Fuel Processing Technology 73 Ž2001.

155–173
www.elsevier.comrlocaterfuproc

Review

Review of literature on catalysts for

biomass gasification
David Sutton ) , Brian Kelleher, Julian R.H. Ross
Centre for EnÕironmental Research, UniÕersity of Limerick, Limerick, Ireland
Received 1 February 2001; received in revised form 22 May 2001; accepted 25 May 2001

Abstract

Biomass gasification is a possible alternative to the direct use of fossil fuel energy. Biomass, a
CO 2 neutral source of renewable fuel, can contribute to the demand for heat, electricity and
synthesis gas. However, there are inefficiencies in the technology, which at present render
biomass gasification economically unviable. The presence of condensable organic compounds and
methane in the product gas renders the gas unsuitable for specific applications. Elimination of the
condensable organic compounds and methane by a suitably cheap technology will enhance the
economic viability of biomass gasification. This paper contains an extensive literature review of
the three main groups of catalysts, which have been evaluated for the elimination of these
hydrocarbons. These three groups of catalysts are dolomite, alkali metals and nickel. q 2001
Elsevier Science B.V. All rights reserved.

Keywords: Biomass gasification; Tar; Syngas; Catalysts; Reforming; Dolomite; Alkali metals and nickel

1. Introduction

Remarkable progress has been achieved in recent years in the design of gasifiers.
However, gas cleaning is still the bottleneck in advanced gas utilisation that limits the
deployment of the use of biomass for electricity generation w1x. The continual build-up
of condensible organic compounds Žoften referred to as tars. present in the produce gas
can cause blockages and corrosion and also reduce overall efficiency. In addition, the
presence of impurities Žsuch as methane. can affect the end usage of the syngas and the
techniques involved in the removal of the impurities in such processes are costly.

)
Corresponding author. Tel.: q353-61-213515; fax: q353-61-202602.
E-mail address: david.sutton@ul.ie ŽD. Sutton..

0378-3820r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 3 8 2 0 Ž 0 1 . 0 0 2 0 8 - 9
156 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

Nitrogen and sulphur are present in many of the by-products and the corresponding
oxides are produced during combustion of the fuel gas; these oxides ŽNO x and SO x . can
have a negative environmental impact.
Since the mid-1980s, interest has grown on the subject of catalysis for biomass
gasification. The advances in this area have been driven by the need to produce a
tar-free product gas from the gasification of biomass, since the removal of tars and the
reduction of the methane content increases the economic viability of the biomass
gasification process. The literature in this area ranges from papers on bench-scale
reactors to those on the use of plant-scale gasifiers. Research on catalysts for use in the
process is often carried out specifically in relation to gasifier design or biomass feed
type. However, the criteria for the catalyst are fundamentally the same and may be
summarised as follows:

1. The catalysts must be effective in the removal of tars.

2. If the desired product is syngas, the catalysts must be capable of reforming
methane.
3. The catalysts should provide a suitable syngas ratio for the intended process.
4. The catalysts should be resistant to deactivation as a result of carbon fouling and
sintering.
5. The catalysts should be easily regenerated.
6. The catalysts should be strong.
7. The catalysts should be inexpensive.

Catalytic decomposition of the unwanted hydrocarbons is also known as hot gas

cleaning. The catalysts employed in this process are responsible both for purification and
bringing about compositional adjustment of the product gas. Hot gas conditioning is
achieved by passing the raw gasifier product gas over a solid catalyst in a fluidised-bed
Žor a fixed-bed. under temperature and pressure conditions that essentially match those
of the gasifier. As the raw gas passes over the catalyst, the hydrocarbons may be
reformed on a catalyst surface with either steam ŽEq. Ž1.. or carbon dioxide or both ŽEq.
Ž2.. to produce additional carbon monoxide and hydrogen:
m
C n H m q nH 2 O m nCO q n q ž / 2
H2 Ž 1.
m
C n H m q nCO 2 m 2CO q ž /
2
H2 Ž 2.

Additional steam or carbon dioxide may be added for the reforming processes w2x.
The use of a catalyst to reform condensible organic compounds and methane can
increase the overall efficiency of the biomass conversion process by 10%. Lindman w3x
investigating air gasification reported higher efficiency being achieved by lower oxygen
consumption, better heat recovery and higher carbon conversion compared to a process
based on non-catalytic techniques. Thermal cracking of the hydrocarbons is also
possible; however, this method is not considered a feasible option as it requires high
temperatures Ž) 1100 8C. to achieve high cleaning efficiency and it also produces soot.
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 157

Table 1
Main chemical reactions of biomass gasification
Reaction no. Reaction DH 298 , kJ moly1
1.3.1 Volatile matter sCH 4 qC Mildly exothermic
1.3.2 Cq0.5O 2 sCO y111
1.3.3 COq0.5O 2 sCO 2 y254
1.3.4 H 2 q0.5O 2 s H 2 O y242
1.3.5 CqH 2 OsCOqH 2 q131
1.3.6 CqCO 2 s 2CO q172
1.3.7 Cq2H 2 sCH 4 y75
1.3.8 COq3H 2 sCH 4 qH 2 O y206
1.3.9 COqH 2 OsCO 2 qH 2 y41
1.3.10 CO 2 q4H 2 sCH 4 q2H 2 O y165

Reforming does not reduce the total chemical energy content of the fuel gas as heating
of the gas to the higher temperatures for thermal cracking is not required.
Catalysts for use in biomass conversion may be divided into two distinct groups
which depend on the position of the catalytic reactor relative to that of the gasifier in the
gasification process. The first group of catalyst Žprimary catalysts. is added directly to
the biomass prior to gasification, which catalyse the reactions in Table 1. The addition is
either by wet impregnation of the biomass material or by dry mixing of the catalyst with
it. These catalysts primarily have the purpose of reducing the tar content and have little
effect on the conversion of methane and C 2 – 3 hydrocarbons in the product gas. They
operate under the same conditions of the gasifier and are usually non-renewable and
consist of cheap disposable material.
The second group of catalysts is placed in a secondary reactor downstream from the
gasifier. Independent of the type of gasifier, they can be operated under different
conditions than those of the gasification unit. The catalysts are active in reforming
hydrocarbons and methane according to Eqs. Ž1. and Ž2..

2. Catalysis for biomass gasification

Three distinct groups of catalysts materials have been the subject of publications for
biomass gasification and are described in the following sections.

2.1. Dolomite catalysts

Dolomite, a magnesium ore with the general formula MgCO 3 P CaCO 3 is used in the
Pidgeon process for the manufacture of magnesium by thermal reduction. The use of
dolomite as a catalyst in biomass gasification has attracted much attention w4–19x since
it is a cheap disposable catalyst that can significantly reduce the tar content of the
product gas from a gasifier. It may be used as a primary catalyst, dry-mixed with the
biomass or, more commonly, in a downstream reactor, in which case it is often referred
to as a guard bed. The chemical composition of dolomite varies from source to source
158 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

but it generally contains 30 wt.% CaO, 21 wt.% MgO and 45 wt.% CO 2 ; it also contains
the trace minerals SiO 2 , Fe 2 O 3 and Al 2 O 3 . The surface areas of the various types also
differ, as do the pore sizes and distributions.
´ et al. w4x investigated four different dolomites Žfrom Norte, Chilches, Malaga
Orıo
and Sevilla. for oxygenrsteam gasification of wood in a downstream catalytic reactor.
The main chemical difference between the various samples was the Fe 2 O 3 content: the
Malaga and Sevilla dolomites had low levels of Fe 2 O 3 compared to those from Norte
and Chilches. These samples were tested as catalysts at varying steam carbon ratios and
temperatures ranging from 805 to 875 8C. Tar conversion was of the order of 95% for
the Norte dolomite and the lowest conversion of 77% was found for the Sevilla
dolomite. The gas yields were increased by the catalyst for all of the dolomites. The
order of activity was: Norte ) Chilches) Malaga) Sevilla. Interestingly, the surface
areas of the Norte and Chilches dolomites were lower than those of the Malaga and
Sevilla materials. The higher activity of the Norte and Chilches dolomites may be
accounted for by their higher Fe 2 O 3 contents and also by their larger pore diameters.
The increase in gas yield was 10–20 vol.%, resulting in an increase of 15% in the
Alower heating valueB ŽLHV. of the gas. The hydrogen content of the gas increased by 4
vol.%, while the content of CO, CO 2 and CH 4 was relatively unchanged.
Aznar et al. w7,17x also investigated the use of Malaga dolomite for steamroxygen
gasification. They reported that the H 2 content of the flue gas increased by 7 vol.%,
while the CO content decreased by 7 vol.%. This effect was due to a greater contribution
of the water–gas shift reaction as a result of a high steam content and high temperature
w7x. In a later paper w17x under different gasification conditions in which the CO 2 content
varied slightly, from y2 to q6 vol.%, an overall decrease in the CO 2 was reported,
indicating that dry reforming also occurred. The methane and steam contents also
decreased by 0.8–2.0 and 3–8 vol.%, respectively, as a result of the water–gas shift
reaction w7x. The LHV of the gas increased by 10% to 21% while the gas yield increased
by 0.15–0.4 m3 w17x. A tar conversion of 96% was obtained at a temperature of 840 8C
w32,42x. Ekstrom¨ et al. w19x also achieved almost 100% conversion of tar at 700–800 8C
using Malaga dolomite under steam reforming conditions. However, they also observed
a marked increase in CH 4 and C 2 H 4 at lower temperatures and showed that calcined
dolomite was 10 times more active than the uncalcined material w19x, in agreement with
the results of several authors w4–7,17x.
Delgado et al. w5x also investigated the use of Norte dolomite and compared it with
calcite ŽCaO. and magnesite ŽMgO. for the steam reforming of biomass tars. They
investigated the effects of temperature, contact time and the particle diameter of the
catalysts and reported that the tar conversion increased on increasing the temperature of
the catalyst bed, complete elimination being observed at 840 8C. The increase in
temperature also resulted in an increase in the gas yield. Delgado et al. w5x reported that
the gas composition was unchanged, in contrast to the results of other groups w7,17,19x.
An increase in the gasrcatalyst contact time gave an increase in the destruction of tars
present in the gas, a maximum being reached at 0.3 kg hrm3 n. The increased contact
time produced more H 2 and CO 2 as a result of the occurrence of tar conversion
reactions and the water–gas shift reaction. An increase in particle size had the same
effect as that of increasing the contact time. Delgado et al. w5x also investigated the effect
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 159

of calcination on all three catalysts. He reported that the order of activity was
dolomite) magnesite) calcite.
Calcined Glanshammar and Sala dolomites were investigated by Vassilatos et al. w6x,
who studied the effects of temperature, catalyst contact time and steamrcarbon ratio.
With the calcined dolomite increasing the temperature, it gave increased gas yields as
with the results published by Orio et al. w4x and by Delgado et al. w5x. The effect on the
gas composition was also similar to that described by Orio et al. w4x, this being an
increase in the H 2 and CO 2 content. The tar content decreased with increased catalyst
loading, for both the Glanshammar and Sala dolomite w6x.
Myren´ et al. w10x and Vassilatos et al. w11x also investigated the use of Sala dolomite
and reported good results for the reforming of tars. The former group reported that the
resulting high conversion of the tars with Sala dolomite leaves almost nothing other than
naphthalene in the condensed tar phase. Vassilatos et al. w6x found that the naphthalene
production seems to be connected with the steam content and temperature. Several
authors w6,10–12x have identified naphthalene as the most abundant condensible product
after reforming the tars over dolomite at 800 to 900 8C. This observation highlights the
limitations of the use of dolomites as catalysts for the complete elimination of tars from
the product gas.
´ et al. w12x and Lammers et al. w18x investigated the catalytic reforming of
Alden
naphthalene over dolomite. The former group w12x reported that the degree of conversion
of naphthalene when passed over calcined dolomite at 800 8C varied with the composi-
tion of the carrier gas. A conversion of 96% was achieved using a carrier gas
composition of 15% CO 2 Žbalance N2 ., while only 79% conversion was achieved with
18% H 2 O in the gas. The resulting product gas compositions are shown in Table 2.
The degree of conversion of naphthalene also varied with the amount of water vapour
in the carrier gas, reaching a maximum for a water concentration of 5–15 vol.%.
Lammers and Beenackers w18x investigated steam gasification and the effect of air on the
reforming of tar over calcined dolomite. The degree of tar conversion at 850 8C over the
calcined dolomite with steam alone was 72% while 96% conversion was achieved in an
airrsteam mix at 850 8C. The gas composition exiting the catalytic reactor was only
slightly affected. From micro-reactor experiments, it was found that the addition of air in
the catalytic reactor introduced an extra parallel naphthalene decomposition reaction.
´ et al. w15x reported an investigation of Glanshammar
In a separate publication, Alden
dolomite for the dry reforming of tars from a biomass gasifier. With calcined dolomite at
800 8C they found a 70% reduction of the tar content and a further 10–15% reduction at
900 8C. Tar yields were further decreased when the pressure was increased to 10 bar.
However, increasing the pressure also had the effect of raising the partial pressure of

Table 2
The product composition Žvol. %. when converting naphthalene over calcined dolomite at 800 8C
Carrier gas comp. balance N2 H2 CO CO 2 CH 4 Dolomite
14% CO 2 1.6 2.8 16 0.67 no
14% CO 2 2.9 15 4.6 0.13 yes
18% H 2 O 11 4.3 1.9 0.22 yes
160 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

CO 2 , leading to carbonisation of the catalyst. At higher pressure, the function of the

catalyst seemed to be more influenced by the partial pressure of the gaseous component
than by the temperature. The increased pressure has the effect of increasing the
residence time in the catalyst bed and hence of achieving a higher tar conversion w15x.
Simell et al. w13x studied a Finnish dolomite catalyst with a relatively high iron
content Žca. 1.5 wt.%.. The surface area of the calcined dolomite was 8.1 m2 gy1. The
gas atmospheres studied were mixtures of nitrogen as carrier, with toluene, and H 2 ,
H 2 O, CO or CO 2 . They showed that the reaction of toluene over dolomite at 900 8C was
faster with carbon dioxide than with steam and that dry reforming was inhibited by the
presence of steam. Tar decomposition was strongly inhibited by the presence of CO. In a
recent publication, Simell et al. w16x investigated the mechanism of the catalytic reaction
of benzene Žas a model compound. over the same Finnish dolomite. Reforming of the
benzene was carried out with carbon dioxide. They developed mechanistic models of the
Langmuir–Hinshelwood type for the reforming of benzene and the adsorption of
benzene on a single site was identified as the rate determining step, the CO 2 adsorbing
non-dissociatively.
Taralas w14x investigated the steam reforming of cyclohexane over calcined dolomites,
testing limestone ŽCaCO 3 ., dolomitic magnesium oxide wMgOx, Sala, Glanshammar and
Larsbo dolomites. Using a first order reaction model, he found that the apparent
activation energies were 177.7 kJ moly1 for MgO, 202.8 kJ moly1 for CaO, 153.8 kJ
moly1 for Glanshammar dolomite and 168.8 kJ moly1 for Larsbo dolomite. The
catalysts should provide a lower activation energy in the transformation of reactants to
products. The lower the activation energy, the more efficient the catalysts. Taralas w14x
also reported that the selected dolomites increased the conversion of the cyclohexane at
700 8C without affecting the product gas composition.
One problem reported by Vassilatos et al. w6x and by other groups w8,9,11x, was the
successive decrease of the mechanical strength of the dolomite with time during all
catalytic runs. This decrease was more evident in fluidised-beds. However, once the
activity was significantly reduced, the catalysts could be replaced without great expense
to the overall process. Deactivation due to carbon deposition was also reported by
several groups w6,8–13,19x. The relatively high amounts of steam used in gasification
were effective in maintaining the activities of the catalysts due to the steam reforming of
any carbon deposited ŽEq. Ž1... Alden ´ et al. w12x also showed that the rate of carbon
deposition was increased if hydrogen was removed from the gas stream. They proposed
that dolomite may catalyse the soot formation and gasification reactions. Taralas w14x
reported that the alkaline earth oxides, produced by the calcination of limestone
ŽCaCO 3 . and dolomites, reduce carbon deposition on the catalyst. Lammers and
Beenackers w18x showed that when dolomite was used, the addition of secondary air to
the catalytic reactor has the important effect of significantly reducing the deactivation
rate of the catalyst.

2.2. Alkali metal and other metal catalysis

Alkali metal catalysts for the elimination of tar and up-grading of the product gas
have also been investigated by several groups w20–28x. The catalysts are often added
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 161

directly to the biomass by dry mixing or wet impregnation. When added in this way, the
catalyst is difficult to recover and this is not always cost effective for the gasification
process. It also gives an increase in the ash content remaining after char gasification, and
the disposal of this is predicted to become a problem for the technology over the coming
years.
Mudge et al. w20x studied the catalytic steam gasification of wood using alkali
carbonates and naturally occurring minerals, which were either impregnated or mixed
with the biomass. They studied the effectiveness of four different primary catalysts and
of different catalyst concentrations at 550 8C, 650 8C and 750 8C. The order of activity
was reported as being potassium carbonate) sodium carbonate) trona ŽNa 3 HŽCO 3 . 2 P
2H 2 O. ) borax ŽNa 2 B 4 O 7 P 10H 2 O.. Impregnated catalysts had little or no carbon
deposition as compared to the mixed catalysts Žcarbon deposition resulted in deactiva-
tion.. Also Mudge et al. reported that the impregnation decreased particle agglomeration.
Wood which had been impregnated with 17 wt.% K 2 CO 3 and was gasified at 750 8C
at atmospheric pressure using a steamrwood ratio of one yielded methanol-quality
synthesis gas w20x. No tar was detected in the exit gas and the synthesis gas yield was
1.23 nm3 kg feedy1 as long as the catalyst remained active. Table 3 shows the average
results of four runs of catalysed and uncatalysed gasification. The potassium carbonate
presented some operational problems as a primary catalyst because of particle agglomer-
ation. These difficulties were overcome by maintaining vigorous fluidisation. Intimate
mixing of the wood and the catalyst was found to contribute to excessive carbon
deposition, which appears to be that main cause of deactivation. The appearance of tars
in the product gas was one of the first signs of loss of catalyst activity.
K 2 CO 3 and Na 2 CO 3 supported on alumina have also been investigated as secondary
catalysts w20x. The gas yields were not as high as when the biomass was directly
impregnated with the catalyst and there was a higher methane content. However, it was
noted that the activity of the alkali carbonate catalysts initially increased and reached a
steady state and that there was no deactivation over the 30-h test carried out. Little or no
carbon was found on the used catalyst.
Hauserman w21x investigated primary alkali catalysts for the production of hydrogen
by the steam gasification of wood and coals. The catalyst was directly added to the feed
by dry mixing. The effect on the rate of gasification was investigated by thermogravi-
metric analysis ŽTGA.. The most active catalysts were further tested in a pilot-plant-scale
fluidised-bed gasifier. Wood ash was investigated as a possible catalyst for the gasifica-

Table 3
Gas composition exiting gasifier for primary catalysed and uncatalysed wood gasification at 750 8C
Gas composition, vol.% Uncatalysed 17 wt.% K 2 CO 3
H2 42.7 52.4
CO 24.6 21.8
CH 4 8.7 3.2
CO 2 21.6 21.4
C 2q 2.4 1.2
162 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

tion reaction, as were some alkali carbonates. Analysis of the wood ash after gasification
showed that there was a high content of alkali metals, the relative amounts being
reported as: CaO, 44.3 wt.%; MgO, 15 wt.%; and K 2 O, 14.5 wt.%. Table 4 shows the
reactivities Žas determined by TGA. for catalysts added to wood and a variety of coal
types.
From the reactivity data above, it can be seen that the addition of 20 wt.% wood ash
increased the reactivity of the bituminous coal at 700 8C by a factor of almost 9 and that
of wood by a factor of 32 relative to uncatalysed results. This is a very significant result
since wood ash is a waste product of the gasification process. At 750 8C, a 20 wt.%
loading of CaCO 3 increased the reactivity of the bituminous coal by a factor of 4.2. The
10 wt.% loading of K 2 CO 3 increased the reactivity of the bituminous and subbitumi-
nous coal Žat 750 8C. by factors of 31.1 and 3.7, respectively. The large difference in the
reactivity of the two coals is a function of the ash content: bituminous coal has a higher
ash content and also a greater wt.% content of potassium and iron. This autocatalytic
effect, due to the minerals present in the ash of the biomass, was also noted by DeGroot
and Shafizaceh w26x.
The wood ash catalyst was tested in a pilot plant where its performance was assessed
by analysis of the exiting gas stream. Wood gasification with a 30 wt.% wood-ash
loading and using as gasification conditions, a temperature of 650 8C, a pressure of
2.43 = 10 5 Pa, a steam carbon ratio of one yielded a gas composition of 52 vol.% H 2 , 6
vol.% CO, 34 vol.% CO 2 , 6 vol.% CH 4 and 2 vol.% C 2q. This gas composition is
suitable for fuel cell applications once the carbon monoxide is removed. On increasing
the pressure to 7.09 = 10 5 Pa, the exiting gas composition altered to H 2 61.5 vol.% and
CO 38.5 vol.%, an ideal gas composition for syngas applications. However, neither the
tar content of gas nor the gas yield per kilogram of biomass was reported for the pilot
plant experiments.
Hallen et al. w22x also investigated the influence of alkali carbonates on biomass
steam gasification. The catalysts ŽNa 2 CO 3 , K 2 CO 3 or CsCO 3 . were dry-mixed with, or

Table 4
Relative reactivity data of gasification for various catalysts directly added to biomass as determined by TGA
Temperature 650 8C 700 8C 750 8C 800 8C
Bituminous coal
Raw – 0.07 0.14 0.33
q20% CaCO 3 – – 0.59 1.45
q10% K 2 CO 3 – – 4.36 –
20% Wood Ash 0.61 3.01 5.15

Subbituminous coal
Raw 0.37 1.16 1.31 3.06
q10% K 2 CO 3 1.25 4.3 8.24 13.48

Wood (hybrid poplar)

Raw – 0.13 0.33 3.10
q10% Wood ash – 4.18 9.84 18.24
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 163

wet-impregnated onto the biomass. In this study two catalyst concentrations were
evaluated in relation to gas production and the results demonstrated only a slight
variation. The presence of catalyst at temperatures of 650 8C and higher decreased the
carbon conversion to gas during the volatilisation of the steam gasification of wood.
There was, however, an increase in the rate and total amount of gas produced during the
gasification stage. The presence of catalyst increased the char yield during the volatilisa-
tion stage but then decreased the char yield during the second stage of the gasification
process. Several groups reported w20–24x that the increase in carbon conversion to gases
using the alkali carbonate results from conversion of the condensible liquids.
The gas phase reactions of the gasification products as well as the model compounds
were investigated by Hallen et al. w22x using selected alkali catalysts. Simple hydrocar-
bon products from gasification did not react over the catalyst in the presence of steam at
600 8C. It was found that at 600 8C the water–gas shift reaction was catalysed by the
catalysts. The methane reforming reaction occurred only at temperatures greater than
800 8C over the alkali catalysts. The alkali carbonates were found to increase the
decomposition of alcohols and carboxylic acid, but had little effect on aldehydes and
ketones Žat 600 8C.. It was noted by Hallen et al. w22x that the alkali carbonates catalyse
the dehydrogenation of alcohols. The order of activities for the catalysts tested was
CsCO 3 ) K 2 CO 3 ) Na 2 CO 3 .
Encinar et al. w23x studied the gasification of grape and olive bagasse under different
experimental conditions. The apparent effects of various additives and of the concentra-
tions were reported. The additives investigated for grape bagasse gasification were LiCl,
NaCl, KCl, AlCl 3 P 6H 2 O and ZnCl 2 at 5 wt.% of cation for gasification temperature of
600 8C in a flow of carbon dioxide. Furthermore, the effect on the solid liquid and gas
fraction were also reported. The most significant effect observed by these authors for all
additives except KCl was the increase of char and gas fraction and the decrease of
liquids. These results are at variance with the findings of previous authors w20–22x,
indicating that potassium in the carbonate form is more active for the increase in the gas
yield at the expense of liquids. Only zinc was reported to increase the specific surface
area and pore volume of the remaining char. The production of methanol and acetone
was enhanced by the potassium chloride whilst the aluminium chloride reduced these
fractions.
Table 5 shows the resultant gas composition for each of the additives investigated by
Encinar et al. w23x. Hydrogen, methane and carbon dioxide were the major gases formed

Table 5
Influence of additives on gas composition
Composition, mol. % None LiCla NaCl a KCl a AlCla3 ZnCl a
H2 14.5 27.7 16.2 15.7 11.5 66.5
CH 4 29.2 36.5 31.8 32.58 33.33 11.1
CO 50.8 27.86 45.8 45.64 49.8 20.4
C2 H6 3.3 5.84 3.9 4.49 4.2 1.3
C2 H4 2.2 2.1 2.21 1.6 1.8 0.7
a
wcatx s 5% by weight cation.
164 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

during gasification. For Na, Al and K catalysts, carbon monoxide made up the largest
fraction in the gas composition. Lithium showed an increase in the hydrogen fraction by
a factor of two over the uncatalysed reaction. The most significant effect reported by
Encinar et al. w23x, however, is that there was hydrogen production when zinc was
present. Compared to the other additives, zinc produced five to eight times more H 2 .
The high fraction of hydrogen can be related to the small rate of production of methane
from reactions of hydrogen and methyl radicals. Sala et al. w24x have also reported that
zinc inhibited methane formation and that increasing the concentration of zinc led to
increasing yields of solid fraction Žchar. and a decrease of liquids and gases other than
for hydrogen w24x.
Li et al. w25x and DeGroot and Shafizaceh w26x studied the kinetics of the catalytic
gasification of biomass. Li et al. w25x investigated the steam gasification of peat
catalysed by K 2 CO 3 Na 2 CO 3 and CaCO 3 , while DeGroot and Shafizaceh w26x investi-
gated the CO 2 and steam gasification of wood. Both authors reported that high
temperature and long gasification time favoured high conversions. The LHV value of the
gas was reduced as a consequence of a reduction in the hydrocarbon fraction. All
catalysts tested increased the gas yield but had no significant effect on the gas
composition w25,26x.
Li et al. w25x also showed that the kinetic behaviour of peat conversion was a first
order process, in agreement with results obtained by Anthony w27x who had investigated
coal gasification. Contrary to this, DeGroot and Shafizaceh w26x reported that the
catalytic gasification of wood by either steam or CO 2 is not a simple first order process
but rather goes through an induction period early in the reaction. After induction, the
reaction rate decayed in a manner intermediate between zero and first order. In the study
of DeGroot and Shafizaceh w26x, the rate of gasification by 0.3 = 10 5 Pa of CO 2 was
similar to that obtained with 0.077 = 10 5 Pa of steam, thus indicating that the rate of
steam gasification to be much greater. This was contrary to results reported by Simell et
al. w16x who investigated steam and dry reforming of biomass tar over nickel and
dolomite catalysts. Further, Li et al. w25x indicated that the differences between catalytic
and non-catalytic reactions become evident only when the reaction time is longer than
120 s. This implies that the catalysts make little contribution to the reactions with lower
activation energies. Analysis of the data of Li et al. w25x yields information suggesting
that the activation energy for catalytic gasification of peat can only be a little smaller
than the activation energy of the non-catalytic process: 125.5 kJ moly1 for K 2 CO 3
versus 127.6 kJ moly1 . DeGroot and Shafizaceh w26x reported that the activation energy
for uncatalysed steam gasification of wood was 258.1 kJ moly1 , the activation energy
for K 2 CO 3 catalysed wood gasification was 178.6 kJ moly1 w26x. The difference in the
results may be due to the biomass type and their inherent mineral content, the latter
being higher for wood biomass.
Gebhard et. al. w28x investigated a catalyst specifically designed for tar destruction.
The catalyst referred to as DN-34, was described as a new inexpensive non-nickel-based
tar-destruction catalyst. It was tested in a micro-reactor-scale experiment with a syn-
thetic feed in a slip stream fluidised-bed reactor attached to a 9 tonne dayy1 gasifier and
was reported to exhibit steam reforming activity for aromatic and polynuclear aromatic
hydrocarbons at temperatures between 650 and 815 8C. However, the catalyst was
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 165

reported to be somewhat less effective for benzene and toluene destruction: the tar
destruction varied from 45% to 90%, depending on the gasification conditions. At 815
8C and with a GHSV of 1500 and 2500 cm3 cmy3 hy1 and with a 40 vol.% steam
content, a conversion of 96% was achieved.
The gasifier experiments produced superior results to those of the micro-reactor and
this may be accounted for by the longer residence time in the catalyst bed. The DN-34
catalyst also exhibited water–gas shift activity: the H 2 :CO ratio varied between 0.6 and
0.7 in the raw syngas from the gasifier but after the steam reforming over DN 34, it
varied between 1.7 and 13. The ratio obtained could be controlled by the percentage of
steam in the raw syngas from the gasifier as well as by variation of the other process
conditions. The catalyst showed relatively poor activity for methane steam reforming,
giving less than 20% conversion of the methane in the raw syngas.

2.3. Nickel catalysts

The most significant body of literature published on the area of hot gas cleaning for
biomass gasification concerns nickel catalysts w2,28–43x. Several groups w2,17–20,28x
have investigated a system of raw gas cleaning that involves a dolomite or alkali catalyst
for the removal of tar Žup to 95%. followed by the adjustment of the gas composition
Žreforming of the methane and the remaining tar. using a nickel steam reforming
catalyst. Steam and dry reforming reactions are catalysed by group VIII metals, and of
these, nickel is the most widely used in the industry w29x. Consequently, the majority of
published work concerns commercially available nickel catalysts designed for steam
reforming of hydrocarbons and also of methane. Using these catalysts at temperatures
greater than 740 8C, there is generally an increase in the hydrogen and carbon monoxide
content of the exiting gas, with elimination or reduction of the hydrocarbon and methane
content. At lower temperatures, the methanation reaction is favoured thermodynamically
and is sometimes optimised in cases when methane is desired as a predominant
component of the exiting gas w20,32x.
Aznar et al. w2,17,30x investigated several commercially available catalysts for the
removal of tars and the adjustment of the product gas composition. Two Haldor Topsøe
catalysts ŽR-67-7H and RKS-1. were investigated for the steam reforming of the tar and
methane w2x. The catalyst was positioned downstream of the gasifier in a secondary
reactor that was maintained at temperatures between 730 and 760 8C; space times of
only 0.1 s were used. Catalyst R-67-7H is described as 12–14% Ni on a MgrAl 2 O 3
support with a free Mg content of less than 0.5 wt.% and a SiO 2 content of - 0.2 wt.%.
Its specific surface area was 12–20 m2 gy1 . The conversions of tar, methane and C 2
and C 3 were greater for the reduced catalyst. The methane was reduced to 0.5 vol.% and
the tar content to only 4 mg my3 . Deactivation of the catalysts occurred due to carbon
fouling and this resulted in short catalyst lifetimes. Reduction in the tar content prior to
the nickel catalysts was identified as a possible way to maintain the activity of the
catalysts. Several authors have also investigated a design using a guard-bed of dolomite
catalyst prior to the nickel catalyst w15,17,30,34x.
Aznar et al. w30x also reported results for several commercial catalysts produced by
BASF, ICI-Katalco, UCI and Haldor Topsøe. Eight catalysts were tested, of which four
166 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

were steam reforming catalysts for light hydrocarbons Žparticularly methane. and the
remaining four were for reforming of heavier hydrocarbons Žnaphthas.. The catalysts
were not individually identified but were referred to as catalysts A, B, . . . , E. The results
demonstrated that commercial steam reforming catalysts designed for the steam reform-
ing of heavy hydrocarbons ŽSRHH. were more active for tar removal than the commer-
cial steam reforming catalysts for light hydrocarbons ŽSRLH.. The individual activities
for the removal of the tars by the SRHH catalysts were not distinguishable. Complete
removal of the tar content was effected by the catalysts bed conditions, depending on
temperature, space time, particle size and the composition of the gas atmosphere. At
relatively high temperatures Ž780–830 8C., no catalyst deactivation was experienced for
45 h on-stream. Interestingly, the SRHH catalysts also showed high activity for the
removal of methane.
Baker et al. w20,31–33x have also investigated several commercial nickel catalysts
and have compared them with specially prepared materials. They reported that a
Harshaw 3266 supported nickel catalyst altered the gas composition to give a methane-
rich composition at low temperatures Ž550–560 8C., but the composition was closer to a
syngas at high temperatures Ž740–760 8C. w20,32x. The conditions used and the resultant
gas compositions are shown in Table 6. The conversion of carbon to gas was not as
efficient at the lower gasification temperatures as at the higher temperatures. No tar was
formed as long as the catalyst was active; deactivation of the catalysts occurred as a
result of carbon fouling and also of thermal sintering Žgrowth of the nickel particles..
Baker et al. w32x also tested G-90C ŽUnited catalyst., C-13-3 ŽGirdler. and a
trimetallic nickel alloy catalyst ŽNi–Co–Mo on silica alumina. developed in their own
organisation ŽPacific Northwest Laboratories ŽPNL... All of these catalysts demonstrated
high activity for the removal of hydrocarbons and methane at temperatures above 740
8C. Deactivation was most severe for the Harshaw catalyst, which could not be
regenerated due to sintering. The PNL catalyst had the longest lifetime and could be
regenerated with only a slight loss of activity due to sintering.
Baker et al. w31x published results which compared another PNL catalyst Ž9.5 wt.%
Ni, 4.25 wt.% CuO, 9.25 wt.% MoO 3 on a SiO 2 –Al 2 O 3 support. with a United Catalyst
Ž 15 wt.% Ni on a CaOrAl 2 O 3 support. and with an ICI catalyst Ž16.5 wt.% Ni on a

Table 6
Production of specific gases from biomass steam gasification
Harshaw 326 Methane rich Syngas
Gasification T, 8C 560–540 740–770
Catalyst T, 8C 550–560 740–760
Steamrwood, grg 0.33 0.7
% Carbon to gas 68 90
H2 29.5 53.1
CO 2 34.3 15.5
CO 10.8 28.3
CH 4 25.5 3.1
C 2q 0.1 0.1
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 167

support containing 14 wt.% SiO 2 , 29 wt.% Al 2 O 3 , 13 wt.% MgO, 13 wt.% CaO and 7
wt.% K 2 O.. Each type of test was carried out in a micro-reactor ŽMR. and a
bench-scale-reactor ŽBSR.. Three different configurations were investigated for each
catalyst Ž1. in a fluid-bed gasifier, Ž2. in a secondary fixed-bed reactor, and Ž3. in a
secondary fluid-bed reactor. The activities of the catalysts in all cases were initially
high, the exit gases consisting primarily of H 2 , CO, CO 2 and trace of CH 4 .
When the catalyst was added directly to the biomass, deactivation was rapid due to
carbon fouling in both test situations ŽMR and BSR.. Micro-reactor tests with the
catalysts in the secondary fixed-bed reactor showed an initial loss of activity but the
activity stabilised after 15–20 h with no further deactivation in a 50-h test period. The
carbon content on the catalyst was less than 1 wt.% after this time period. After about
1200 h of exposure in the micro reactor, the PNL catalyst contained only 4 wt.% carbon.
For the BSR tests carried out, all catalysts continued to deactivate. The United
Catalyst contained 6 wt.% carbon after 16 h of exposure. When each of the three
catalysts were tested in the secondary fluidised-bed, there was some activity lost during
the first 8–10 h, but there was no further loss in activity during an additional 34 h of
exposure. The gas compositions produced by all three catalysts were apparently similar:
ca. 55 vol.% H 2 , 14 vol.% CO 2 , 1 vol.% C 2q, 6 vol.% CH 4 and 24 vol.% CO.
Baker et al. w33x and Roy et al. w35x also investigated a range of steam reforming
methanation and hydrogenation catalysts for the Thermochemical Environmental Energy
System ŽTEES w .. The TEES w process is described as a low temperature catalytic
gasification system Ž360 8C and 200 atm. designed for a wide variety of feedstocks
ranging from dilute organics in water to waste sludges from food processing. The
catalysts were not specifically identified and were indicated as catalysts A-E; see Table
7.
The conversions of various waterrorganic mixtures were determined; these included
p-Cresol Ž2% in water., cheese whey and lactose Ž5% in water.. The steam reforming
catalyst was found to be unsuccessful at reducing the chemical oxygen demand of the
water after the process and this corresponded to low gas yields. The methanation catalyst
reduced the Chemical Oxygen Demand ŽCOD. by 99.1% and resulted in a gas
composition of 53 vol.% CH 4 , 36 vol.% CO 2 and 11 vol.% H 2 . The hydrogenation
catalyst C also gave 99% reduction in the COD of the effluent and a slightly different
gas composition of 47 vol.% CH 4 , 49 vol.% CO 2 and 4 vol.% H 2 . The crystalline

Table 7
Nickel catalysts tested in the TEES w process
Property Catalyst
A B C D E
Type M H H R R
Surface area Moderate High High Low Low
Nickel content Moderate High High Moderate Low
Thermal stability Moderate Low Low High High

M s Methanation, H s Hydrogenation, R sReforming.

168 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

growth of the nickel metal was the cause of deactivation, while carbon fouling was
identified as the main consequence of this crystallite growth.
Yamaguchi et al. w40x carried out durability testing of alumina-supported nickel
catalysts for steam gasification of wood at 700 8C. The activity of the catalyst decreased
with the gasification time as a result of carbon deposition and sintering of the nickel
metal particles in the catalyst. Catalyst regeneration was investigated under three
different flows ŽH 2 , H 2 O, O 2 . at 600 8C. The O 2 treatment was reported to return the
nickel specific area to its fresh value; however, transmission electron micrographs of
nickel particles in the fresh and used catalyst clearly showed sintering of the nickel.
Analysis of the coked catalyst by XRD showed the presence carbon in the form of
graphite.
Kinoshita et al. w34x performed parametric tests on the catalytic reforming of tars,
which are produced during the gasification of biomass using a bench-scale fluid-bed
catalytic reformer and a commercial nickel catalyst ŽUnited catalyst G-90B 11% nickel
on a ceramic support.. They reported that tar conversion was increased at higher
temperatures or longer space times. For each temperature, a critical space time existed at
which point all the tars were reformed into essentially gaseous species. The gas yields
were dependent on the space time and temperature: with an increase in temperature or
space time, the content of hydrogen and carbon monoxide increased and there was a
subsequent decrease in the contents of carbon dioxide, methane and hydrocarbons.
Increasing the steam to biomass ratio increased the H 2 :CO ratio; however, it decreased
the gross heating value of the gas. Kinoshita et al. w34x concluded that all tar and
methane could be reformed over nickel catalysts under the correct operating conditions.
These conditions were referred only to a varying temperature, equivalence ratio and
residence time. The findings are in keeping with those of several other groups relating to
both nickel and dolomite catalysts. Kinoshita et al. w34x, however, pointed out that nickel
catalysts have potential drawbacks vis-a-vis` cost, intolerance to oxygen breakthrough
and disposal.
Minowa et al. w36x investigated two nickel catalysts for the gasification of wood and
cellulose at 350 8C and 17 MPa in an autoclave. The first was a precipitated 50 wt.%
nickel catalyst on kieselguhr containing sodium carbonate as a promoter and the second
was a commercial nickel catalyst ŽEngelhard Ni-3288 containing about 50 wt.% nickel
on silica–alumina.. The catalysts were directly mixed with the biomass and water in the
autoclave. An increase in the reaction pressure Žfrom 8 to 17 MPa. was shown to result
in an increased gas yield. The concentration of hydrogen and carbon dioxide in the
product gas was greater at increased pressures while the methane concentration was
reduced as the pressure increased. On increasing the catalyst loading, the gas yield
increased at the expense of the liquid fraction and the hydrogen content also increased
significantly. Table 8 summarises the results.
The 50% NirKieselguhr catalyst gave a higher gas yield and a greater mol fraction of
hydrogen in the gas phase than did the commercial catalyst. Minowa et al. w36x attributed
the greater activity of the 50% NirKieselguhr catalyst to the support material, although
no direct evidence was reported.
Wang et. al. w37x have reported on the use of commercial nickel catalysts that
generate hydrogen by the thermocatalytic processing of a variety of bio-oils. Catalysts
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 169

Table 8
Comparison of 50% NirKieselguhr catalyst and commercial catalyst for gasification at 350 8C for 1 h and 18
MPa
50% NirKieselguhr Ni-3288
Gas yield, wt.% 93.7 79.4
H 2 , mmol 124.9 35.2
CO 2 , mmol 91.7 70.3
CH 4 , mmol 20.5 40.4
Carbon to gas, % 63.2 68.9
Hydrogen to gas, % 102.4 78.4
Oxygen to gas, % 115.5 88.3

investigated were G-90C G-91 ŽUnited Catalyst., G-125S ŽBASF. and 46-4 ŽICI
Katalco.. The conditions used for reforming were as follows: temperature, 700 8C; steam
to carbon ratio, 5; GHSV, 6725 hy1 ; and residence time, 0.1 s. The hydrogen yield for
all the catalysts tested was above 90% of the stoichiometric value. The ICI catalyst used
in commercial reforming of naphthas reduced coke formation, and extended the catalyst
life over that of the United Catalysts sample. Both catalysts could be easily regenerated
by the removal of carbon deposits with steam or CO 2 . The CO content of the gas could
be increased by lowering the steam to carbon ratio; this decreased the H 2 and CO 2
yields by lowering the degree of which the water–gas shift reaction occurred.
Simell et al. w16x studied the steam and dry reforming of toluene over a 13 wt.%
nickel-on-alumina catalyst at 900 8C and 20 MPa. The conversion was very high and the
equilibrium composition was almost achieved; however, carbon deposition was rapid
and resulted in deactivation of the catalyst. They reported that the dry reforming was
slightly faster than steam reforming, a result which is at variance with the results
obtained by Rostrup-Nielsen and Bak Hansen w29x. The contradictory results may have
arisen because of the lower temperature range investigated by Rostrup-Nielsen and Bak
Hansen w29x. In addition, the two sets of authors w16,29x used different catalysts.
Furthermore, the study by Simell et al. w16x was unlikely to have been carried out in the
kinetic regime.
Several groups w38–43x prepared and tested nickel catalysts for the elimination of tars
and methane present in the exit gas of a gasifier. Garcia et al. w38,39x reported results for
a co-precipitated nickel–alumina catalyst Žof molar ratio 1:2. investigating the removal
of tars from the product gas of a wood pyrolysis unit operating at 650 8C and
atmospheric pressure. Calcination of the catalyst was carried out at three different
temperatures: 650 8C, 750 8C and 850 8C. They obtained lower tar yields and higher gas
yields using a catalyst reduction time of 2 h rather than one. The ratio of H 2 :CO was
higher for catalysts reduced for 2 h. The different calcination temperatures of the
prepared catalyst was shown to affect the tar conversion. At the lower calcination
temperatures, higher tar conversion was accompanied by higher gas yields.
Garcia et al. w39x in a separate paper, reported results for the same catalyst calcined at
750 8C for 3 h but not reduced prior to use. They reported on the influence of the
catalyst weightrbiomass flow rate ratio ŽWrmb. on the product distribution and on the
170 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

quality of the gas product. The Wrmb ratio had a significant influence on the yield of
various gases in the pyrolysis process. When the ratio was increased, the total gas, H 2
and CO yields all increased, and the CO 2 CH 4 and C 2q yields decreased. For a given
Wrmb ratio, the H 2 and CO yields decreased with reaction time while the CO 2 CH 4
and C 2q yields increased.
Modification of nickel catalysts by addition of promoters have also been investigated
w41–43x. Arauzo et al. w41x reported on the addition of magnesium and potassium to a
nickel alumina catalyst. The magnesium substitution was made at two different levels
resulting in two catalysts of Ni 2 MgAl 8 O 16 and NiMgAl 4 O 8 . The magnesium addition
was made to increase the physical strength of the catalyst and its resistance to attrition.
Partial replacement of nickel by magnesium improved the strength of the resultant
catalyst but gave a 14% lower gas yield and also resulted in an increase in the char
yield. The CO and H 2 yields decreased slightly with nickel content. The magnesium
modified the catalyst structure and pore size distribution, the unmodified catalyst
containing a higher fraction of wider macropores. It was furthermore proposed that the
magnesium may inhibit the reduction of the nickel. Carbon deposition was reported as
the cause of deactivation.
Richardson and Gray w42x reported results for the hydro-processing of biomass
obtained using a nickel catalyst doped with alkali metals. A commercial hydro-processing
catalyst consisting of alumina supporting 12.5 wt.% MoO 3 and 3.5 wt.% NiO was used.
The catalyst was doped with KNO 3 , KOH, NaOH or LiOH by incipient wetness
technique and calcined at 450 8C for 4 h. Doping was carried out in order to reduce the
surface acidity, which in turn was expected to reduce coking and increase the activity.
Doping with KNO 3 was not effective in neutralising the surface acidity at low
concentrations and resulted in catalyst poisoning at high concentrations. Doping of the
catalyst with KOH, NaOH and LiOH did reduce the acidity of the surface of the
catalysts; however, only a slight reduction of coke deposition was reported w43x.
Poisoning of the catalysts occurred when high concentrations of these additives were
presented.
Bangala et al. w43x investigated the steam reforming of naphthalene on NirAl 2 O 3
catalysts doped with MgO, TiO 2 or La 2 O 3 . The catalyst supports were prepared by
mixing Al 2 O 3 and La 2 O 3 Žor TiO 2 . and MgO, this being followed by calcination at 800
8C. The active element ŽNi. and the promoters were introduced by wet impregnation.
The effect of the metal loading of the nickel on alumina material was reported for
concentrations of 5, 10, 15, 20 wt.% Ni. The conversions and gas yields increased with
metal loading up to 15 wt.% and declined at higher loading. The decrease was attributed
to decrease in the metal dispersion at higher loading. The maxima in the selectivities to
H 2 and CO was also reported to occur with 15 wt.% Ni. The coke yield increased
continuously as metal loading increased, demonstrating that the coking of a catalyst is
related to its metal loading. The effect on the 15 wt.% catalyst of the calcination
temperature used was also investigated. The conversion and gas yield decreased with
increasing calcination temperature and the catalysts calcined at lower temperatures had
low stability. The selectivities for H 2 and CO decreased with increasing calcination
temperature. The effect of doping the alumina support with MgO was carried out at
various loadings Ž5, 10, 20, 25 wt.%. w43x. The conversions and gas yields decreased for
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 171

samples containing more than 15 wt.% MgO while the degree of coke deposition
increased steadily with increasing Ni loading. The Al 2 O 3 P MgO support was also mixed
and calcined with TiO 2 and La 2 O 3 . The material based on the La 2 O 3 doped support was
reported to give the highest conversions and gas yields of all the catalysts examined w43x.
This material showed a definite decrease in carbon deposition without affecting conver-
sion.

3. Conclusions

Dolomite is a suitable catalyst for the removal of hydrocarbons, which are evolved in
the gasification of biomass. Dolomites increase gas yields at the expense of liquid
products. With suitable ratios of biomass feed to oxidant, almost 100% elimination of
tars can be achieved. The dolomite catalyst deactivates due to carbon deposition and
attrition; however, dolomite is cheap and easily replaced. The catalyst is most active if
calcined and placed downstream of the gasifier in a fluidised-bed at temperatures above
800 8C. The reforming reaction of tars over dolomite occurs at a higher rate with carbon
dioxide than steam. Dolomite activity can be directly related to the pore size and
distribution. A higher activity is also observed when iron oxide is present in significant
amounts. Dolomites are not active for reforming the methane present in the product gas
and hence they are not suitable catalysts if syngas is required. The main function of
dolomite is to act as a guard bed for the removal of heavy hydrocarbons prior to the
reforming of the lighter hydrocarbons to produce a product gas of syngas quality.
Alkali catalysts directly added to the biomass by wet impregnation or dry mixing
reduce tar content significantly and also reduce the methane content of the product gas.
When directly added these catalysts increase the rate of gasification dramatically.
However, the recovery of the catalyst is difficult and costly. Inherent in the ash of
several biomass types are high concentrations of alkali metals. Ash is an effective
catalyst for the removal of tar when mixed with the biomass. Alkali catalysts directly
added to the biomass in a fluidised-bed gasifier are subject to particle agglomeration.
Alkali metal catalysts are also active as secondary catalysts. Potassium carbonate
supported on alumina is more resistant to carbon deposition although not as active as
nickel. The catalyst is not suitable as a secondary catalyst since the hydrocarbon
conversion rarely exceeds 80%.
Commercially available nickel reforming catalysts are highly effective at the removal
of hydrocarbons and adjustment of the gas composition to syngas quality. The nickel
catalysts act best as secondary catalysts located in a downstream reactor, which can be
operated under different conditions than those of the gasifier. The catalysts are most
active and have longer lifetime when operated at 780 8C in a fluidised-bed. Operated at
lower temperatures Ž600 8C., the catalysts can produce a methane-rich gas. The activity
of the catalysts is sensitive to nickel loading and metal dispersion. Deactivation is
primarily due to carbon deposition and nickel particle growth. Deactivation due to
coking may be reduced by the introduction of a guard bed of dolomite. The addition of
dopants such as lanthanum can also reduce carbon deposition. The catalysts are
effective, commercially available and relatively cheap.
172 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173

References
w1x M. Kaltschmitt, C. Rosch,
¨ Research Report AIR-CT94-2284, IER, Universitat ¨ Stuttgart, 1998, p. 47.
w2x M.P. Aznar, J. Corella, J. Delgado, J. Lahoz, Ind. Eng. Chem. Res. 32 Ž1993. 1.
w3x N. Lindman, Proceedings of Conference on Energy from Biomass and Wastes V, Lake Buena Vista,
Elsevier, Amsterdam, 1981, p. 89.
w4x A. Orıo,
´ J. Corella, I. Narvaez,
´ Proceedings of Conference on Developments in Thermochemical Biomass
Conversion, Banff, Canada, 1996, p. 1144.
w5x J. Delgado, M.P. Aznar, J. Corella, Ind. Eng. Chem. Res. 36 Ž1997. 1535.
w6x V. Vassilatos, G. Taralas, K. Sjostrom,
¨ ¨ E. Bjornbom,
¨ Can. J. Chem. Eng. 70 Ž1992. 1008.
w7x P. Perez,
´ P.M. Aznar, M.A. Caballero, J. Gill, J.A. Martin, J. Corella, Energy Fuel 11 Ž1997. 1194.
w8x J. Corella, J. Herguido, J. Gonzales-Saiz,
´ J. Trujillo, in: A.V. Bridgwater, J.L. Kuester ŽEds.., Research in
Thermochemical Biomass Conversion, Elsevier Applied Science, London, 1988, p. 754.
w9x I. Nanaez,
´ A. Orıo,´ M.P. Aznar, J. Corrella, Ind. Eng. Chem. Res. 35 Ž1996. 2110.
w10x C. Myren,´ C. Hornell,
¨ ¨ ¨ Q. Yu, C. Brage, E. Bjornbom,
K. Sjostrom, ¨ Proceedings of Conference on
Developments in Thermochemical Biomass Conversion, Banff, Canada, 1996, p. 1170.
w11x V. Vassilatos, G. Taralas, K. Sjostrom,
¨ ¨ E. Bjornbom,
¨ Can. J. Chem. Eng. 70 Ž1992..
w12x H. Alden,´ E. Bjorkman,
¨ M. Carlsson, L. Waldheim, Proceedings of Conference on AAdvances in
Thermochemical Biomass ConversionB, Interlaken, Switzerland May 11th–15th, 1992, p. 216.
w13x P.A. Simell, N. Hakala, H.E. Haario, Ind. Eng. Chem. Res. 36 Ž1997. 42.
w14x G. Taralas, Proceedings of Conference on AAdvances in Thermochemical Biomass ConversionB, Inter-
laken, Switzerland May 11th–15th, 1992, p. 296.
w15x H. Alden,´ P. Hagstrom,¨ A. Hallgren, L. Waldheim, Proceedings of Conference on Developments in
Thermochemical Biomass Conversion, Banff, Canada, 1996, p. 1131.
w16x P.A. Simell, J.O. Hepola, A. Krause, Fuel 76 Ž1997. 1117.
w17x M.P. Aznar, J. Corella, J. Gill, J.A. Martin, M.A. Caballero, A. Olivares, E. Frances, ´ Proceedings of
Conference on Developments in Thermochemical Biomass Conversion, Banff, Canada, 1996, p. 1117.
w18x G. Lammers, A.A.C.M. Beenackers, Proceedings of Conference on Developments in Thermochemical
Biomass Conversion, Banff, Canada, 1996, p. 1179.
w19x C. Ekstrom,
¨ N. Lindman, R. Pettersson, Proceedings of Conference on Developments in Thermochemical
Biomass Conversion, Banff, Canada, 1996, p. 601.
w20x L.K. Mudge, E.G Baker, D.H. Mitchell, M.D. Brown, J. Solar Energy Eng. 107 Ž1985. 89.
w21x W.B. Hauserman, Int. J. Hydrogen Energy 19 Ž1994. 413.
w22x R.T. Hallen, L.T. Sealock, R. Cuello, A.V. Bridgwater, in: J.L. Kuester ŽEd.., Research in Thermochemi-
cal Biomass Conversion, Elsevier Applied Science, London, 1988, p. 157.
w23x J.M. Encinar, F.J. Beltran,
´ A. Ramiro, J.F. Gonzalez,´ Fuel Process. Technol. 55 Ž1998. 219.
w24x E. Sala, H. Kamazawa, M. Kuds, Ind. Eng. Chem. Res. 31 Ž1992. 612.
w25x T.C. Li, Y.J. Yan, Z. Ren, Fuel Sci. Technol. 14 Ž1996. 879.
w26x W.F. DeGroot, F. Shafizaceh, Proceedings of Conference on Developments in Thermochemical Biomass
Conversion, Banff, Canada, 1996, p. 275.
w27x D.B. Anthony, AIChE J. 22 Ž1976. 625.
w28x S.C. Gebhard, D. Wang, R.P. Overend, M.A. Paisley, Biomass Bioenergy 7 Ž1994. 307.
w29x J.R. Rostrup-Nielsen, J.H. Bak Hansen, J. Catal. 144 Ž1993. 38.
w30x M.P. Aznar, M.A. Caballero, J. Gil, J.A. Martin, J. Corella, Ind. Eng. Chem. Res. 37 Ž1998. 37.
w31x E.G. Baker, L.K. Mudge, M.D. Brown, Ind. Eng. Chem. Res. 26 Ž1987. 1335.
w32x E.G. Baker, L.K. Mudge, W.A. Wilcox, Proceedings of Conference on Developments in Thermochemical
Biomass Conversion, Banff, Canada, 1996, p. 305.
w33x E.G. Baker, D.C. Elliott, R.S. Butner, L.J. Sealock, J. Solar Energy Eng. 115 Ž1993. 52.
w34x C.M. Kinoshita, Y. Wang, J. Zhou, Ind. Eng. Chem. Res. 34 Ž1995. 2949.
w35x C. Roy, H. Pakdel, H.G. Zhang, D.C. Elliott, Can. J. Chem. Eng. 72 Ž1994. 98.
w36x T. Minowa, T. Ogi, S. Yokoyama, Fuel 7 Ž1996. 1342.
w37x D. Wang, S. Czernik, E. Chornet, Energy Fuel 12 Ž1998. 19.
w38x L. Garcia, M.L. Salvador, J. Arauzo, R. Bilbao, J. Arauzo, Proceedings of Conference on Developments
in Thermochemical Biomass Conversion, Banff, Canada, 1996, p. 315.
 D. Sutton et al.r Fuel Processing Technology 73 (2001) 155–173 173

w39x L. Garcia, M.L. Salvador, J. Arauzo, R. Bilbao, Ind. Eng. Chem. Res. 37 Ž1998. 3812.
w40x T. Yamaguchi, K. Yamasaki, O. Yoshida, Y. Kanai, A. Ueno, Y. Kotera, Ind. Eng. Chem. Prod. Res.
Dev. 25 Ž1986. 239.
w41x J. Arauzo, D. Radlein, J. Piskorz, D.S. Scott, Ind. Eng. Chem. Res. 36 Ž1997. 67.
w42x S.M. Richardson, M.R. Gray, Energy Fuels 11 Ž1997. 1119.
w43x D.N. Bangala, N. Abatzoglou, E. Chornet, AIChE J. 44 Ž1998. 927.