Fuel Processing Technology 90 (2009) 677–685

Contents lists available at ScienceDirect

Fuel Processing Technology

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / f u p r o c

Air–steam gasification of biomass in a fluidised bed: Process optimisation by

enriched air
Manuel Campoy, Alberto Gómez-Barea ⁎, Fernando B. Vidal, Pedro Ollero
Bioenergy Group, Chemical and Environmental Engineering Department, Escuela Superior de Ingenieros (University of Seville), Camino de los Descubrimientos s/n. 41092 - Seville, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The effect of oxygen concentration in the gasification agent was studied by enriched–air–steam biomass
Received 8 August 2008 gasification tests in a bubbling fluidised-bed gasification (FBG) plant. The oxygen content in the enriched air was
Received in revised form 4 December 2008 varied from 21% (v/v, i.e. air) to 40% (v/v), aiming at simulating FBG where enriched air is produced by membranes.
Accepted 12 December 2008
The stoichiometric ratio (ratio of actual to stoichiometric oxygen flow rates) and steam-to-biomass ratio (ratio of
steam to biomass, dry and ash-free, flow rates) were varied from 0.24 to 0.38 and from 0 to 0.63, respectively.
Keywords:
The tests were conducted under simulated adiabatic and autothermal conditions, to reproduce the behaviour
Gasification
Biomass
of larger industrial FBG. The temperature of the inlet gasification mixture was fixed consistently at 400 °C for
Fluidised-bed all tests, a value that can be achieved by energy recovery from the off-gas in large FBG without tar condensation.
Scaling-up It was shown that the enrichment of air from 21 to 40% v/v made it possible to increase the gasification
Enriched air–steam mixtures efficiency from 54% to 68% and the lower heating value of the gas from 5 to 9.3 MJ/Nm3, while reaching a
maximum carbon conversion of 97%. The best conditions were found at intermediate values of steam-to-biomass
ratio, specifically within the range 0.25–0.35. The enriched-air–steam gasification concept explored in this work
seems to be an interesting option for the improvement of standalone direct air–blown FBG because it considerably
improves the process efficiency while maintaining the costs relatively low as compared to oxygen-steam
gasification.
© 2008 Elsevier B.V. All rights reserved.

1. Introduction In steam-oxygen gasification the biomass is partially oxidised to

Gasification is a promising technology for biomass and waste
utilisation with low environmental impact, reducing global CO2
emissions. Standalone air-blown bubbling fluidised-bed gasification
(FBG) is the simplest and probably the most cost-effective concept for
medium-scale thermal and electricity applications (b5–10 MWe). It
has been successfully demonstrated connected to a large coal-fired
boiler in power plants, resulting in high efficiency when burning
biomass. For small-scale power production a unit comprised of a
gasifier and a compression-ignition engine is less expensive than a
boiler-based power cycle, thus providing an attractive option for
remote locations [1].
The use of air in gasification yields a fuel gas highly diluted by
nitrogen. The lower heating value (LHV) of the gas is therefore
typically below 6 MJ/Nm3. The use of steam has proven effective for
achieving a medium heating value (up to 14 MJ/Nm3), but the process
becomes more complex. Two different concepts have been developed,
for full-scale steam gasification: steam-oxygen mixtures and indirect
gasification based on twin-bed reactors [2].
⁎ Corresponding author. Tel.: +34 95 4487223; fax: +34 95 4461775.
E-mail address: agomezbarea@esi.us.es (A. Gómez-Barea).
provide the heat necessary to make the process self-sufficient. The gas
produced has a high hydrogen content and the dilution of nitrogen is
avoided. This concept has been extensively tested at laboratory scale
[3,4], but the high cost of pure oxygen (for instance, based on
distillation units) makes the implementation of this process uncertain
at industrial scale [5].
Indirect biomass gasification is based on the separation of the
gasification and combustion stages in two different chambers. The
heat supply to the gasification process, which is fed with steam, is
generated in the combustion chamber, in which the char is burnt out
using air. The heat transfer is achieved by circulation of the bed
inventory between the two stages. The essential contribution of this
concept is that it allows ‘autothermal’ steam gasification without the
need of oxygen (only air is used), producing a gas with medium
heating value, i.e., not diluted by nitrogen. This technology has
achieved semi-commercial status today [6,7].
Much experimental work has been done on FB biomass gasification
using different gasification agents. Air [8], pure steam [9–11], oxygen-
steam [3,4] and air–steam [12–14] tests have been carried out in
various lab-scale facilities, generating useful knowledge for under-
standing the process. Relatively little work has been found, however,
using enriched–air–steam mixtures as gasification agent [15]. More-
over, most studies at laboratory or pilot scale have been conducted
allothermally; the temperature, air/oxygen-to-biomass ratio and

0378-3820/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.fuproc.2008.12.007
678 M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685
steam-to-biomass ratio have been varied independently because the
temperature of the gasifier is controlled by external heat addition
using electric heaters. However, this method of supplying heat is
neither technically nor economically feasible for large-scale imple-
mentation, so the results from “allothermal” lab-scale rigs must be
interpreted with caution.
In previous work [16], we have studied the process improvement
by using air–steam mixtures under simulated adiabatic and auto-
thermal conditions. We have shown how the optimisation of air–
steam mixtures in autothermal FBG increases the gasification
efficiency from 40% to 60%, while maintaining the low heating value
of the gas at around 5 MJ/Nm3. The process was, however, limited to
using air as the source of oxygen to the system, i.e., 21% (v/v).
In this study we have extended the previous work by allowing the
oxygen to vary in purity up to 40% (v/v) in the enriched air. This higher
figure was chosen because it is the oxygen concentration that can be
produced using commercial air separators based on membrane
technologies. This way of producing oxygen would keep the invest-
ment and operational costs relatively low compared to processes that
need pure oxygen, usually based on distillation units. Consequently,
the size of plants where this concept is expected to be feasible is not
necessarily large. Therefore, the concept investigated in this work
could be convenient for medium-scale electricity production (b5–10
MWe), provided the gas cleaning, particularly tars, is done properly.
This study looked at gas quality and the impact on carbon conversion
and process efficiency. Tar production and cleaning strategies will be
addressed in future work.
The tests were conducted varying the flow rates of biomass, steam
and enriched air and the oxygen concentration in the enriched air,
making it possible to explore the effects of these variables on the
quality and composition of the produced gas, aiming at process
optimisation. All the tests were conducted with wood pellets in a
pilot-scale FBG simulating autothermal and adiabatic conditions. In
addition, the temperature of the gasification agent was fixed at 400 °C
for all tests, a value that can be achieved by energy recovery from the
off-gas without tar condensation. With this experimental setup, we
expect that the results will be useful for scaling up the data to design
industrial FBG, i.e., large standalone direct gasification units without
external heat supply and with low wall-heat loss.
2. Experimental
2.1. Materials
The biomass used was wood pellets with the empirical formula
CH1.4O0.64 (dry and ash-free, d.a.f.) [16]. The moisture and ash content
were 6.3% and 0.5% (w/w), and the lower heating value of the fuel (as
received) was 17.1 MJ/kg. The pellets were cylindrically shaped with a
mean diameter of 6 mm and 5–10 mm long. The apparent density of
the pellets and the bulk density were 1300 and 600 kg/m3. The bed
material used was ofite, a sub-volcanic rock composed mainly of
feldspar, pyroxene and limestone, whose complete characterisation is
given in [17].
2.2. Facility
Fig. 1 shows the layout of the facility and Table 1 gives the main
parameters of the pilot plant. The rig has been described in detail in a
previous publication [16]. The only modifications made in the plant
with respect to previous work were the erection of a plant for oxygen
production and the removal of the tar scrubber to avoid tar
condensation in the pipes between the cyclones and the combustion
chamber. In addition, the pipes were maintained at a temperature
above 400 °C by heating elements and insulation blankets. The oxygen
plant comprised four 10-m3 oxygen bottles, making it possible to
produce 10 Nm3/h with a purity higher than 99%.
2.3. Test procedure
The protocol for the tests was similar to that used in [16].
Therefore, only a brief description is given here.
At the beginning of each test a batch of bed material (around 8 kg)
was added to the reactor. The bed was heated up with the hot air and
the electric heater. The bed and the freeboard were rapidly heated up
to approximately 700 °C. It was necessary, however, to wait longer
before starting the biomass feed in order to avoid tar deposition on the
pipes between the cyclones and the combustion chamber. Once the
temperature upstream of the combustion chamber was higher than
300 °C, the facility was considered ready for biomass, and biomass was
fed slowly into the reactor. In these conditions of oxygen excess, the
biomass was oxidised completely and the reactor was rapidly heated
to the desired process temperature. From the beginning of the test, a
computer-based data acquisition system monitored and recorded the
temperatures, pressures, gas composition (H2, CO, CO2, CH4, O2),
power supplied to the heating equipment and the flow rates of gas and
solid. The transition from combustion to gasification was made by
increasing the biomass flow rate to decrease the air-to-biomass ratio.
The steam addition started once a steady-state condition was
established. The mixture of steam and air was heated in a preheater
and then fed to the FBG. Once the process condition was stable, the
oxygen was added to the steam before mixing with air, with the flow
rate reduced to an appropriate value to obtain the desired level of
oxygen in the final mixture. The final air–steam–oxygen mixture was
heated in the preheater to the process temperature (400 °C). In all
tests an initial transitory period of 3 to 4 h was followed by a steady-
state period of 5 to 7 h. The operation was finalised by taking a sample
from the bed inventory and combusting the remaining char in the bed.
After each test, the two cyclone bins and the extraction-ash bin were
sampled and analysed.
2.4. Operating conditions
The mode of operation we used for conducting the tests (adiabatic
tests and temperature of inlet gasification mixture fixed at 400 °C) for
a given biomass (wood pellets) consisted of four variables that could
be independently varied: the flow rates of biomass, air, steam and
oxygen. The flow rates of air and oxygen determined the oxygen
content in the enriched air inlet, i.e., the O2 purity (hereafter referred
to as OP, in % v/v).
For a given biomass flow rate, two ratios can be defined for the
analysis of the process: (1) the stoichiometric ratio (SR), defined as
the mass ratio between the amount of total oxygen fed in and the
stoichiometric amount of oxygen required for combustion, and (2) the
steam-to-biomass ratio (SBR), defined as the flow rate of steam fed to
the reactor divided by the biomass flow rate (dry and ash-free). Two
additional variables must be taken into account: (3) the oxygen
percentage of the enriched air (OP), which is an indication of the
nitrogen dilution of the produced gas, and (4) the biomass throughput.
However, for a limited range of this parameter in an FB, the system can
be analysed approximately by a mean of (1), (2) and (3), as the biomass
flow rate is expected to have a minor influence on the results.
The experimental program comprised tests with air–oxygen–
steam mixtures in different proportions. In the first set of tests, air was
used as oxygen supplier, so the OP was 21% v/v for all tests and the
oxygen flow rate was nil. Some of these tests have been reported in
previous work [16]. The second set of tests involved air–oxygen-steam
mixtures in which the OP of the enriched air was set at 30, 35 and 40%
v/v. Some of these tests were conducted twice to ensure the
reproducibility of the results.
The experimental programme comprised tests varying SR, SBR and
OP between the indicated ranges: SR from 0.24 to 0.38, SBR from 0 to
0.63 and OP from 21 to 40%. To establish these ratios, the following
flow rates were varied between the indicated ranges: the steam flow
 M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685 679

Fig. 1. Pilot plant layout.

rate from 0 to 7.4 kg/h, the biomass flow rate from 10 to 21.6 kg/h, the
air flow rate from 5.6 to 17 Nm3/h and the oxygen flow rate from 0 to
3.7 Nm3/h. An experimental matrix was constructed with four values
of OP (21, 30, 35 and 40) and two levels of SR and SBR, representing
low and high values: 0.24–0.27 (low SR) and 0.33−0.38 (high SR) for
SR and 0.22–0.36 (low SBR) and 0.43–0.63 (high SBR) for SBR. The
experimental conditions were selected to assess the operation at low
and high values of both SR and SBR, while changing the OP to keep the
temperature high enough. Table 2 groups the tests in four combina-
tions on the basis of the level of the SR and SBR used. In total, the
matrix of experiments comprised 16 independent experiments. The
main results of the tests are given in Table 3.
Table 3 includes 20 tests because four additional tests were
conducted to study the effect of steam in more detail. These additional
tests were conducted at a fixed value of SR ≈ 0.35 and two OP values
(21% in test 1 and 40% in tests 14, 16 and 17).
The gas residence times in the bed and the freeboard are important
operating data, and should therefore be specified clearly. Ideally, the
gas residence time in the bed and the freeboard should take into
account the gas from biomass devolatilisation as well as that from the
gasification of char. In this work, we calculated the residence time in
the bed based on the total amount of gasification agent fed to the
reactor at the mean bed temperature. For the gas residence time in the
freeboard, the calculation was made by considering the outlet dry gas
and the average freeboard temperature (water content was not
considered because of the uncertainty of this value).The gas residence
times for the bed and the freeboard were calculated taking into
account the dimensions given in Table 1 and ranged from 0.99 to 1.43 s
and between 3.7 and 6.1 s, respectively.
One significant aspect of the tests described in this work is that
they were conducted nearly adiabatically: the heating system was
controlled to provide just the necessary amount of heat to compensate
for the heat losses. This operation was achieved by keeping the tem-
perature of the furnace slightly lower (5–10 °C) than the temperature
measured inside the reactor. This operational procedure has been
explained in [16]. In addition, the temperature of the gasification
agent was fixed at 400 °C for all tests, a value which can be achieved by
heat recovery in standalone FB systems without tar condensation.
3. Results and discussion
The variables analysed included gas composition, gas yield, heating
value, gasification efficiency and carbon conversion.

Table 1
Technical and operating data for the facility and the main test operating data.
Table 2
Inside Bed Diameter 0.15 m Definition of the four combinations used for the analysis as a function of the levels of SR
Inside Freeboard Diameter 0.25 m and SBR tested.
Freeboard Height 2.15 m
SR SBR
Bed Material Ofite
Fuel Wood Pellets Combination High Low High Low
Fuel Flow Rate 10–21.6 kg/h 1 0.33–0.38 – – 0.22–0.36
Gasification Agent Air + Oxygen + Steam 2 0.33–0.38 – 0.45–0.60 –
Operation Temperature 755–840 °C 3 – 0.25–0.27 – 0.23–0.31
Fluidisation Regime Bubbling 4 – 0.24–0.27 0.43–0.63 –
 680
Table 3
Test results.

Test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Operational conditions
Biomass Flow Rate (kg/h) 11.5 12.2 12.2 15 15 12.4 10.0 16.2 12.0 14 11.8 16.8 12.6 21.6 16.2 14.8 13.2 12.0 18.8 14.0
Air Flow Rate (Nm3/h) 17 17 17 17 17 11.9 9.1 10.6 7.7 11.6 8.3 9.5 7 11.5 9.3 8.9 7.3 6.8 8.1 5.6
Oxygen Flow Rate (Nm3/h) 0 0 0 0 0 1.5 1.2 1.4 1.0 2.3 1.8 2 1.5 3.7 2.9 2.8 2.3 2.1 2.6 1.8
Steam Flow Rate (kg/h) 0 2.5 5.1 3.2 6 3.7 5.6 4.7 6.5 4.3 6.2 4.9 7.4 2.1 4.4 4.9 6.6 6.4 5.3 7.3
Mean Bed Temperature, Tb (°C) 812 804 789 786 755 808 790 781 765 820 795 800 757 840 829 830 813 806 803 766
Furnace Set-point (°C) 800 800 780 780 750 800 780 770 750 810 780 790 750 825 815 815 800 800 790 760

M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685

Mean Freeboard Temperature(°C) 716 721 709 708 709 715 715 716 695 715 725 725 720 724 708 727 725 727 725 722
tbed (s)⁎ 1.3 1.1 1.0 1.1 1.0 1.2 1.3 1.3 1.4 1.2 1.3 1.3 1.3 1.2 1.2 1.2 1.3 1.3 1.3 1.4
tfreeboard (s)⁎ 4.3 4.0 4.0 3.9 3.7 4.7 6.1 4.5 6.0 4.2 5.7 4.5 5.8 3.4 4.4 4.7 5.5 5.8 4.4 6.0

Main variables for analysis

OP (%) 21 21 21 21 21 30 30 30 30 35 35 35 35 40 40 40 40 40 40 40
SR 0.35 0.33 0.33 0.27 0.27 0.36 0.35 0.25 0.24 0.38 0.34 0.27 0.26 0.32 0.33 0.35 0.32 0.33 0.26 0.24
SBR 0 0.22 0.45 0.23 0.43 0.32 0.60 0.31 0.58 0.33 0.56 0.31 0.63 0.10 0.29 0.36 0.54 0.57 0.30 0.56
Combination (see Table 2) – 1 2 3 4 1 2 3 4 1 2 3 4 – 1 – – 2 3 4

Gas: composition and yields

CO (%v/v, dry) 15.8 15.4 13.8 15.0 11.9 18.9 15.7 20.8 15.3 20.0 17.5 23.9 19.3 27.4 25.1 23.9 20.2 19.3 28.5 23.5
H2 (%v/v, dry) 8.7 11.9 13.3 14.0 16.2 16.4 18.3 20.0 22.3 17.5 21.8 22.4 25.1 18.3 23.1 22.3 24.5 25.7 25.7 27.5
CO2 (%v/v, dry) 15.1 15.9 17.0 16.2 18.6 17.6 18.8 15.8 20.3 16.8 18.0 12.6 16.2 16.2 13.7 14.6 16.7 17.0 9.2 14.6
CH4 (%v/v, dry) 5.1 4.8 4.6 4.7 5.3 5.5 5.7 6.7 7.1 5.6 6.1 7.3 7.4 7.3 6.5 6.7 6.9 6.7 8.1 7.7
CO (%v/v, dry, N2 free) 35.3 32.1 28.3 30.1 22.9 32.3 26.9 32.9 23.5 33.4 27.6 36.1 28.4 39.6 36.7 35.4 29.6 28.1 39.8 32.1
H2 (%v/v, dry, N2 free) 19.5 24.8 27.3 28.1 31.2 28.1 31.2 31.6 34.3 29.2 34.4 33.9 36.9 26.5 33.8 33.0 35.9 37.4 36.0 37.5
CO2 (%v/v, dry, N2 free) 33.8 33.1 34.9 32.5 35.8 30.2 32.2 24.9 31.3 28.0 28.4 19.0 23.8 23.4 20.0 21.6 24.4 24.8 12.9 19.9
CH4 (%v/v, dry, N2 free) 11.4 10.0 9.4 9.4 10.2 9.4 9.7 10.6 10.9 9.4 9.6 11.0 10.9 10.5 9.5 9.9 10.1 9.7 11.3 10.5
CO yield (g/kg d.a.f.b.) 453.4 443.4 402.7 365.7 302.8 456.1 362.8 405.0 305.6 475.1 366.1 438.1 369.3 520.1 499.4 483.5 386.7 384.7 476.2 391.5
H2 yield (g/kg d.a.f.b.) 17.8 24.5 27.7 24.4 29.4 28.3 30.1 27.8 31.9 29.7 32.6 29.4 34.3 24.9 32.9 32.2 33.5 36.6 30.8 32.7
CO2 yield (g/kg/d.a.f.b.) 680.9 719.3 779.6 620.6 743.7 669.7 681.6 482.0 639.3 626.7 591.5 361.8 487.0 481.9 427.6 464.4 501.5 532.7 242.0 381.3
CH4 yield (g/kg d.a.f.b.) 83.6 79.0 76.7 65.5 77.1 76.0 75.1 74.5 81.2 76.1 72.9 76.5 80.7 79.2 73.9 77.5 75.4 76.2 77.5 73.2

Process variables
Gas Flow Rate (Nm3/h, dry)⁎ 24.6 26.1 26.5 27.2 28.4 22.3 17.2 23.5 17.9 24.8 18.4 22.9 17.9 30.5 24.0 22.3 18.8 17.8 23.4 17.3
Gas yield (GY) (Nm3 dry gas, N2 free/kg d.a.f.b.) 1.03 1.11 1.14 0.97 1.06 1.13 1.08 0.98 1.04 1.14 1.06 0.97 1.04 1.05 1.09 1.09 1.05 1.10 0.96 0.98
Low healing value (LHV) (MJ/Nm3 dry gas) 4.76 4.95 4.83 5.09 5.15 6.12 6.00 7.19 6.88 6.41 6.75 8.06 7.81 8.06 8.00 7.83 7.67 7.62 9.28 8.70
Cold Gasification Efficiency 0.59 0.62 0.61 0.54 0.57 0.64 0.60 0.61 0.60 0.66 0.61 0.64 0.65 0.67 0.69 0.69 0.64 0.66 0.68 0.63
Carbon Conversion 0.93 0.90 0.92 0.90 0.91 0.94 0.95 0.96 0.96 0.95 0.95 0.96 0.96 0.97 0.96 0.96 0.97 0.96 0.97 0.97

⁎ Not measured (calculated from other measurements).

 M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685
Fig. 2. (a) Bed temperature as a function of OP for different SR and SBR levels. : SR= High,
SBR = Low (combination 1); : SR= High, SBR = High (combination 2); : SR= Low,
SBR = Low (combination 3); SR= Low, SBR = High (combination 4). (See Table 2 for
the definition of the four combinations). (b) Comparison between mean bed temperature
(Tb, ) and the theoretical bed temperature (T⁎, b solid line and ). (c) Bed temperature as
a function of SBR for SR= 0.35 and different levels of OP: : OP= 21%; : OP= 30%; :
OP= 35%; OP= 40%.
The effect of OP on the gasification temperature is shown in Fig. 2(a),
indicating that the temperature in the bed increased with OP in the
enriched air for all SR and SBR levels analysed. For fixed values of SR
and SBR, the increase in OP means that the same amount of combus-
tible matter is burned whilst the nitrogen flow that had to be heated
decreased, leading to a higher bed temperature. However, our mea-
surements did not identify such as increase in temperature that would
be expected from merely a decrease in the nitrogen dilution. We
demonstrate this observation by means of Fig. 2(b). This figure shows
the measured (average) bed temperature, Tb, for four tests (solid
diamonds) and Tb⁎ (blank squares over the solid line) as a function of OP,
where Tb⁎ is the bed temperature for those tests considering that the
only effect of decreasing the nitrogen flow rate was to raise the bed
temperature. Specifically, Fig. 2(b) presents the values of Tb and Tb⁎ for
combination 1 (high SR and low SBR). For the calculation of Tb⁎ for a
test with OPN 21%, say test 6, a heat balance was made by assuming that
681
the difference between the heat for heating up the fluidisation agent in
that test (6) as compared to that in test 2 (reference), was entirely in the
form of sensitive energy of the outlet gas. The same assumption was
made for the calculation of the corresponding Tb⁎ for tests 10 and 15 in
Fig. 2(b).
Then, the fact that, for a fixed OP, in Fig. 2(b) Tb is lower than Tb⁎
is interpreted as an enhancement of endothermic processes such
as tar and hydrocarbon reforming as well as char gasification with
CO2 and H2O. These processes are kinetically-limited under the
operating conditions tested, so a rise in the temperature enhances the
rate of these reactions. The actual bed temperature in a test with
OP N 21% is the net result of two opposing processes, one promoting an
increase in the bed temperature while the other making the tem-
perature lower. The difference between Tb and Tb⁎ for the tests with
OP N 21% in Fig. 2(b) confirmed that the aforementioned endothermic
processes are significant. This is confirmed below by observing an
increase of the CO and H2 concentration, expressed in nitrogen-free
basis, with OP.
Fig. 2(c) shows the effect of SBR on gasification temperature using
various degrees of OP for a fixed value of SR ≈ 0.35. As expected, the
temperature in the bed decreased as the steam flow rate increased.
This occurred for the four values of OP studied.
The effect of OP in the enriched air on the gas composition is given
in Fig. 3 (% v/v of dry gas) and Fig. 4 (yields per kilogram of dry and
ash-free biomass, d.a.f.b.). The volume fraction of the main gas
species measured in the outlet gas is shown in Table 3, expressed in
dry basis and also in dry nitrogen-free basis. As seen the composition
of CO, H2, CO2 and CH4 presents substantial differences depending
on the basis used. For a given combination, the molar composition
of CO and H2 increased with OP, expressed in both dry and dry and
nitrogen-free basis. This fact supports the conclusion made above
that the nitrogen dilution is not the only effect caused by the increasing
of OP.
Fig. 3(a) shows that, for all values of SR and SBR, the carbon
monoxide increased as OP increased. An increase of carbon monoxide
from 12 up to 28 was observed when OP was changed from 21 to 40%.
The combination of low SR and low SBR (combination 3 defined in
Table 2), seemed to provide the largest amount of CO, whilst
combination 2 (high SR and SBR) led to lower CO concentration in
the gas. The CO content increased as the OP rose. The H2 content in the
gas is shown in Fig. 3(b). It is clearly seen that combination 4 (low SR
and high SBR; see Table 2) increased the hydrogen content in the gas
for all values of OP tested.
Fig. 4 shows the yields of carbon monoxide and hydrogen per
kilogram of d.a.f.b. of CO and H2, respectively, as a function of the OP
for the different combinations of SR and SBR. We note that both
yields increased with OP. The yield of methane was roughly constant
(75 g/kg d.a.f.b.)
These findings are quite reasonable because, in general, as SR is
increased, a smaller amount of CO ends up in the produced gas and
more nitrogen is added to the system. In low-temperature gasification
the devolatilisation of the biomass determines the resulting gas
composition to a large extent. CO is oxidised to a larger extent,
resulting in lower CO yields. At a fixed SR, increasing OP means less
nitrogen in the gas and hence higher CO concentration for the same
CO yields. In addition, the rising temperature can increase CO yields
for a given SR and SBR (as seen in Fig. 4(a)), in this way making the
effect of increasing CO concentration even greater.
The fact that SBR increased the H2 content is also reasonable, but
the behaviour with respect to SR was not known a priori; less
hydrogen was probably burned as the SR decreased. However, for a
constant OP and SBR, the thermal level in the gasifier was lower,
resulting in a lower gas yield of H2 and possibly less secondary gas
phase reaction, supporting the presence of H2, such as in cracking and
reforming of hydrocarbons and tars. Secondary gas reactions such as
the water-gas-shift-reaction (WGSR) can alter the final composition if
682 M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685

and oxidation of carbon monoxide. In fact, the situation is more

complex since the CO2 is also a reactant for heterogeneous reactions
with char and possibly with tar; hence there is competition for the
oxygen among different species present in the gasifier: hydrogen,
char, tar, methane, light hydrocarbons, etc. It is most likely that
hydrogen and carbon monoxide react with oxygen to a greater extent
since they are more reactive species. This conjecture is also sup-
ported by the small variation detected for methane in Fig. 3(d). A
methane increase of only up to 3% was observed as OP increased from
21 to 40%.
The complex processes occurring simultaneously can explain the
trend shown in Fig. 3(c) for low values of OP, typically representing
conditions at lower temperature. As seen, the CO2 concentration was
slightly affected by the OP at low values of this parameter.
From an overall comparison of different graphs in Fig. 3, it seems
that devolatilisation and subsequent hydrogen and carbon monoxide
oxidation provide a good explanation for the gas distribution
measured at the exit of the gasifier. The four graphs in Fig. 3 establish
that CO and H2 increased, CO2 decreased and CH4 decreased slightly
with the OP in the enriched air. The WGSR does not explain the gas
phase distribution observed since it seems that the hydrogen and
carbon monoxide increased for all values of OP. Other homogeneous,
and to a lesser extent heterogeneous, reactions probably determined
the gas product distribution and the equilibrium of the WGSR was not
attained. In general, the reduction of the nitrogen dilution effect that
took place when increasing the flow rate of oxygen also explains the
trends observed.
The gas yield (GY), defined as the ratio of the volumetric flow rate
of the dry and nitrogen-free product gas (without considering light
hydrocarbons and tars) and the mass flow rate of biomass (d.a.f.), is
shown in Fig. 5. As observed, there was not a clear increase in gas yield
with OP for any of the SR-SBR tested. It would be expected, however,
that the gas yield would increase with OP, since for the same SR
and SBR, the temperature attained in the gasifier was higher and,

Fig. 3. Gas composition as a function of OP for different SR and SBR levels. : SR=High,
SBR=Low (combination 1); : SR=High, SBR=High (combination 2); : SR=Low,
SBR=Low (combination 3); SR=Low, SBR=High (combination 4). (a) carbon monoxide,
(b) hydrogen, (c) carbon dioxide and (d) methane.

the temperature and residence time of the gas are sufficiently high.
This is analysed below by comparing the simultaneous effects on all
the species.
Fig. 3(c) shows the effect of OP on the carbon dioxide, evidencing Fig. 4. Gas yield as a function of OP for different SR and SBR levels. : SR=High, SBR=Low
behaviour compatible with the behaviour of carbon monoxide shown (combination 1); : SR=High, SBR=High (combination 2); : SR=Low, SBR=Low (com-
above; carbon dioxide is the result of both biomass devolatilisation bination 3); SR=Low, SBR=High (combination 4). (a) carbon monoxide and (b) hydrogen.
 M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685 683

of oxygen) is intended to be used for burning in thermal applications

Fig. 5. Gas yield as a function of OP for different SR and SBR levels. : SR = High,
SBR = Low (combination 1); : SR = High, SBR = High (combination 2); : SR = Low,
SBR = Low (combination 3); SR = Low, SBR = High (combination 4).
or, if the gas is properly cleaned, for power production. The LHV of the
gas must consistently include the gross gas, including nitrogen. This is
not the usual way to calculate the LHV in most of air–steam
gasification works, where the LHV is expressed on a nitrogen-free
basis because the gas is mainly aimed at medium-heating-value gas
production or for high-hydrogen gas production.
Gasification efficiency, defined as the ratio of combustion heat of
the produced gas calculated using the LHV defined above, to
combustion heat of the biomass, is shown in Fig. 7. Cold gasification
efficiency assumes a temperature of the product gases at 25 °C, so the
sensitive heat of the gas is not taken into account. Fig. 7(a) displays the
cold gasification efficiency as a function of OP for the two levels of SR
and SBR investigated, showing that the increase of OP in the enriched
air increased the efficiency from 55% up to 70%, the latter
corresponding to the higher SR tested. Fig. 7(b) depicts the effect of

consequently, more gas was produced via more intense devolatilisa-

tion. In addition, enhanced tar decomposition and char gasification
would be expected, which would contribute to further increasing the
gas yield. The behaviour observed in Fig. 5, however, probably has to
do with the way that the gas yield was calculated. It included the
contribution of the four main species measured online: CO, CO2, H2
and CH4.
Fig. 6 shows the lower heating value of the product gas (LHV),
defined as the heat of combustion at 25 °C of 1 Nm3 of gas, as a
function of the OP for the four combinations of SBR and SR tested.
As expected, the higher the OP, the higher the LHV of the gas. The
analysis of gas composition in Figs. 3 and 4 explains the behaviour
observed for the LHV in Fig. 6. Low SR leads to more combustible gas
species in the gas produced, whereas higher OP makes it possible to
reach a higher temperature in the gasifier with the same degree of
fuel oxidation, i.e., SR level. Low SR and high OP therefore lead to
enhancement of reforming reactions. In addition, it is expected that,
up to some values of SBR, not known a priori, the higher concentration
of steam leads to higher LHV, because the presence of steam enhances
the reforming reactions even more if a sufficiently high temperature
is maintained in the reactor. We can clearly envisage an optimised
operation window by proper synchronisation of the various input
parameters.
The following are a few comments on the numerical values of LHV
given in this work. On the one hand, the LHV is calculated taking into
account the energy content of H2, CO and CH4. Other fuel gases such as
C2 and other light hydrocarbons (which were not measured), as well
as tar, were not considered in the computed LHV, so the actual LHV of
the gas should be higher than reported here. On the other hand, the
LHV of the dry gas, includes the presence of nitrogen, because the gas
produced in this process (an autothermal gasifier using air as a source

Fig. 6. LHV of the product gas as a function of OP for different SR and SBR levels. : SR=High,
SBR=Low (combination 1); : SR=High, SBR=High (combination 2); : SR=Low,
SBR=Low (combination 3); SR=Low, SBR=High (combination 4).
Fig. 7. Cold gasification efficiency under different process conditions. (a) Cold gasification
efficiency as a function OP for different SR and SBR levels. : SR=High, SBR=Low (combination
1); : SR=High, SBR=High (combination 2); : SR=Low, SBR=Low (combination 3);
SR=Low, SBR=High (combination 4); (b) Cold gasification efficiency as a function of OP for
SR≈0.35 at three levels of SBR: : 0-0.2, : 0.2-0.4 and : 0.4-0.6; (c) Cold gasification efficiency
as a function of SBR for SR≈0.35 at four OP: : OP=21%; : OP=30%; : OP=35%;
OP=40%.
684 M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685
Fig. 8. Carbon conversion under different process conditions. (a) Carbon conversion as a
function of OP for different SR and SBR levels. : SR= High, SBR = Low (combination 1);
: SR= High, SBR = High (combination 2); : SR= Low, SBR = Low (combination 3);
SR= Low, SBR = High (combination 4); (b) Carbon conversion as a function of SBR for
SR≈ 0.35 at four OP: : OP= 21%; : OP= 30%; : OP= 35%; OP= 40%.
the OP over cold gasification efficiency for a given value of SR ≈ 0.35 at
three SBR levels: low (0–0.2), medium (0.2–0.4) and high (0.4–0.6).
The increase in the cold gasification efficiency for all levels of SBR
tested is clearly shown. The average value of efficiency goes from 60%
(for air, i.e., OP = 21%) to 67% (OP = 40%).
Cold gasification efficiency as a function of SBR with different OP is
given in Fig. 7(c), which shows that the use of steam increased the
efficiency up to 70% for OP = 40%. Although there are not enough tests
to draw more precise trend curves, it is observed that the efficiency
reached a maximum. This was argued previously when analysing the
effects of process conditions on gas composition and yields. In
particular, a maximum is observed in Fig. 7 (c) for OP = 40%. The cases
of OP = 21, 30 and 35% suggest similar behaviour, though the lack of
points does not allow us to define clear trends. Despite this
uncertainty, Fig. 7(c) suggests that there is a threshold of SBR at
which the presence of steam seems to lower the efficiency. The
appearance of maximum efficiency at intermediate values of SBR is
explained by the compromise between the enhancement of the
gasification reactions and the reduction of the bed temperature that
occurs as steam addition is increased. This result suggests that there is
an optimal steam flow rate, which seems to be in the range 0.25–0.35
for SR = 0.35 and OP of 40%.
Fig. 8(a) illustrates carbon conversion, calculated as the difference
between the carbon flow rate in the feed and cyclone ash, divided by
the flow rate of carbon in the feed. The flow rate of carbon for both
streams is computed by the product of the stream (fuel and ash) flow
rate and the total carbon content determined by ASTM analysis.
It is shown that carbon conversion increases with OP for all SBR
and SR levels, ranging from 91 to 97%. The behaviour shown in Fig. 8
is explained by the improvement of the gasification reactions at
high temperature, leading to greater char conversion. This is sup-
ported by the analysis of the ash collected in the cyclones. The increase
in carbon conversion in Fig. 8 was higher in the lower range of OP,
from 21% to 30%.
Fig. 8(b) shows the effect of SBR on carbon conversion for the four
OP studied, demonstrating that the use of steam affects the carbon
conversion only slightly. This result agrees with the analysis of the
efficiency optimum given in Fig. 7.
4. Conclusions
The effect of oxygen concentration in the gasification agent was
studied experimentally in a bubbling fluidised-bed wood gasifier
(FBG). In this study we have extended previous work on steam-air
gasification in which the oxygen content was supplied by air (21%
(v/v)), by allowing the oxygen in this case to vary in purity up to
40% (v/v) in the enriched air. The choice of greater oxygen purity
(40% (v/v)) was made because that is the oxygen concentration that
can be produced using commercial air separators based on mem-
brane technologies. Two different levels of SR and SBR were tested
(high and low), in order to evaluate the effect of enriched air for
different process conditions.
The use of enriched air reduces the nitrogen dilution effect,
increasing the gasification temperature. This allows the addition of
low-quality steam while maintaining the thermal level in the gasifier.
The appropriate combination of temperature and steam leads to
higher CO and H2 yields, heating value, carbon conversion and
gasification efficiency. The optimisation of steam addition using
enriched air leads to a maximum efficiency of 70% for steam-to-
biomass ratio around 0.3 at 40% oxygen purity.
The gasification concept studied in this work seems to be an
interesting option for standalone direct air–blown FBG, considerably
improving the process efficiency while keeping the costs relatively
low as compared to oxygen-steam gasification.
Acknowledgments
This work was partly financed by the European Commission, the
Commission of Science and Technology of Spain and the Regional
Government of Andalusia (Junta de Andalucía). The authors also
acknowledge Antonio Albea and Antonio Cabello for their assistance
in performing the experimental work.
References
[1] E. Kurkela, M. Nieminen, P. Simell, Development and Commercialisation of
Biomass and Waste Gasification Technologies from Reliable and Robust Co-firing
Plants Towards Synthesis Gas Production and Advanced Power Cycles, Proceedings
of the 2nd World Conference on Biomass for Energy, Industry and Climate
Protection, Rome, Italy, 10-14 May 2004.
[2] J. Corella, J.M. Toledo, G. Molina, A review on dual fluidized-bed biomass gasifiers,
Ind. Eng. Chem. Res. 46 (2007) 6831–6839.
[3] J. Gil, M.P. Aznar, M.A. Caballero, E. Francés, J. Corella, Biomass gasification in
fluidized bed at pilot scale with steam-oxygen mixtures. product distribution for
very different operating conditions, Energy Fuels 11 (1997) 1109–1118.
[4] Y. Wang, C.M. Kinoshita, Experimental analysis of biomass gasification with steam
and oxygen, Sol. Energy 49 (1992) 153–158.
[5] K. Maniatis, Progress in biomass gasification: an overview, in: A.V. Bridgwater (Ed.),
Progress in Thermochemical Biomass Conversion, Blackwell Science, London, 2001,
pp. 1–31.
[6] M.A. Paisley, R.P. Overend, The SylvaGas Process from Future Energy Resources - A
Commercialization Success, 12th European Biomass Conference, Amsterdam, The
Netherlands, 17-21 June 2002.
[7] R. Rauch, Steam Gasification of Biomass at CHP Plant in Güssing - Status of the
Demonstration Plant, Proceedings of the 2nd World Conference on Biomass for
Energy, Industry and Climate Protection, Rome, Italy, 10-14 May 2004.
[8] I. Narváez, A. Orio, J. Corella, M.P. Aznar, Biomass gasification with air in an
atmospheric bubbling fluidized bed. effect of six operational variables on the
quality of the produced raw gas, Ind. Eng. Chem. Res. 34 (1996) 2110–2120.
[9] W.P. Walawender, D.A. Hoveland, L.T. Fan, Steam gasification of pure cellulose.
1. Uniform temperature profile, Ind. Eng. Chem. Process Des. Dev. 24 (1985)
813–817.
[10] J. Herguido, J. Corella, J. González-Sanz, Steam gasification of lignocellulosic
residues in a fluidized bed at a small pilot scale. Effect of the type of feedstock, Ind.
Eng. Chem. Res. 31 (1992) 1274–1282.
 M. Campoy et al. / Fuel Processing Technology 90 (2009) 677–685 685

[11] C. Franco, F. Pinto, I. Gulyurtlu, I. Cabrita, The study of reactions influencing the Glossary
biomass steam gasification process, Fuel 82 (2003) 835–842.
[12] S.S. Sadaka, A.E. Ghaly, M.A. Sabbah, Two phase biomass air–steam gasification
d.a.f. dry and ash-free
model for fluidized bed reactors: part III-model validation, Biomass Bioenergy 22
d.a.f.b. dry and ash-free biomass
(2002) 479–487.
FB fluidised bed
[13] P. Lv, J. Chang, Z. Xiong, H. Huang, C. Wu, Y. Chen, Biomass air–steam gasification in
FBG fluidised bed gasifier (or gasification)
a fluidized bed to produce hydrogen-rich gas, Energy Fuels 17 (2003) 677–682.
GY Gas yield, Nm3 dry gas free of nitrogen/kg d.a.f.b.
[14] P.M. Lv, Z.H. Xiong, J. Chang, C.Z. Wu, Y. Chen, J.X. Xhu, An experimental study on
LHV low heating value, MJ/Nm3 dry gas
biomass air–steam gasification in a fluidized bed, Bioresour. Technol. 95 (2004)
OP oxygen purity, % v/v
95–101.
SBR steam-to-biomass ratio, kg/kg d.a.f.b.
[15] S. Turn, C. Kinoshita, Z. Zhang, D. Ishimura, J. Zhou, An experimental investigation of
SR stoichiometric ratio, kg/kg
hydrogen production from biomass gasification, Int. J. Hydrogen Energy 8 (1998)
Tb mean bed temperature, °C
641–648.
T⁎b theoretically-calculated mean bed temperature, °C
[16] M. Campoy, A. Gómez-Barea, A.L. Villanueva, P. Ollero, Air–steam gasification of
tbed gas residence time in the bed, s
biomass in a fluidized bed under simulated autothermal and adiabatic conditions,
tfreeboard gas residence time in the freeboard, s
Ind. Eng. Chem. Res. 47 (2008) 5957–5965.
WGSR water-gas-shift reaction
[17] A. Gómez-Barea, L.F. Vilches, C. Leiva, M. Campoy, C. Fernández-Pereira, Plant
optimisation and ash recycling in fluidised bed waste gasification, Chem. Eng. J. 146
(2009) 227–236.