et

Ar; CO2
He; N2
N 2 + H2
al. (1990a,b, 1991a,b), Lara-Marquez et al. (1992), and

Cassanello et al. (1996).
H2
{ {
Case 1: no gas flow, all pressure levels. This case

corresponds to the pure trickle flow regime. While the
gas is stagnant, the dimensionless pressure gradient is
1993, 1994), Wild et al. (1991),
Larachi et al. (1990, 1991, 1992,
zero and the liquid flow is exclusively driven by its

Lara-Marquez et al. (1992),
Cassanello et al. (1996)v,V
weight. Under these circumstances, reactor pressure

Vergel et al. (1993, 1995)V,v
within the range mentioned, i.e., up to 30 MPa, is not

important and TBR fluid dynamics is determined by the
Stuber et al. (1996)v
liquid flow rate and liquid properties and by the bed

characteristics. As a result, at a given liquid superficial
velocity, wetting efficiency and gas-liquid interfacial
area are the smallest while liquid holdup is the largest.
In other words, liquid fills the major pore space readily
but does not spread uniformly across the reactor diam-
eter and over the external particles surface.
with pressure and gas velocity and can still be neglected.
TBR fluid dynamics can to a good approximation be seen
as gravity driven and gas-phase independent. This
situation holds for ∆P/FLgZ < 0.1-0.2 (Al-Dahhan and
Dudukovic, 1994, 1995; Wammes et al., 1991a,b;
Wammes and Westerterp, 1990).
Case 3: high gas superficial velocity, UG, and low
pressure, P: UG > 7 cm/s and P < 0.35 MPa for
nitrogen and for gases with equivalent density at
this range of pressure. Case 3 together with case 2
corresponds to the conditions at which the vast majority
of experiments on TBRs have been conducted. At nearly
atmospheric pressure and high gas velocity, the pres-
sure gradient increases in comparison to the gravita-
tional force. Consequently, the dimensionless pressure
gradient (∆P/FLgZ) increases, which causes a decrease
in liquid holdup and an increase in the catalyst wetting
efficiency and gas-liquid interfacial area. This is due
to the increased gas-liquid interfacial drag, decreased
liquid film, and, hence, improved liquid spreading over
the external particles surface and across the reactor
diameter caused by larger gas flow rate. The effects of
gas flow rate in this case are more noticeable at high
liquid flow rates than at lower liquid flow rates.
Case 4: low gas superficial velocity, UG, and high
pressure, P: UG < 2 cm/s and P > 3.5 MPa for
nitrogen and for gases with equivalent density at
this range of pressure. As a result of increased
pressure or gas density, the pressure drop increases and
so does the dimensionless pressure gradient ∆P/FLgZ.
This causes liquid holdup, wetting efficiency, and gas-
liquid interfacial area to behave as in case 3 but in a
less pronounced manner since the pressure gradient is
more sensitive to velocity changes than to gas density
changes (Al-Dahhan and Dudukovic, 1994, 1995). Nev-
ertheless, the changes are relatively larger than those
in case 2 at both low and high liquid flow rates. Case
4 arises for ∆P/FLgZ larger than about 0.25 (Wammes
et al., 1991b; Wammes and Westerterp, 1990; Al-
Dahhan and Dudukovic, 1994, 1995).
Case 5: high gas superficial velocity, UG, and
high pressure, P: UG > 7 cm/s and P > 3.5 MPa
for nitrogen and for gases with equivalent density
at this range of pressure. This is the most important
case in terms of sensitivity of TBR fluid dynamics to
pressure. As reactor pressure or gas density and gas
flow rate increase, the dimensionless pressure gradient
(∆P/FLgZ) increases dramatically and liquid holdup
decreases significantly. Hence, liquid film thickness at
a constant liquid flow rate decreases, while the shear
stress on the gas-liquid interface increases. These
result in a better spreading of the liquid film over the
external packing area and across the reactor diameter.
Accordingly, catalyst wetting efficiency and gas-liquid
interfacial area improve noticeably. The effects of high
pressure and high gas velocity at high liquid flow rates
are greater than those at low liquid flow rates. The
trickle-to-pulse flow transition at given superficial gas
Figure 2. Effect of high pressure and gas flow rate on pressure
drop (a), liquid holdup (b), and catalyst wetting efficiency (c) in a velocity shifts toward higher liquid flow rate. Qualita-
bed of extrudates: system, hexane/nitrogen; particles size, 0.157 tively, cases 3 and 4 can be viewed as the limiting cases
× 0.43 cm; reactor diameter, 2.2 cm (after Al-Dahhan and of case 5, the first one as the limit at low pressure and
Dudukovic, 1995). the second one as the limit at low gas superficial
Case 2: low gas superficial velocity, UG, and low velocity.
pressure, P: UG < 2 cm/s and P < 0.35 MPa for If the gas density is increased via molecular weight
nitrogen and for gases with equivalent density at using different gases (such as helium, nitrogen, argon,
this range of pressure. At low pressure, or nearly carbon dioxide, etc.) at constant pressure, the effects on
atmospheric pressure, and at low gas velocity the the fluid dynamic parameters follow the same trend of
dimensionless pressure gradient changes only slightly the five limiting cases discussed above. When the
pressures of gases of different molecular weights are set dispersed bubble flow. The pulsing flow regime may be
to result in equal densities, they bring about identical approached either from the gas-continuous trickle flow
effects on the fluid dynamics as shown in Figure 5a-c or from the liquid-continuous coalesced bubble flow
(Al-Dahhan and Dudukovic, 1994; Wammes et al., regime. Such a pulsing regime is observed for moderate
1991b, Larachi et al., 1991a, 1994). liquid flow rates and moderate to high gas flow rates
Model II. In model II (Larachi et al., 1992; Lara- and can be depicted as a macroscopic combination of
Marquez et al., 1992; Cassanello et al., 1996) it has been dispersed bubble flow occurring in the liquid-rich slugs
suggested that beyond a critical value of superficial gas and trickle flow in the gas-rich slugs, both propagating
velocity (≈1-2 cm/s) pressure (or gas density) intensifies along the bed. At still higher gas flow rates, a mist or
the gas shear over the trickling liquid films. Below this spray flow is eventually observed for which the gas
critical velocity, pressure (or gas density) effects are becomes again the continuous phase and in which the
marginal so that a pressurized TBR behaves as if liquid is entrained as droplets. The above high interac-
operated at 1 atm or close to it. This operation below tion flow patterns apply for coalescing liquids, while for
critical superficial gas velocity encompasses cases 1, 2, coalescence-inhibiting (or foaming) liquids, two more
and 4 of the above-discussed model I. Above UG ) 1-2 flow regimes are encountered, namely, the foaming flow
cm/s the high-pressure momentum transfer through the and the foaming-pulsing flow (Charpentier and Favier,
gas-liquid interface becomes large enough to entrain 1975).
the gas into the liquid. As a result the gas disperses in In order to properly design TBRs based on laboratory
the liquid, forming tiny bubbles. The size of these data, it is important to predict in which flow regime the
minuscule bubbles is dictated by competition between reactor is operating for a given set of conditions. It is
the liquid viscous shear stress, which tends to deform also important to know if pressure effects can be well
and break the bubbles apart, and the force induced by accounted for in order to predict whether the same flow
interfacial tension, which tends to stabilize them. The regime will be preserved once scaleup or scaledown is
bubbles thus formed are assumed to flow without slip performed. In the literature, a number of studies have
with the liquid (i.e., form a pseudohomogeneous mixture been devoted to this task for which Table 2 may be
with the liquid) and lead with increased gas density to consulted. Many flow regime charts and attempts at
enhanced gas-liquid interfacial area and gas-side and modeling flow regime transitions have been proposed
liquid-side mass-transfer coefficients. Proliferation of thus far. Table 4 summarizes all the empirical correla-
such small bubbles with increased gas density contrib- tions and theoretical models that have been either based
utes to reduced liquid holdup and to increased ∆P/Z. or tested on high-pressure transition data. This table
∆P/Z also increases with gas density, as the bed tortu- includes the two empirical correlations (Wammes et al.,
ous passages render the gas flow more dissipative at 1990a; Larachi et al., 1993a) and the three phenomeno-
higher momentum flow rates; the additional frictional logical/semitheoretical models; the predictions of all of
interface contributed by bubbles is another factor that these have been tested using high pressure transition
increases ∆P/Z. The situation depicted here for UG > data (Dudukovic et al., 1991; Holub et al., 1992, 1993;
1-2 cm/s corresponds to case 5 of model I. Dimenstein et al., 1984; Dimenstein and Ng, 1986; Ng,
1986; Ng and Chu, 1987; Sundaresan, 1987; Grosser et
al., 1988; Dankworth et al., 1990a,b; Dankworth and
3. Flow Regime Transitions at High Pressure Sundaresan, 1992). A thorough evaluation of available
models and empirical correlations for the prediction of
Although a number of flow regimes can be identified trickle-to-pulse flow transition boundaries with pressure
in trickle-beds, there is a consensus to classify them all was made by Wild et al. (1991), Larachi (1991), and
into two broad regimes, namely, a low interaction Larachi et al. (1993a). These studies concluded that the
regime (LIR; trickle flow regime) and a high interaction use of available phenomenological and semitheoretical
regime (HIR, pulse, spray, bubble, and dispersed bubble models for prediction of transitions at high pressure still
flow regimes) (Charpentier and Favier, 1975; Midoux leads to very large errors, and no single approach can
et al., 1976). The LIR is observed at low gas and liquid be recommended.
flow rates and is characterized by a weak gas-liquid The knowledge of several transitions which delineate
interfacial activity and a gravity-driven liquid flow. the above-depicted flow regimes is indispensable in the
Gas-liquid interaction in the trickle flow regime would design and scaleup of trickle-beds. Although the effect
increase at high gas and liquid flow rates and at of fluid flow rates, liquid physical properties, and
elevated pressure (i.e., close to the transition to the HIR packing properties on the flow regime transitions has
regimes). The liquid, which may be foaming or not, been thoroughly documented in the literature over the
trickles down the packing in the form of droplets, films, past 30 years (see Table 2), scarcity of measurements
and rivulets, while the continuous gas phase occupies and prediction tools for transition boundaries at el-
the remaining porous space and flows separately. The evated pressures is noticeable. Table 3 illustrates that
HIR is characterized by a moderate to intense gas- only three types of transitions have been studied as a
liquid shear due to moderate to high flow rate of one or function of pressure (or gas density). One was the
both of the fluids. As a result, various flow patterns trickle-to-pulse flow transition for nonfoaming gas-
arise depending on the gas-to-liquid holdup ratio and liquid systems (Hasseni et al., 1987; Wild et al., 1991;
liquid tendency to foam. Low gas flow rates, and Wammes et al., 1990a,b, 1991b; Wammes and Wester-
sufficiently high liquid flow rates, lead to the bubble flow terp, 1991; Larachi et al., 1993a), the other was the
regime with a continuous liquid phase which contains pulsing-to-dispersed bubble flow for nonfoaming liquids
small spherical bubbles. At medium gas flow rates but (Wammes et al., 1990a), and the third was the foaming-
with still high liquid flow rates, the liquid phase to-foaming-pulsing transition for foaming liquids (Lara-
remains continuous but the bubbles coalesce and the chi, 1991). Moreover, only models or correlations con-
gas flows in the form of elongated bubbles (Charpentier cerned with trickle-to-pulse flow transition at high
et al., 1972). This flow regime is referred to as the pressure were tested.
Table 4. Correlations for Prediction of Trickle-to-Pulse Flow Regime Transition in High-Pressure Trickle-Bed Reactors
author flow configuration P (MPa) approach
Wammes et al. (1990a) trickle bed 0.2-2.0 empirical
βe,trans )
c
UG0.26FG0.04
c) { 0.27 water/N2
0.32 water + 40% ETG/N2
(1)
1
1+
UG FrGL
tested on Talmor diagram (1977) versus for water/N2 system
UL 1
WeGL +
ReGL
Larachi et al. (1993a) trickle bed 0.1-7.0 empirical
LλψΦ G
()
-1.25
)
G λ
x
σw µL
()()
1/3 2/3
FG FL Fw 1
λ) ψ) Φ) (2)
Fa Fw σL µw FL FG
4.76 + 0.5
Fa
Holub et al. (1992, 1993) trickle bed tested on data of semiempirical
Wammes et al. (1990a)
2.9ReLE5/11
1
e1 (3)
ΨL0.17GaL0.41Ka1/11
2.8 < GaL < 6.3 × 105; 1 < ΨL < 55; 1.2 < L < 12 kg/m2‚s; 0.004 < G < 2.2 kg/m2‚s
Ng (1986) trickle bed tested on data of phenomenological
Hasseni et al. (1987) and
x
1 d 4 σL 1 8σL
F V 21-
2 G G ( ( ) ) ) 4d
d′ p
- FLgdp;
2
dp,max )
FLg
(4)
Grosser et al. (1998) trickle bed tested on data of phenomenological

Hasseni et al. (1987) and
FLL (
∂UL
+ UL
∂UL
+ L
∂PL
)- gFLL - FL(z) +
∂ Lµ*L
∂UL
∂z
)0
( )
∂t ∂z ∂z ∂z
FGG (
∂UG
+ UG
∂UG
+ G
∂PG
)
- gFGG - FG(z) +
∂ Gµ*G
∂UG
∂z
)0
( )
∂t ∂z ∂z ∂z
FG(z) ) -
[ AµG(1 - )21.8
dp {G}
2 2.8
+
BFG(1 - )1.8
dp{G}1.8 | |]
UG - UL (UG - UL)
[AµL(1 - )2L2 BFL(1 - )L3 2 - Ls

]( )
2.43
FL(z) ) - U L + UL - FG(z) (5)
dp23 dp 3 L - Ls
From the experimental studies of the three types of (c) At high pressures (or gas densities > 2.3 kg/m3)
flow regime boundaries at elevated pressure, the fol- and high gas flow rates, the trickle-to-pulse flow transi-
lowing current understanding emerges: tion at a given liquid (respectively gas) superficial
(a) For any range of gas density, when the pressures velocity shifts toward higher gas (respectively liquid)
of gases of different molecular weights are set to have superficial velocities, thus making the trickle flow
the same densities, the trickle-to-pulse flow transition operating region wider at elevated pressure. This
occurs at the same gas and liquid superficial velocities. enlargement of the trickle flow operating region at high
Thus, if a reactor is to operate at 40 MPa pure hydrogen, pressure holds true even if, instead of drawing super-
its trickle-to-pulse flow transition boundary can be ficial velocity UG versus UL boundaries, mass superficial
simulated by a 1.6 MPa argon pressure for the same velocity G versus L boundaries are plotted (Figure 3)
temperature, with the densities of argon at 1.6 MPa and (Hasseni et al., 1987; Wild et al., 1991).
hydrogen at 40 MPa being equal. Also the liquid holdup (d) At any pressure level the transition from trickling
and the pressure drop keep the same values (Wammes to pulsing regime occurs at smaller fluid throughputs
et al., 1991b; Al-Dahhan and Dudukovic, 1994). for viscous liquids than for less viscous liquids (Figure
(b) Irrespective of the gas used, no effect of the gas 3b) (Wild et al., 1991).
density on the trickle-to-pulse flow boundary is observed (e) At liquid velocities corresponding to the dispersed
as long as gas density lies below 2.3 kg/m3 as shown in bubble flow regime, with increasing pressure, increased
Figure 3a. Hydrogen-pressurized TBRs at a few mega- gas velocities are necessary to bring pulse flow.
pascals can thus be simulated by air or nitrogen at (f) Foaming liquids behave similarly to nonfoaming
pressures close to atmospheric. liquids in the LIR regardless of pressure. When a liquid
pressure drops at high pressure in porous media was

provided by Larachi (1991). Several gases and liquids
were used, and the results were successfully represented
by an Ergun-type equation (Ergun, 1952). Pressure,
density, and molecular weight, which primarily affect
the fluid inertia, are well accounted for by the Reynolds
number. Analysis of the Blake-Kozeny-Carman hK
and Burke-Plummer hB coefficients (which are known
as Ergun constants E1 and E2, respectively) showed no
evidence of clear dependence upon pressure or gas or
liquid density (Hasseni et al., 1987; Larachi et al., 1990;
Al-Dahhan, 1993). However, hK (E1) values were slightly
different depending on whether the working fluid was
a liquid or a gas.
5. Two-Phase Pressure Drop

The influence of pressure on the two-phase pressure
drop in trickle-beds was first studied by Hasseni et al.
(1987). Operating pressures extending from atmo-
spheric pressure up to 10.1 MPa using nitrogen allowed
pressure drops to be obtained in both the trickle flow
and pulse flow and at their corresponding transition.
Wammes et al. (1990a,b, 1991a,b) and Wammes and
Westerterp (1990, 1991) measured the pressure drops
at pressures up to 7.5 MPa in the trickling, pulsing,
Figure 3. Effect of pressure and physical properties on trickle- bubbly dispersed flow regimes and trickling/pulsing and
to-pulse flow regime transition: (a) water/nitrogen and water/ pulsing/bubbly dispersed transitions. Larachi et al.
helium systems (after Wammes et al., 1991b); (b) water/nitrogen (1991a, 1990) and Wild et al. (1991) conducted extensive
system (continuous lines) and ethyleneglycol/nitrogen system experiments (over 1500 measurements) on the effect of
(dotted lines) (after Wild et al., 1991). Lines represent the trends. pressure (from 0.2 to 8.1 MPa) by testing more than 30
systems. Experiments were performed using pure or
foams at atmospheric pressure, it keeps foaming at mixed gases and pure or mixed aqueous or organic
higher pressures. In the foaming-pulsing flow regime, solutions. These latter systems were chosen to explore
the foam plugs travel down the bed, with frequencies behaviors such as foaming, coalescing, viscous nonfoam-
and amplitudes increasing with the gas mass flux. The ing, and viscous foaming flows. The operating condi-
amplitude in the pressure oscillations in this regime tions covered trickling, pulsing, bubbly dispersed re-
increases with pressure (Larachi, 1991). Also the gas gime, foaming pulsing regime, and their respective
superficial velocity corresponding to the transition transition boundaries. Recently, Al-Dahhan (1993) and
between foaming and foaming-pulsing decreases as Al-Dahhan and Dudukovic (1994) measured two-phase
pressure increases. The inception of foaming-pulsing pressure drops in the trickling flow regime using organic
flow is delayed as the liquid throughput increases and aqueous nonfoaming liquids at pressures up to 5.0
regardless of pressure (Larachi, 1991). MPa (see Figure 2a). The pressure dependence of the
two-phase pressure drop was discussed in light of the
4. Single-Phase Pressure Drop five limiting cases of model I described above. Tables
3 and 5 summarize the experimental conditions ex-
Knowledge of the single-phase pressure drop for the plored and the empirical correlations and theoretical
gas and the liquid flowing individually throughout the models proposed so far by the different investigators,
packed bed under high pressure conditions is important. respectively. Al-Dahhan and Dudukovic (1994) showed
Such information constitutes the limiting cases of TBRs that Holub et al.’s (1992, 1993) phenomenological model
operated either at trace liquid flow rates, such as in mist predicts both pressure drop and liquid holdup better
or deep in trickle flow regimes, or at trace gas flow rate, than the current empirical high-pressure correlations
as in the dispersed bubble flow regime. Besides, single- for the range of operating conditions studied. Al-
phase pressure drop plays a direct role in the two-phase Dahhan et al. (1996) extended this model to improve
separated flow models developed for the quantification its prediction at high pressure and high gas flow rates
of liquid holdup and two-phase pressure drops in TBRs by accounting for the gas-liquid interactions.
(Midoux et al., 1976; Morsi et al., 1982; Holub et al., Figures 2a and 4a-d illustrate some of the experi-
1992, 1993). mental data at high-pressure operation. The five limit-
Single-phase pressure drop measurements have been ing cases discussed in section 2 describe well the
carried out at pressures up to 18 MPa using either experimental observations as shown in Figure 2a.
gases, liquids, or supercritical fluids in various bed sizes When the pressures of gases of different molecular
and geometries and various particle shapes and diam- weights are set to have equal densities at constant liquid
eters (Hasseni et al., 1987; Larachi et al., 1990; Olowson throughputs, they bring about identical pressure drops
and Almstedt, 1991; Salatino and Massimila, 1990; as illustrated in Figure 5a,c (Al-Dahhan and Dudukovic,
Molerus and Schweintzer, 1989; Perrut, 1987; Naka- 1994; Wammes et al., 1991b, Larachi et al., 1991a, 1994)
mura et al., 1985; Rowe, 1984; Rowe et al., 1984; for the case of He at 2.1 MPa and N2 at 0.3 MPa which
Kawabata et al., 1981; Babu et al., 1978; Saxena and have the same density.
Vogel, 1976; Handley and Heggs, 1968; Simpson and It is noteworthy that for coalescing to weakly foaming
Rodger, 1961). An extensive review of single-phase systems and gases of sufficient inertia, i.e., FGUG2 g 1%
Table 5. Correlations and Models for Prediction of Two-Phase Pressure Drop in High-Pressure Trickle- and
Flooded-Bed Reactors
Ellman et al. (1988) trickle bed 0.1-10.1 empirical
high interaction regime
ReL0.25WeL0.2
f ) A(XGξ1)j + B(XGξ1)k where ξ1 )
(1 + 3.17ReL1.65WeL1.2)0.1
low interaction regime
ReL2
f ) C(XGξ2)m + D(XGξ2)n where ξ2 )
0.001 + ReL1.5
A ) 6.96; B ) 53.27; C ) 200; D ) 85; j ) -2; k ) -1.5; m ) -1.2; n ) -0.5 (6)

Wammes et al. (1991b) trickle bed 0.2-7.5 empirical
∆P dp 1 - FGUGdp
( )
-0.37
) 155 (7)
Z 1 2 G µG(1 - )
F U
2 G G
f)
1
(XG(ReLWeL)1/4)3/2 [
31.3 +
17.3
xXG(ReLWeL)1/4 ] (8)
Larachi et al. (1994) flooded bed 0.1-5.1 empirical
f)
1
(XG(ReLWeL)1/4)3/2 [
45.6 +
15.9
xXG(ReLWeL)1/4 ] (9)
Holub et al. (1992, 1993), trickle bed 0.31-5 phenomenological

Al-Dahhan and Dudukovic (1994)
( )[ ]
3
∆P E1ReL E2ReL2
ΨL ) +1) +
FLgZ L (1 - )GaL (1 - )2GaL
( )[ ]
3
∆P E1ReG E2ReG2
ΨG ) +1) +
FGgZ - L (1 - )GaG (1 - )2GaG
FG
ΨL ) 1 + (Ψ - 1) (10)
FL G
a E and E are Ergun constants that characterize the bed and are determined from single-phase flow measurements. The third equation
1 2
is first solved for liquid holdup, and then the first or the second equation is solved for pressure drop.
FLUL2, it was found (Larachi et al., 1990, 1991a, 1994; away. Liquid holdup also affects the catalyst wetting
Wild et al., 1991) that the gas momentum rate can be efficiency, which, in turn, affects the reaction selectivity
used as a similarity criterion to evaluate pressure drops depending on whether the reaction takes place solely
at high-pressure operation from atmospheric pressure on the wetted catalyst area or on dry and wetted
drop experiments. Thus, if the flow regime is controlled catalyst areas alike. Table 3 presents the experimental
by inertia and the liquid mass flux is the same, conditions explored by the various investigators, whereas
the correlations and models devoted to liquid holdup are
[FGUG2 ) G2/FG]1 atm ) [FGUG2 ) G2/FG]P>1 atm listed in Table 6. Figures 2b and 6 present some of the
experimental data for liquid holdup at high-pressure
then operation. The phenomenological analysis of Al-Dahhan
and Dudukovic (1994), presented in section 2, describes
[∆P/Z]1 atm ) [∆P/Z]P>1 atm well the effect of pressure and gas flow rates on liquid
holdup. Systems with gases of different molecular
This is demonstrated by Figure 4d, in which a plot of weights that are maintained at the same densities, by
pressure gradient to gas inertia ratio as a function of adjusting the reactor pressure, render the same liquid
gas momentum rate is able to merge all pressure holdup and pressure drop at the same liquid mass flux
gradient lines at four different gas densities (Figure 4c) as illustrated in Figure 5b. Hence, increasing gas
and a unique line describes pressure drop irrespective density via molecular weight by using different gases
of gas density. (helium, nitrogen, argon, carbon dioxide) at constant
pressure duplicates the trends due to pressure (Al-
6. Liquid Holdup Dahhan and Dudukovic, 1994; Wammes et al., 1991b,
Larachi et al., 1991a, 1994). The static liquid holdup
Liquid holdup plays an important role in TBR hydro- measured by the stop-flow technique was found not to
dynamics and mass and heat transfer. In highly be affected by the reactor pressure and gas flow rate
exothermic reactions, knowledge of holdup is essential (Al-Dahhan, 1993; Al-Dahhan and Dudukovic, 1994;
for avoiding hot spots and for preventing reactor run- Wammes, 1990; Wammes et al., 1991a).
Figure 4. Effect of gas density on the two-phase pressure Figure 5. Effect of gas density on pressure drop (a) and liquid
gradient: (a) pressure gradient-to-liquid gravitational force ratio holdup (b) in trickle-bed reactors after Al-Dahhan and Dudukovic
as a function of superficial gas velocity; (b) pressure gradient-to- (1994). System: hexane/nitrogen and helium. Particles: 1.14 mm
liquid gravitational force ratio as a function of gas mass flux; (c) glass beads. Reactor diameter: 2.2 cm (c) after Larachi et al. (1994)
pressure drop-to-gas inertia force ratio as a function of gas mass for trickle-bed and upflow reactors. For cocurrent downflow
flux; (d) pressure drop-to-gas inertia force ratio as a function of (TBR): system, polypropylene carbonate (PC)/nitrogen and helium;
gas momentum flow rate. System: ethylene glycol/nitrogen. L ) particles, 1.4 mm glass beads; reactor diameter, 2.3 cm. For
13.2 kg/m2‚s. Particles: 2 mm glass beads. Reactor diameter: 2.3 cocurrent upflow: system, water/nitrogen and helium; particles,
cm (after Larachi, 1991). 3.4 mm polypropylene extrudates; reactor diameter, 2.3 cm. Lines
Liquid holdup at high pressure was first reported by represent the trends.
Abbott et al. (1979), who measured liquefied butane technique, Kohler and Richarz (1984, 1985) measured
holdup under nitrogen pressure in the range 2.4-3.8 capillary (static) and total liquid holdups up to 1 MPa.
MPa. Larkins-type correlations (Larkins et al., 1961), Water, nonviscous coalescing organic solutions, nitro-
involving the Lockhart-Martinelli dimensionless num- gen, and hydrogen were the test fluids used within the
ber (Lockhart and Martinelli, 1949) successfully de- operating range restricted to the trickle flow regime up
scribed their data. Later, by means of a tracer injection to the transition to pulsing. Under zero gas flow and
Table 6. Correlations and Models for Prediction of Liquid Holdup in High-Pressure Trickle- and Flooded-Bed Reactors
Abbott et al. (1967) trickle bed 2.41-3.79 empirical
log10(βnc) ) -0.44 + 0.40log10(χ) - 0.12(log10(χ))2 (11)
Saada (1975) flooded bed 1.36 empirical
( ) ReL d
βe ) c
ReG
() ( ) ()
dp 0.38
c 0.48
) if ReG < 0.44ReL2
d 0.25 Dc
() ( ) ()
dp 0.38
c 0.32
) if ReG > 0.44ReL2 (12)
d 0.07 Dc
Kohler and Richarz (1985) trickle bed 0.1-1.0 empirical
( )( ) ( )( ) FL2gdp3
0.65 0.53
avdp -0.42
FLULdp FGUGdp -0.31
βnc ) 0.71 (13)
µL2 µL µG
Ellman et al. (1990) trickle bed 0.1-8.0 empirical
log βnc ) -RXGmReLnWeLp ( )

avdK
1-
q
(14)
high interaction regime
R ) 0.16; m ) 0.325; n ) 0.163; p ) -0.13; q ) -0.163
low interaction regime
R ) 0.42; m ) 0.24; n ) 0.14; p ) 0; q ) -0.14
van Gelder and Westerterp (1990) flooded bed 0.2-1.2 empirical
( )
UG 0.125
βe ) 0.354 + 0.143UL0.405 - 0.206 (15)
UL
Ring and Missen (1991) trickle bed 10 empirical
βa ) 15.6UL0.679 (16)
Wammes et al. (1991a) trickle bed 0.3-6.0 empirical

trickling, no gas flow
( )( ) FL2gdp3
c d
FLULdp
βnc ) 16.3
µL µL2
c) { 0.36 if Re < 11
0.55 if Re > 15
d) {
-0.39 if Re < 11
-0.42 if Re > 15
(17)
gas-liquid flow (Specchia and Baldi (1977) correlation)
( )( ( FL2gdp3
)) ( )
0.55 0.65
FLULdp ∆P
-0.42
avdp
βnc ) 3.8 1+
µL µL2 FLgZ
1.22WeL0.15
log(1 - βe) ) - (18)
ReL0.20XG0.15
Larachi et al. (1991b) trickle/flooded bed 0.2-8.1 empirical
Jdf - UG
βe ) 1 + (19)
UG + UL
where Jdf ) AUGb (a, b) depend on the gas-liquid-solid system

Lara-Marquez et al. (1992) flooded bed 0.3-5.1 empirical
0.93WeL0.08
log(1 - βe) ) - (20)
ReL0.20XG0.15
Table 6 (Continued)
Holub et al. (1992, 1993), trickle bed 0.31-5 phenomenological

Al-Dahhan and Dudukovic (1994)a
( )[ ]
3
∆P E1ReL E2ReL2
ΨL ) +1) +
FLgZ L (1 - )GaL (1 - )2GaL
( )[ ]
3
∆P E1ReG E2ReG2
ΨG ) +1) +
FGgZ - L (1 - )GaG (1 - )2GaG
FG
ΨL ) 1 + (Ψ - 1) (21)
FL G
Tsamatsoulis and Papayannakos (1994) trickle bed 5.0 empirical
βe ) aULb where (a, b) depend on boundary conditions (22)

a E1 and E2 are Ergun constants that characterize the bed and are determined from single-phase flow measurements. The third equation
is first solved for liquid holdup, and then the first or the second equation is solved for pressure drop.
(Satterfield and Way, 1972; Goto and Smith, 1975),

whereas for ReL > 15, the Specchia and Baldi correlation
(1977) for zero gas flow rate led to a good agreement.
Larachi et al. (1990, 1991a-d) and Wild et al. (1991)
used the tracer impulse technique to investigate total
liquid holdups within the same operating range as used
in two-phase pressure drop measurements. The main
conclusions drawn from their study are consistent with
the discussion of section 2. They found that liquid
holdup is insensitive to pressure for UG < 1-2 cm/s, and
the corresponding density rule was proposed for the
estimation of liquid holdup in the presence of a light
gas at high pressure knowing the holdup at atmospheric
pressure in the presence of a heavy gas. A drift flux
Figure 6. External ethylene glycol holdup as a function of approach and a correlation were also developed for
superficial nitrogen velocity in bubble flow and pulse flow regimes coalescing to weakly foaming systems (Table 6). Figure
in upflow (FBR) operation and trickle flow and pulse flow regimes 6 shows some of Larachi et al.’s (1991a-d) liquid holdup
in TBR operation. P ) 0.3, 2.1 MPa (after Larachi, 1991). data at high-pressure operation.
Ring and Missen (1991) investigated total, capillary,
in the trickling flow regime, the data were well de-
scribed by the low interaction correlation of Specchia and dynamic liquid holdups during the catalytic hy-
and Baldi (1977). For two-phase flow, their data were drodesulfurization of dibenzothiophene in trickle flow
correlated using the modified high interaction correla- at 10 MPa and 330-370 °C. The results showed the
tion of Specchia and Baldi (1977). independence of liquid holdup upon temperature and
Based on ENSIC’s Nancy data bank of liquid holdups pressure for the very low hydrogen mass fluxes tested
at atmospheric pressure, Ellman et al. (1990) derived (53 × 10-5 < G < 32 × 10-4 kg/m2‚s) and found that
two correlations for liquid holdup, one for the high and the available correlations derived for cold conditions (1
the other for the low interaction regimes. Good agree- atm and 25 °C) can be useful for estimation purposes.
ment was found between their high interaction regime Al-Dahhan and Dudukovic (1994) measured liquid
correlation and water holdup under high-pressure ni- holdups using the stop-flow technique and bed drainage
trogen (up to 8.1 MPa) (Wild et al., 1991). Wammes et for the same conditions as for the pressure drops, i.e.,
al. (1990a,b, 1991a,b) and Wammes and Westerterp trickle flow regime. They interpreted the evolution of
(1990) investigated static and noncapillary liquid hold- liquid holdup as a function of gas density according to
ups by the stop-flow technique and bed drainage for the the five cases of their phenomenological analysis pre-
same operating conditions as used in the two-phase sented in section 2. By plotting liquid holdup as a
pressure drop studies. They reported that liquid holdup
function of superficial mass liquid velocity, they con-
decreased when the pressure was increased for given
firmed the observations made by Wammes et al. (1990a,b,
gas and liquid superficial velocities for each of the three
liquids tested (see Table 3). Such a decrease was 1991a,b), Wammes and Westerterp (1990, 1991), and
interpreted as due to a shift in the reactor fluid Larachi et al. (1990, 1991a,b). Their high-pressure data
dynamics from a state predominantly controlled by were well described with the parameter-free phenom-
gravity (trickle flow with zero gas flow rate) to a state enological model of Holub et al. (Holub et al., 1992, 1993)
controlled by gas-liquid shear stress (or pressure drop). and its extension (Al-Dahhan et al., 1996) (see Table
The Specchia and Baldi correlation (1977) for the low 6). They found, within the operating conditions that
interaction regime described correctly their two-phase they have studied, that Holub et al. (1992, 1993)
flow liquid holdup data. For trickling flow and zero gas mechanistic model and its extended version for high
flow (case 1 above of section 2) and ReL < 11, liquid pressure predict liquid holdup and pressure drop better
holdups were well described by available correlations than the available high-pressure correlations.
Table 7. Correlations and Models for Prediction of Gas-Liquid Interfacial Areas in High-Pressure Trickle-Bed Reactors
Wild et al. (1992) trickle bed atmospheric and high pressure empirical
low interaction
a
av [
) 10 XGReL-1/2WeL ( )]
avdK
1-
3/2 0.7
high interaction
a°
av [
) 1550 XGReL-1/2WeL
1-( )]
avdK -5 0.7
transition regime
a
av [
) 21.3 XGReL-1/2WeL ( )]
avdK
1-
-2 0.5
(23)
Larachi (1991) trickle bed 0.3-3.2 empirical

a
) 584XG0.86ReL1.48Ka-0.5 (24)
av
Cassanello et al. (1996) trickle bed 0.3-3.2 phenomenological
{ () ( ( ))( )}
µG 1/6 We βe
L 1 1
a ) a° 1 + 4.9 × 104 1 + 2.5 1- - (25)
µL ReL βe° βe βe°
Figure 8. Effect of pressure (gas density) and gas flow rates on

the volumetric gas-liquid mass-transfer coefficient at L ) 3.7
kg/m2‚s: system, ETG and 0.05 mol/L DEA/0.05 mole fraction CO2
in a mixture of nitrogen and CO2; particle diameter, 3.4 mm;
Figure 7. Effect of pressure and gas and liquid flow rates on gas- reactor diameter, 2.3 cm. Lines represent the trend (after Lara-
liquid interfacial areas in TBRs: system, 1.5 mol/L DEA (dietha- Marquez et al., 1992).
nolamine) in water/0.05 mole fraction CO2 in a mixture of nitrogen
and CO2; equivalent particle diameter, 3.1 mm PVC extrudates; transfer coefficients. These parameters were deter-
reactor diameter, 2.3 cm. Lines represent the trends (after mined in the trickling flow regime solely by the chemical
Cassanello et al., 1996). method using CO2 carbamation of diethanolamine aque-
ous or organic solutions. These studies report enhance-
7. Gas-Liquid Mass-Transfer and Interfacial ment of gas-liquid mass-transfer with increasing pres-
Area sures.
In three-phase gas-liquid-solid systems such as Based on the experimental study of interfacial areas
TBRs, the gas-liquid mass-transfer resistance can have and liquid-side volumetric mass-transfer coefficients at
a detrimental effect on the overall reactor performance. elevated pressure, the following observations can be
Therefore, accurate estimation of the gas-liquid mass- made:
transfer parameters is important for achieving success- (a) Liquid-side volumetric mass-transfer coefficients
ful reactor design or scaleup. In spite of the vast and gas-liquid interfacial areas follow a trend quali-
information found in the literature on gas-liquid mass- tatively similar to pressure drops with regard to the
transfer characteristics of trickle beds at atmospheric mass fluxes or superficial velocities for a given pressure
pressure (Ellman, 1988), only a few researchers have or gas density.
studied how they evolve at elevated pressures (Larachi (b) For gas superficial velocity less than 1-2 cm/s,
et al., 1992; Lara-Marquez et al., 1992; Wammes et al., neither of the two mass-transfer parameters depends
1991b; Wild et al., 1992). From these studies it was on gas density. There is still not enough information
established that gas-liquid interfacial areas and volu- to assess the effect of pressure on the liquid-side mass-
metric mass-transfer coefficients are influenced by transfer coefficient kL alone.
pressure in TBRs as shown in Figures 7 and 8, respec- (c) For gas velocities larger than 1-2 cm/s and at a
tively. Table 3 shows the experimental ranges explored given liquid velocity, the gas-liquid interfacial area
by these investigators, and Tables 7 and 8 list the increases when pressure is increased. To interpret this
models and correlations recently published on gas- effect, Cassanello et al. (1996) postulated that due to
liquid interfacial areas and liquid-side volumetric mass- shear intensification the liquid trickling films are
Table 8. Correlations and Models for Prediction of Liquid-Side Gas-Liquid Volumetric Mass-Transfer Coefficient in
High-Pressure Trickle-Bed Reactors
Wild et al. (1992) trickle bed atmospheric empirical
low interaction
kLa°dK2
DAL [
) 2.8 × 10-4 XG1/4ReL1/5WeL1/5ScL1/2
avdK
1- ( )] 1/4 3.4
high interaction
kLa°dK2
DAL [
) 0.45 XG1/2ReL4/5WeL1/5ScL1/2
avdK
1- ( )] 1/4 1.3
transition regime
kLa°dK2
DAL [
) 0.091 XG1/4ReL1/5WeL1/5ScL3/10
avdK
1- ( )] 1/4 3.8
(26)
Cassanello et al. (1996) trickle bed 0.3-3.2 phenomenological
kLa ) (kLa)° + kL(Sh)2)(a - a°) (27)
Table 9. Correlations and Models for Prediction of Liquid-Solid Contacting Efficiency in High-Pressure Trickle-Bed
Reactors
Ruecker and Agkerman (1987) trickle bed 5.2 empirical
2
ηCE ) 1 + 0.14πLV - 1.17πLV (28)
Ring and Missen (1991) trickle bed 10 empirical
ηCE ) 1 - exp(-644UL0.964) based on diffusivity ratio
ηCE ) 1 - exp(-118UL0.635) based on diffusivity square root ratio (29)

Al-Dahhan and Dudukovic (1995) trickle bed 0.31-5 semiempirical
[ ]
∆P 1/9
+1
FLgZ
ηCE ) 1.104(Re)1/3 (30)
GaL
invaded by tiny bubbles whose size results from a Missen (1991) and Ruecker and Akgerman (1987)
balance between viscosity and surface tension forces. measured the contacting efficiency at 10 and 5.2 MPa,
It is the formation of such small bubbles that enhances they did not investigate the effect of high pressure.
the interfacial area and the volumetric mass-transfer Table 9 lists the wetting efficiency correlations devel-
coefficient. A simple model was then derived in which oped for the high pressure/temperature laboratory-scale
the interfacial area at high pressure is split into two trickle-beds. Al-Dahhan and Dudukovic (1995) have
contributions, that of liquid spreading on the solids and studied the catalyst wetting efficiency at high-pressure
that due to transported gas bubbles which give rise to operation via a tracer technique (see Table 3). In their
the pressure dependence of the gas-liquid mass- work, a phenomenological analysis was developed to
transfer. This model is given in Tables 7 and 8. relate the wetting efficiency with operating conditions
According to the phenomenological analysis for the such as reactor pressure and gas and liquid flow rate
effects of pressure proposed by Al-Dahhan and Dudu- which resulted in the five limiting cases as shown in
kovic (1994, 1995), the liquid film thickness at a fixed Figure 2c. For a fixed liquid mass velocity, at high
liquid mass velocity decreases with pressure and gas pressure and high gas flow rates, wetting efficiency
velocity where the shear stress on the gas-liquid improves noticeably whereas pressure drop increases
interface increases. This results in an improved spread- significantly and liquid holdup decreases considerably.
ing of the liquid film over the external packing area, As liquid flow rate increases, contacting efficiency
making them more wet and at the same time increasing improves further due to an increase in both pressure
the gas-liquid interfacial area. drop and liquid holdup. The effect of gas flow rate on
the wetting efficiency, pressure drop, and liquid holdup
is more pronounced at elevated pressure (see Figure
8. Catalyst Wetting Efficiency 2a-c). The improvement in wetting efficiency with
increased gas flow rate is due to the improved spreading
External catalyst wetting efficiency is an important of the liquid holdup over the external packing area. This
design and scale-up parameter in determining the is supported by the finding of Larachi et al. (1992) where
degree of catalyst utilization in trickle-bed reactors (the the gas-liquid interfacial area increases at high pres-
internal contacting is usually equal to unity due to sure and high gas flow rate operation. Al-Dahhan and
capillary effects). The available predictions of wetting Dudukovic (1995) developed a correlation for wetting
efficiency rest on data collected at atmospheric pressure efficiency for high-pressure operation given by eq 30 of
(Shulman et al., 1955; Onda et al., 1967; Krauze and Table 9. This correlation is also in good agreement with
Serwinski, 1971; Puranik and Vogelpohl, 1974; Mills the atmospheric data collected previously and provides
and Dudukovic, 1981; El-Hisnawi, 1981; El-Hisnawi et the means for assessing liquid-catalyst contacting at
al., 1981; Lazzaronni et al., 1988). Although Ring and high operating pressures.
9. Catalyst Dilution with Inert Fines in
Laboratory-Scale TBRs
Since laboratory-scale reactors need to match the
space velocity of commercial units, the actual velocities
in them are much lower, which leads to incomplete
catalyst wetting. Moreover, the same catalyst sizes,
types, and shapes employed in the commercial reactors
are used in the small-diameter laboratory units where
the criterion Dc/dp > 20, desirable for avoiding wall
effects, usually cannot be met. Such conditions can
cause liquid maldistribution and insufficient use of the
catalyst bed for liquid-limited reactions by reducing
liquid-solid contacting efficiency (Al-Dahhan and Dudu-
kovic, 1995; Wu et al., 1996a; Sie, 1991; Van Klinken
and Van Dongen, 1980; Gierman, 1988). As a remedy
to the above problems, dilution of the laboratory beds
with fines (small, inert, and nonporous particles of about
0.1 catalyst diameter) has been recommended and
utilized for decades to overcome the shortcomings of
laboratory-scale reactors and/or to provide better tem-
perature uniformity in the reactor particularly for
highly exothermic reactions. In this technique, the
hydrodynamics is largely dictated by the packing of
fines, whereas the catalytic phenomena are governed
by the catalyst particle of the same shape, size, and form
as used in the commercial reactors. It is noteworthy
that the utility of the dilution techniques would be
negated if the packing method of mixing the bed of
catalyst and fines failed to produce reproducible results.
Al-Dahhan et al. (1995) developed a reproducible pro-
cedure for packing small-diameter packed beds with a
mixture of fines and catalyst which relies on filling the
original bed voids with fines; the step-by-step procedure
is reported in Al-Dahhan et al. (1995) and Al-Dahhan
and Dudukovic (1996). Al-Dahhan and Dudukovic
(1996) also measured the liquid-catalyst contacting
efficiency in a diluted bed, with fines and compared it
to that obtained in a nondiluted bed, both beds being
operated under high pressure at the same set of liquid
mass velocity. They reported that fines improve cata-
lyst wetting efficiency, which in a diluted bed is strongly
related to both pressure drop and liquid holdup as
shown in Figure 9a-c. Thus, in a diluted bed, while
wetting efficiency improves at high pressure and high
gas flow rates, pressure drop increases and liquid
holdup decreases.
10. Liquid-Limited and Gas-Limited Reactions

in High-Pressure TBRs Figure 9. Effect of fines on catalyst wetting efficiency (a),
Wu et al. (1996a,b) and Khadilkar et al. (1996) studied pressure drop (b), and liquid holdup (c) in a diluted bed of
the effect of catalyst wetting at high pressure on the extrudates: vol. fines/vol. cat. ) 0.54; particle size, 0.157 × 0.43
cm; fines, 0.2 mm silicon carbide; reactor diameter, 2.2 cm (packed
performance of TBRs. Comparisons have been made according to Al-Dahhan et al. (1995) procedure) (after Al-Dahhan
between the performance of trickle flow (partially wet- and Dudukovic, 1996).
ted catalyst) and upflow, flooded-bed reactor (FBR;
completely wetted catalyst) without fines and for both pressure and high feed concentrations of R-MS but
TBR and FBR with fines (completely wetted catalyst in became liquid-limited at high pressure (>200 psig) and
each reactor). Hydrogenation of R-methylstyrene (in low R-MS feed concentrations. Also it was found (Wu
hexane solvent) to cumene over 2.5% alumina-supported et al., 1996a) that a TBR performs better than the
palladium catalyst extrudates was used as a test reac- upflow reactor (FBR) with fully wetted catalyst at low
tion (A(g) + bB(l,nonvolatile) f product(l)) (Wu et al., pressures and high feed concentrations of R-MS when
1996a). A simple and usable criterion for identifying the reaction is gas limited, due to ready access of the
gas or liquid limitation has been suggested based on the gas to the incompletely wetted external catalyst area.
relative availability of the reactant species at the In contrast the upflow reactor (FBR) performs better
reaction site. For the hydrogenation of R-methylstyrene at high pressure when liquid reactant limitation con-
(R-MS), a value of DeBCBi/b(DeAC*A) . 1 would imply trols the rate, due to the completely wetted catalyst in
gaseous reactant limitation, while DeBCBi/b(DeAC*A) , the upflow reactor. Since low molecular weight gas
1 indicates liquid reactant limitation (Wu et al., 1996a). (hydrogen) is used, the effect of pressure and gas
Hence, this reaction was found to be gas-limited at low velocity employed in their study on the catalyst wetting
Figure 10. Comparison between trickle-bed and upflow perfor- Figure 11. Effect of inert fines on the performance of trickle-bed
mance at actual gas velocity ) 4.4 cm/s. X: conversion. tsp: space and upflow reactors at actual gas velocity ) 4.4 cm/s. X: conversion.
time (s) (3600/LHSV). (a) Gas-limited reaction; (b) liquid-limited tsp: space time (s) (3600/LHSV). (a) Gas-limited reaction; (b) liquid-
reaction. System: hydrogenation of R-methylstyrene (in hexane limited reaction. System: hydrogenation of R-methylstyrene (in
solvent) to cumene over 2.5% alumina-supported palladium cata- hexane solvent) to cumene over 2.5% alumina-supported palladium
lyst extrudates. Particle size: 0.13 × 0.56 cm. Reactor diameter: catalyst extrudates. Particle size: 0.13 × 0.56 cm. Fines: 0.2 mm
2.2 cm (after Wu et al., 1996a). silicon carbide. Reactor diameter: 2.2 cm (after Wu et al., 1996a).
efficiency in TBR is insignificant. As liquid mass of conversion with liquid reactant feed concentration
velocity is increased (i.e., space time is decreased), the (typical of zero-order behavior) was observed. The effect
performance of TBR approaches that of FBR due to the of gas velocity on the reactor performance for both
increase in the wetting efficiency of TBR toward com- modes of operation (i.e., TBR and FBR) at all the feed
plete wetting. Figure 10 illustrates the performance concentrations tested was found to be insignificant in
comparison of TBR and FBR without inert fines. The the range of hydrogen velocities (3.8-14.4 cm/s) and
conversions are higher in both TBR (downflow) and FBR reactor pressures (30-300 psig) studied (Wu et al.,
(upflow) reactors operated under liquid-limited reaction 1996a; Khadilkar et al., 1996).
than those obtained in these reactors operated under In beds diluted with fines, it was found that both
gas-limited reactions. The effects of reactor pressure, trickle bed and upflow perform identically under both
liquid reactant feed concentration, and gas velocity on gas- and liquid-limited conditions as shown in Figure
the performance of these reactors for both gas- and 11, corroborating the fact that the fluid dynamics and
liquid-limited conditions that explain such observations kinetics can be decoupled by using fines. This implies
are discussed in detail by Wu et al. (1996a). However, that the performance of the diluted bed is not dependent
the performance of both downflow and upflow improves on the reactant limitation and flow mode used. This
with the reactor pressure due to an increase in the gas conclusion is important in establishing the use of fines
solubility which helps the rate of transport to the wetted in laboratory-scale reactors as an effective and viable
catalyst (in both modes) and improves the driving force scale-up tool possibly to be preferred to upflow reactors.
for gas to catalyst mass-transfer to the inactively wetted Indeed, in the later case, complete wetting occurs at the
catalyst in the downflow mode. At low feed concentra- expense of increased liquid holdup, which in return can
tion of the liquid reactant (R-methylstyrene (CBi) ) 3.1% alter the relative extent of homogeneous and heteroge-
v/v) and high pressure (>100 psig), the reaction becomes neous reactions occurring in more complex systems and
liquid reactant limited, where no further enhancement should therefore be avoided. In contrast, in the bed
is observed when pressure is increased from 100 to 200 packed with fines, wetting is increased without a
psig (where the ratio DeBCBi/b(DeAC*A) drops from 1.5 significant increase in liquid holdup. Wu et al.’s (1996a)
at 100 psig to 0.8 at 200 psig). This means that any study also demonstrated that the advantage of upflow
further increase in the reactor pressure, and hence or downflow depends on whether liquid or gas reactant
liquid phase hydrogen concentration, will have a mini- is rate-limiting and that a single criterion proposed for
mal effect since hydrogen is not the limiting reactant identifying the limiting reactant can explain most of the
anymore (i.e., the effect of pressure diminishes when data reported in the literature on these two modes of
liquid limitation is approached). Moreover, when liquid operation.
limitation is observed, there is no effect of liquid
reactant feed concentration (CBi) on its conversion in 11. Evaluation of TBR Models
either mode of operation. This is a result of the liquid
reactant transport or intrinsic rate limitation which Most of the reported investigations in the literature
shows up as a first-order dependence, making conver- have suggested a plug flow model for the liquid phase
sion independent of feed concentration. At high initial modified by consideration of some other factors such as
liquid reactant feed concentration (CBi ) 7.8% v/v) and external liquid holdup, external contacting, catalyst
low pressure (30 psig) (relatively close to the atmo- effectiveness, etc. Wu et al. (1996b) have evaluated the
spheric pressure), the hydrogenation of R-methylstyrene current plug flow models (Dudukovic, 1977; Henry and
becomes gas limited and is known to exhibit a zeroth- Gilbert, 1973; Mears, 1974) for first-order reactions in
order behavior with respect to R-methylstyrene and first high-pressure TBRs using the isothermal decomposition
order with respect to hydrogen (Beaudry et al., 1986; of hydrogen peroxide on a Cu-Cr catalyst as shown in
Wu et al., 1996a,b). Hence, an inverse proportionality Figure 12. Comparison of model predictions and ex-
Figure 12. Comparison of model predictions and experimental

data for trickle-bed reactors. X: conversion. tsp: space time (s)
(3600/LHSV). ηCE: catalyst wetting efficiency. System: liquid
phase decomposition of hydrogen peroxide in water (first order
reaction). Particles: copper chromite 1/16 in. extrudates. Reactor
diameter: 2.2 cm (after Wu et al., 1996b).
perimental data indicates that both external mass-

transfer effects and incomplete external catalyst wetting
need to be accounted for. Dudukovic (1977) and Beaudry
et al. (1987) pellet-scale approximate models for the
catalyst effectiveness factor adequately simulate both
effects (Wu et al., 1996b). Khadilkar et al. (1996)
evaluated the performance prediction for both TBR and
upflow reactors at high pressure using the reactor-scale
model of El-Hisnawi et al. (1981) and pellet-scale models
(Beaudry et al., 1987; Harold and Watson, 1993). Both
gas- and liquid-limited reactions were used for such an
evaluation. Hydrogenation of R-methylstyrene in hex-
ane to cumene over 2.5% Pd alumina catalyst particles Figure 13. Comparison between model predictions and experi-
was used as a test reaction for both the gas-limited case, mental data for the performance of trickle-bed and upflow reac-
at low pressure and high feed R-MS concentration, and tors: (a) gas limited reaction; (b) liquid limited reactions. X:
the liquid-limited case, at high pressure and low feed conversion. tsp: space time (s) (3600/LHSV). System: hydrogena-
R-MS concentration as mentioned above (Wu et al., tion of R-methylstyrene (in hexane solvent) to cumene over 2.5%
alumina-supported palladium catalyst extrudates. Particle size:
1996a). They found, as shown in Figure 13a,b, that the 0.13 × 0.56 cm. Fines: 0.2 mm silicon carbide. Reactor diameter:
predictions of the reactor-scale and pellet-scale models 2.2 cm (after Khadilkar et al., 1996).
are satisfactory for the current conditions. These
predictions could be improved by using reliable high- the pressure drop at high-pressure operation is equal
pressure correlations for mass-transfer coefficient and to the pressure drop at atmospheric pressure. However,
interfacial area, especially in cases where the rate is at high pressure and high gas flow rate (case 5, section
significantly affected by external mass transfer. 2), the pressure drop increases considerably, liquid
holdup decreases significantly, catalyst wetting ef-
12. Concluding Remarks ficiency improves, gas-liquid interfacial area and volu-
metric mass-transfer coefficient increase, and the trickle-
Based on our review of the available recent investiga- to-pulse transition shifts toward higher liquid flow rates
tions conducted in high-pressure TBRs, the effect of (i.e., the trickle flow regime becomes wider at elevated
pressure on the design and scale-up parameters, such pressure). At liquid velocities corresponding to the
as flow regime transition, pressure drop, liquid holdup, dispersed bubble flow regime, higher gas velocities are
catalyst wetting efficiency, and gas-liquid interfacial required to bring pulse flow.
area and mass-transfer, can be described based on the The improvement in catalyst wetting efficiency due
five limiting cases proposed by Al-Dahhan and Dudu- to diluting the bed with fines is enhanced further at high
kovic (1994, 1995) discussed in section 2. The effect of pressure and high gas flow rate, which result in higher
pressure arises due to an increase in gas density and, pressure drop and lower liquid holdup.
hence, can be simulated by the increase in molecular Trickle-bed reactor (TBR) outperforms the upflow
weight of the gas. When the pressures of gases of reactor (FBR) for gas-limited reaction, while the latter
different molecular weight are set in such a manner so performs better than TBR for a liquid-limited reaction.
as to provide equal gas densities, the effects on the TBR The performance of these reactors becomes similar only
fluid dynamics are about the same. at high liquid mass velocities and/or high pressure and
If the superficial gas velocity is below 2 cm/s (cases 2 high gas velocity when the catalyst in TBR becomes
and 4, section 2), pressure (or gas density) effects are completely wetted similar to that usually encountered
minimal so that a pressurized TBR behaves as if in the upflow reactor (FBR). However, in a diluted bed
operated at about 1 atm. For the operating conditions with fines both the trickle-bed and upflow reactor
in which gas has sufficient inertia, FGUG2 g 1% FLUL2, perform identically under both gas- and liquid-limited
conditions, confirming the fact that the fluid dynamics Ka ) Kapitza dimensionless number,
and kinetics can be decoupled by using fines. The
performance predictions of the current trickle-bed reac- σL3FL
tor models for the investigations presented in the review
are satisfactory for both gas- and liquid-limited reac- µL4g
tions.
It is evident that atmospheric data and models or kLa ) volumetric liquid-side mass-transfer coefficient, s-1
correlations cannot, in general, be extrapolated to L ) liquid mass flux, kg/m2‚s
operation at elevated pressures particularly conditions LHSV ) liquid hourly space velocity (superficial velocity/
of high gas flow rate and high gas molecular weight. bed length), 1/s
P ) pressure, MPa
This can only be accomplished in some special cases
Q ) volumetric flow rate, mL/s
mentioned in this paper. Re ) Reynolds number, FUdp/µ
It is noteworthy to mention that it is valuable to
compare the high-pressure laboratory data and the Re ) Reynolds number (modified),
correlations prediction with large-scale trickle-bed reac- FUdp
tor data to demonstrate the applicability of the high-
pressure findings and correlations and models for µ(1 - )
scaleup or scaledown. Unfortunately, this is not in-
cluded in this paper due to the unavailability of the S ) selectivity
high-pressure large-scale industrial trickle-bed reactor Sc ) Schmidt number, µ/FD
data for the parameters discussed in this paper. T ) temperature, K
t ) time, s
Clearly, a lot of work remains to be done in providing
tsp ) space time (3600/LHSV), s
a fundamentally based description of the effect of
U ) superficial velocity, m/s
pressure on the parameters of importance in TBR V ) interstitial velocity, m/s
operation, design, and scaleup or scaledown. WeL ) liquid Weber number,
Nomenclature FLUL2dp
a ) gas-liquid interfacial area, m2/m3 σL
av ) bed-specific surface area, m2/m3
C* ) gas solubility X ) conversion
C*A ) concentration of gaseous reactant in the liquid phase, XG ) modified Lockhart-Martinelli ratio,
mol/m3
CBi ) concentration of liquid reactant in liquid, mol/m3 UGxFG
Cp ) heat capacity at constant pressure, kcal/kg/K
D ) molecular diffusivity, m2/s or cm2/s ULxFL
DeA ) effective diffusivity of gaseous reactant in the
catalyst, m2/s Z ) bed length, cm
DeB ) effective diffusivity of liquid reactant in the catalyst,
m2/s Greek Letters
Dc ) column diameter, m Rt ) bed-to-wall heat transfer coefficient, W/m2‚K
Dz ) axial dispersion coefficient, m2/s β ) liquid saturation
dK ) Krischer-Kast hydraulic diameter, m ∆P/Z ) pressure drop
) bed porosity
3 L ) liquid holdup
dp
x 163
9π(1 - )2
Φ ) density correction in Charpentier diagram, see eq 2
ηCE ) wetting efficiency
κ ) thermal conductivity, J/m‚s‚K
λ ) dimensionless parameter in eq 2,
dp ) particle diameter, m
E1, E2 ) Ergun constants, eq 21
x
f ) two-phase flow friction factor,
FGFL
FaFw
∆P/ZdK
µ ) dynamic viscosity, Pa‚s
2FGUG2 πLV ) vapor-to-feed molar ratio
F ) density, kg/m3
Fi ) volume-averaged forces exerted on phase i by the other σ ) surface tension, N/m
phases, N/m3 τLG ) gas-liquid interfacial shear stress, Pa
Fr ) Froude number, U2/gdp ψ ) dimensionless parameter in eq 2,
()()
G ) gas mass flux, kg/m2‚s 1/3 2/3
g ) gravitational acceleration, m/s2
σw µL Fw
Ga ) Galileo number, σL µw FL
dp3F2g3 Subscripts
AL ) gaseous reactant dissolved in liquid
µ2(1 - )3 a ) active liquid holdup, or air
e ) external liquid holdup
H ) Henry’s constant G ) gas
Jdf ) drift flux, m/s GL ) two-phase
L ) liquid Beaudry, E. G.; Mills, P. L.; Dudukovic, M. P. Gas-Liquid
mix ) mixture Downflow, Upflow, and Countercurrent Flow in Packed
nc ) noncapillary BedssReactor Performance for Gas Limiting Reaction. World
Congress III of Chemical Engineering, Tokyo, Japan, Sept 1986;
r ) reduced Paper 9b-152, p IV-173.
trans ) transition Beaudry, E. G.; Dudukovic, M. P.; Mills, P. L. Trickle-Bed
w ) water Reactors: Liquid Diffusional Effects in a Gas-Limited Reaction.
AIChE J. 1987, 33, 1435.
Superscript Belhaj, M. S. Design, Analysis and Modeling of an Enzymatic
° ) atmospheric Oxidation Three-Phase Reactor Using Immobilized Enzymes
(Conception, Analyse et Modélisation d’un Réacteur d’Oxydation
Acronyms Enzymatique Triphase á Enzymes Immobilisées). Ph.D. Dis-
sertation, Université de Technologie de Compiegne, Compiegne,
DBT ) dibenzothiophene France, 1984.
DEA ) diethanolamine Cassanello, M.; Larachi, F.; Laurent, A.; Wild, G.; Midoux, N. Gas-
ETG ) ethylene glycol Liquid Mass Transfer in High Pressure Trickle-Bed Reactors:
FBR ) flooded-bed reactor Experiments and Modeling. In 3rd International Symposium
HIR ) high interaction regime on High Pressure Chemical Engineering, Zürich, Switzerland,
Oct 7-9, 1996; Rudolf von Rohr, Ph., Trepp, Ch., Eds.; High
HVGO ) heavy gas oil Pressure Chemical Engineering; Amsterdam Process Technol-
LIR ) low interaction regime ogy Proceedings; Elsevier: Amsterdam, The Netherlands, 1996;
PC ) propylene carbonate Vol. 12, p 493.
TBR ) trickle-bed reactor Charpentier, J. C. Recent Progress in Two-Phase Gas-Liquid
V ) downflow Mass Transfer in Packed Beds. Chem. Eng. J. 1976, 11, 161.
v ) upflow Charpentier, J. C. Gas-Liquid Reactors. In Chemical Reaction
Engineering Reviews; Luss, D., Weekman, V. W., Eds.; ACS
Symposium Series 72; American Chemical Society: Washington,
Literature Cited DC, 1978; p 223.
Charpentier, J. C. Hydrodynamics of Two-Phase Flow Through
Abbott, M. D.; Moore, G. R.; Ralph, J. L. Proceedings of the 7th
Porous Media. In Chemical Engineering of Gas-Liquid-Solid
ABG Conference. In Gas-liquid-solid reactor design; Shah, Y.
Catalyst Reactions; L’Homme, G. A., Ed.; CEBEDOC: Liège,
T., Ed.; McGraw-Hill: New York, 1979; Paper S-4.
Belgium, 1979; p. 78.
Al-Dahhan, M. H. Effects of High Pressure and Fines on the
Charpentier, J. C. What’s New in Absorption with Chemical
Hydrodynamics of Trickle-Bed Reactors. D.Sc. Dissertation,
Reaction. Trans. Inst. Chem. Eng. 1982, 60, 131.
Washington University, St. Louis, MO, 1993.
Charpentier, J. C. Mass Transfer in Fixed Bed Reactors. In
Al-Dahhan, M. H.; Dudukovic, M. P. Pressure Drop and Liquid
Multiphase Chemical ReactorssTheory, Design, Scale-up; Gi-
Hold-up in High Pressure Trickle-Bed Reactors. Chem. Eng. Sci.
anetto, A., Silveston, P. L., Eds.; Hemisphere Publ. Co.: Wash-
1994, 49, 5681.
ington, 1986; p 289.
Al-Dahhan, M. H.; Dudukovic, M. P. Catalyst Wetting Efficiency
Charpentier, J. C.; Favier, M. Some Liquid Hold-up Experimental
in Trickle-Bed Reactors at High Pressure. Chem. Eng. Sci. 1995,
Data in Trickle-Bed Reactors for Foaming and Nonfoaming
50, 2377.
Hydrocarbons. AIChE J. 1975, 21, 1213.
Al-Dahhan, M. H.; Dudukovic, M. P. Catalyst Bed Dilution for
Improving Catalyst Wetting in Laboratory Trickle-Bed Reactors. Charpentier, J. C.; Bakos, M.; LeGoff, P. Question IV, Rapport
AIChE J. 1996, 42, 2594. 10, Société Hydrotechnique de France, XII èmes Journées de
l’Hydraulique, Paris, 1972.
Al-Dahhan, M. H.; Wu, Y. X.; Dudukovic, M. P. Reproducible
Technique for Packing Laboratory-Scale Trickle-Bed Reactors Chen, S.; Chuang, K. T. Simultaneous Methanol Removal and
with a Mixture of Catalyst and Fines. Ind. Eng. Chem. Res. Destruction from Wastewater in a Trickle-Bed Reactor. Can.
1995, 34, 741. J. Chem. Eng. 1992, 70, 727.
Al-Dahhan, M. H.; Wu, Y. X.; Khadilkar, M. R.; Dudukovic, M. P. Crine, M.; L’Homme, G. Recent Trends in the Modelling of
Improved Prediction of Pressure Drop and Liquid Hold-up in Catalytic Trickle-Bed Reactors. In Mass Transfer with Chemical
High Pressure Trickle-Bed Reactors. 5th World Congress of Reaction in Multiphase Systems: Three-Phase Systems; Alper,
Chemical Engineering; Symposium 3: Advanced Fundamentals, E., Ed.; Martinus Nijhoff Publ.: The Hague, The Netherlands,
Session: Catalysis, Kinetics and Reaction Engineering, San 1983; Vol. II; p 99.
Diego, CA, July 14-18, 1996, American Institute of Chemical Dankworth, D. C.; Sundaresan, S. Time-Dependent Vertical Gas-
Engineering (AIChE): New York, 1996; Vol. 1, p 209. Liquid Flow in Packed Beds. Chem. Eng. Sci. 1992, 47, 337.
Babu, S. P.; Shah, B.; Talwalkar, A. Fluidization Correlations for Dankworth, D. C.; Kevrekidis, I. G.; Sundaresan, S. Dynamics of
Coal Gasification MaterialssMinimum Fluidization Velocity Pulsing Flow in Trickle Beds. AIChE J. 1990a, 36, 605.
and Fluidized Bed Expansion Ratio. AIChE Symp. Ser. 1978, Dankworth, D. C.; Kevrekidis, I. G.; Sundaresan, S. Time-
74, 176. Dependent Hydrodynamics in Multiphase Flow. Chem. Eng. Sci.
Bailey, J. E.; Ollis, D. F. Biochemical Engineering Fundamentals, 1990b, 45, 2239.
2nd ed.; McGraw-Hill Book Co. New York, 1986. Diks, R. M.; Ottengraf, S. P. P. Verification of a Simplified Model
Baker, B., III. Determination of the Extent of Catalyst Utilization for Removal of Dichloromethane from Waste Gases Using a
in a Trickle-Flow Reactor. ACS Symp. Ser. 1978, 65, 425. Biological Trickling Filter. Bioprocess Eng. 1991, 6, 131.
Baldi, G. Design and Scale-up of Trickle-Bed Reactors. Solid- Dimenstein, D. M.; Ng, K. M. A Model for Pulsing Flow in
Liquid Contacting Effectiveness. In Multiphase Chemical Reac- Cocurrent Downflow Trickle Bed Reactors. Chem. Eng. Com-
tors; Rodrigues, A. E., Calo, J. M., Sweed, N. H., Eds.; NATO mun. 1986, 41, 215.
Advanced Study Institite Series E; Sitjhoff & Noordhoff: Aal- Dimenstein, D. M.; Zimmerman, S. P.; Ng, K. M. Trickling and
phen aan den Rijn, The Netherlands, 1981a; Vol. 52, p 323. Pulsing Transition in Cocurrent Downflow Trickle-Bed Reactors
Baldi, G. Heat Transfer in Gas-Liquid-Solid Reactors. In Mul- with Special Reference to Large-Scale Columns. ACS Symp. Ser.
tiphase Chemical Reactors; Rodrigues, A. E., Calo, J. M., Sweed, 1984, 234, 3.
N. H., Eds.; NATO Advanced Study Institute Series E; Sitjhoff Dirkx, J. M. H. De Trickle-Bed Reactor Part I. Polytech. Tijd-
& Noordhoff: Aalphen aan den Rijn, The Netherlands, 1981b; schr.: Procestech. 1979a, 34, 227.
Vol. 52, p 307. Dirkx, J. M. H. De Trickle-Bed Reactor Part II. Polytech. Tijdschr.
Baldi, G. Hydrodynamics of Multiphase Reactors. In Multiphase Procestech. 1979b, 34, 316.
Chemical Reactors; Rodrigues, A. E., Calo, J. M., Sweed, N. H., Dudukovic, M. P. Catalyst Effectiveness Factor and Contacting
Eds.; NATO Adv. Study Inst. Series E, Sitjhoff & Noordhoff: Efficiency in Trickle-Bed Reactors. AIChE J. 1977, 23, 940.
Aalphen aan den Rijn, The Netherlands, 1981c; Vol. 52, p 271. Dudukovic, M. P.; Mills, P. L. Catalyst Effectiveness Factors in
Baldi, G.; Gianetto, A.; Sicardi, S.; Specchia, V.; Mazzarino, I. Trickle-Bed Reactors. ACS Symp. Ser. 1978, 65, 387.
Oxidation of Ethyl Alcohol in Trickle-Bed Reactors: Analysis Dudukovic, M. P.; Mills, P. L. Contacting and Hydrodynamics in
of the Conversion Rate. Can. J. Chem. Eng. 1985, 63, 62. Trickle-Bed Reactors. In Encyclopedia of Fluid Mechanics:
Gas-Liquid Flows; Cheremisinoff, N. P., Eds.; Gulf Publ. Hasseni, W.; Laurent, A.; Midoux, N.; Charpentier, J. C. Flow
Corp.: Houston, TX, 1986; Vol. 3, p 969. Regimes in Trickle Beds at High Pressure (0.1-10 MPa)
Dudukovic, M. P.; Devanathan, N.; Holub, R. Multiphase Reac- (Régimes d’EÄ coulement dans un Réacteur Catalytique à Lit Fixe
tors: Models and Experimental Verification. Rev. Inst. Fr. Pet. Arrosé Fonctionnant sous Pression (0.1-10 MPa) à Co-courant
1991, 46, 439. de Gaz et de Liquide vers le Bas). Entropie 1987, 137/138, 127.
Dwivedi, P. N.; Upadhyay, S. N. Particle-Fluid Mass Transfer in Haure, P. M.; Hudgins, R. R.; Silveston, P. L. Steady-State Models
Fixed and Fluidized Beds. Ind. Eng. Chem. Process Des. Dev. for SO2 Oxidation in Trickle-Bed Reactors. Chem. Eng. J. 1990,
1977, 16, 157. 43, 121.
El-Hisnawi, A. A. Tracer and Reaction Studies in Trickle-Bed Henry, H. C.; Gilbert, J. B. Scale-up of Pilot Plant Data for
Reactors. D.Sc. Dissertation, Washington University, St. Louis, Catalytic Hydroprocessing. Ind. Eng. Chem. Process Des. Dev.
MO, 1981. 1973, 12, 328.
El-Hisnawi, A. A.; Dudukovic, M. P.; Mills, P. L. Trickle-Bed Herskowitz, M.; Smith, J. M. Trickle-Bed Reactors: A Review.
Reactors: Dynamic Tracer Tests, Reaction Studies and Model- AIChE J. 1983, 29, 1.
ing of Reactor Performance. ACS Symp. Ser. 1981, 196, 421. Hofmann, H. Hydrodynamics, Transport Phenomena, and Math-
Ellman, M. J. Characteristics of CoCurrent Downflow Three-Phase ematical Models in Trickle-Bed Reactors. Int. Chem. Eng. 1977,
Fixed-Bed Reactors (Caractéristiques des Réacteurs Tripha- 17, 19. Also in Chem. Ing. Technol. 1975, 47, 823.
siques à Lit Fixe Fonctionnant à Co-courant vers le Bas). Ph.D. Hofmann, H. Multiphase Catalytic Packed Bed Reactors. Catal.
Dissertation, Institut National Polytechnique de Lorraine, Rev. Sci. Eng. 1978, 17, 71.
Nancy, France, 1988. Hofmann, H. Fluid Dynamics, Mass Transfer and Chemical
Ellman, M. J.; Midoux, N.; Laurent, A.; Charpentier, J. C. A New, Reaction in Multiphase Catalytic Fixed Bed Reactors. In Mass
Improved Pressure Drop Correlation for Trickle-Bed Reactors. Transfer with Chemical Reaction in Multiphase Systems: Three-
Chem. Eng. Sci. 1988, 43, 2201. Phase Systems; Alper, E., Ed.; Martinus Nijhoff Publ.: The
Ellman, M. J.; Midoux, N.; Wild, G.; Laurent, A.; Charpentier, J. Hague, The Netherlands, 1983; Vol. II, p 73.
C. A New, Improved Liquid Hold-up Correlation for Trickle- Holub, R. A.; Dudukovic, M. P.; Ramachandran, P. A. A Phenom-
Bed Reactors. Chem. Eng. Sci. 1990, 45, 1677. enological Model of Pressure Drop, Liquid Hold-up and Flow
Ergun, S. Fluid Flow through Packed Columns. Chem. Eng. Prog. Regime Transition in Gas-Liquid Trickle Flow. Chem. Eng. Sci.
1952, 48, 89. 1992, 47, 2343-2348.
Euzen, J. P.; Trambouze, P.; Wauquier, J. P. Methodology for the Holub, R. A.; Dudukovic, M. P.; Ramachandran, P. A. Pressure
Scale-up of Chemical Processes (Méthodologie pour l’extrapolation Drop, Liquid Hold-up and Flow Regime Transition in Trickle
des procédés chimiques). Publications de l’Institut Français du Flow. AIChE J. 1993, 39, 302-321.
Pétrole; E
Ä ditions Technip Institute of French Studies: New Kawabata, J.; Yumiyama, M.; Tazaki, Y.; Honma, S.; Chiba, T.;
York, 1993. Sumiya, T.; Endo, K. Characteristics of Gas Fluidized Beds
Froment, G. F.; Depauw, G. A.; Vanrysselberghe, V. Kinetic under Pressure. J. Chem. Eng. Jpn. 1981, 14, 85-89.
Modelling and Reactor Simulation in Hydrodesulfurization of Khadilkar, M. R.; Wu, Y. X.; Al-Dahhan, M. H.; Dudukovic, M.
Oil Fractions. Ind. Eng. Chem. Res. 1994, 33, 2975. P.; Colakyan, M. Comparison of Trickle-Bed and Upflow Reactor
Germain, A. Industrial Applications of Three-Phase Catalytic Performance at High Pressure: Model Predictions and Experi-
Fixed Bed Reactors. In Mass Transfer with Chemical Reaction mental Observations. Chem. Eng. Sci. 1996, 51, 2139-2148.
in Multiphase Systems: Three-Phase Systems; Alper, E., Ed.; Kiared, K.; Zoulalian, A. Study and Modeling of Catalytic Sulfur
Martinus Nijhoff Publ.: The Hague, The Netherlands, 1983; Dioxide Oxidation in Verlifix Three-Phase Reactor. Chem. Eng.
Vol. II, p 19. Sci. 1992, 47, 3705-3712.
Germain, A.; L’Homme, G.; Lefebvre, A. The Trickle Flow and Kohler, M.; Richarz, W. Untersuchungen des Flussigkeitshold-up
Bubble Flow Reactors in Chemical Processing. In Chemical in Rieselbettreaktoren. Chem. Eng. Technol. 1984, 56, 1-19.
Engineering of Gas-Liquid-Solid Catalyst Reactions; L’Homme, Kohler, M.; Richarz, W. Investigation of Liquid Hold-up in Trickle
G. A., Ed.; CEBEDOC: Liege, Belgium, 1979; p 265. Bed Reactors. Ger. Chem. Eng. 1985, 8, 295-300.
Gianetto, A.; Berruti, F. Modelling of Trickle-Bed Reactors. In Koros, R. M. Scale-up Considerations for Mixed Phase Catalytic
Chemical Reactor Design and Technology; DeLasa, H. I., Ed.; Reactors. In Multiphase Chemical Reactors; Rodrigues, A. E.,
Nijhoff: The Hague, The Netherlands, 1986; p 631. Calo, J. M., Sweed, N. H., Eds.; NATO Advanced Study Institute
Gianetto, A.; Specchia, V. Trickle-Bed Reactors: State of Art and Series E; Sitjhoff & Noordhoff: Aalphen aan den Rijn, The
Perspectives. Chem. Eng. Sci. 1992, 47, 3197. Netherlands, 1981; Vol. 52, p 429.
Gianetto, A.; Baldi, G.; Specchia, V.; Sicardi, S. Hydrodynamics Krauze, R.; Serwinski, M. Moisted Surface and Fractional Wetted
and Solid-Liquid Contacting Effectiveness in Trickle-Bed Reac- Area of Ceramic Raschig Rings. Inz. Chem. 1971, 1, 415.
tors. AIChE J. 1978, 24, 1087. L’Air Liquide. Encyclopédie des Gaz; Elsevier: Amsterdam, The
Gierman, H. Design of Laboratory Hydrotreating Reactors Scaling Netherlands, 1976.
Down Trickle-Flow Reactors. Appl. Catal. 1988, 43, 277. Larachi, F. Three-Phase Trickle and Flooded Fixed-Bed Reac-
Goto, S.; Smith, J. M. Trickle-Bed Reactor Performance. Part II: tors: Influence of Pressure on the Hydrodynamics and the Gas-
Reaction Studies. AIChE J. 1974, 21, 714. Liquid Mass Transfer (Les Réacteurs Triphasiques à Lit Fixe
Goto, S.; Smith, J. M. Trickle-Bed Reactor Performance. AIChE à E
Ä coulement à Co-courant vers le Bas et vers le Haut de Gaz
J. 1975, 21, 706. et de Liquide. E Ä tude de l’Influence de la Pression sur
Goto, S.; Mabuchi, K. Oxidation of Ethanol in Gas-Liquid Upflow l’Hydrodynamique et le Transfert de Matière Gaz-Liquide).
and Downflow Reactors. Can. J. Chem. Eng. 1984, 62, 865. Ph.D. Dissertation, Institut National Polytechnique de Lorraine,
Goto, S.; Levec, J.; Smith, J. M. Trickle-Bed Oxidation Reactors. Nancy, France, 1991.
Catal. Rev. Sci. Eng. 1977, 15, 187. Larachi, F.; Midoux, N.; Laurent, A. Hydrodynamic Study of a
Grosser, K.; Carbonell, R. G.; Sundaresan, S. Onset of Pulsing in Pressurized Trickle-Bed Reactor. Liquid Saturation, Single-
Two-Phase Cocurrent Downflow Through a Packed Bed. AIChE Phase and Two-Phase Pressure Drop. In 2nd International
J. 1988, 34, 1850. Symposium on High Pressure Chem. Engineering, Erlangen,
Gupta, R. In Handbook of Fluids in Motion; Cheremisinoff, N. P., Germany, Sept 24-26, 1990; Vetter G., Ed.; DECHEMA:
Gupta, R., Eds.; Ann Arbor Science: Ann Arbor, MI, 1985. Frankfurt, Germany, 1990; p 441.
Handley, D.; Heggs, P. J. Momentum and Heat Transfer Mecha- Larachi, F.; Laurent, A.; Midoux, N.; Wild, G. Experimental study
nisms in Regular Shaped Packings. Trans. Inst. Chem. Eng. of a Trickle-Bed Reactor Operating at High Pressure: Two-
1968, 46, T251. Phase Pressure Drop and Liquid Saturation. Chem. Eng. Sci.
Hanika, J.; Stanek, V. Operation and Design of Trickle-Bed 1991a, 46, 1233.
Reactors. In Handbook of Heat and Mass Transfer: Mass Larachi, F.; Laurent, A.; Wild, G.; Midoux, N. Some Experimental
Transfer and Reactor Design; Cheremisinoff, N. P., Ed.; Gulf Liquid Saturation Results in Fixed-Bed Reactors Operated
Publ. Corp.: Houston, TX, 1986; Vol. 2, p 1029. under Elevated Pressure in Cocurrent Upflow and Downflow
Harold, M. P.; Watson, P. C. Bimolecular Exothermic Reaction of the Gas and the Liquid. Ind. Eng. Chem. Res. 1991b, 30,
With Vaporization in the Half-Slab Catalyst. Chem. Eng. Sci. 2404.
1993, 48, 981. Larachi, F.; Laurent, A.; Midoux, N.; Wild, G. Liquid Saturation
Hartman, M.; Coughlin, R. W. Oxidation of SO2 in Trickle-Bed Data in Trickle Beds Operating Under Elevated Pressure.
Packed with Carbon. Chem. Eng. Sci. 1971, 27, 867. AIChE J. 1991c, 37, 1109.
Larachi, F.; Laurent, A.; Midoux, N. Pressure Effects on the Liquid Ng, K. M.; Chu, C. F. Trickle-Bed Reactors. Chem. Eng. Prog. 1987,
Saturation of a Trickle-Bed Reactor. Technol. Today 1991d, 1, Nov. 55.
146. Olowson, P. A.; Almstedt, A. E. Influence of Pressure on the
Larachi, F.; Laurent, A.; Wild, G.; Midoux, N. Pressure Effects on Minimum Fluidization Velocity. Chem. Eng. Sci. 1991, 46, 637.
Gas-Liquid Interfacial Areas in Cocurrent Trickle-Flow Reac- Onda, K.; Takeuchi, H.; Kayama, Y. Effect of Packing Materials
tors. Chem. Eng. Sci. 1992, 47, 2325. on Wetted Surface Area. Kagaku Kogaku 1967, 31, 126.
Larachi, F.; Laurent, A.; Wild, G.; Midoux, N. Effect of Pressure Østergaard, K. Gas-Liquid-Particle Operations in Chemical
on the Trickle-Pulsed Transition in Irrigated Fixed Bed Cata- Reaction Engineering. In Advances in Chemical Engineering;
lytic Reactors (Effet de la Pression sur la Transition Ruisselant- Academic Press: Drew, T. B., Cokelet, G. R., Hoopes, J. W.,-
Pulsé dans les Réacteurs Catalytiques à Lit Fixe Arrosé). Can. Vermeulen, T., Eds.; New York, 1968; Vol. 7, p 71.
J. Chem. Eng. 1993a, 71, 319. Oyevaar, M. H.; de la Rie, T.; van der Sluijs, C. L.; Westerterp, K.
Larachi, F.; Laurent, A.; Wild, G.; Midoux, N. Influence of Pressure R. Interfacial Areas and Gas Hold-ups in Bubble columns and
on the Hydrodynamics of Flooded-Fixed Beds (Contribution à Packed Bubble columns at Elevated Pressures. Chem. Eng.
l’E
Ä tude de l’Influence de la Pression sur l’Hydrodynamique des Proc. 1989, 26, 1.
Lits Fixes Noyés). Récents Prog. Génie Procédés 1993b, 30, 79. Papayannakos, N. G.; Galtier, P. A.; Bigeard, P. H.; Kasztelan, S.
Larachi, F.; Wild, G.; Laurent, A.; Midoux, N. Influence of Gas Hydrodynamic Effects in Bench Scale Hydrotreaters Operating
Density on the Hydrodynamics of Cocurrent Gas-Liquid Upflow in Cocurrent Gas-Liquid Upflow Mode. Chem. Eng. Sci. 1992,
Fixed-Bed Reactors. Ind. Eng. Chem. Res. 1994, 33, 519. 47, 2275.
Lara-Márquez, A.; Larachi, F.; Wild, G.; Laurent, A. Mass Tranfer Perrut, M. Flow of Supercritical Fluids Through Porous Media
Characteristics of Fixed Beds with Cocurrent Upflow and (EÄ coulement des Fluides en EÄ tat Supercritique dans les Milieux
Downflow. A Special Reference to the Effect of Pressure. Chem. Poreux). Récents Prog. Génie Procedes 1987, 2, 260.
Eng. Sci. 1992, 47, 3485. Puranik, S. S.; Vogelpohl, V. Effective Interfacial Area in Irrigated
Larkins, R. P.; White, R. R.; Jeffrey, D. W. Two-Phase Cocurrent Packed Columns. Chem. Eng. Sci. 1974, 29, 501.
Upflow and Downflow. AIChE J. 1961, 7, 231. Ramachandran, P. A.; Chaudhari, R. V. Three-Phase Catalytic
Lazzaronni, C. L.; Keselman, H. R.; Figoli, H. S. Calorimetric Reactors; Gordon & Breach: New York, 1983.
Evaluation of the Efficiency of Liquid-Solid Contacting in Ramachandran, P. A.; Dudukovic, M. P.; Mills, P. L. Recent
Trickle Flow. Ind. Eng. Chem. Res. 1988, 27, 1132. Advances in the Analysis and Design of Trickle-Bed Reactors.
Lemcoff, N. O.; Cukierman, A. L.; Martinez, O. M. Effectiveness Sadhana 1987, 10, 269.
Factor of Partially Wetted Catalyst Particles: Evaluation and Rao, V. G.; Drinkenburg, A. A. H. Solid-Liquid Mass Transfer in
Application to the Modeling of Trickle Bed Reactors. Catal. Rev. Packed Beds with Cocurrent Gas-Liquid Downflow. AIChE J.
Sci. Eng. 1988, 30, 393. 1985, 31, 1059.
Levec, J.; Smith, J. M. Oxidation of Acetic Acid Solutions in a Reid, R. C.; Prausnitz, J. M.; Poling, B. E. The Properties of Gases
Trickle-Bed Reactor. AIChE J. 1976, 22, 159. and Liquids, 4th ed.; McGraw-Hill: New York, 1987.
Levec, J.; Lakota, A. Liquid-Solid Mass Transfer in Packed Beds
Ring, Z. E.; Missen, R. W. Trickle-Bed Reactors: Tracer Study of
with Cocurrent Downward Two-Phase Flow. In Heat and Mass
Liquid Hold-up and Wetting Efficiency at High Temperature
Transfer in Porous Media; Quintard, M., Todorovic, M., Eds.;
and Pressure. Can. J. Chem. Eng. 1991, 69, 1016.
Elsevier: Amsterdam, The Netherlands, 1992, p 663.
Rowe, P. N. The Effect of Pressure on Minimum Fluidisation
Lockhart, R. W.; Martinelli, R. C. Proposed Correlation of Data
Velocity. Chem. Eng. Sci. 1984, 39, 173.
for Isothermal Two-Phase, Two-Component Flow in Pipes.
Chem. Eng. Prog. 1949, 45, 39. Rowe, P. N.; Foscolo, P. U.; Hoffmann, A. C.; Yates, J. G. X-ray
Observations of Gas-Fluidised Beds under Pressure. In Fluidi-
Martinez, O. M.; Cassanello, M. C.; Cukierman, A. L. Three-Phase
zation IV; Kunii, D., Toei, R., Eds.; Engineering Foundation:
Fixed Bed Catalytic Reactors: Application to Hydrotreatment
New York, 1984; p 53.
Processes. Trends Chem. Eng. 1994, 2, 393.
Ruecker, C. M.; Akgerman, A. Determination of Wetting Efficien-
Mata, A. R., Smith, J. M. Oxidation of Sulfur Oxide in Trickle-
cies for a Trickle-Bed Reactor at High Temperatures and
Bed Reactor. Chem. Eng. J. 1981, 22, 229.
Pressures. Ind. Eng. Chem. Res. 1987, 26, 164.
McManus, R. L.; Funk, G. A.; Harold, M. P.; Ng, K. M. Experi-
mental Study of Reaction in Trickle-Bed Reactors with Liquid Saada, M. Y. Fluid Mechanics of Cocurrent Two-Phase Flow in
Maldistribution. Ind. Eng. Chem. Res. 1993, 32, 570. Packed Beds: Pressure Drop and Liquid Hold-up Studies.
Periodica Polym. Chem. Eng. 1975, 19, 317.
Mears, D. E. The Role of Axial Dispersion in Trickle Flow
Laboratory Reactors. Chem. Eng. Sci. 1974, 26, 1361. Salatino, P.; Massimila, L. Pressure Drop in Flow of Nearly
Critical Fluid through Packed Beds of Spheres. Chem. Eng.
Meyers, R. A. Handbook of Petroleum Refining Processes, 2nd ed.;
Commun. 1990, 93, 101.
McGraw-Hill: New York, 1996.
Midoux, N.; Favier, M.; Charpentier, J. C. Flow Pattern, Pressure Saroha, A. K.; Nigam, K. D. P. Trickle-Bed Reactors. Reviews in
Loss and Liquid Hold-up Data in Gas-Liquid Downflow Packed Chemical Engineering; 1996; Vol. 12, Chapters 3-4, p 207.
Beds with Foaming and Nonfoaming Hydrocarbons. J. Chem. Satterfield, C. N. Trickle-Bed Reactors. AIChE J. 1975, 21, 209.
Eng. Jpn. 1976, 9, 350. Satterfield, C. N.; Way, P. F. The Role of the Liquid Phase in the
Mills, P. L.; Dudukovic, M. P. Evaluation of Liquid-Solid Contact- Performance of a Trickle-Bed Reactor. AIChE J. 1972, 18, 305.
ing in Trickle Beds by Tracer Methods. AIChE J. 1981, 27, 893. Satterfield, C. N.; Van Eek, M. W.; Bliss, G. S. Liquid-Solid Mass
Mills, P. L.; Dudukovic, M. P. A Comparison of Current Models Transfer in Packed Bed with Downward Cocurrent Gas-Liquid
for Isothermal Trickle-Bed Reactors. Application of a Model Flow. AIChE J. 1978, 24, 709.
Reaction System. ACS Symp. Ser. 1984, 234, 37. Saxena, S. C.; Vogel, G. J. The Properties of a Dolomite Bed of a
Molerus, O.; Schweintzer, J. Resistance of Particles Beds at Range of Particle Sizes and Shapes at Minimum Fluidization.
Reynolds Numbers up to Re ∼ 104. Chem. Eng. Sci. 1989, 44, Can. J. Chem. Eng. 1976, 54, 453.
1071. Schwartz, J. G.; Weger, E.; Dudukovic, M. P. A New Tracer Method
Morsi, B. I.; Laurent, A.; Midoux, N.; Barthole-Delauney, G.; for Determination of Solid-Liquid Contacting Efficiency in
Storck, A.; Charpentier, J. C. Hydrodynamics and Gas-Liquid- Trickle-Bed Reactors. AIChE J. 1976, 22, 894.
Solid Interfacial Parameters of Co-Current Downward Two- Sebastian, H. M.; Lin, H. M.; Chao, K. C. Correlation of Solubility
Phase Flow in Trickle-Bed Reactors. AIChE Annual Meeting, of Hydrogen in Hydrocarbon Solvent. AIChE J. 1981, 27, 138.
New Orleans, LA, Nov 8-12, 1981. Shah, Y. T. Gas-Liquid-Solid Reactor Design; McGraw-Hill: New
Morsi, B. I.; Midoux, N.; Laurent, A.; Charpentier, J. C. Hydro- York, 1979.
dynamics and Interfacial Areas in Downward Cocurrent Gas- Shulman, H. L.; Ullrich, C. F.; Proulx, A. Z.; Zimmerman, J. O.
Liquid Flow Through Fixed Beds. Influence of the Nature of Performance of Packed Column II: Wetted and Effective
the Liquid. Int. Chem. Eng. 1982, 22, 141. Interfacial Area, Gas and Liquid Mass Trasnfer Rates. AIChE
Nakamura, M.; Hamada, Y.; Toyama, S.; Fouda, A. E.; Capes, C. J. 1955, 1, 353.
E. An Experimental Investigation of Minimum Fluidization Sie, S. T. Scale Effects in Laboratory and Pilot-Plant Reactors for
Velocity at Elevated Temperatures and Pressures. Can. J. Trickle-Flow Processes. Rev. Inst. Fr. Pet. 1991, 46, 501.
Chem. Eng. 1985, 63, 8. Simpson, H. C.; Rodger, B. W. The Fluidization of Light Solids by
Ng, K. M. A Model for Flow Regime Transitions in Cocurrent Gases under Pressure and Heavy Solids by Water. Chem. Eng.
Downflow Trickle-Bed Reactors. AIChE J. 1986, 32, 115. Sci. 1961, 16, 153.
Specchia, V.; Baldi, G. Pressure Drop and Liquid Hold-up for Two- Vergel, H. C. A.; Euzen, J. P.; Trambouze, P.; Wauquier, J. P. Two-
Phase Cocurrent Flow in Packed Beds. Chem. Eng. Sci. 1977, Phase Flow Catalytic Reactor. Influence of Hydrodynamics on
32, 515. Selectivity. Chem. Eng. Sci. 1995, 50, 3303.
Specchia, V.; Baldi, G.; Gianetto, L. Solid-Liquid Mass Transfer Wammes, W. J. A. Hydrodynamics in cocurrent gas-liquid trickle-
in Cocurrent Two-Phase Flow Through Packed Beds. Ind. Eng. bed reactor at elevated pressure. Ph.D. Thesis, University of
Chem. Process Des. Dev. 1978, 17, 362. Twente, Enschede, The Netherlands, 1990.
Stuber, F.; Wilhelm, A. M.; Delmas, H. Modelling of Three-Phase Wammes, W. J. A.; Westerterp, K. R. Hydrodynamics in a
Catalytic Upflow Reactor: A Significant Chemical Determina- Cocurrent Gas-Liquid Trickle-Bed Reactor at Elevated Pres-
tion of Liquid-Solid and Gas-Liquid Mass Transfer Coef- sures. In 2nd International Symposium on High Pressure
ficients. Chem. Eng. Sci. 1996, 51, 2161. Chemical Engineering, Erlangen, Germany, Sept 24-26, 1990;
Sundaresan, S. Mathematical Modeling of Pulsing Flow in Large Vetter, G., Ed.; DECHEMA: Frankfurt, Germany, 1990; p 435.
Trickle Beds. AIChE J. 1987, 33, 455.
Tahraoui, K. Hydrodynamics, Mass Transfer, Implementation and Wammes, W. J. A.; Westerterp, K. R. Hydrodynamics in a
Modeling of a Catalytic Reaction in a Three-Phase Verlifix Pressurized Cocurrent Gas-Liquid Trickle-Bed Reactor. Chem.
Reactor Equipped with a Jet Venturi (Hydrodynamique, Trans- Eng. Technol. 1991, 14, 406.
ferts de Matiere, Mise en Æuvre et Modelisation d’une Reaction Wammes, W. J. A.; Mechielsen, S. J.; Westerterp, K. R. The
Catalytique dans un Reacteur Triphase Verlifix Muni d’un Transition Between Trickle Flow and Pulse Flow in a Co-
Venturi a jet). Ph.D. Dissertation, Institut National Polytech- current Gas-Liquid Trickle-Bed Reactor at Elevated Pressures.
nique de Lorraine, Nancy, France, 1990. Chem. Eng. Sci. 1990a, 45, 3149.
Talmor, E. Two-Phase Downflow Through Catalyst Beds. AIChE Wammes, W. J. A.; Mechielsen, S. J.; Westerterp, K. R. The
J. 1977, 23, 868. Influence of the Reactor Pressure on the Hydrodynamics in a
Tan, C. S.; Smith, J. M. A Dynamic Method for Liquid-Particle Cocurrent Gas-Liquid Trickle-Bed Reactor. Chem. Eng. Sci.
Mass Transfer in Trickle Beds. AIChE J. 1982, 28, 190. 1990b, 45, 2247.
Tarhan, M. O. Catalytic Reactor Design; McGraw-Hill: New York, Wammes, W. J. A.; Mechielsen, S. J.; Westerterp, K. R. The
1983. Influence of Pressure on the Liquid Hold-up in a Cocurrent
Thanos, A. M.; Galtier, P. A.; Papayannakos, N. G. Liquid Flow Gas-Liquid Trickle-Bed Reactor Operating at low Gas Veloci-
Non-Idealities and Hold-Up in a Pilot Scale Packed Bed Reactor ties. Chem. Eng. Sci. 1991a, 46, 409.
with Cocurrent Gas-Liquid Upflow. Chem. Eng. Sci. 1996, 51, Wammes, W. J. A.; Middelkamp, J.; Huisman, W. J.; de Baas, C.
2709. M.; Westerterp, K. R. Hydrodynamics in a Cocurrent Gas-
Trambouze, P. Multiphase Catalytic Reactors in the Oil Industry, Liquid Trickle Bed at Elevated Pressures. AIChE J. 1991b, 37,
an Introduction. Rev. Inst. Fr. Pet. 1991, 46, 433. 1849.
Trambouze, P. Engineering of Hydrotreating Processes. In Chemi-
cal Reactor Technology for Environmentally Safe Reactors and Weekman, V. W. Hydroprocessing Reaction Engineering. In
Products; De Lasa, H. I., Dogu, G., Ravella A., Eds.; NATO Proceedings of the 4th International/6th European Symposium
Advanced Study Institute Series E; Plenum: New York, 1993; on Chemical Reaction Engineering, VDI: Heidelberg, Germany,
p 425. 1976.
Tsamatsoulis, D.; Papayannakos, N. Axial Dispersion Model and Wild, G.; Larachi, F.; Laurent, A. The Hydrodynamic Character-
Hold-up in a Bench-Scale Trickle-Bed Reactor at Operating istics of Cocurrent Downflow and Cocurrent Upflow Gas-
Conditions. Chem. Eng. Sci. 1994, 49, 523. Liquid-Solid Catalytic Fixed Bed Reactors: The Effect of
Tsamatsoulis, D.; Papayannakos, N. Simulation of Non-Ideal Flow Pressure. Rev. Inst. Fr. Pet. 1991, 46, 467.
in a Trickle-Bed Hydrotreater by Crossflow Model. Chem. Eng. Wild, G.; Larachi, F.; Charpentier, J. C. Heat and Mass Transfer
Sci. 1995, 50, 3685. in Gas-Liquid-Solid Fixed Bed Reactors. In Heat and Mass
Turek, F.; Lange, R. Mass Transfer in Trickle-Bed Reactors at Low Transfer in Porous Media; Quintard M., Todorovic, M., Eds.;
Reynolds Number. Chem. Eng. Sci. 1981, 36, 569-579. Elsevier: Amsterdam, The Netherlands, 1992; p 616.
Van de Vusse, J. G.; Wesselingh, J. A. Multiphase Reactors. Wu, Y.; Khadilkar, M. R.; Al-Dahhan, M. H.; Dudukovic, M. P.
Proceedings of the 4th International/6th European Symposium Comparison of Upflow and Downflow Two-Phase Flow Packed-
on Chemical Reaction Engineering; VDI: Heidelberg, Germany, Bed Reactors with and without Fines: Experimental Observa-
1976. tions. Ind. Eng. Chem. Res. 1996a, 35, 397.
van Gelder, K. B.; Westerterp, K. R. Residence Time Distribution Wu, Y.; Al-Dahhan, M. A.; Khadilkar, M. R.; Dudukovic, M. P.
and Hold-up in a Co-current Upflow Packed Bed Reactor at Evaluation of Trickle Bed Reactor Models for a Liquid Limited
Elevated Pressure. Chem. Eng. Technol. 1990, 13, 27-40. Reaction. Chem. Eng. Sci. 1996b, 51, 2721.
Van Klinken, J.; Van Dongen, R. H. Catalyst Dilution for Improved
Zhukova, T.; Pisarenko, B. V. N.; Kafarov, V. V. Modeling and
Performance of Laboratory Trickle-Flow Reactors. Chem. Eng.
Design of Industrial Reactors with a Stationary Bed of Catalyst
Sci. 1980, 35, 59.
and Two-Phase Gas-Liquid FlowsA Review. Int. Chem. Eng.
Van Landeghem, H. Multiphase Reactors: Mass Transfer and
1990, 30, 57.
Modeling. Chem. Eng. Sci. 1980, 35, 1912-1949.
Vanrysselberghe, V.; Froment, G. F. Hydrodesulfurization of
Dibenzothiophene on CoMo/Al2O3 Catalyst: Reaction Network Received for review January 29, 1997
and Kinetics. Ind. Eng. Chem. Res. 1996, 35, 3311-3318. Revised manuscript received April 23, 1997
Vergel, H. C. A. Gas-Liquid Catalytic Fixed Bed Reactors. Accepted April 24, 1997X
Experimental and Theoretical Comparison of the Performances
of the Reactor for Different Flow Directions (Les Réacteurs IE9700829
Catalytiques à Lit Fixe avec EÄ coulement de Gaz et de Liquide.
Comparaison sur le Plan Théorique et Expérimental de la
Performance du Réacteur dans Différents Sens d’E Ä coulement).
X Abstract published in Advance ACS Abstracts, July 1,
Ph.D. Dissertation, Institut National Polytechnique de Lorraine,
Nancy, France, 1993. 1997.

Trickle Bed Reactor

Uploaded by

Copyright:

Available Formats

Trickle Bed Reactor

Uploaded by

Document Information

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

Trickle Bed Reactor

Uploaded by

Copyright:

Available Formats

What are trickle-bed reactors used for?

What are trickle-bed reactors used for?

What are some advantages and disadvantages of trickle-bed reactors?

What are some advantages and disadvantages of trickle-bed reactors?

3292 Ind. Eng. Chem. Res.

1997, 36, 3292-3314

High-Pressure Trickle-Bed Reactors: A Review

A concise review of relevant experimental observations and modeling of high-pressure trickle-

Introduction are their impracticality for reactions with rapidly

∆P/Z; βe; a; kLa; trickle-to-pulse flow &

X; Dz; Rt; kLa

UL: 3 × 10-4-2 × 10-3

of gaseous hydrogen. Temperature ≈ 410 K. Curve 1: density, kg/

m3 (L’Air Liquide, 1976; experimental data). Curve 2: heat

of H2 dissolved in tetralin/mole fraction of H2 in the vapor phase)

(Sebastian et al., 1981; experimental data). Curve 7: thermal

conductivity cal/cm/s/K × 105 (L’Air Liquide, 1976; experimental

gravitational force depends on liquid density and is not

column diameter/height (mm):

affected by pressure in the usual operating range of

packing diameter (mm)

0.5% Pd/Al2O3: 3 × 3.2

pressure drop can be split into an effect of the superficial

gas velocity and another due to gas density. Increased

packing size (mm)

gas density may result either from increased reactor

operating pressure or from use of gases of higher

leads to increased gas-liquid interaction and higher

superficial velocity on the pressure gradient, liquid

pressure drop, liquid holdup, and catalyst wetting

efficiency. The criteria (the values of the reactor pres-

below that represent these cases are based on the

experimental observations of Al-Dahhan and Dudukovic

(1994, 1995), Larachi et al. (1990, 1992), Wammes et

al. (1990a,b, 1991a,b), Lara-Marquez et al. (1992), and

Case 1: no gas flow, all pressure levels. This case

zero and the liquid flow is exclusively driven by its

weight. Under these circumstances, reactor pressure

within the range mentioned, i.e., up to 30 MPa, is not

liquid flow rate and liquid properties and by the bed

Grosser et al. (1998) trickle bed tested on data of phenomenological

[AµL(1 - )2L2 BFL(1 - )L3 2  - Ls

pressure drops at high pressure in porous media was

5. Two-Phase Pressure Drop

A ) 6.96; B ) 53.27; C ) 200; D ) 85; j ) -2; k ) -1.5; m ) -1.2; n ) -0.5 (6)

Larachi et al. (1994) flooded bed 0.1-5.1 empirical

Holub et al. (1992, 1993), trickle bed 0.31-5 phenomenological

log10(βnc) ) -0.44 + 0.40log10(χ) - 0.12(log10(χ))2 (11)

Saada (1975) flooded bed 1.36 empirical

Kohler and Richarz (1985) trickle bed 0.1-1.0 empirical

Ellman et al. (1990) trickle bed 0.1-8.0 empirical

log βnc ) -RXGmReLnWeLp ( )

Ring and Missen (1991) trickle bed 10 empirical

βa ) 15.6UL0.679 (16)

Wammes et al. (1991a) trickle bed 0.3-6.0 empirical

gas-liquid flow (Specchia and Baldi (1977) correlation)

Larachi et al. (1991a) trickle bed 0.2-8.1 empirical

Larachi et al. (1991b) trickle/flooded bed 0.2-8.1 empirical

where Jdf ) AUGb (a, b) depend on the gas-liquid-solid system

Holub et al. (1992, 1993), trickle bed 0.31-5 phenomenological

[AµL(1 - )2L2 BFL(1 - )L3 2 - Ls

βa ) 15.6UL0.679 (16)

βe ) aULb where (a, b) depend on boundary conditions (22)